Ideas for Getting the word out

[David Blevins <da...@gmail.com>]

Some IRC chat resulted in some neat ideas on ways to dramatically improve how easy it is to consume and learn about OpenEJB.

# Examples

One thing that hit home is that we have now 46 examples in trunk!!!  Amazing.  However the benefit of that is dramatically reduced as most of them are only available in zip file form.  Links to svn aren't that effective.  And examples change too as the technology improves so keeping wiki pages up to date is hard -- which version of the example do you show?  What if you really want all versions available to see?

So the idea was to use README files that are formatted in Markdown and use that to generate a page for each example.  No more having part of the example in svn and part of it in confluence and then always breaking.  A stackoverflow inspired solution -- Markdown + Google's prettyprint.

All the examples are in a zip file that is published in the maven repository, so we could use a maven program to pull them down and extract them to the target directory where we can then do our little page generation.  A stub with those ideas in comments:

  http://svn.apache.org/repos/asf/openejb/trunk/sandbox/tools/src/main/java/org/apache/openejb/tools/examples/GenerateIndex.java

# Retweeting

We should monitor this feed http://twitter.com/#!/OpenEJB/contributors and retweet anything that mentions OpenEJB.

So if anyone in the contributors list tweeted about OpenEJB, the OpenEJB twitter account would retweet it.

Two things will happen as a result:
  -  The more activity on the OpenEJB twitter account the more followers it will get
  -  The more @joe and other contributors are seen on the account, the more followers they will get

The OpenEJB twitter account has more followers than most everyone else so getting it to retweet is a good way to expose people to all our wonderful contributors and get them some followers and help the project at the same time.

The result is we as a community will have more ability overall to get the word out!

Twitterfeed.com was the obvious first idea, but it turns out twitter does not allow you to post content from twitter back onto twitter.  Not unless you use their API http://dev.twitter.com/doc .  So we could maybe hack together some tool we run hourly and retweet things that contributors tweet.  A little stub with comments here:

  http://svn.apache.org/repos/asf/openejb/trunk/sandbox/tools/src/main/java/org/apache/openejb/tools/twitter/Retweet.java


Both are up for grabs!  If you're looking for something to do, either one would be excellent ways to improve the project!



-David


  

Re: Ideas for Getting the word out -- Examples documentation

[stratwine <to...@gmail.com>]

Looks cool ! I'll join in and do some CSS styling tomorrow.

-Vishwa

--
View this message in context: http://openejb.979440.n4.nabble.com/Ideas-for-Getting-the-word-out-tp3558206p3657785.html
Sent from the OpenEJB Dev mailing list archive at Nabble.com.

Re: Ideas for Getting the word out -- Examples documentation

[Romain Manni-Bucau <rm...@gmail.com>]

for the theme i'll maybe have a look monday evening but before i want to add
the scan of .xml/properties which can be cool too.

well, if anybody wants to make a better template we can work together to
make it better than it is today :)

- Romain

2011/7/10 David Blevins <da...@gmail.com>

>
> On Jul 9, 2011, at 3:42 PM, Marius Kruger wrote:
>
> > (sorry for resurrecting the topic and maybe talking about stuff that
> > was already dealt with)
> >
> > On 30 May 2011 01:35, David Blevins <da...@gmail.com> wrote:
> > ...
> >> We could create a new space for 4.0 and copy the examples there and
> update them there, but our track record on getting any content at all up in
> the wiki for each example is incredibly poor already.  Currently only about
> 13 of 46 examples are documented in the wiki.  No way we'd do any better
> with double the pages to maintain.
> >
> > I count 15 examples on http://openejb.apache.org/examples.html
> > It is terrible that so few are effectively visible to the public with
> > no mention that there are actually more available. (I know I wish I
> > realised that before now, I assumed they were all there)
> > I think an easy first step is to at least mention that more examples
> > are available here:
> > http://svn.apache.org/viewvc/openejb/trunk/openejb3/examples/
> > (I tried to do that but it seems I don't have permission (username:
> amanica))
>
> Added you to the openejb-contributors group so you can edit docs.
>
> >> # Better possible solution
> >>
> >> It's a bit of a nightmare.  Seems if we ditched the wiki part of it
> completely and just made the README the sole source of info instead of
> splitting it, converted it to markdown which looks fine in plain text, we
> might stand a better chance of keeping things current.
> >
> > +1 for that
> > * stay DRY, don't make extra work, things should just appear :-)
> > * configure once to maybe get updated by the continuous integration
> > server and have them updated forever.
> > * I'd say keep them separate for each version. That is how bazaar dvcs
> > does it [1]
> >  (disclaimer: I've worked on bazaar a bit, having to write eg.
> > changelogs in rest syntax[2],
> >  I think it works quite nice, they use sphinx[3] to process it)
> > * maybe I can somehow lend a hand with some of this, time permitting.
> >  (not promising anything :-)
>
> Any help is more than welcome :)  Romain did some great work in this area
> on a stackoverflow inspired example listing:
>
>  http://ci.apache.org/projects/openejb/examples-generated/
>
> Still very lacking in the actual README files but is a concrete step in the
> right direction.  At this point I think it's waiting for someone with some
> better CSS skills to make it look pretty.  Would be great if we could use
> the same look and feel as the website.  All templates and things are here:
>
>  http://svn.apache.org/repos/asf/openejb/trunk/sandbox/tools/
>
> Any hacking at all in this area is totally welcome.  Any features or bells
> and whistles are great.  Anything that makes digesting such a large quantity
> of examples that much easier.  Once we get CDI examples in as well, we'll
> easily reach 100 examples.
>
>
> -David
>
>

Re: Ideas for Getting the word out -- Examples documentation

[David Blevins <da...@gmail.com>]

The real page is here:

  https://cwiki.apache.org/confluence/pages/editpage.action?spaceKey=OPENEJBx30&title=Examples+Table

Try that one.  I think we have it setup so that the main site (OPENEJB) as stricter perms than the documentation space (OPENEJBx30).

Hopefully it will all be a matter of simple patches pretty soon :)


-David


On Jul 11, 2011, at 3:42 PM, Karan Malhi wrote:

> Hi Marius,
> 
> This is a bit strange. I just checked and you are in the OpenEJB
> contributors group, which means you should be able to add content.
> 
> On Sun, Jul 10, 2011 at 1:15 PM, Marius Kruger <am...@gmail.com> wrote:
>> On 10 July 2011 01:46, David Blevins <da...@gmail.com> wrote:
>> ...
>> 
>>> Added you to the openejb-contributors group so you can edit docs.
>>> 
>> 
>> thank you and Karan for trying to add me, but it did not seem to work.
>> If I try to edit the examples page:
>> https://cwiki.apache.org/confluence/pages/editpage.action?spaceKey=OPENEJB&title=Examples
>>  I still get "You are not permitted to perform this operation."
>> 
>> --
>> <>< Marius ><>
>> 
> 
> 
> 
> -- 
> 
> Karan Singh Malhi
> twitter.com/KaranSinghMalhi

Re: Ideas for Getting the word out -- Examples documentation

[Karan Malhi <ka...@gmail.com>]

Hi Marius,

This is a bit strange. I just checked and you are in the OpenEJB
contributors group, which means you should be able to add content.

On Sun, Jul 10, 2011 at 1:15 PM, Marius Kruger <am...@gmail.com> wrote:
> On 10 July 2011 01:46, David Blevins <da...@gmail.com> wrote:
> ...
>
>> Added you to the openejb-contributors group so you can edit docs.
>>
>
> thank you and Karan for trying to add me, but it did not seem to work.
> If I try to edit the examples page:
> https://cwiki.apache.org/confluence/pages/editpage.action?spaceKey=OPENEJB&title=Examples
>  I still get "You are not permitted to perform this operation."
>
> --
> <>< Marius ><>
>



-- 

Karan Singh Malhi
twitter.com/KaranSinghMalhi

Re: Ideas for Getting the word out -- Examples documentation

[Marius Kruger <am...@gmail.com>]

On 10 July 2011 01:46, David Blevins <da...@gmail.com> wrote:
...

> Added you to the openejb-contributors group so you can edit docs.
>

thank you and Karan for trying to add me, but it did not seem to work.
If I try to edit the examples page:
https://cwiki.apache.org/confluence/pages/editpage.action?spaceKey=OPENEJB&title=Examples
 I still get "You are not permitted to perform this operation."

-- 
<>< Marius ><>

Re: Ideas for Getting the word out -- Examples documentation

[David Blevins <da...@gmail.com>]

On Jul 11, 2011, at 9:09 AM, stratwine wrote:

> 
> David Blevins-2 wrote:
>> 
>> 
>> Anything that makes digesting such a large quantity of examples that much
>> easier.  
>> 
>> 
> 
> An option to view examples grouped as trails maybe ? Something like the 
> http://download.oracle.com/javase/tutorial/java/data/characters.html  java
> trail   ?

That could be really good.   Shouldn't be too hard to generate that.  Could even use a trails.properties file:

  learning_ejb = simple-stateless, simple-stateful, simple-singleton

Then just link them together with "previous" and "next".

Or maybe something else entirely... just brainstorming :)


-David

Re: Ideas for Getting the word out -- Examples documentation

[stratwine <to...@gmail.com>]

David Blevins-2 wrote:
> 
> 
> Anything that makes digesting such a large quantity of examples that much
> easier.  
> 
> 

An option to view examples grouped as trails maybe ? Something like the 
http://download.oracle.com/javase/tutorial/java/data/characters.html  java
trail   ?


--
View this message in context: http://openejb.979440.n4.nabble.com/Ideas-for-Getting-the-word-out-tp3558206p3660003.html
Sent from the OpenEJB Dev mailing list archive at Nabble.com.

Re: Ideas for Getting the word out -- Examples documentation

[David Blevins <da...@gmail.com>]

On Jul 9, 2011, at 3:42 PM, Marius Kruger wrote:

> (sorry for resurrecting the topic and maybe talking about stuff that
> was already dealt with)
> 
> On 30 May 2011 01:35, David Blevins <da...@gmail.com> wrote:
> ...
>> We could create a new space for 4.0 and copy the examples there and update them there, but our track record on getting any content at all up in the wiki for each example is incredibly poor already.  Currently only about 13 of 46 examples are documented in the wiki.  No way we'd do any better with double the pages to maintain.
> 
> I count 15 examples on http://openejb.apache.org/examples.html
> It is terrible that so few are effectively visible to the public with
> no mention that there are actually more available. (I know I wish I
> realised that before now, I assumed they were all there)
> I think an easy first step is to at least mention that more examples
> are available here:
> http://svn.apache.org/viewvc/openejb/trunk/openejb3/examples/
> (I tried to do that but it seems I don't have permission (username: amanica))

Added you to the openejb-contributors group so you can edit docs.

>> # Better possible solution
>> 
>> It's a bit of a nightmare.  Seems if we ditched the wiki part of it completely and just made the README the sole source of info instead of splitting it, converted it to markdown which looks fine in plain text, we might stand a better chance of keeping things current.
> 
> +1 for that
> * stay DRY, don't make extra work, things should just appear :-)
> * configure once to maybe get updated by the continuous integration
> server and have them updated forever.
> * I'd say keep them separate for each version. That is how bazaar dvcs
> does it [1]
>  (disclaimer: I've worked on bazaar a bit, having to write eg.
> changelogs in rest syntax[2],
>  I think it works quite nice, they use sphinx[3] to process it)
> * maybe I can somehow lend a hand with some of this, time permitting.
>  (not promising anything :-)

Any help is more than welcome :)  Romain did some great work in this area on a stackoverflow inspired example listing:

  http://ci.apache.org/projects/openejb/examples-generated/

Still very lacking in the actual README files but is a concrete step in the right direction.  At this point I think it's waiting for someone with some better CSS skills to make it look pretty.  Would be great if we could use the same look and feel as the website.  All templates and things are here:

  http://svn.apache.org/repos/asf/openejb/trunk/sandbox/tools/

Any hacking at all in this area is totally welcome.  Any features or bells and whistles are great.  Anything that makes digesting such a large quantity of examples that much easier.  Once we get CDI examples in as well, we'll easily reach 100 examples.


-David

Re: Ideas for Getting the word out -- Examples documentation

[Karan Malhi <ka...@gmail.com>]

On Sat, Jul 9, 2011 at 6:42 PM, Marius Kruger <am...@gmail.com> wrote:
> (sorry for resurrecting the topic and maybe talking about stuff that
> was already dealt with)
I would say thank you for bringing our attention back to it.
>
> On 30 May 2011 01:35, David Blevins <da...@gmail.com> wrote:
> ...
>> We could create a new space for 4.0 and copy the examples there and update them there, but our track record on getting any content at all up in the wiki for each example is incredibly poor already.  Currently only about 13 of 46 examples are documented in the wiki.  No way we'd do any better with double the pages to maintain.
>
> I count 15 examples on http://openejb.apache.org/examples.html
> It is terrible that so few are effectively visible to the public with
> no mention that there are actually more available. (I know I wish I
> realised that before now, I assumed they were all there)
> I think an easy first step is to at least mention that more examples
> are available here:
> http://svn.apache.org/viewvc/openejb/trunk/openejb3/examples/
> (I tried to do that but it seems I don't have permission (username: amanica))
>
You should now have access to edit the page
>> # Better possible solution
>>
>> It's a bit of a nightmare.  Seems if we ditched the wiki part of it completely and just made the README the sole source of info instead of splitting it, converted it to markdown which looks fine in plain text, we might stand a better chance of keeping things current.
>
> +1 for that
> * stay DRY, don't make extra work, things should just appear :-)
> * configure once to maybe get updated by the continuous integration
> server and have them updated forever.
> * I'd say keep them separate for each version. That is how bazaar dvcs
> does it [1]
>  (disclaimer: I've worked on bazaar a bit, having to write eg.
> changelogs in rest syntax[2],
>  I think it works quite nice, they use sphinx[3] to process it)
> * maybe I can somehow lend a hand with some of this, time permitting.
>  (not promising anything :-)
All ideas are welcome. Anything to make it simpler and up-to-date.
>
> --
> [1] http://doc.bazaar.canonical.com/en/
> [2] http://docutils.sourceforge.net/rst.html
> [3] http://sphinx.pocoo.org/
> <>< Marius ><>
>



-- 

Karan Singh Malhi
twitter.com/KaranSinghMalhi

Re: Ideas for Getting the word out -- Examples documentation

[Marius Kruger <am...@gmail.com>]

On 10 July 2011 00:42, Marius Kruger <am...@gmail.com> wrote:
> I think an easy first step is to at least mention that more examples
> are available

[done]

I added the following to
https://cwiki.apache.org/confluence/display/OPENEJBx30/Examples+Table
, please tweak if you can do better:)

But I don't see the main examples page[1] updating, I assume there is
some reviewing or scheduled update that will make that happen later.
Or is the pages actually maintained separately?

==
Unfortunately all the examples are not documented here yet, more
examples can be [browsed in version
control|http://svn.apache.org/viewvc/openejb/trunk/openejb3/examples/]
or check it out with subversion ({{svn co
http://svn.apache.org/repos/asf/openejb/trunk/openejb3/examples}}).
==

[1] http://openejb.apache.org/examples.html

-- 
<>< Marius ><>

Re: Ideas for Getting the word out -- Examples documentation

[Marius Kruger <am...@gmail.com>]

(sorry for resurrecting the topic and maybe talking about stuff that
was already dealt with)

On 30 May 2011 01:35, David Blevins <da...@gmail.com> wrote:
...
> We could create a new space for 4.0 and copy the examples there and update them there, but our track record on getting any content at all up in the wiki for each example is incredibly poor already.  Currently only about 13 of 46 examples are documented in the wiki.  No way we'd do any better with double the pages to maintain.

I count 15 examples on http://openejb.apache.org/examples.html
It is terrible that so few are effectively visible to the public with
no mention that there are actually more available. (I know I wish I
realised that before now, I assumed they were all there)
I think an easy first step is to at least mention that more examples
are available here:
http://svn.apache.org/viewvc/openejb/trunk/openejb3/examples/
(I tried to do that but it seems I don't have permission (username: amanica))

> # Better possible solution
>
> It's a bit of a nightmare.  Seems if we ditched the wiki part of it completely and just made the README the sole source of info instead of splitting it, converted it to markdown which looks fine in plain text, we might stand a better chance of keeping things current.

+1 for that
* stay DRY, don't make extra work, things should just appear :-)
* configure once to maybe get updated by the continuous integration
server and have them updated forever.
* I'd say keep them separate for each version. That is how bazaar dvcs
does it [1]
  (disclaimer: I've worked on bazaar a bit, having to write eg.
changelogs in rest syntax[2],
  I think it works quite nice, they use sphinx[3] to process it)
* maybe I can somehow lend a hand with some of this, time permitting.
  (not promising anything :-)

-- 
[1] http://doc.bazaar.canonical.com/en/
[2] http://docutils.sourceforge.net/rst.html
[3] http://sphinx.pocoo.org/
<>< Marius ><>

Re: Ideas for Getting the word out

[Romain Manni-Bucau <rm...@gmail.com>]

+1

for the design i like "fashion styles" = either very flat either 3D. I'll
try to have a look next week.

- Romain

2011/5/31 David Blevins <da...@gmail.com>

>
> On May 30, 2011, at 12:14 AM, Romain Manni-Bucau wrote:
>
> > are we talking about a openejb communication refactoring?
>
> Wasn't originally thinking so broad, but it has turned into that... which
> is pretty cool :)
>
> > from my side i know there are the website, an apache blog etc...
> >
> > couldn't it be merged into a kind of portal with a blog part, a tutorial
> > part, a community a wiki part etc...? All these pages with a new layout
> and
> > template?
> >
> > IMHO OpenEJB is sexy to use but its communication needs to be more
> > centralized and more attractive.
>
> I'm open to that.  Karan had some ideas on IRC for redoing the front page.
>  My response to that was that maybe we need to redo the whole site.  Perhaps
> a fresh new look in celebration of OpenEJB 4 and TomEE.
>
> I just posted about the CMS.  Might be a good opportunity for some playing.
>
> Sounds like you have some ideas.  Any interest in mocking up something?
>
> The only thoughts I've had on it are that we should go with a standard
> fixed width "one middle column" approach.  Maybe reserve the purple to just
> a top band.
>
>
> -David
>
>

Re: Ideas for Getting the word out

[David Blevins <da...@gmail.com>]

On May 30, 2011, at 12:14 AM, Romain Manni-Bucau wrote:

> are we talking about a openejb communication refactoring?

Wasn't originally thinking so broad, but it has turned into that... which is pretty cool :)

> from my side i know there are the website, an apache blog etc...
> 
> couldn't it be merged into a kind of portal with a blog part, a tutorial
> part, a community a wiki part etc...? All these pages with a new layout and
> template?
> 
> IMHO OpenEJB is sexy to use but its communication needs to be more
> centralized and more attractive.

I'm open to that.  Karan had some ideas on IRC for redoing the front page.  My response to that was that maybe we need to redo the whole site.  Perhaps a fresh new look in celebration of OpenEJB 4 and TomEE.

I just posted about the CMS.  Might be a good opportunity for some playing.

Sounds like you have some ideas.  Any interest in mocking up something?

The only thoughts I've had on it are that we should go with a standard fixed width "one middle column" approach.  Maybe reserve the purple to just a top band.


-David

Re: Ideas for Getting the word out -- Examples documentation

[Romain Manni-Bucau <rm...@gmail.com>]

are we talking about a openejb communication refactoring?

from my side i know there are the website, an apache blog etc...

couldn't it be merged into a kind of portal with a blog part, a tutorial
part, a community a wiki part etc...? All these pages with a new layout and
template?

IMHO OpenEJB is sexy to use but its communication needs to be more
centralized and more attractive.

- Romain

2011/5/30 David Blevins <da...@gmail.com>

>
> On May 28, 2011, at 4:50 PM, David Blevins wrote:
>
> > Some IRC chat resulted in some neat ideas on ways to dramatically improve
> how easy it is to consume and learn about OpenEJB.
> >
> > # Examples
> >
> > One thing that hit home is that we have now 46 examples in trunk!!!
>  Amazing.  However the benefit of that is dramatically reduced as most of
> them are only available in zip file form.  Links to svn aren't that
> effective.  And examples change too as the technology improves so keeping
> wiki pages up to date is hard -- which version of the example do you show?
>  What if you really want all versions available to see?
> >
> > So the idea was to use README files that are formatted in Markdown and
> use that to generate a page for each example.  No more having part of the
> example in svn and part of it in confluence and then always breaking.  A
> stackoverflow inspired solution -- Markdown + Google's prettyprint.
> >
> > All the examples are in a zip file that is published in the maven
> repository, so we could use a maven program to pull them down and extract
> them to the target directory where we can then do our little page
> generation.  A stub with those ideas in comments:
> >
> >
> http://svn.apache.org/repos/asf/openejb/trunk/sandbox/tools/src/main/java/org/apache/openejb/tools/examples/GenerateIndex.java
>
> Haven't seen any discussion specifically about this, so breaking it out as
> a subtopic so we can hopefully get someone to work on it.  If we keep it
> simple, it's maybe 500 line mashup of the right libraries.
>
> # The problem
>
> The 3.1.4 and trunk examples are pretty different now, so the confluence
> content now doesn't line up with what we have:
>
>  http://openejb.apache.org/3.0/simple-stateless-example.html
>
> # Possible fixes
>
> We could update that webpage to reflect the trunk version, but then we'd
> lose the documentation for the currently supported release which is 3.1.4
> and all the pages would reflect stuff that isn't released.
>
> We could create a new space for 4.0 and copy the examples there and update
> them there, but our track record on getting any content at all up in the
> wiki for each example is incredibly poor already.  Currently only about 13
> of 46 examples are documented in the wiki.  No way we'd do any better with
> double the pages to maintain.
>
> # Previous approaches
>
> At first we started out adding a README.txt to each example and including
> that in the wiki page.  But that got hard to keep up as to update an example
> doc you had to tweak both the svn and wiki page at the same time.  So the
> README.txt files started to get neglected and now most of them are flat out
> wrong.
>
> # Better possible solution
>
> It's a bit of a nightmare.  Seems if we ditched the wiki part of it
> completely and just made the README the sole source of info instead of
> splitting it, converted it to markdown which looks fine in plain text, we
> might stand a better chance of keeping things current.
>
> There are already generic markdown+svn solutions like Maven doxia markdown
> and the Apache CMS which is markdown based.  We could potentially use those.
>  If we wanted to go the "build it" route instead of the "reuse" route, it
> would probably be about 500 lines of code with the right libraries.
>
> # Proof of concept
>
>
> http://svn.apache.org/repos/asf/openejb/trunk/openejb3/examples/simple-stateless/README.md
>  http://people.apache.org/~dblevins/simple-stateless.html
>
> Side note is that they also happen to look nice on github (hit refresh
> twice on this page to load up the readme):
>
>
> https://github.com/apache/openejb/tree/trunk/openejb3/examples/simple-stateless
>
>
> -David
>
>
>
>
>
>

Re: Ideas for Getting the word out -- Examples documentation

[David Blevins <da...@gmail.com>]

On May 28, 2011, at 4:50 PM, David Blevins wrote:

> Some IRC chat resulted in some neat ideas on ways to dramatically improve how easy it is to consume and learn about OpenEJB.
> 
> # Examples
> 
> One thing that hit home is that we have now 46 examples in trunk!!!  Amazing.  However the benefit of that is dramatically reduced as most of them are only available in zip file form.  Links to svn aren't that effective.  And examples change too as the technology improves so keeping wiki pages up to date is hard -- which version of the example do you show?  What if you really want all versions available to see?
> 
> So the idea was to use README files that are formatted in Markdown and use that to generate a page for each example.  No more having part of the example in svn and part of it in confluence and then always breaking.  A stackoverflow inspired solution -- Markdown + Google's prettyprint.
> 
> All the examples are in a zip file that is published in the maven repository, so we could use a maven program to pull them down and extract them to the target directory where we can then do our little page generation.  A stub with those ideas in comments:
> 
>  http://svn.apache.org/repos/asf/openejb/trunk/sandbox/tools/src/main/java/org/apache/openejb/tools/examples/GenerateIndex.java

Haven't seen any discussion specifically about this, so breaking it out as a subtopic so we can hopefully get someone to work on it.  If we keep it simple, it's maybe 500 line mashup of the right libraries.

# The problem

The 3.1.4 and trunk examples are pretty different now, so the confluence content now doesn't line up with what we have:

  http://openejb.apache.org/3.0/simple-stateless-example.html

# Possible fixes

We could update that webpage to reflect the trunk version, but then we'd lose the documentation for the currently supported release which is 3.1.4 and all the pages would reflect stuff that isn't released.

We could create a new space for 4.0 and copy the examples there and update them there, but our track record on getting any content at all up in the wiki for each example is incredibly poor already.  Currently only about 13 of 46 examples are documented in the wiki.  No way we'd do any better with double the pages to maintain.

# Previous approaches

At first we started out adding a README.txt to each example and including that in the wiki page.  But that got hard to keep up as to update an example doc you had to tweak both the svn and wiki page at the same time.  So the README.txt files started to get neglected and now most of them are flat out wrong.

# Better possible solution

It's a bit of a nightmare.  Seems if we ditched the wiki part of it completely and just made the README the sole source of info instead of splitting it, converted it to markdown which looks fine in plain text, we might stand a better chance of keeping things current.

There are already generic markdown+svn solutions like Maven doxia markdown and the Apache CMS which is markdown based.  We could potentially use those.  If we wanted to go the "build it" route instead of the "reuse" route, it would probably be about 500 lines of code with the right libraries.

# Proof of concept

  http://svn.apache.org/repos/asf/openejb/trunk/openejb3/examples/simple-stateless/README.md
  http://people.apache.org/~dblevins/simple-stateless.html

Side note is that they also happen to look nice on github (hit refresh twice on this page to load up the readme):

  https://github.com/apache/openejb/tree/trunk/openejb3/examples/simple-stateless


-David

Re: Ideas for Getting the word out -- IDE projects examples

[David Blevins <da...@gmail.com>]

On May 29, 2011, at 5:21 AM, Karan Malhi wrote:

> The assumption that everybody knows maven is a dangerous one.

You made a bit of a jump there :)  But important subject so splitting it out.

> Examples
> should also be packaged so that they could be imported into IDE's i.e. zip
> files for eclipse, idea and netbeans. People typically have their favorite
> IDE's already installed on their machines and giving the ability to import
> examples directly into the IDE and "Get Started " is very effective.
> The thing with only relying on maven is that if somebody does not have it
> installed, they first have to install it, then they also need to know a bit
> of it. I know that you can always tell them to simply run mvn clean install
> to run the example, but for people not exposed to maven, there is always a
> feeling of "I think maven did some magic to make it work. Wonder what it
> would take to get this example working in my IDE" - and this is where we can
> potentially leave a bad taste. "First impression is a lasting one" - lets
> try and make that first impression a great one

I wonder if more videos could help here?  We have one, but we could have many more.

  http://vimeo.com/6149008

Probably one of those good ideas lack of time as prevented from happening.


-David

> 
> On Sun, May 29, 2011 at 4:04 AM, Aldrin Leal <al...@leal.eng.br> wrote:
> 
>> For the examples, some could be turned into archetypes (which answers the
>> previous question "What are archetypes useful for?") as a means to reach a
>> wide audience: No need to download sources, leave it up to the archetype
>> (Actually Wicket Docs does just like that)
>> 
>> As for the feeds, I suggest either a planet setup or simply a Twitter User
>> List at first
>> 
>> --
>> -- Aldrin Leal, <al...@leal.eng.br> / http://www.leal.eng.br/mnemetica/
>> 
>> 
>> On Sat, May 28, 2011 at 8:50 PM, David Blevins <david.blevins@gmail.com
>>> wrote:
>> 
>>> Some IRC chat resulted in some neat ideas on ways to dramatically improve
>>> how easy it is to consume and learn about OpenEJB.
>>> 
>>> # Examples
>>> 
>>> One thing that hit home is that we have now 46 examples in trunk!!!
>>> Amazing.  However the benefit of that is dramatically reduced as most of
>>> them are only available in zip file form.  Links to svn aren't that
>>> effective.  And examples change too as the technology improves so keeping
>>> wiki pages up to date is hard -- which version of the example do you
>> show?
>>> What if you really want all versions available to see?
>>> 
>>> So the idea was to use README files that are formatted in Markdown and
>> use
>>> that to generate a page for each example.  No more having part of the
>>> example in svn and part of it in confluence and then always breaking.  A
>>> stackoverflow inspired solution -- Markdown + Google's prettyprint.
>>> 
>>> All the examples are in a zip file that is published in the maven
>>> repository, so we could use a maven program to pull them down and extract
>>> them to the target directory where we can then do our little page
>>> generation.  A stub with those ideas in comments:
>>> 
>>> 
>>> 
>> http://svn.apache.org/repos/asf/openejb/trunk/sandbox/tools/src/main/java/org/apache/openejb/tools/examples/GenerateIndex.java
>>> 
>>> # Retweeting
>>> 
>>> We should monitor this feed http://twitter.com/#!/OpenEJB/contributorsand
>>> retweet anything that mentions OpenEJB.
>>> 
>>> So if anyone in the contributors list tweeted about OpenEJB, the OpenEJB
>>> twitter account would retweet it.
>>> 
>>> Two things will happen as a result:
>>> -  The more activity on the OpenEJB twitter account the more followers
>> it
>>> will get
>>> -  The more @joe and other contributors are seen on the account, the
>> more
>>> followers they will get
>>> 
>>> The OpenEJB twitter account has more followers than most everyone else so
>>> getting it to retweet is a good way to expose people to all our wonderful
>>> contributors and get them some followers and help the project at the same
>>> time.
>>> 
>>> The result is we as a community will have more ability overall to get the
>>> word out!
>>> 
>>> Twitterfeed.com was the obvious first idea, but it turns out twitter does
>>> not allow you to post content from twitter back onto twitter.  Not unless
>>> you use their API http://dev.twitter.com/doc .  So we could maybe hack
>>> together some tool we run hourly and retweet things that contributors
>> tweet.
>>> A little stub with comments here:
>>> 
>>> 
>>> 
>> http://svn.apache.org/repos/asf/openejb/trunk/sandbox/tools/src/main/java/org/apache/openejb/tools/twitter/Retweet.java
>>> 
>>> 
>>> Both are up for grabs!  If you're looking for something to do, either one
>>> would be excellent ways to improve the project!
>>> 
>>> 
>>> 
>>> -David
>>> 
>>> 
>>> 
>>> 
>>> 
>>> 
>>> 
>>> 
>>> 
>> 
> 
> 
> 
> -- 
> Karan Singh Malhi

Re: Ideas for Getting the word out

[Aldrin Leal <al...@leal.eng.br>]

Now we are inventing Spring Roo :)

--
-- Aldrin Leal, <al...@leal.eng.br> / http://www.leal.eng.br/mnemetica/


On Sun, May 29, 2011 at 9:34 AM, dsh <da...@googlemail.com> wrote:

> One thought concerning IDEs - If you export sample projects from an
> IDE and store them in SVN to be imported by someone else you have to
> maintain that IDE-specific samples too cause IDEs of course are
> changing over time and hence you have to tweak such projects every now
> and then. What I like more is having the build tool creating
> IDE-specific projects from the meta descriptors provided by a build
> system cause it could be suspected that such meta descriptors are
> always up-to-date.
>
> Such a build system could be Maven for Java-specific projects or CMake
> for C/C++ specific projects.
>
> Cheers
> Daniel
>
> On Sun, May 29, 2011 at 2:21 PM, Karan Malhi <ka...@gmail.com>
> wrote:
> > The assumption that everybody knows maven is a dangerous one. Examples
> > should also be packaged so that they could be imported into IDE's i.e.
> zip
> > files for eclipse, idea and netbeans. People typically have their
> favorite
> > IDE's already installed on their machines and giving the ability to
> import
> > examples directly into the IDE and "Get Started " is very effective.
> > The thing with only relying on maven is that if somebody does not have it
> > installed, they first have to install it, then they also need to know a
> bit
> > of it. I know that you can always tell them to simply run mvn clean
> install
> > to run the example, but for people not exposed to maven, there is always
> a
> > feeling of "I think maven did some magic to make it work. Wonder what it
> > would take to get this example working in my IDE" - and this is where we
> can
> > potentially leave a bad taste. "First impression is a lasting one" - lets
> > try and make that first impression a great one
> >
> > On Sun, May 29, 2011 at 4:04 AM, Aldrin Leal <al...@leal.eng.br> wrote:
> >
> >> For the examples, some could be turned into archetypes (which answers
> the
> >> previous question "What are archetypes useful for?") as a means to reach
> a
> >> wide audience: No need to download sources, leave it up to the archetype
> >> (Actually Wicket Docs does just like that)
> >>
> >> As for the feeds, I suggest either a planet setup or simply a Twitter
> User
> >> List at first
> >>
> >> --
> >> -- Aldrin Leal, <al...@leal.eng.br> /
> http://www.leal.eng.br/mnemetica/
> >>
> >>
> >> On Sat, May 28, 2011 at 8:50 PM, David Blevins <david.blevins@gmail.com
> >> >wrote:
> >>
> >> > Some IRC chat resulted in some neat ideas on ways to dramatically
> improve
> >> > how easy it is to consume and learn about OpenEJB.
> >> >
> >> > # Examples
> >> >
> >> > One thing that hit home is that we have now 46 examples in trunk!!!
> >> >  Amazing.  However the benefit of that is dramatically reduced as most
> of
> >> > them are only available in zip file form.  Links to svn aren't that
> >> > effective.  And examples change too as the technology improves so
> keeping
> >> > wiki pages up to date is hard -- which version of the example do you
> >> show?
> >> >  What if you really want all versions available to see?
> >> >
> >> > So the idea was to use README files that are formatted in Markdown and
> >> use
> >> > that to generate a page for each example.  No more having part of the
> >> > example in svn and part of it in confluence and then always breaking.
>  A
> >> > stackoverflow inspired solution -- Markdown + Google's prettyprint.
> >> >
> >> > All the examples are in a zip file that is published in the maven
> >> > repository, so we could use a maven program to pull them down and
> extract
> >> > them to the target directory where we can then do our little page
> >> > generation.  A stub with those ideas in comments:
> >> >
> >> >
> >> >
> >>
> http://svn.apache.org/repos/asf/openejb/trunk/sandbox/tools/src/main/java/org/apache/openejb/tools/examples/GenerateIndex.java
> >> >
> >> > # Retweeting
> >> >
> >> > We should monitor this feed
> http://twitter.com/#!/OpenEJB/contributorsand
> >> > retweet anything that mentions OpenEJB.
> >> >
> >> > So if anyone in the contributors list tweeted about OpenEJB, the
> OpenEJB
> >> > twitter account would retweet it.
> >> >
> >> > Two things will happen as a result:
> >> >  -  The more activity on the OpenEJB twitter account the more
> followers
> >> it
> >> > will get
> >> >  -  The more @joe and other contributors are seen on the account, the
> >> more
> >> > followers they will get
> >> >
> >> > The OpenEJB twitter account has more followers than most everyone else
> so
> >> > getting it to retweet is a good way to expose people to all our
> wonderful
> >> > contributors and get them some followers and help the project at the
> same
> >> > time.
> >> >
> >> > The result is we as a community will have more ability overall to get
> the
> >> > word out!
> >> >
> >> > Twitterfeed.com was the obvious first idea, but it turns out twitter
> does
> >> > not allow you to post content from twitter back onto twitter.  Not
> unless
> >> > you use their API http://dev.twitter.com/doc .  So we could maybe
> hack
> >> > together some tool we run hourly and retweet things that contributors
> >> tweet.
> >> >  A little stub with comments here:
> >> >
> >> >
> >> >
> >>
> http://svn.apache.org/repos/asf/openejb/trunk/sandbox/tools/src/main/java/org/apache/openejb/tools/twitter/Retweet.java
> >> >
> >> >
> >> > Both are up for grabs!  If you're looking for something to do, either
> one
> >> > would be excellent ways to improve the project!
> >> >
> >> >
> >> >
> >> > -David
> >> >
> >> >
> >> >
> >> >
> >> >
> >> >
> >> >
> >> >
> >> >
> >>
> >
> >
> >
> > --
> > Karan Singh Malhi
> >
>

Re: Ideas for Getting the word out

[Karan Malhi <ka...@gmail.com>]

I was thinking on the same lines. Something like an all-in-one bundle zip
file which one can import into the IDE and expands to multiple projects (one
project per example) in the IDE would be super.

On Mon, May 30, 2011 at 8:33 PM, David Blevins <da...@gmail.com>wrote:

>
> On May 30, 2011, at 5:29 PM, David Blevins wrote:
>
> > Excellent statement of the problem space.  I don't think I ever thought
> of things like this.
> >
> > Any ideas on solutions?  Not sure I see what we need to code up.
>
> Referring specifically to the idea of an examples zip that contains 100% of
> the binaries required and the magic IDE files.  Not sure how this is done.
>
>
> Just thinking loosely... there's a bunch of 'IDE' code in some maven
> plugins.   Is this task matter of grabbing a copy of that code and modifying
> the parts that generate the project files so that they point to libraries in
> the zip?
>
>
>
> -David
>
>
> >
> > On May 30, 2011, at 4:42 PM, Karan Malhi wrote:
> >
> >> Absolutely, IDE files should not be checked into SVN, thats why I am
> talking
> >> about "generating IDE specific examples zip " and not about "checking in
> IDE
> >> files".
> >>
> >> Here is a perspective on why we might want to generate IDE examples zip
> >> separately, "in addition to " the stuff you mentioned about maven
> >> archetypes, ant samples.
> >>
> >> We also need to keep the user in mind - specially the one who does most
> of
> >> their learning at work. I teach for a living and at client sites, I
> cannot
> >> run maven - downloading jars is an issue. Nobody uses subversion - a
> very
> >> famous commercial SCM is used . Automated downloading of stuff from
> central
> >> repo is totally blocked and you can get into a lot of trouble for
> attempting
> >> to bypass security. These are financial and insurance firms and due to
> legal
> >> requirements, they cannot have a desktop configuration of their choice,
> they
> >> have to meet certain very stringent security parameters. If i need to
> show
> >> openejb examples, I have to show them on my laptop. Prior to coming to
> >> class, I have to run the examples on my laptop and download
> dependencies,
> >> since i am not allowed to connect my laptop to the network or plugin a
> >> CD/floppy/USB on the client machine.
> >>
> >> What I can do though is get a pre-approved set of zip files and put them
> on
> >> a share, so that they can access it and use it. It becomes even easier
> to
> >> utilize that zip if it is known that it is for education and learning
> and
> >> none of those jars are permitted to be used in the real project until
> that
> >> particular software/jar is formally approved.
> >>
> >> Getting a pre-approved zip also sometimes takes its own sweet time,
> thats
> >> why having the ability to browse our examples online is just fantastic.
> >> People might not get to run them immediately, but they add a great value
> to
> >> folks who cannot download stuff on their work machines.
> >>
> >> So imagine we putting up all these maven archetypes, maven based
> projects
> >> and quite a few folks do not get to use our examples at work. Maven, ant
> ,
> >> subversion are great ways to share examples, no doubt about that. I
> fully
> >> agree with you on the demerits of putting IDE files into subversion
> >> (although in some cases - specially in a learning environment it is
> easier
> >> and does make sense to put them in svn - wont go there at this moment).
> >>
> >> As for support for maven in IDE's, thats great, but again, 100% of the
> >> companies I go to use eclipse based IDE's. How many are open to
> installing a
> >> plugin of your choice like m2eclipse? -- 10% probably, others are pretty
> >> locked down. So even though Intellij and Netbeans have fantastic support
> and
> >> m2eclipse is awesome, the truth out there is that all the above software
> can
> >> only be installed on a personal laptop, in a corporate environment -
> >> specially certain verticals, its a challenge.
> >>
> >>
> >>
> >>
> >> On Mon, May 30, 2011 at 2:32 AM, Mark Struberg <st...@yahoo.de>
> wrote:
> >>
> >>> IDE files must never, _never_, NEVER be checked in into any SCM!
> >>>
> >>> Most times IDE config files contain lots of absolute paths, and
> personal
> >>> configs. In ancient times Eclipse config files did also contain CVS
> >>> passwords...
> >>>
> >>> Trust me, you do not like to have such a waste into your Source Code
> >>> Management system!
> >>>
> >>> There is usually a
> >>> $> svn propedit svn:ignore .
> >>> which contains the following items:
> >>> target
> >>> .settings
> >>> .project
> >>> .classpath
> >>> *.iml
> >>> *.ipr
> >>> *.iws
> >>>
> >>> Id suggest to use either maven or ant for the samples. Of course I
> prefer
> >>> maven, because you dont need to checkin any 3rd party dependencies. And
> the
> >>> IDE integration in Idea and Netbeans ist first class! The integration
> in
> >>> Eclipse is pretty good also.
> >>>
> >>>
> >>> Regarding the Archetypes:
> >>>
> >>> People like to see and browse samples online (svn browser). But they
> also
> >>> like to have something quick they can start hacking on. This is exactly
> what
> >>> archetypes provide!
> >>>
> >>> In Apache MyFaces we have a few archetypes which generate simple
> projects
> >>> skeletons. There are different types of them and I'd think of extending
> the
> >>> CODI-JSF2.0 archetype and add OpenEJB on top.
> >>>
> >>> http://wiki.apache.org/myfaces/MyFaces_Archetypes_for_Maven
> >>>
> >>> The CODI JSF20 archetype should get updated
> (webbeansconfigurationlistener
> >>> removed from web.xml and instead using webbeans-tomcat7 plugin), but
> its a
> >>> good starting point. Just add openejb and a @Stateless backend service.
> >>>
> >>> LieGrue,
> >>> strub
> >>>
> >>> --- On Sun, 5/29/11, Karan Malhi <ka...@gmail.com> wrote:
> >>>
> >>>> From: Karan Malhi <ka...@gmail.com>
> >>>> Subject: Re: Ideas for Getting the word out
> >>>> To: dev@openejb.apache.org
> >>>> Date: Sunday, May 29, 2011, 2:26 PM
> >>>> Exactly my thought. Apart from
> >>>> releasing maven based projects, also have
> >>>> maven generate projects/zip files specific to IDE's and
> >>>> make them available
> >>>> for download from our website. Anytime we update the
> >>>> examples, the IDE
> >>>> specific zips should be one of the artifacts which would
> >>>> automatically get
> >>>> generated and posted to the website.
> >>>>
> >>>> On Sun, May 29, 2011 at 8:34 AM, dsh <da...@googlemail.com>
> >>>> wrote:
> >>>>
> >>>>> One thought concerning IDEs - If you export sample
> >>>> projects from an
> >>>>> IDE and store them in SVN to be imported by someone
> >>>> else you have to
> >>>>> maintain that IDE-specific samples too cause IDEs of
> >>>> course are
> >>>>> changing over time and hence you have to tweak such
> >>>> projects every now
> >>>>> and then. What I like more is having the build tool
> >>>> creating
> >>>>> IDE-specific projects from the meta descriptors
> >>>> provided by a build
> >>>>> system cause it could be suspected that such meta
> >>>> descriptors are
> >>>>> always up-to-date.
> >>>>>
> >>>>> Such a build system could be Maven for Java-specific
> >>>> projects or CMake
> >>>>> for C/C++ specific projects.
> >>>>>
> >>>>> Cheers
> >>>>> Daniel
> >>>>>
> >>>>> On Sun, May 29, 2011 at 2:21 PM, Karan Malhi <ka...@gmail.com>
> >>>>> wrote:
> >>>>>> The assumption that everybody knows maven is a
> >>>> dangerous one. Examples
> >>>>>> should also be packaged so that they could be
> >>>> imported into IDE's i.e.
> >>>>> zip
> >>>>>> files for eclipse, idea and netbeans. People
> >>>> typically have their
> >>>>> favorite
> >>>>>> IDE's already installed on their machines and
> >>>> giving the ability to
> >>>>> import
> >>>>>> examples directly into the IDE and "Get Started "
> >>>> is very effective.
> >>>>>> The thing with only relying on maven is that if
> >>>> somebody does not have it
> >>>>>> installed, they first have to install it, then
> >>>> they also need to know a
> >>>>> bit
> >>>>>> of it. I know that you can always tell them to
> >>>> simply run mvn clean
> >>>>> install
> >>>>>> to run the example, but for people not exposed to
> >>>> maven, there is always
> >>>>> a
> >>>>>> feeling of "I think maven did some magic to make
> >>>> it work. Wonder what it
> >>>>>> would take to get this example working in my IDE"
> >>>> - and this is where we
> >>>>> can
> >>>>>> potentially leave a bad taste. "First impression
> >>>> is a lasting one" - lets
> >>>>>> try and make that first impression a great one
> >>>>>>
> >>>>>> On Sun, May 29, 2011 at 4:04 AM, Aldrin Leal
> >>>> <al...@leal.eng.br>
> >>>> wrote:
> >>>>>>
> >>>>>>> For the examples, some could be turned into
> >>>> archetypes (which answers
> >>>>> the
> >>>>>>> previous question "What are archetypes useful
> >>>> for?") as a means to reach
> >>>>> a
> >>>>>>> wide audience: No need to download sources,
> >>>> leave it up to the archetype
> >>>>>>> (Actually Wicket Docs does just like that)
> >>>>>>>
> >>>>>>> As for the feeds, I suggest either a planet
> >>>> setup or simply a Twitter
> >>>>> User
> >>>>>>> List at first
> >>>>>>>
> >>>>>>> --
> >>>>>>> -- Aldrin Leal, <al...@leal.eng.br>
> >>>> /
> >>>>> http://www.leal.eng.br/mnemetica/
> >>>>>>>
> >>>>>>>
> >>>>>>> On Sat, May 28, 2011 at 8:50 PM, David
> >>>> Blevins <david.blevins@gmail.com
> >>>>>>>> wrote:
> >>>>>>>
> >>>>>>>> Some IRC chat resulted in some neat
> >>>> ideas on ways to dramatically
> >>>>> improve
> >>>>>>>> how easy it is to consume and learn
> >>>> about OpenEJB.
> >>>>>>>>
> >>>>>>>> # Examples
> >>>>>>>>
> >>>>>>>> One thing that hit home is that we have
> >>>> now 46 examples in trunk!!!
> >>>>>>>> Amazing.  However the benefit
> >>>> of that is dramatically reduced as most
> >>>>> of
> >>>>>>>> them are only available in zip file
> >>>> form.  Links to svn aren't that
> >>>>>>>> effective.  And examples change too
> >>>> as the technology improves so
> >>>>> keeping
> >>>>>>>> wiki pages up to date is hard -- which
> >>>> version of the example do you
> >>>>>>> show?
> >>>>>>>> What if you really want all
> >>>> versions available to see?
> >>>>>>>>
> >>>>>>>> So the idea was to use README files that
> >>>> are formatted in Markdown and
> >>>>>>> use
> >>>>>>>> that to generate a page for each
> >>>> example.  No more having part of the
> >>>>>>>> example in svn and part of it in
> >>>> confluence and then always breaking.
> >>>>> A
> >>>>>>>> stackoverflow inspired solution --
> >>>> Markdown + Google's prettyprint.
> >>>>>>>>
> >>>>>>>> All the examples are in a zip file that
> >>>> is published in the maven
> >>>>>>>> repository, so we could use a maven
> >>>> program to pull them down and
> >>>>> extract
> >>>>>>>> them to the target directory where we
> >>>> can then do our little page
> >>>>>>>> generation.  A stub with those
> >>>> ideas in comments:
> >>>>>>>>
> >>>>>>>>
> >>>>>>>>
> >>>>>>>
> >>>>>
> >>>
> http://svn.apache.org/repos/asf/openejb/trunk/sandbox/tools/src/main/java/org/apache/openejb/tools/examples/GenerateIndex.java
> >>>>>>>>
> >>>>>>>> # Retweeting
> >>>>>>>>
> >>>>>>>> We should monitor this feed
> >>>>> http://twitter.com/#!/OpenEJB/contributorsand
> >>>>>>>> retweet anything that mentions OpenEJB.
> >>>>>>>>
> >>>>>>>> So if anyone in the contributors list
> >>>> tweeted about OpenEJB, the
> >>>>> OpenEJB
> >>>>>>>> twitter account would retweet it.
> >>>>>>>>
> >>>>>>>> Two things will happen as a result:
> >>>>>>>> -  The more activity on the
> >>>> OpenEJB twitter account the more
> >>>>> followers
> >>>>>>> it
> >>>>>>>> will get
> >>>>>>>> -  The more @joe and other
> >>>> contributors are seen on the account, the
> >>>>>>> more
> >>>>>>>> followers they will get
> >>>>>>>>
> >>>>>>>> The OpenEJB twitter account has more
> >>>> followers than most everyone else
> >>>>> so
> >>>>>>>> getting it to retweet is a good way to
> >>>> expose people to all our
> >>>>> wonderful
> >>>>>>>> contributors and get them some followers
> >>>> and help the project at the
> >>>>> same
> >>>>>>>> time.
> >>>>>>>>
> >>>>>>>> The result is we as a community will
> >>>> have more ability overall to get
> >>>>> the
> >>>>>>>> word out!
> >>>>>>>>
> >>>>>>>> Twitterfeed.com was the obvious first
> >>>> idea, but it turns out twitter
> >>>>> does
> >>>>>>>> not allow you to post content from
> >>>> twitter back onto twitter.  Not
> >>>>> unless
> >>>>>>>> you use their API http://dev.twitter.com/doc .  So we could maybe
> >>>>> hack
> >>>>>>>> together some tool we run hourly and
> >>>> retweet things that contributors
> >>>>>>> tweet.
> >>>>>>>> A little stub with comments here:
> >>>>>>>>
> >>>>>>>>
> >>>>>>>>
> >>>>>>>
> >>>>>
> >>>
> http://svn.apache.org/repos/asf/openejb/trunk/sandbox/tools/src/main/java/org/apache/openejb/tools/twitter/Retweet.java
> >>>>>>>>
> >>>>>>>>
> >>>>>>>> Both are up for grabs!  If you're
> >>>> looking for something to do, either
> >>>>> one
> >>>>>>>> would be excellent ways to improve the
> >>>> project!
> >>>>>>>>
> >>>>>>>>
> >>>>>>>>
> >>>>>>>> -David
> >>>>>>>>
> >>>>>>>>
> >>>>>>>>
> >>>>>>>>
> >>>>>>>>
> >>>>>>>>
> >>>>>>>>
> >>>>>>>>
> >>>>>>>>
> >>>>>>>
> >>>>>>
> >>>>>>
> >>>>>>
> >>>>>> --
> >>>>>> Karan Singh Malhi
> >>>>>>
> >>>>>
> >>>>
> >>>>
> >>>>
> >>>> --
> >>>> Karan Singh Malhi
> >>>>
> >>>
> >>
> >>
> >>
> >> --
> >> Karan Singh Malhi
> >
>
>


-- 
Karan Singh Malhi

Re: Ideas for Getting the word out

[David Blevins <da...@gmail.com>]

On May 30, 2011, at 5:29 PM, David Blevins wrote:

> Excellent statement of the problem space.  I don't think I ever thought of things like this.
> 
> Any ideas on solutions?  Not sure I see what we need to code up.

Referring specifically to the idea of an examples zip that contains 100% of the binaries required and the magic IDE files.  Not sure how this is done.


Just thinking loosely... there's a bunch of 'IDE' code in some maven plugins.   Is this task matter of grabbing a copy of that code and modifying the parts that generate the project files so that they point to libraries in the zip?



-David


> 
> On May 30, 2011, at 4:42 PM, Karan Malhi wrote:
> 
>> Absolutely, IDE files should not be checked into SVN, thats why I am talking
>> about "generating IDE specific examples zip " and not about "checking in IDE
>> files".
>> 
>> Here is a perspective on why we might want to generate IDE examples zip
>> separately, "in addition to " the stuff you mentioned about maven
>> archetypes, ant samples.
>> 
>> We also need to keep the user in mind - specially the one who does most of
>> their learning at work. I teach for a living and at client sites, I cannot
>> run maven - downloading jars is an issue. Nobody uses subversion - a very
>> famous commercial SCM is used . Automated downloading of stuff from central
>> repo is totally blocked and you can get into a lot of trouble for attempting
>> to bypass security. These are financial and insurance firms and due to legal
>> requirements, they cannot have a desktop configuration of their choice, they
>> have to meet certain very stringent security parameters. If i need to show
>> openejb examples, I have to show them on my laptop. Prior to coming to
>> class, I have to run the examples on my laptop and download dependencies,
>> since i am not allowed to connect my laptop to the network or plugin a
>> CD/floppy/USB on the client machine.
>> 
>> What I can do though is get a pre-approved set of zip files and put them on
>> a share, so that they can access it and use it. It becomes even easier to
>> utilize that zip if it is known that it is for education and learning and
>> none of those jars are permitted to be used in the real project until that
>> particular software/jar is formally approved.
>> 
>> Getting a pre-approved zip also sometimes takes its own sweet time, thats
>> why having the ability to browse our examples online is just fantastic.
>> People might not get to run them immediately, but they add a great value to
>> folks who cannot download stuff on their work machines.
>> 
>> So imagine we putting up all these maven archetypes, maven based projects
>> and quite a few folks do not get to use our examples at work. Maven, ant ,
>> subversion are great ways to share examples, no doubt about that. I fully
>> agree with you on the demerits of putting IDE files into subversion
>> (although in some cases - specially in a learning environment it is easier
>> and does make sense to put them in svn - wont go there at this moment).
>> 
>> As for support for maven in IDE's, thats great, but again, 100% of the
>> companies I go to use eclipse based IDE's. How many are open to installing a
>> plugin of your choice like m2eclipse? -- 10% probably, others are pretty
>> locked down. So even though Intellij and Netbeans have fantastic support and
>> m2eclipse is awesome, the truth out there is that all the above software can
>> only be installed on a personal laptop, in a corporate environment -
>> specially certain verticals, its a challenge.
>> 
>> 
>> 
>> 
>> On Mon, May 30, 2011 at 2:32 AM, Mark Struberg <st...@yahoo.de> wrote:
>> 
>>> IDE files must never, _never_, NEVER be checked in into any SCM!
>>> 
>>> Most times IDE config files contain lots of absolute paths, and personal
>>> configs. In ancient times Eclipse config files did also contain CVS
>>> passwords...
>>> 
>>> Trust me, you do not like to have such a waste into your Source Code
>>> Management system!
>>> 
>>> There is usually a
>>> $> svn propedit svn:ignore .
>>> which contains the following items:
>>> target
>>> .settings
>>> .project
>>> .classpath
>>> *.iml
>>> *.ipr
>>> *.iws
>>> 
>>> Id suggest to use either maven or ant for the samples. Of course I prefer
>>> maven, because you dont need to checkin any 3rd party dependencies. And the
>>> IDE integration in Idea and Netbeans ist first class! The integration in
>>> Eclipse is pretty good also.
>>> 
>>> 
>>> Regarding the Archetypes:
>>> 
>>> People like to see and browse samples online (svn browser). But they also
>>> like to have something quick they can start hacking on. This is exactly what
>>> archetypes provide!
>>> 
>>> In Apache MyFaces we have a few archetypes which generate simple projects
>>> skeletons. There are different types of them and I'd think of extending the
>>> CODI-JSF2.0 archetype and add OpenEJB on top.
>>> 
>>> http://wiki.apache.org/myfaces/MyFaces_Archetypes_for_Maven
>>> 
>>> The CODI JSF20 archetype should get updated (webbeansconfigurationlistener
>>> removed from web.xml and instead using webbeans-tomcat7 plugin), but its a
>>> good starting point. Just add openejb and a @Stateless backend service.
>>> 
>>> LieGrue,
>>> strub
>>> 
>>> --- On Sun, 5/29/11, Karan Malhi <ka...@gmail.com> wrote:
>>> 
>>>> From: Karan Malhi <ka...@gmail.com>
>>>> Subject: Re: Ideas for Getting the word out
>>>> To: dev@openejb.apache.org
>>>> Date: Sunday, May 29, 2011, 2:26 PM
>>>> Exactly my thought. Apart from
>>>> releasing maven based projects, also have
>>>> maven generate projects/zip files specific to IDE's and
>>>> make them available
>>>> for download from our website. Anytime we update the
>>>> examples, the IDE
>>>> specific zips should be one of the artifacts which would
>>>> automatically get
>>>> generated and posted to the website.
>>>> 
>>>> On Sun, May 29, 2011 at 8:34 AM, dsh <da...@googlemail.com>
>>>> wrote:
>>>> 
>>>>> One thought concerning IDEs - If you export sample
>>>> projects from an
>>>>> IDE and store them in SVN to be imported by someone
>>>> else you have to
>>>>> maintain that IDE-specific samples too cause IDEs of
>>>> course are
>>>>> changing over time and hence you have to tweak such
>>>> projects every now
>>>>> and then. What I like more is having the build tool
>>>> creating
>>>>> IDE-specific projects from the meta descriptors
>>>> provided by a build
>>>>> system cause it could be suspected that such meta
>>>> descriptors are
>>>>> always up-to-date.
>>>>> 
>>>>> Such a build system could be Maven for Java-specific
>>>> projects or CMake
>>>>> for C/C++ specific projects.
>>>>> 
>>>>> Cheers
>>>>> Daniel
>>>>> 
>>>>> On Sun, May 29, 2011 at 2:21 PM, Karan Malhi <ka...@gmail.com>
>>>>> wrote:
>>>>>> The assumption that everybody knows maven is a
>>>> dangerous one. Examples
>>>>>> should also be packaged so that they could be
>>>> imported into IDE's i.e.
>>>>> zip
>>>>>> files for eclipse, idea and netbeans. People
>>>> typically have their
>>>>> favorite
>>>>>> IDE's already installed on their machines and
>>>> giving the ability to
>>>>> import
>>>>>> examples directly into the IDE and "Get Started "
>>>> is very effective.
>>>>>> The thing with only relying on maven is that if
>>>> somebody does not have it
>>>>>> installed, they first have to install it, then
>>>> they also need to know a
>>>>> bit
>>>>>> of it. I know that you can always tell them to
>>>> simply run mvn clean
>>>>> install
>>>>>> to run the example, but for people not exposed to
>>>> maven, there is always
>>>>> a
>>>>>> feeling of "I think maven did some magic to make
>>>> it work. Wonder what it
>>>>>> would take to get this example working in my IDE"
>>>> - and this is where we
>>>>> can
>>>>>> potentially leave a bad taste. "First impression
>>>> is a lasting one" - lets
>>>>>> try and make that first impression a great one
>>>>>> 
>>>>>> On Sun, May 29, 2011 at 4:04 AM, Aldrin Leal
>>>> <al...@leal.eng.br>
>>>> wrote:
>>>>>> 
>>>>>>> For the examples, some could be turned into
>>>> archetypes (which answers
>>>>> the
>>>>>>> previous question "What are archetypes useful
>>>> for?") as a means to reach
>>>>> a
>>>>>>> wide audience: No need to download sources,
>>>> leave it up to the archetype
>>>>>>> (Actually Wicket Docs does just like that)
>>>>>>> 
>>>>>>> As for the feeds, I suggest either a planet
>>>> setup or simply a Twitter
>>>>> User
>>>>>>> List at first
>>>>>>> 
>>>>>>> --
>>>>>>> -- Aldrin Leal, <al...@leal.eng.br>
>>>> /
>>>>> http://www.leal.eng.br/mnemetica/
>>>>>>> 
>>>>>>> 
>>>>>>> On Sat, May 28, 2011 at 8:50 PM, David
>>>> Blevins <david.blevins@gmail.com
>>>>>>>> wrote:
>>>>>>> 
>>>>>>>> Some IRC chat resulted in some neat
>>>> ideas on ways to dramatically
>>>>> improve
>>>>>>>> how easy it is to consume and learn
>>>> about OpenEJB.
>>>>>>>> 
>>>>>>>> # Examples
>>>>>>>> 
>>>>>>>> One thing that hit home is that we have
>>>> now 46 examples in trunk!!!
>>>>>>>> Amazing.  However the benefit
>>>> of that is dramatically reduced as most
>>>>> of
>>>>>>>> them are only available in zip file
>>>> form.  Links to svn aren't that
>>>>>>>> effective.  And examples change too
>>>> as the technology improves so
>>>>> keeping
>>>>>>>> wiki pages up to date is hard -- which
>>>> version of the example do you
>>>>>>> show?
>>>>>>>> What if you really want all
>>>> versions available to see?
>>>>>>>> 
>>>>>>>> So the idea was to use README files that
>>>> are formatted in Markdown and
>>>>>>> use
>>>>>>>> that to generate a page for each
>>>> example.  No more having part of the
>>>>>>>> example in svn and part of it in
>>>> confluence and then always breaking.
>>>>> A
>>>>>>>> stackoverflow inspired solution --
>>>> Markdown + Google's prettyprint.
>>>>>>>> 
>>>>>>>> All the examples are in a zip file that
>>>> is published in the maven
>>>>>>>> repository, so we could use a maven
>>>> program to pull them down and
>>>>> extract
>>>>>>>> them to the target directory where we
>>>> can then do our little page
>>>>>>>> generation.  A stub with those
>>>> ideas in comments:
>>>>>>>> 
>>>>>>>> 
>>>>>>>> 
>>>>>>> 
>>>>> 
>>> http://svn.apache.org/repos/asf/openejb/trunk/sandbox/tools/src/main/java/org/apache/openejb/tools/examples/GenerateIndex.java
>>>>>>>> 
>>>>>>>> # Retweeting
>>>>>>>> 
>>>>>>>> We should monitor this feed
>>>>> http://twitter.com/#!/OpenEJB/contributorsand
>>>>>>>> retweet anything that mentions OpenEJB.
>>>>>>>> 
>>>>>>>> So if anyone in the contributors list
>>>> tweeted about OpenEJB, the
>>>>> OpenEJB
>>>>>>>> twitter account would retweet it.
>>>>>>>> 
>>>>>>>> Two things will happen as a result:
>>>>>>>> -  The more activity on the
>>>> OpenEJB twitter account the more
>>>>> followers
>>>>>>> it
>>>>>>>> will get
>>>>>>>> -  The more @joe and other
>>>> contributors are seen on the account, the
>>>>>>> more
>>>>>>>> followers they will get
>>>>>>>> 
>>>>>>>> The OpenEJB twitter account has more
>>>> followers than most everyone else
>>>>> so
>>>>>>>> getting it to retweet is a good way to
>>>> expose people to all our
>>>>> wonderful
>>>>>>>> contributors and get them some followers
>>>> and help the project at the
>>>>> same
>>>>>>>> time.
>>>>>>>> 
>>>>>>>> The result is we as a community will
>>>> have more ability overall to get
>>>>> the
>>>>>>>> word out!
>>>>>>>> 
>>>>>>>> Twitterfeed.com was the obvious first
>>>> idea, but it turns out twitter
>>>>> does
>>>>>>>> not allow you to post content from
>>>> twitter back onto twitter.  Not
>>>>> unless
>>>>>>>> you use their API http://dev.twitter.com/doc .  So we could maybe
>>>>> hack
>>>>>>>> together some tool we run hourly and
>>>> retweet things that contributors
>>>>>>> tweet.
>>>>>>>> A little stub with comments here:
>>>>>>>> 
>>>>>>>> 
>>>>>>>> 
>>>>>>> 
>>>>> 
>>> http://svn.apache.org/repos/asf/openejb/trunk/sandbox/tools/src/main/java/org/apache/openejb/tools/twitter/Retweet.java
>>>>>>>> 
>>>>>>>> 
>>>>>>>> Both are up for grabs!  If you're
>>>> looking for something to do, either
>>>>> one
>>>>>>>> would be excellent ways to improve the
>>>> project!
>>>>>>>> 
>>>>>>>> 
>>>>>>>> 
>>>>>>>> -David
>>>>>>>> 
>>>>>>>> 
>>>>>>>> 
>>>>>>>> 
>>>>>>>> 
>>>>>>>> 
>>>>>>>> 
>>>>>>>> 
>>>>>>>> 
>>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> --
>>>>>> Karan Singh Malhi
>>>>>> 
>>>>> 
>>>> 
>>>> 
>>>> 
>>>> --
>>>> Karan Singh Malhi
>>>> 
>>> 
>> 
>> 
>> 
>> -- 
>> Karan Singh Malhi
> 

Re: Ideas for Getting the word out

[David Blevins <da...@gmail.com>]

Excellent statement of the problem space.  I don't think I ever thought of things like this.

Any ideas on solutions?  Not sure I see what we need to code up.


-David


On May 30, 2011, at 4:42 PM, Karan Malhi wrote:

> Absolutely, IDE files should not be checked into SVN, thats why I am talking
> about "generating IDE specific examples zip " and not about "checking in IDE
> files".
> 
> Here is a perspective on why we might want to generate IDE examples zip
> separately, "in addition to " the stuff you mentioned about maven
> archetypes, ant samples.
> 
> We also need to keep the user in mind - specially the one who does most of
> their learning at work. I teach for a living and at client sites, I cannot
> run maven - downloading jars is an issue. Nobody uses subversion - a very
> famous commercial SCM is used . Automated downloading of stuff from central
> repo is totally blocked and you can get into a lot of trouble for attempting
> to bypass security. These are financial and insurance firms and due to legal
> requirements, they cannot have a desktop configuration of their choice, they
> have to meet certain very stringent security parameters. If i need to show
> openejb examples, I have to show them on my laptop. Prior to coming to
> class, I have to run the examples on my laptop and download dependencies,
> since i am not allowed to connect my laptop to the network or plugin a
> CD/floppy/USB on the client machine.
> 
> What I can do though is get a pre-approved set of zip files and put them on
> a share, so that they can access it and use it. It becomes even easier to
> utilize that zip if it is known that it is for education and learning and
> none of those jars are permitted to be used in the real project until that
> particular software/jar is formally approved.
> 
> Getting a pre-approved zip also sometimes takes its own sweet time, thats
> why having the ability to browse our examples online is just fantastic.
> People might not get to run them immediately, but they add a great value to
> folks who cannot download stuff on their work machines.
> 
> So imagine we putting up all these maven archetypes, maven based projects
> and quite a few folks do not get to use our examples at work. Maven, ant ,
> subversion are great ways to share examples, no doubt about that. I fully
> agree with you on the demerits of putting IDE files into subversion
> (although in some cases - specially in a learning environment it is easier
> and does make sense to put them in svn - wont go there at this moment).
> 
> As for support for maven in IDE's, thats great, but again, 100% of the
> companies I go to use eclipse based IDE's. How many are open to installing a
> plugin of your choice like m2eclipse? -- 10% probably, others are pretty
> locked down. So even though Intellij and Netbeans have fantastic support and
> m2eclipse is awesome, the truth out there is that all the above software can
> only be installed on a personal laptop, in a corporate environment -
> specially certain verticals, its a challenge.
> 
> 
> 
> 
> On Mon, May 30, 2011 at 2:32 AM, Mark Struberg <st...@yahoo.de> wrote:
> 
>> IDE files must never, _never_, NEVER be checked in into any SCM!
>> 
>> Most times IDE config files contain lots of absolute paths, and personal
>> configs. In ancient times Eclipse config files did also contain CVS
>> passwords...
>> 
>> Trust me, you do not like to have such a waste into your Source Code
>> Management system!
>> 
>> There is usually a
>> $> svn propedit svn:ignore .
>> which contains the following items:
>> target
>> .settings
>> .project
>> .classpath
>> *.iml
>> *.ipr
>> *.iws
>> 
>> Id suggest to use either maven or ant for the samples. Of course I prefer
>> maven, because you dont need to checkin any 3rd party dependencies. And the
>> IDE integration in Idea and Netbeans ist first class! The integration in
>> Eclipse is pretty good also.
>> 
>> 
>> Regarding the Archetypes:
>> 
>> People like to see and browse samples online (svn browser). But they also
>> like to have something quick they can start hacking on. This is exactly what
>> archetypes provide!
>> 
>> In Apache MyFaces we have a few archetypes which generate simple projects
>> skeletons. There are different types of them and I'd think of extending the
>> CODI-JSF2.0 archetype and add OpenEJB on top.
>> 
>> http://wiki.apache.org/myfaces/MyFaces_Archetypes_for_Maven
>> 
>> The CODI JSF20 archetype should get updated (webbeansconfigurationlistener
>> removed from web.xml and instead using webbeans-tomcat7 plugin), but its a
>> good starting point. Just add openejb and a @Stateless backend service.
>> 
>> LieGrue,
>> strub
>> 
>> --- On Sun, 5/29/11, Karan Malhi <ka...@gmail.com> wrote:
>> 
>>> From: Karan Malhi <ka...@gmail.com>
>>> Subject: Re: Ideas for Getting the word out
>>> To: dev@openejb.apache.org
>>> Date: Sunday, May 29, 2011, 2:26 PM
>>> Exactly my thought. Apart from
>>> releasing maven based projects, also have
>>> maven generate projects/zip files specific to IDE's and
>>> make them available
>>> for download from our website. Anytime we update the
>>> examples, the IDE
>>> specific zips should be one of the artifacts which would
>>> automatically get
>>> generated and posted to the website.
>>> 
>>> On Sun, May 29, 2011 at 8:34 AM, dsh <da...@googlemail.com>
>>> wrote:
>>> 
>>>> One thought concerning IDEs - If you export sample
>>> projects from an
>>>> IDE and store them in SVN to be imported by someone
>>> else you have to
>>>> maintain that IDE-specific samples too cause IDEs of
>>> course are
>>>> changing over time and hence you have to tweak such
>>> projects every now
>>>> and then. What I like more is having the build tool
>>> creating
>>>> IDE-specific projects from the meta descriptors
>>> provided by a build
>>>> system cause it could be suspected that such meta
>>> descriptors are
>>>> always up-to-date.
>>>> 
>>>> Such a build system could be Maven for Java-specific
>>> projects or CMake
>>>> for C/C++ specific projects.
>>>> 
>>>> Cheers
>>>> Daniel
>>>> 
>>>> On Sun, May 29, 2011 at 2:21 PM, Karan Malhi <ka...@gmail.com>
>>>> wrote:
>>>>> The assumption that everybody knows maven is a
>>> dangerous one. Examples
>>>>> should also be packaged so that they could be
>>> imported into IDE's i.e.
>>>> zip
>>>>> files for eclipse, idea and netbeans. People
>>> typically have their
>>>> favorite
>>>>> IDE's already installed on their machines and
>>> giving the ability to
>>>> import
>>>>> examples directly into the IDE and "Get Started "
>>> is very effective.
>>>>> The thing with only relying on maven is that if
>>> somebody does not have it
>>>>> installed, they first have to install it, then
>>> they also need to know a
>>>> bit
>>>>> of it. I know that you can always tell them to
>>> simply run mvn clean
>>>> install
>>>>> to run the example, but for people not exposed to
>>> maven, there is always
>>>> a
>>>>> feeling of "I think maven did some magic to make
>>> it work. Wonder what it
>>>>> would take to get this example working in my IDE"
>>> - and this is where we
>>>> can
>>>>> potentially leave a bad taste. "First impression
>>> is a lasting one" - lets
>>>>> try and make that first impression a great one
>>>>> 
>>>>> On Sun, May 29, 2011 at 4:04 AM, Aldrin Leal
>>> <al...@leal.eng.br>
>>> wrote:
>>>>> 
>>>>>> For the examples, some could be turned into
>>> archetypes (which answers
>>>> the
>>>>>> previous question "What are archetypes useful
>>> for?") as a means to reach
>>>> a
>>>>>> wide audience: No need to download sources,
>>> leave it up to the archetype
>>>>>> (Actually Wicket Docs does just like that)
>>>>>> 
>>>>>> As for the feeds, I suggest either a planet
>>> setup or simply a Twitter
>>>> User
>>>>>> List at first
>>>>>> 
>>>>>> --
>>>>>> -- Aldrin Leal, <al...@leal.eng.br>
>>> /
>>>> http://www.leal.eng.br/mnemetica/
>>>>>> 
>>>>>> 
>>>>>> On Sat, May 28, 2011 at 8:50 PM, David
>>> Blevins <david.blevins@gmail.com
>>>>>>> wrote:
>>>>>> 
>>>>>>> Some IRC chat resulted in some neat
>>> ideas on ways to dramatically
>>>> improve
>>>>>>> how easy it is to consume and learn
>>> about OpenEJB.
>>>>>>> 
>>>>>>> # Examples
>>>>>>> 
>>>>>>> One thing that hit home is that we have
>>> now 46 examples in trunk!!!
>>>>>>> Amazing.  However the benefit
>>> of that is dramatically reduced as most
>>>> of
>>>>>>> them are only available in zip file
>>> form.  Links to svn aren't that
>>>>>>> effective.  And examples change too
>>> as the technology improves so
>>>> keeping
>>>>>>> wiki pages up to date is hard -- which
>>> version of the example do you
>>>>>> show?
>>>>>>> What if you really want all
>>> versions available to see?
>>>>>>> 
>>>>>>> So the idea was to use README files that
>>> are formatted in Markdown and
>>>>>> use
>>>>>>> that to generate a page for each
>>> example.  No more having part of the
>>>>>>> example in svn and part of it in
>>> confluence and then always breaking.
>>>> A
>>>>>>> stackoverflow inspired solution --
>>> Markdown + Google's prettyprint.
>>>>>>> 
>>>>>>> All the examples are in a zip file that
>>> is published in the maven
>>>>>>> repository, so we could use a maven
>>> program to pull them down and
>>>> extract
>>>>>>> them to the target directory where we
>>> can then do our little page
>>>>>>> generation.  A stub with those
>>> ideas in comments:
>>>>>>> 
>>>>>>> 
>>>>>>> 
>>>>>> 
>>>> 
>> http://svn.apache.org/repos/asf/openejb/trunk/sandbox/tools/src/main/java/org/apache/openejb/tools/examples/GenerateIndex.java
>>>>>>> 
>>>>>>> # Retweeting
>>>>>>> 
>>>>>>> We should monitor this feed
>>>> http://twitter.com/#!/OpenEJB/contributorsand
>>>>>>> retweet anything that mentions OpenEJB.
>>>>>>> 
>>>>>>> So if anyone in the contributors list
>>> tweeted about OpenEJB, the
>>>> OpenEJB
>>>>>>> twitter account would retweet it.
>>>>>>> 
>>>>>>> Two things will happen as a result:
>>>>>>> -  The more activity on the
>>> OpenEJB twitter account the more
>>>> followers
>>>>>> it
>>>>>>> will get
>>>>>>> -  The more @joe and other
>>> contributors are seen on the account, the
>>>>>> more
>>>>>>> followers they will get
>>>>>>> 
>>>>>>> The OpenEJB twitter account has more
>>> followers than most everyone else
>>>> so
>>>>>>> getting it to retweet is a good way to
>>> expose people to all our
>>>> wonderful
>>>>>>> contributors and get them some followers
>>> and help the project at the
>>>> same
>>>>>>> time.
>>>>>>> 
>>>>>>> The result is we as a community will
>>> have more ability overall to get
>>>> the
>>>>>>> word out!
>>>>>>> 
>>>>>>> Twitterfeed.com was the obvious first
>>> idea, but it turns out twitter
>>>> does
>>>>>>> not allow you to post content from
>>> twitter back onto twitter.  Not
>>>> unless
>>>>>>> you use their API http://dev.twitter.com/doc .  So we could maybe
>>>> hack
>>>>>>> together some tool we run hourly and
>>> retweet things that contributors
>>>>>> tweet.
>>>>>>> A little stub with comments here:
>>>>>>> 
>>>>>>> 
>>>>>>> 
>>>>>> 
>>>> 
>> http://svn.apache.org/repos/asf/openejb/trunk/sandbox/tools/src/main/java/org/apache/openejb/tools/twitter/Retweet.java
>>>>>>> 
>>>>>>> 
>>>>>>> Both are up for grabs!  If you're
>>> looking for something to do, either
>>>> one
>>>>>>> would be excellent ways to improve the
>>> project!
>>>>>>> 
>>>>>>> 
>>>>>>> 
>>>>>>> -David
>>>>>>> 
>>>>>>> 
>>>>>>> 
>>>>>>> 
>>>>>>> 
>>>>>>> 
>>>>>>> 
>>>>>>> 
>>>>>>> 
>>>>>> 
>>>>> 
>>>>> 
>>>>> 
>>>>> --
>>>>> Karan Singh Malhi
>>>>> 
>>>> 
>>> 
>>> 
>>> 
>>> --
>>> Karan Singh Malhi
>>> 
>> 
> 
> 
> 
> -- 
> Karan Singh Malhi

Re: Ideas for Getting the word out

[Karan Malhi <ka...@gmail.com>]

Absolutely, IDE files should not be checked into SVN, thats why I am talking
about "generating IDE specific examples zip " and not about "checking in IDE
files".

Here is a perspective on why we might want to generate IDE examples zip
separately, "in addition to " the stuff you mentioned about maven
archetypes, ant samples.

We also need to keep the user in mind - specially the one who does most of
their learning at work. I teach for a living and at client sites, I cannot
run maven - downloading jars is an issue. Nobody uses subversion - a very
famous commercial SCM is used . Automated downloading of stuff from central
repo is totally blocked and you can get into a lot of trouble for attempting
to bypass security. These are financial and insurance firms and due to legal
requirements, they cannot have a desktop configuration of their choice, they
have to meet certain very stringent security parameters. If i need to show
openejb examples, I have to show them on my laptop. Prior to coming to
class, I have to run the examples on my laptop and download dependencies,
since i am not allowed to connect my laptop to the network or plugin a
CD/floppy/USB on the client machine.

What I can do though is get a pre-approved set of zip files and put them on
a share, so that they can access it and use it. It becomes even easier to
utilize that zip if it is known that it is for education and learning and
none of those jars are permitted to be used in the real project until that
particular software/jar is formally approved.

Getting a pre-approved zip also sometimes takes its own sweet time, thats
why having the ability to browse our examples online is just fantastic.
People might not get to run them immediately, but they add a great value to
folks who cannot download stuff on their work machines.

 So imagine we putting up all these maven archetypes, maven based projects
and quite a few folks do not get to use our examples at work. Maven, ant ,
subversion are great ways to share examples, no doubt about that. I fully
agree with you on the demerits of putting IDE files into subversion
(although in some cases - specially in a learning environment it is easier
and does make sense to put them in svn - wont go there at this moment).

As for support for maven in IDE's, thats great, but again, 100% of the
companies I go to use eclipse based IDE's. How many are open to installing a
plugin of your choice like m2eclipse? -- 10% probably, others are pretty
locked down. So even though Intellij and Netbeans have fantastic support and
m2eclipse is awesome, the truth out there is that all the above software can
only be installed on a personal laptop, in a corporate environment -
specially certain verticals, its a challenge.




On Mon, May 30, 2011 at 2:32 AM, Mark Struberg <st...@yahoo.de> wrote:

> IDE files must never, _never_, NEVER be checked in into any SCM!
>
> Most times IDE config files contain lots of absolute paths, and personal
> configs. In ancient times Eclipse config files did also contain CVS
> passwords...
>
> Trust me, you do not like to have such a waste into your Source Code
> Management system!
>
> There is usually a
> $> svn propedit svn:ignore .
> which contains the following items:
> target
> .settings
> .project
> .classpath
> *.iml
> *.ipr
> *.iws
>
> Id suggest to use either maven or ant for the samples. Of course I prefer
> maven, because you dont need to checkin any 3rd party dependencies. And the
> IDE integration in Idea and Netbeans ist first class! The integration in
> Eclipse is pretty good also.
>
>
> Regarding the Archetypes:
>
> People like to see and browse samples online (svn browser). But they also
> like to have something quick they can start hacking on. This is exactly what
> archetypes provide!
>
> In Apache MyFaces we have a few archetypes which generate simple projects
> skeletons. There are different types of them and I'd think of extending the
> CODI-JSF2.0 archetype and add OpenEJB on top.
>
> http://wiki.apache.org/myfaces/MyFaces_Archetypes_for_Maven
>
> The CODI JSF20 archetype should get updated (webbeansconfigurationlistener
> removed from web.xml and instead using webbeans-tomcat7 plugin), but its a
> good starting point. Just add openejb and a @Stateless backend service.
>
> LieGrue,
> strub
>
> --- On Sun, 5/29/11, Karan Malhi <ka...@gmail.com> wrote:
>
> > From: Karan Malhi <ka...@gmail.com>
> > Subject: Re: Ideas for Getting the word out
> > To: dev@openejb.apache.org
> > Date: Sunday, May 29, 2011, 2:26 PM
> > Exactly my thought. Apart from
> > releasing maven based projects, also have
> > maven generate projects/zip files specific to IDE's and
> > make them available
> > for download from our website. Anytime we update the
> > examples, the IDE
> > specific zips should be one of the artifacts which would
> > automatically get
> > generated and posted to the website.
> >
> > On Sun, May 29, 2011 at 8:34 AM, dsh <da...@googlemail.com>
> > wrote:
> >
> > > One thought concerning IDEs - If you export sample
> > projects from an
> > > IDE and store them in SVN to be imported by someone
> > else you have to
> > > maintain that IDE-specific samples too cause IDEs of
> > course are
> > > changing over time and hence you have to tweak such
> > projects every now
> > > and then. What I like more is having the build tool
> > creating
> > > IDE-specific projects from the meta descriptors
> > provided by a build
> > > system cause it could be suspected that such meta
> > descriptors are
> > > always up-to-date.
> > >
> > > Such a build system could be Maven for Java-specific
> > projects or CMake
> > > for C/C++ specific projects.
> > >
> > > Cheers
> > > Daniel
> > >
> > > On Sun, May 29, 2011 at 2:21 PM, Karan Malhi <ka...@gmail.com>
> > > wrote:
> > > > The assumption that everybody knows maven is a
> > dangerous one. Examples
> > > > should also be packaged so that they could be
> > imported into IDE's i.e.
> > > zip
> > > > files for eclipse, idea and netbeans. People
> > typically have their
> > > favorite
> > > > IDE's already installed on their machines and
> > giving the ability to
> > > import
> > > > examples directly into the IDE and "Get Started "
> > is very effective.
> > > > The thing with only relying on maven is that if
> > somebody does not have it
> > > > installed, they first have to install it, then
> > they also need to know a
> > > bit
> > > > of it. I know that you can always tell them to
> > simply run mvn clean
> > > install
> > > > to run the example, but for people not exposed to
> > maven, there is always
> > > a
> > > > feeling of "I think maven did some magic to make
> > it work. Wonder what it
> > > > would take to get this example working in my IDE"
> > - and this is where we
> > > can
> > > > potentially leave a bad taste. "First impression
> > is a lasting one" - lets
> > > > try and make that first impression a great one
> > > >
> > > > On Sun, May 29, 2011 at 4:04 AM, Aldrin Leal
> > <al...@leal.eng.br>
> > wrote:
> > > >
> > > >> For the examples, some could be turned into
> > archetypes (which answers
> > > the
> > > >> previous question "What are archetypes useful
> > for?") as a means to reach
> > > a
> > > >> wide audience: No need to download sources,
> > leave it up to the archetype
> > > >> (Actually Wicket Docs does just like that)
> > > >>
> > > >> As for the feeds, I suggest either a planet
> > setup or simply a Twitter
> > > User
> > > >> List at first
> > > >>
> > > >> --
> > > >> -- Aldrin Leal, <al...@leal.eng.br>
> > /
> > > http://www.leal.eng.br/mnemetica/
> > > >>
> > > >>
> > > >> On Sat, May 28, 2011 at 8:50 PM, David
> > Blevins <david.blevins@gmail.com
> > > >> >wrote:
> > > >>
> > > >> > Some IRC chat resulted in some neat
> > ideas on ways to dramatically
> > > improve
> > > >> > how easy it is to consume and learn
> > about OpenEJB.
> > > >> >
> > > >> > # Examples
> > > >> >
> > > >> > One thing that hit home is that we have
> > now 46 examples in trunk!!!
> > > >> >  Amazing.  However the benefit
> > of that is dramatically reduced as most
> > > of
> > > >> > them are only available in zip file
> > form.  Links to svn aren't that
> > > >> > effective.  And examples change too
> > as the technology improves so
> > > keeping
> > > >> > wiki pages up to date is hard -- which
> > version of the example do you
> > > >> show?
> > > >> >  What if you really want all
> > versions available to see?
> > > >> >
> > > >> > So the idea was to use README files that
> > are formatted in Markdown and
> > > >> use
> > > >> > that to generate a page for each
> > example.  No more having part of the
> > > >> > example in svn and part of it in
> > confluence and then always breaking.
> > >  A
> > > >> > stackoverflow inspired solution --
> > Markdown + Google's prettyprint.
> > > >> >
> > > >> > All the examples are in a zip file that
> > is published in the maven
> > > >> > repository, so we could use a maven
> > program to pull them down and
> > > extract
> > > >> > them to the target directory where we
> > can then do our little page
> > > >> > generation.  A stub with those
> > ideas in comments:
> > > >> >
> > > >> >
> > > >> >
> > > >>
> > >
> http://svn.apache.org/repos/asf/openejb/trunk/sandbox/tools/src/main/java/org/apache/openejb/tools/examples/GenerateIndex.java
> > > >> >
> > > >> > # Retweeting
> > > >> >
> > > >> > We should monitor this feed
> > > http://twitter.com/#!/OpenEJB/contributorsand
> > > >> > retweet anything that mentions OpenEJB.
> > > >> >
> > > >> > So if anyone in the contributors list
> > tweeted about OpenEJB, the
> > > OpenEJB
> > > >> > twitter account would retweet it.
> > > >> >
> > > >> > Two things will happen as a result:
> > > >> >  -  The more activity on the
> > OpenEJB twitter account the more
> > > followers
> > > >> it
> > > >> > will get
> > > >> >  -  The more @joe and other
> > contributors are seen on the account, the
> > > >> more
> > > >> > followers they will get
> > > >> >
> > > >> > The OpenEJB twitter account has more
> > followers than most everyone else
> > > so
> > > >> > getting it to retweet is a good way to
> > expose people to all our
> > > wonderful
> > > >> > contributors and get them some followers
> > and help the project at the
> > > same
> > > >> > time.
> > > >> >
> > > >> > The result is we as a community will
> > have more ability overall to get
> > > the
> > > >> > word out!
> > > >> >
> > > >> > Twitterfeed.com was the obvious first
> > idea, but it turns out twitter
> > > does
> > > >> > not allow you to post content from
> > twitter back onto twitter.  Not
> > > unless
> > > >> > you use their API http://dev.twitter.com/doc .  So we could maybe
> > > hack
> > > >> > together some tool we run hourly and
> > retweet things that contributors
> > > >> tweet.
> > > >> >  A little stub with comments here:
> > > >> >
> > > >> >
> > > >> >
> > > >>
> > >
> http://svn.apache.org/repos/asf/openejb/trunk/sandbox/tools/src/main/java/org/apache/openejb/tools/twitter/Retweet.java
> > > >> >
> > > >> >
> > > >> > Both are up for grabs!  If you're
> > looking for something to do, either
> > > one
> > > >> > would be excellent ways to improve the
> > project!
> > > >> >
> > > >> >
> > > >> >
> > > >> > -David
> > > >> >
> > > >> >
> > > >> >
> > > >> >
> > > >> >
> > > >> >
> > > >> >
> > > >> >
> > > >> >
> > > >>
> > > >
> > > >
> > > >
> > > > --
> > > > Karan Singh Malhi
> > > >
> > >
> >
> >
> >
> > --
> > Karan Singh Malhi
> >
>



-- 
Karan Singh Malhi

Re: Ideas for Getting the word out

[dsh <da...@googlemail.com>]

You know in 2007 I was like you and try to educate my kind colleagues
that this is supposed to be a bad habit. Finally after 4 years I gave
up cause you can't educate ignorant peeps... probably trying to
educate ignorant peeps should be considered a bad habit too ;)

Cheers
Daniel

On Mon, May 30, 2011 at 3:33 PM, Mark Struberg <st...@yahoo.de> wrote:
> Yea, bad habits die hard ... ;)
>
> You know, there's  an old saying about fixing something 'quick and dirty': "Dirty remains, while quick is long forgotten..."
>
> Just a simple question: how often do you have all those IDE settings in your SCM dirty?
> This is just as broken as to not agree on a distinct space vs tabs policy, etc
>
> LieGrue,
> strub
>
>
> --- On Mon, 5/30/11, Stephen Connolly <st...@gmail.com> wrote:
>
>> From: Stephen Connolly <st...@gmail.com>
>> Subject: Re: Ideas for Getting the word out
>> To: dev@openejb.apache.org
>> Date: Monday, May 30, 2011, 1:20 PM
>> On 30 May 2011 14:16, dsh <da...@googlemail.com>
>> wrote:
>> > On Mon, May 30, 2011 at 8:32 AM, Mark Struberg <st...@yahoo.de>
>> wrote:
>> >> IDE files must never, _never_, NEVER be checked in
>> into any SCM!
>> >>
>> >
>> > LOL couldn't resist... guess what we do each day at
>> work???
>>
>> Glad I don't work for your company so....
>>
>> >
>> > Cheers
>> > Daniel
>> >
>>
>

Re: Ideas for Getting the word out

[Mark Struberg <st...@yahoo.de>]

Yea, bad habits die hard ... ;)

You know, there's  an old saying about fixing something 'quick and dirty': "Dirty remains, while quick is long forgotten..."

Just a simple question: how often do you have all those IDE settings in your SCM dirty? 
This is just as broken as to not agree on a distinct space vs tabs policy, etc

LieGrue,
strub


--- On Mon, 5/30/11, Stephen Connolly <st...@gmail.com> wrote:

> From: Stephen Connolly <st...@gmail.com>
> Subject: Re: Ideas for Getting the word out
> To: dev@openejb.apache.org
> Date: Monday, May 30, 2011, 1:20 PM
> On 30 May 2011 14:16, dsh <da...@googlemail.com>
> wrote:
> > On Mon, May 30, 2011 at 8:32 AM, Mark Struberg <st...@yahoo.de>
> wrote:
> >> IDE files must never, _never_, NEVER be checked in
> into any SCM!
> >>
> >
> > LOL couldn't resist... guess what we do each day at
> work???
> 
> Glad I don't work for your company so....
> 
> >
> > Cheers
> > Daniel
> >
> 

Re: Ideas for Getting the word out

[Stephen Connolly <st...@gmail.com>]

On 30 May 2011 14:16, dsh <da...@googlemail.com> wrote:
> On Mon, May 30, 2011 at 8:32 AM, Mark Struberg <st...@yahoo.de> wrote:
>> IDE files must never, _never_, NEVER be checked in into any SCM!
>>
>
> LOL couldn't resist... guess what we do each day at work???

Glad I don't work for your company so....

>
> Cheers
> Daniel
>

Re: Ideas for Getting the word out

[dsh <da...@googlemail.com>]

On Mon, May 30, 2011 at 8:32 AM, Mark Struberg <st...@yahoo.de> wrote:
> IDE files must never, _never_, NEVER be checked in into any SCM!
>

LOL couldn't resist... guess what we do each day at work???

Cheers
Daniel

Re: Ideas for Getting the word out

[David Jencks <da...@yahoo.com>]

dunno anything about netbeans, but IDEA consumes maven based projects pretty much natively and eclipse + m2e does too.  I'd say the maven idea plugin is obsolete for anything except removing idea project files.

david jencks

On May 29, 2011, at 7:26 AM, Karan Malhi wrote:

> Exactly my thought. Apart from releasing maven based projects, also have
> maven generate projects/zip files specific to IDE's and make them available
> for download from our website. Anytime we update the examples, the IDE
> specific zips should be one of the artifacts which would automatically get
> generated and posted to the website.
> 
> On Sun, May 29, 2011 at 8:34 AM, dsh <da...@googlemail.com> wrote:
> 
>> One thought concerning IDEs - If you export sample projects from an
>> IDE and store them in SVN to be imported by someone else you have to
>> maintain that IDE-specific samples too cause IDEs of course are
>> changing over time and hence you have to tweak such projects every now
>> and then. What I like more is having the build tool creating
>> IDE-specific projects from the meta descriptors provided by a build
>> system cause it could be suspected that such meta descriptors are
>> always up-to-date.
>> 
>> Such a build system could be Maven for Java-specific projects or CMake
>> for C/C++ specific projects.
>> 
>> Cheers
>> Daniel
>> 
>> On Sun, May 29, 2011 at 2:21 PM, Karan Malhi <ka...@gmail.com>
>> wrote:
>>> The assumption that everybody knows maven is a dangerous one. Examples
>>> should also be packaged so that they could be imported into IDE's i.e.
>> zip
>>> files for eclipse, idea and netbeans. People typically have their
>> favorite
>>> IDE's already installed on their machines and giving the ability to
>> import
>>> examples directly into the IDE and "Get Started " is very effective.
>>> The thing with only relying on maven is that if somebody does not have it
>>> installed, they first have to install it, then they also need to know a
>> bit
>>> of it. I know that you can always tell them to simply run mvn clean
>> install
>>> to run the example, but for people not exposed to maven, there is always
>> a
>>> feeling of "I think maven did some magic to make it work. Wonder what it
>>> would take to get this example working in my IDE" - and this is where we
>> can
>>> potentially leave a bad taste. "First impression is a lasting one" - lets
>>> try and make that first impression a great one
>>> 
>>> On Sun, May 29, 2011 at 4:04 AM, Aldrin Leal <al...@leal.eng.br> wrote:
>>> 
>>>> For the examples, some could be turned into archetypes (which answers
>> the
>>>> previous question "What are archetypes useful for?") as a means to reach
>> a
>>>> wide audience: No need to download sources, leave it up to the archetype
>>>> (Actually Wicket Docs does just like that)
>>>> 
>>>> As for the feeds, I suggest either a planet setup or simply a Twitter
>> User
>>>> List at first
>>>> 
>>>> --
>>>> -- Aldrin Leal, <al...@leal.eng.br> /
>> http://www.leal.eng.br/mnemetica/
>>>> 
>>>> 
>>>> On Sat, May 28, 2011 at 8:50 PM, David Blevins <david.blevins@gmail.com
>>>>> wrote:
>>>> 
>>>>> Some IRC chat resulted in some neat ideas on ways to dramatically
>> improve
>>>>> how easy it is to consume and learn about OpenEJB.
>>>>> 
>>>>> # Examples
>>>>> 
>>>>> One thing that hit home is that we have now 46 examples in trunk!!!
>>>>> Amazing.  However the benefit of that is dramatically reduced as most
>> of
>>>>> them are only available in zip file form.  Links to svn aren't that
>>>>> effective.  And examples change too as the technology improves so
>> keeping
>>>>> wiki pages up to date is hard -- which version of the example do you
>>>> show?
>>>>> What if you really want all versions available to see?
>>>>> 
>>>>> So the idea was to use README files that are formatted in Markdown and
>>>> use
>>>>> that to generate a page for each example.  No more having part of the
>>>>> example in svn and part of it in confluence and then always breaking.
>> A
>>>>> stackoverflow inspired solution -- Markdown + Google's prettyprint.
>>>>> 
>>>>> All the examples are in a zip file that is published in the maven
>>>>> repository, so we could use a maven program to pull them down and
>> extract
>>>>> them to the target directory where we can then do our little page
>>>>> generation.  A stub with those ideas in comments:
>>>>> 
>>>>> 
>>>>> 
>>>> 
>> http://svn.apache.org/repos/asf/openejb/trunk/sandbox/tools/src/main/java/org/apache/openejb/tools/examples/GenerateIndex.java
>>>>> 
>>>>> # Retweeting
>>>>> 
>>>>> We should monitor this feed
>> http://twitter.com/#!/OpenEJB/contributorsand
>>>>> retweet anything that mentions OpenEJB.
>>>>> 
>>>>> So if anyone in the contributors list tweeted about OpenEJB, the
>> OpenEJB
>>>>> twitter account would retweet it.
>>>>> 
>>>>> Two things will happen as a result:
>>>>> -  The more activity on the OpenEJB twitter account the more
>> followers
>>>> it
>>>>> will get
>>>>> -  The more @joe and other contributors are seen on the account, the
>>>> more
>>>>> followers they will get
>>>>> 
>>>>> The OpenEJB twitter account has more followers than most everyone else
>> so
>>>>> getting it to retweet is a good way to expose people to all our
>> wonderful
>>>>> contributors and get them some followers and help the project at the
>> same
>>>>> time.
>>>>> 
>>>>> The result is we as a community will have more ability overall to get
>> the
>>>>> word out!
>>>>> 
>>>>> Twitterfeed.com was the obvious first idea, but it turns out twitter
>> does
>>>>> not allow you to post content from twitter back onto twitter.  Not
>> unless
>>>>> you use their API http://dev.twitter.com/doc .  So we could maybe
>> hack
>>>>> together some tool we run hourly and retweet things that contributors
>>>> tweet.
>>>>> A little stub with comments here:
>>>>> 
>>>>> 
>>>>> 
>>>> 
>> http://svn.apache.org/repos/asf/openejb/trunk/sandbox/tools/src/main/java/org/apache/openejb/tools/twitter/Retweet.java
>>>>> 
>>>>> 
>>>>> Both are up for grabs!  If you're looking for something to do, either
>> one
>>>>> would be excellent ways to improve the project!
>>>>> 
>>>>> 
>>>>> 
>>>>> -David
>>>>> 
>>>>> 
>>>>> 
>>>>> 
>>>>> 
>>>>> 
>>>>> 
>>>>> 
>>>>> 
>>>> 
>>> 
>>> 
>>> 
>>> --
>>> Karan Singh Malhi
>>> 
>> 
> 
> 
> 
> -- 
> Karan Singh Malhi

Re: Ideas for Getting the word out

[Mark Struberg <st...@yahoo.de>]

IDE files must never, _never_, NEVER be checked in into any SCM!

Most times IDE config files contain lots of absolute paths, and personal configs. In ancient times Eclipse config files did also contain CVS passwords...

Trust me, you do not like to have such a waste into your Source Code Management system!

There is usually a 
$> svn propedit svn:ignore .
which contains the following items:
target
.settings
.project
.classpath
*.iml
*.ipr
*.iws

Id suggest to use either maven or ant for the samples. Of course I prefer maven, because you dont need to checkin any 3rd party dependencies. And the IDE integration in Idea and Netbeans ist first class! The integration in Eclipse is pretty good also.


Regarding the Archetypes:

People like to see and browse samples online (svn browser). But they also like to have something quick they can start hacking on. This is exactly what archetypes provide!

In Apache MyFaces we have a few archetypes which generate simple projects skeletons. There are different types of them and I'd think of extending the CODI-JSF2.0 archetype and add OpenEJB on top.

http://wiki.apache.org/myfaces/MyFaces_Archetypes_for_Maven

The CODI JSF20 archetype should get updated (webbeansconfigurationlistener removed from web.xml and instead using webbeans-tomcat7 plugin), but its a good starting point. Just add openejb and a @Stateless backend service. 

LieGrue,
strub

--- On Sun, 5/29/11, Karan Malhi <ka...@gmail.com> wrote:

> From: Karan Malhi <ka...@gmail.com>
> Subject: Re: Ideas for Getting the word out
> To: dev@openejb.apache.org
> Date: Sunday, May 29, 2011, 2:26 PM
> Exactly my thought. Apart from
> releasing maven based projects, also have
> maven generate projects/zip files specific to IDE's and
> make them available
> for download from our website. Anytime we update the
> examples, the IDE
> specific zips should be one of the artifacts which would
> automatically get
> generated and posted to the website.
> 
> On Sun, May 29, 2011 at 8:34 AM, dsh <da...@googlemail.com>
> wrote:
> 
> > One thought concerning IDEs - If you export sample
> projects from an
> > IDE and store them in SVN to be imported by someone
> else you have to
> > maintain that IDE-specific samples too cause IDEs of
> course are
> > changing over time and hence you have to tweak such
> projects every now
> > and then. What I like more is having the build tool
> creating
> > IDE-specific projects from the meta descriptors
> provided by a build
> > system cause it could be suspected that such meta
> descriptors are
> > always up-to-date.
> >
> > Such a build system could be Maven for Java-specific
> projects or CMake
> > for C/C++ specific projects.
> >
> > Cheers
> > Daniel
> >
> > On Sun, May 29, 2011 at 2:21 PM, Karan Malhi <ka...@gmail.com>
> > wrote:
> > > The assumption that everybody knows maven is a
> dangerous one. Examples
> > > should also be packaged so that they could be
> imported into IDE's i.e.
> > zip
> > > files for eclipse, idea and netbeans. People
> typically have their
> > favorite
> > > IDE's already installed on their machines and
> giving the ability to
> > import
> > > examples directly into the IDE and "Get Started "
> is very effective.
> > > The thing with only relying on maven is that if
> somebody does not have it
> > > installed, they first have to install it, then
> they also need to know a
> > bit
> > > of it. I know that you can always tell them to
> simply run mvn clean
> > install
> > > to run the example, but for people not exposed to
> maven, there is always
> > a
> > > feeling of "I think maven did some magic to make
> it work. Wonder what it
> > > would take to get this example working in my IDE"
> - and this is where we
> > can
> > > potentially leave a bad taste. "First impression
> is a lasting one" - lets
> > > try and make that first impression a great one
> > >
> > > On Sun, May 29, 2011 at 4:04 AM, Aldrin Leal
> <al...@leal.eng.br>
> wrote:
> > >
> > >> For the examples, some could be turned into
> archetypes (which answers
> > the
> > >> previous question "What are archetypes useful
> for?") as a means to reach
> > a
> > >> wide audience: No need to download sources,
> leave it up to the archetype
> > >> (Actually Wicket Docs does just like that)
> > >>
> > >> As for the feeds, I suggest either a planet
> setup or simply a Twitter
> > User
> > >> List at first
> > >>
> > >> --
> > >> -- Aldrin Leal, <al...@leal.eng.br>
> /
> > http://www.leal.eng.br/mnemetica/
> > >>
> > >>
> > >> On Sat, May 28, 2011 at 8:50 PM, David
> Blevins <david.blevins@gmail.com
> > >> >wrote:
> > >>
> > >> > Some IRC chat resulted in some neat
> ideas on ways to dramatically
> > improve
> > >> > how easy it is to consume and learn
> about OpenEJB.
> > >> >
> > >> > # Examples
> > >> >
> > >> > One thing that hit home is that we have
> now 46 examples in trunk!!!
> > >> >  Amazing.  However the benefit
> of that is dramatically reduced as most
> > of
> > >> > them are only available in zip file
> form.  Links to svn aren't that
> > >> > effective.  And examples change too
> as the technology improves so
> > keeping
> > >> > wiki pages up to date is hard -- which
> version of the example do you
> > >> show?
> > >> >  What if you really want all
> versions available to see?
> > >> >
> > >> > So the idea was to use README files that
> are formatted in Markdown and
> > >> use
> > >> > that to generate a page for each
> example.  No more having part of the
> > >> > example in svn and part of it in
> confluence and then always breaking.
> >  A
> > >> > stackoverflow inspired solution --
> Markdown + Google's prettyprint.
> > >> >
> > >> > All the examples are in a zip file that
> is published in the maven
> > >> > repository, so we could use a maven
> program to pull them down and
> > extract
> > >> > them to the target directory where we
> can then do our little page
> > >> > generation.  A stub with those
> ideas in comments:
> > >> >
> > >> >
> > >> >
> > >>
> > http://svn.apache.org/repos/asf/openejb/trunk/sandbox/tools/src/main/java/org/apache/openejb/tools/examples/GenerateIndex.java
> > >> >
> > >> > # Retweeting
> > >> >
> > >> > We should monitor this feed
> > http://twitter.com/#!/OpenEJB/contributorsand
> > >> > retweet anything that mentions OpenEJB.
> > >> >
> > >> > So if anyone in the contributors list
> tweeted about OpenEJB, the
> > OpenEJB
> > >> > twitter account would retweet it.
> > >> >
> > >> > Two things will happen as a result:
> > >> >  -  The more activity on the
> OpenEJB twitter account the more
> > followers
> > >> it
> > >> > will get
> > >> >  -  The more @joe and other
> contributors are seen on the account, the
> > >> more
> > >> > followers they will get
> > >> >
> > >> > The OpenEJB twitter account has more
> followers than most everyone else
> > so
> > >> > getting it to retweet is a good way to
> expose people to all our
> > wonderful
> > >> > contributors and get them some followers
> and help the project at the
> > same
> > >> > time.
> > >> >
> > >> > The result is we as a community will
> have more ability overall to get
> > the
> > >> > word out!
> > >> >
> > >> > Twitterfeed.com was the obvious first
> idea, but it turns out twitter
> > does
> > >> > not allow you to post content from
> twitter back onto twitter.  Not
> > unless
> > >> > you use their API http://dev.twitter.com/doc .  So we could maybe
> > hack
> > >> > together some tool we run hourly and
> retweet things that contributors
> > >> tweet.
> > >> >  A little stub with comments here:
> > >> >
> > >> >
> > >> >
> > >>
> > http://svn.apache.org/repos/asf/openejb/trunk/sandbox/tools/src/main/java/org/apache/openejb/tools/twitter/Retweet.java
> > >> >
> > >> >
> > >> > Both are up for grabs!  If you're
> looking for something to do, either
> > one
> > >> > would be excellent ways to improve the
> project!
> > >> >
> > >> >
> > >> >
> > >> > -David
> > >> >
> > >> >
> > >> >
> > >> >
> > >> >
> > >> >
> > >> >
> > >> >
> > >> >
> > >>
> > >
> > >
> > >
> > > --
> > > Karan Singh Malhi
> > >
> >
> 
> 
> 
> -- 
> Karan Singh Malhi
> 

Re: Ideas for Getting the word out

[Karan Malhi <ka...@gmail.com>]

Exactly my thought. Apart from releasing maven based projects, also have
maven generate projects/zip files specific to IDE's and make them available
for download from our website. Anytime we update the examples, the IDE
specific zips should be one of the artifacts which would automatically get
generated and posted to the website.

On Sun, May 29, 2011 at 8:34 AM, dsh <da...@googlemail.com> wrote:

> One thought concerning IDEs - If you export sample projects from an
> IDE and store them in SVN to be imported by someone else you have to
> maintain that IDE-specific samples too cause IDEs of course are
> changing over time and hence you have to tweak such projects every now
> and then. What I like more is having the build tool creating
> IDE-specific projects from the meta descriptors provided by a build
> system cause it could be suspected that such meta descriptors are
> always up-to-date.
>
> Such a build system could be Maven for Java-specific projects or CMake
> for C/C++ specific projects.
>
> Cheers
> Daniel
>
> On Sun, May 29, 2011 at 2:21 PM, Karan Malhi <ka...@gmail.com>
> wrote:
> > The assumption that everybody knows maven is a dangerous one. Examples
> > should also be packaged so that they could be imported into IDE's i.e.
> zip
> > files for eclipse, idea and netbeans. People typically have their
> favorite
> > IDE's already installed on their machines and giving the ability to
> import
> > examples directly into the IDE and "Get Started " is very effective.
> > The thing with only relying on maven is that if somebody does not have it
> > installed, they first have to install it, then they also need to know a
> bit
> > of it. I know that you can always tell them to simply run mvn clean
> install
> > to run the example, but for people not exposed to maven, there is always
> a
> > feeling of "I think maven did some magic to make it work. Wonder what it
> > would take to get this example working in my IDE" - and this is where we
> can
> > potentially leave a bad taste. "First impression is a lasting one" - lets
> > try and make that first impression a great one
> >
> > On Sun, May 29, 2011 at 4:04 AM, Aldrin Leal <al...@leal.eng.br> wrote:
> >
> >> For the examples, some could be turned into archetypes (which answers
> the
> >> previous question "What are archetypes useful for?") as a means to reach
> a
> >> wide audience: No need to download sources, leave it up to the archetype
> >> (Actually Wicket Docs does just like that)
> >>
> >> As for the feeds, I suggest either a planet setup or simply a Twitter
> User
> >> List at first
> >>
> >> --
> >> -- Aldrin Leal, <al...@leal.eng.br> /
> http://www.leal.eng.br/mnemetica/
> >>
> >>
> >> On Sat, May 28, 2011 at 8:50 PM, David Blevins <david.blevins@gmail.com
> >> >wrote:
> >>
> >> > Some IRC chat resulted in some neat ideas on ways to dramatically
> improve
> >> > how easy it is to consume and learn about OpenEJB.
> >> >
> >> > # Examples
> >> >
> >> > One thing that hit home is that we have now 46 examples in trunk!!!
> >> >  Amazing.  However the benefit of that is dramatically reduced as most
> of
> >> > them are only available in zip file form.  Links to svn aren't that
> >> > effective.  And examples change too as the technology improves so
> keeping
> >> > wiki pages up to date is hard -- which version of the example do you
> >> show?
> >> >  What if you really want all versions available to see?
> >> >
> >> > So the idea was to use README files that are formatted in Markdown and
> >> use
> >> > that to generate a page for each example.  No more having part of the
> >> > example in svn and part of it in confluence and then always breaking.
>  A
> >> > stackoverflow inspired solution -- Markdown + Google's prettyprint.
> >> >
> >> > All the examples are in a zip file that is published in the maven
> >> > repository, so we could use a maven program to pull them down and
> extract
> >> > them to the target directory where we can then do our little page
> >> > generation.  A stub with those ideas in comments:
> >> >
> >> >
> >> >
> >>
> http://svn.apache.org/repos/asf/openejb/trunk/sandbox/tools/src/main/java/org/apache/openejb/tools/examples/GenerateIndex.java
> >> >
> >> > # Retweeting
> >> >
> >> > We should monitor this feed
> http://twitter.com/#!/OpenEJB/contributorsand
> >> > retweet anything that mentions OpenEJB.
> >> >
> >> > So if anyone in the contributors list tweeted about OpenEJB, the
> OpenEJB
> >> > twitter account would retweet it.
> >> >
> >> > Two things will happen as a result:
> >> >  -  The more activity on the OpenEJB twitter account the more
> followers
> >> it
> >> > will get
> >> >  -  The more @joe and other contributors are seen on the account, the
> >> more
> >> > followers they will get
> >> >
> >> > The OpenEJB twitter account has more followers than most everyone else
> so
> >> > getting it to retweet is a good way to expose people to all our
> wonderful
> >> > contributors and get them some followers and help the project at the
> same
> >> > time.
> >> >
> >> > The result is we as a community will have more ability overall to get
> the
> >> > word out!
> >> >
> >> > Twitterfeed.com was the obvious first idea, but it turns out twitter
> does
> >> > not allow you to post content from twitter back onto twitter.  Not
> unless
> >> > you use their API http://dev.twitter.com/doc .  So we could maybe
> hack
> >> > together some tool we run hourly and retweet things that contributors
> >> tweet.
> >> >  A little stub with comments here:
> >> >
> >> >
> >> >
> >>
> http://svn.apache.org/repos/asf/openejb/trunk/sandbox/tools/src/main/java/org/apache/openejb/tools/twitter/Retweet.java
> >> >
> >> >
> >> > Both are up for grabs!  If you're looking for something to do, either
> one
> >> > would be excellent ways to improve the project!
> >> >
> >> >
> >> >
> >> > -David
> >> >
> >> >
> >> >
> >> >
> >> >
> >> >
> >> >
> >> >
> >> >
> >>
> >
> >
> >
> > --
> > Karan Singh Malhi
> >
>



-- 
Karan Singh Malhi

Re: Ideas for Getting the word out

[dsh <da...@googlemail.com>]

One thought concerning IDEs - If you export sample projects from an
IDE and store them in SVN to be imported by someone else you have to
maintain that IDE-specific samples too cause IDEs of course are
changing over time and hence you have to tweak such projects every now
and then. What I like more is having the build tool creating
IDE-specific projects from the meta descriptors provided by a build
system cause it could be suspected that such meta descriptors are
always up-to-date.

Such a build system could be Maven for Java-specific projects or CMake
for C/C++ specific projects.

Cheers
Daniel

On Sun, May 29, 2011 at 2:21 PM, Karan Malhi <ka...@gmail.com> wrote:
> The assumption that everybody knows maven is a dangerous one. Examples
> should also be packaged so that they could be imported into IDE's i.e. zip
> files for eclipse, idea and netbeans. People typically have their favorite
> IDE's already installed on their machines and giving the ability to import
> examples directly into the IDE and "Get Started " is very effective.
> The thing with only relying on maven is that if somebody does not have it
> installed, they first have to install it, then they also need to know a bit
> of it. I know that you can always tell them to simply run mvn clean install
> to run the example, but for people not exposed to maven, there is always a
> feeling of "I think maven did some magic to make it work. Wonder what it
> would take to get this example working in my IDE" - and this is where we can
> potentially leave a bad taste. "First impression is a lasting one" - lets
> try and make that first impression a great one
>
> On Sun, May 29, 2011 at 4:04 AM, Aldrin Leal <al...@leal.eng.br> wrote:
>
>> For the examples, some could be turned into archetypes (which answers the
>> previous question "What are archetypes useful for?") as a means to reach a
>> wide audience: No need to download sources, leave it up to the archetype
>> (Actually Wicket Docs does just like that)
>>
>> As for the feeds, I suggest either a planet setup or simply a Twitter User
>> List at first
>>
>> --
>> -- Aldrin Leal, <al...@leal.eng.br> / http://www.leal.eng.br/mnemetica/
>>
>>
>> On Sat, May 28, 2011 at 8:50 PM, David Blevins <david.blevins@gmail.com
>> >wrote:
>>
>> > Some IRC chat resulted in some neat ideas on ways to dramatically improve
>> > how easy it is to consume and learn about OpenEJB.
>> >
>> > # Examples
>> >
>> > One thing that hit home is that we have now 46 examples in trunk!!!
>> >  Amazing.  However the benefit of that is dramatically reduced as most of
>> > them are only available in zip file form.  Links to svn aren't that
>> > effective.  And examples change too as the technology improves so keeping
>> > wiki pages up to date is hard -- which version of the example do you
>> show?
>> >  What if you really want all versions available to see?
>> >
>> > So the idea was to use README files that are formatted in Markdown and
>> use
>> > that to generate a page for each example.  No more having part of the
>> > example in svn and part of it in confluence and then always breaking.  A
>> > stackoverflow inspired solution -- Markdown + Google's prettyprint.
>> >
>> > All the examples are in a zip file that is published in the maven
>> > repository, so we could use a maven program to pull them down and extract
>> > them to the target directory where we can then do our little page
>> > generation.  A stub with those ideas in comments:
>> >
>> >
>> >
>> http://svn.apache.org/repos/asf/openejb/trunk/sandbox/tools/src/main/java/org/apache/openejb/tools/examples/GenerateIndex.java
>> >
>> > # Retweeting
>> >
>> > We should monitor this feed http://twitter.com/#!/OpenEJB/contributorsand
>> > retweet anything that mentions OpenEJB.
>> >
>> > So if anyone in the contributors list tweeted about OpenEJB, the OpenEJB
>> > twitter account would retweet it.
>> >
>> > Two things will happen as a result:
>> >  -  The more activity on the OpenEJB twitter account the more followers
>> it
>> > will get
>> >  -  The more @joe and other contributors are seen on the account, the
>> more
>> > followers they will get
>> >
>> > The OpenEJB twitter account has more followers than most everyone else so
>> > getting it to retweet is a good way to expose people to all our wonderful
>> > contributors and get them some followers and help the project at the same
>> > time.
>> >
>> > The result is we as a community will have more ability overall to get the
>> > word out!
>> >
>> > Twitterfeed.com was the obvious first idea, but it turns out twitter does
>> > not allow you to post content from twitter back onto twitter.  Not unless
>> > you use their API http://dev.twitter.com/doc .  So we could maybe hack
>> > together some tool we run hourly and retweet things that contributors
>> tweet.
>> >  A little stub with comments here:
>> >
>> >
>> >
>> http://svn.apache.org/repos/asf/openejb/trunk/sandbox/tools/src/main/java/org/apache/openejb/tools/twitter/Retweet.java
>> >
>> >
>> > Both are up for grabs!  If you're looking for something to do, either one
>> > would be excellent ways to improve the project!
>> >
>> >
>> >
>> > -David
>> >
>> >
>> >
>> >
>> >
>> >
>> >
>> >
>> >
>>
>
>
>
> --
> Karan Singh Malhi
>

Re: Ideas for Getting the word out

[Romain Manni-Bucau <rm...@gmail.com>]

even if i'm using maven (or build tool compatible with maven repos), i don't
like archetypes, i find them useless, to be able to browse examples online
is really better => we fastly see what we need.

Another solution is to provide a tool for each ide...and one for people not
using ide...so i think what is done today is good.

>From maven we can generate ant scripts.

I think all IDEs accept at least ant or maven build script...people using vi
know what to do with code ;).

- Romain

2011/5/29 Karan Malhi <ka...@gmail.com>

> The assumption that everybody knows maven is a dangerous one. Examples
> should also be packaged so that they could be imported into IDE's i.e. zip
> files for eclipse, idea and netbeans. People typically have their favorite
> IDE's already installed on their machines and giving the ability to import
> examples directly into the IDE and "Get Started " is very effective.
> The thing with only relying on maven is that if somebody does not have it
> installed, they first have to install it, then they also need to know a bit
> of it. I know that you can always tell them to simply run mvn clean install
> to run the example, but for people not exposed to maven, there is always a
> feeling of "I think maven did some magic to make it work. Wonder what it
> would take to get this example working in my IDE" - and this is where we
> can
> potentially leave a bad taste. "First impression is a lasting one" - lets
> try and make that first impression a great one
>
> On Sun, May 29, 2011 at 4:04 AM, Aldrin Leal <al...@leal.eng.br> wrote:
>
> > For the examples, some could be turned into archetypes (which answers the
> > previous question "What are archetypes useful for?") as a means to reach
> a
> > wide audience: No need to download sources, leave it up to the archetype
> > (Actually Wicket Docs does just like that)
> >
> > As for the feeds, I suggest either a planet setup or simply a Twitter
> User
> > List at first
> >
> > --
> > -- Aldrin Leal, <al...@leal.eng.br> / http://www.leal.eng.br/mnemetica/
> >
> >
> > On Sat, May 28, 2011 at 8:50 PM, David Blevins <david.blevins@gmail.com
> > >wrote:
> >
> > > Some IRC chat resulted in some neat ideas on ways to dramatically
> improve
> > > how easy it is to consume and learn about OpenEJB.
> > >
> > > # Examples
> > >
> > > One thing that hit home is that we have now 46 examples in trunk!!!
> > >  Amazing.  However the benefit of that is dramatically reduced as most
> of
> > > them are only available in zip file form.  Links to svn aren't that
> > > effective.  And examples change too as the technology improves so
> keeping
> > > wiki pages up to date is hard -- which version of the example do you
> > show?
> > >  What if you really want all versions available to see?
> > >
> > > So the idea was to use README files that are formatted in Markdown and
> > use
> > > that to generate a page for each example.  No more having part of the
> > > example in svn and part of it in confluence and then always breaking.
>  A
> > > stackoverflow inspired solution -- Markdown + Google's prettyprint.
> > >
> > > All the examples are in a zip file that is published in the maven
> > > repository, so we could use a maven program to pull them down and
> extract
> > > them to the target directory where we can then do our little page
> > > generation.  A stub with those ideas in comments:
> > >
> > >
> > >
> >
> http://svn.apache.org/repos/asf/openejb/trunk/sandbox/tools/src/main/java/org/apache/openejb/tools/examples/GenerateIndex.java
> > >
> > > # Retweeting
> > >
> > > We should monitor this feed
> http://twitter.com/#!/OpenEJB/contributorsand
> > > retweet anything that mentions OpenEJB.
> > >
> > > So if anyone in the contributors list tweeted about OpenEJB, the
> OpenEJB
> > > twitter account would retweet it.
> > >
> > > Two things will happen as a result:
> > >  -  The more activity on the OpenEJB twitter account the more followers
> > it
> > > will get
> > >  -  The more @joe and other contributors are seen on the account, the
> > more
> > > followers they will get
> > >
> > > The OpenEJB twitter account has more followers than most everyone else
> so
> > > getting it to retweet is a good way to expose people to all our
> wonderful
> > > contributors and get them some followers and help the project at the
> same
> > > time.
> > >
> > > The result is we as a community will have more ability overall to get
> the
> > > word out!
> > >
> > > Twitterfeed.com was the obvious first idea, but it turns out twitter
> does
> > > not allow you to post content from twitter back onto twitter.  Not
> unless
> > > you use their API http://dev.twitter.com/doc .  So we could maybe hack
> > > together some tool we run hourly and retweet things that contributors
> > tweet.
> > >  A little stub with comments here:
> > >
> > >
> > >
> >
> http://svn.apache.org/repos/asf/openejb/trunk/sandbox/tools/src/main/java/org/apache/openejb/tools/twitter/Retweet.java
> > >
> > >
> > > Both are up for grabs!  If you're looking for something to do, either
> one
> > > would be excellent ways to improve the project!
> > >
> > >
> > >
> > > -David
> > >
> > >
> > >
> > >
> > >
> > >
> > >
> > >
> > >
> >
>
>
>
> --
> Karan Singh Malhi
>

Re: Ideas for Getting the word out

[Karan Malhi <ka...@gmail.com>]

The assumption that everybody knows maven is a dangerous one. Examples
should also be packaged so that they could be imported into IDE's i.e. zip
files for eclipse, idea and netbeans. People typically have their favorite
IDE's already installed on their machines and giving the ability to import
examples directly into the IDE and "Get Started " is very effective.
The thing with only relying on maven is that if somebody does not have it
installed, they first have to install it, then they also need to know a bit
of it. I know that you can always tell them to simply run mvn clean install
to run the example, but for people not exposed to maven, there is always a
feeling of "I think maven did some magic to make it work. Wonder what it
would take to get this example working in my IDE" - and this is where we can
potentially leave a bad taste. "First impression is a lasting one" - lets
try and make that first impression a great one

On Sun, May 29, 2011 at 4:04 AM, Aldrin Leal <al...@leal.eng.br> wrote:

> For the examples, some could be turned into archetypes (which answers the
> previous question "What are archetypes useful for?") as a means to reach a
> wide audience: No need to download sources, leave it up to the archetype
> (Actually Wicket Docs does just like that)
>
> As for the feeds, I suggest either a planet setup or simply a Twitter User
> List at first
>
> --
> -- Aldrin Leal, <al...@leal.eng.br> / http://www.leal.eng.br/mnemetica/
>
>
> On Sat, May 28, 2011 at 8:50 PM, David Blevins <david.blevins@gmail.com
> >wrote:
>
> > Some IRC chat resulted in some neat ideas on ways to dramatically improve
> > how easy it is to consume and learn about OpenEJB.
> >
> > # Examples
> >
> > One thing that hit home is that we have now 46 examples in trunk!!!
> >  Amazing.  However the benefit of that is dramatically reduced as most of
> > them are only available in zip file form.  Links to svn aren't that
> > effective.  And examples change too as the technology improves so keeping
> > wiki pages up to date is hard -- which version of the example do you
> show?
> >  What if you really want all versions available to see?
> >
> > So the idea was to use README files that are formatted in Markdown and
> use
> > that to generate a page for each example.  No more having part of the
> > example in svn and part of it in confluence and then always breaking.  A
> > stackoverflow inspired solution -- Markdown + Google's prettyprint.
> >
> > All the examples are in a zip file that is published in the maven
> > repository, so we could use a maven program to pull them down and extract
> > them to the target directory where we can then do our little page
> > generation.  A stub with those ideas in comments:
> >
> >
> >
> http://svn.apache.org/repos/asf/openejb/trunk/sandbox/tools/src/main/java/org/apache/openejb/tools/examples/GenerateIndex.java
> >
> > # Retweeting
> >
> > We should monitor this feed http://twitter.com/#!/OpenEJB/contributorsand
> > retweet anything that mentions OpenEJB.
> >
> > So if anyone in the contributors list tweeted about OpenEJB, the OpenEJB
> > twitter account would retweet it.
> >
> > Two things will happen as a result:
> >  -  The more activity on the OpenEJB twitter account the more followers
> it
> > will get
> >  -  The more @joe and other contributors are seen on the account, the
> more
> > followers they will get
> >
> > The OpenEJB twitter account has more followers than most everyone else so
> > getting it to retweet is a good way to expose people to all our wonderful
> > contributors and get them some followers and help the project at the same
> > time.
> >
> > The result is we as a community will have more ability overall to get the
> > word out!
> >
> > Twitterfeed.com was the obvious first idea, but it turns out twitter does
> > not allow you to post content from twitter back onto twitter.  Not unless
> > you use their API http://dev.twitter.com/doc .  So we could maybe hack
> > together some tool we run hourly and retweet things that contributors
> tweet.
> >  A little stub with comments here:
> >
> >
> >
> http://svn.apache.org/repos/asf/openejb/trunk/sandbox/tools/src/main/java/org/apache/openejb/tools/twitter/Retweet.java
> >
> >
> > Both are up for grabs!  If you're looking for something to do, either one
> > would be excellent ways to improve the project!
> >
> >
> >
> > -David
> >
> >
> >
> >
> >
> >
> >
> >
> >
>



-- 
Karan Singh Malhi

Re: Ideas for Getting the word out -- Maven support

[David Blevins <da...@gmail.com>]

On May 29, 2011, at 1:04 AM, Aldrin Leal wrote:

> For the examples, some could be turned into archetypes (which answers the
> previous question "What are archetypes useful for?") as a means to reach a
> wide audience: No need to download sources, leave it up to the archetype
> (Actually Wicket Docs does just like that)

That's a pretty cool idea.  The copy/edit approach is the one I usually take when creating examples and I admit even I find that tedious sometimes.  After 46 examples or so it gets old.

As well I can see how with so many examples, it can get increasingly difficult for people to know which to copy.

Off the top of my head, it seems like there's probably a dozen examples that would make great archetypes.  A moviefun/JPA one.  Security. Transactions.  Webserivces. MDBs.  Singletons.

Maybe even a multiproject one that has a couple modules and a war that is built and tested against TomEE.

> As for the feeds, I suggest either a planet setup or simply a Twitter User
> List at first

We have a list -- we should add you too it as well -- but we have some very chatty committers :)  Something more selective would be great

  https://twitter.com/#!/OpenEJB/contributors

-David


> 
> --
> -- Aldrin Leal, <al...@leal.eng.br> / http://www.leal.eng.br/mnemetica/
> 
> 
> On Sat, May 28, 2011 at 8:50 PM, David Blevins <da...@gmail.com>wrote:
> 
>> Some IRC chat resulted in some neat ideas on ways to dramatically improve
>> how easy it is to consume and learn about OpenEJB.
>> 
>> # Examples
>> 
>> One thing that hit home is that we have now 46 examples in trunk!!!
>> Amazing.  However the benefit of that is dramatically reduced as most of
>> them are only available in zip file form.  Links to svn aren't that
>> effective.  And examples change too as the technology improves so keeping
>> wiki pages up to date is hard -- which version of the example do you show?
>> What if you really want all versions available to see?
>> 
>> So the idea was to use README files that are formatted in Markdown and use
>> that to generate a page for each example.  No more having part of the
>> example in svn and part of it in confluence and then always breaking.  A
>> stackoverflow inspired solution -- Markdown + Google's prettyprint.
>> 
>> All the examples are in a zip file that is published in the maven
>> repository, so we could use a maven program to pull them down and extract
>> them to the target directory where we can then do our little page
>> generation.  A stub with those ideas in comments:
>> 
>> 
>> http://svn.apache.org/repos/asf/openejb/trunk/sandbox/tools/src/main/java/org/apache/openejb/tools/examples/GenerateIndex.java
>> 
>> # Retweeting
>> 
>> We should monitor this feed http://twitter.com/#!/OpenEJB/contributors and
>> retweet anything that mentions OpenEJB.
>> 
>> So if anyone in the contributors list tweeted about OpenEJB, the OpenEJB
>> twitter account would retweet it.
>> 
>> Two things will happen as a result:
>> -  The more activity on the OpenEJB twitter account the more followers it
>> will get
>> -  The more @joe and other contributors are seen on the account, the more
>> followers they will get
>> 
>> The OpenEJB twitter account has more followers than most everyone else so
>> getting it to retweet is a good way to expose people to all our wonderful
>> contributors and get them some followers and help the project at the same
>> time.
>> 
>> The result is we as a community will have more ability overall to get the
>> word out!
>> 
>> Twitterfeed.com was the obvious first idea, but it turns out twitter does
>> not allow you to post content from twitter back onto twitter.  Not unless
>> you use their API http://dev.twitter.com/doc .  So we could maybe hack
>> together some tool we run hourly and retweet things that contributors tweet.
>> A little stub with comments here:
>> 
>> 
>> http://svn.apache.org/repos/asf/openejb/trunk/sandbox/tools/src/main/java/org/apache/openejb/tools/twitter/Retweet.java
>> 
>> 
>> Both are up for grabs!  If you're looking for something to do, either one
>> would be excellent ways to improve the project!
>> 
>> 
>> 
>> -David
>> 
>> 
>> 
>> 
>> 
>> 
>> 
>> 
>> 

Re: Ideas for Getting the word out

[Aldrin Leal <al...@leal.eng.br>]

For the examples, some could be turned into archetypes (which answers the
previous question "What are archetypes useful for?") as a means to reach a
wide audience: No need to download sources, leave it up to the archetype
(Actually Wicket Docs does just like that)

As for the feeds, I suggest either a planet setup or simply a Twitter User
List at first

--
-- Aldrin Leal, <al...@leal.eng.br> / http://www.leal.eng.br/mnemetica/


On Sat, May 28, 2011 at 8:50 PM, David Blevins <da...@gmail.com>wrote:

> Some IRC chat resulted in some neat ideas on ways to dramatically improve
> how easy it is to consume and learn about OpenEJB.
>
> # Examples
>
> One thing that hit home is that we have now 46 examples in trunk!!!
>  Amazing.  However the benefit of that is dramatically reduced as most of
> them are only available in zip file form.  Links to svn aren't that
> effective.  And examples change too as the technology improves so keeping
> wiki pages up to date is hard -- which version of the example do you show?
>  What if you really want all versions available to see?
>
> So the idea was to use README files that are formatted in Markdown and use
> that to generate a page for each example.  No more having part of the
> example in svn and part of it in confluence and then always breaking.  A
> stackoverflow inspired solution -- Markdown + Google's prettyprint.
>
> All the examples are in a zip file that is published in the maven
> repository, so we could use a maven program to pull them down and extract
> them to the target directory where we can then do our little page
> generation.  A stub with those ideas in comments:
>
>
> http://svn.apache.org/repos/asf/openejb/trunk/sandbox/tools/src/main/java/org/apache/openejb/tools/examples/GenerateIndex.java
>
> # Retweeting
>
> We should monitor this feed http://twitter.com/#!/OpenEJB/contributors and
> retweet anything that mentions OpenEJB.
>
> So if anyone in the contributors list tweeted about OpenEJB, the OpenEJB
> twitter account would retweet it.
>
> Two things will happen as a result:
>  -  The more activity on the OpenEJB twitter account the more followers it
> will get
>  -  The more @joe and other contributors are seen on the account, the more
> followers they will get
>
> The OpenEJB twitter account has more followers than most everyone else so
> getting it to retweet is a good way to expose people to all our wonderful
> contributors and get them some followers and help the project at the same
> time.
>
> The result is we as a community will have more ability overall to get the
> word out!
>
> Twitterfeed.com was the obvious first idea, but it turns out twitter does
> not allow you to post content from twitter back onto twitter.  Not unless
> you use their API http://dev.twitter.com/doc .  So we could maybe hack
> together some tool we run hourly and retweet things that contributors tweet.
>  A little stub with comments here:
>
>
> http://svn.apache.org/repos/asf/openejb/trunk/sandbox/tools/src/main/java/org/apache/openejb/tools/twitter/Retweet.java
>
>
> Both are up for grabs!  If you're looking for something to do, either one
> would be excellent ways to improve the project!
>
>
>
> -David
>
>
>
>
>
>
>
>
>

Re: Ideas for Getting the word out

[Romain Manni-Bucau <rm...@gmail.com>]

Hi,

It will sound stupid but the first thing i thought when i saw the template
website and logo was "is this project dead" because it is very flat. Maybe
we could improve it and the 4.0.0 is the release to do it no?

- Romain

Le 29 mai 2011 08:51, "Mark Struberg" <st...@yahoo.de> a écrit :
> Hi David!
>
> As this is true for a lot Apache projects, we might also discuss this on a
foundation level?
>
> I think there is already such a small incubator project which contains
'inhouse' tools as far as I remember.
>
> LieGrue,
> strub
>
> --- On Sat, 5/28/11, David Blevins <da...@gmail.com> wrote:
>
>> From: David Blevins <da...@gmail.com>
>> Subject: Ideas for Getting the word out
>> To: dev@openejb.apache.org
>> Date: Saturday, May 28, 2011, 11:50 PM
>> Some IRC chat resulted in some neat
>> ideas on ways to dramatically improve how easy it is to
>> consume and learn about OpenEJB.
>>
>> # Examples
>>
>> One thing that hit home is that we have now 46 examples in
>> trunk!!!  Amazing.  However the benefit of that is
>> dramatically reduced as most of them are only available in
>> zip file form.  Links to svn aren't that
>> effective.  And examples change too as the technology
>> improves so keeping wiki pages up to date is hard -- which
>> version of the example do you show?  What if you really
>> want all versions available to see?
>>
>> So the idea was to use README files that are formatted in
>> Markdown and use that to generate a page for each
>> example.  No more having part of the example in svn and
>> part of it in confluence and then always breaking.  A
>> stackoverflow inspired solution -- Markdown + Google's
>> prettyprint.
>>
>> All the examples are in a zip file that is published in the
>> maven repository, so we could use a maven program to pull
>> them down and extract them to the target directory where we
>> can then do our little page generation.  A stub with
>> those ideas in comments:
>>
>>
http://svn.apache.org/repos/asf/openejb/trunk/sandbox/tools/src/main/java/org/apache/openejb/tools/examples/GenerateIndex.java
>>
>> # Retweeting
>>
>> We should monitor this feed http://twitter.com/#!/OpenEJB/contributorsand retweet
>> anything that mentions OpenEJB.
>>
>> So if anyone in the contributors list tweeted about
>> OpenEJB, the OpenEJB twitter account would retweet it.
>>
>> Two things will happen as a result:
>>   -  The more activity on the OpenEJB twitter
>> account the more followers it will get
>>   -  The more @joe and other contributors are
>> seen on the account, the more followers they will get
>>
>> The OpenEJB twitter account has more followers than most
>> everyone else so getting it to retweet is a good way to
>> expose people to all our wonderful contributors and get them
>> some followers and help the project at the same time.
>>
>> The result is we as a community will have more ability
>> overall to get the word out!
>>
>> Twitterfeed.com was the obvious first idea, but it turns
>> out twitter does not allow you to post content from twitter
>> back onto twitter.  Not unless you use their API
http://dev.twitter.com/doc .  So we could maybe
>> hack together some tool we run hourly and retweet things
>> that contributors tweet.  A little stub with comments
>> here:
>>
>>
http://svn.apache.org/repos/asf/openejb/trunk/sandbox/tools/src/main/java/org/apache/openejb/tools/twitter/Retweet.java
>>
>>
>> Both are up for grabs!  If you're looking for
>> something to do, either one would be excellent ways to
>> improve the project!
>>
>>
>>
>> -David
>>
>>
>>
>>
>>
>>
>>
>>
>>

Re: Ideas for Getting the word out

[David Blevins <da...@gmail.com>]

On May 28, 2011, at 11:51 PM, Mark Struberg wrote:

> Hi David!
> 
> As this is true for a lot Apache projects, we might also discuss this on a foundation level?

Good idea.  Sent a note too community dev see if anyone had a twitter bot to share.

In terms of developing something at the foundation level, that's a terrible way[1] to build a bike shed :)  Apache is huge.  If I had the time to read and respond to all that email, I'd just write the tools myself :)

Each task is probably 500 lines of the right code using the right libraries.  More if you get fancy, but the heart of each is tiny.

It's tempting to write them myself, but we have a handful of people looking to get involved with the project and bite-sized mini-projects like these are a great way to get people started.  They're fun and chewable! :)


-David

[1] http://en.wikipedia.org/wiki/Parkinson's_Law_of_Triviality

> --- On Sat, 5/28/11, David Blevins <da...@gmail.com> wrote:
> 
>> From: David Blevins <da...@gmail.com>
>> Subject: Ideas for Getting the word out
>> To: dev@openejb.apache.org
>> Date: Saturday, May 28, 2011, 11:50 PM
>> Some IRC chat resulted in some neat
>> ideas on ways to dramatically improve how easy it is to
>> consume and learn about OpenEJB.
>> 
>> # Examples
>> 
>> One thing that hit home is that we have now 46 examples in
>> trunk!!!  Amazing.  However the benefit of that is
>> dramatically reduced as most of them are only available in
>> zip file form.  Links to svn aren't that
>> effective.  And examples change too as the technology
>> improves so keeping wiki pages up to date is hard -- which
>> version of the example do you show?  What if you really
>> want all versions available to see?
>> 
>> So the idea was to use README files that are formatted in
>> Markdown and use that to generate a page for each
>> example.  No more having part of the example in svn and
>> part of it in confluence and then always breaking.  A
>> stackoverflow inspired solution -- Markdown + Google's
>> prettyprint.
>> 
>> All the examples are in a zip file that is published in the
>> maven repository, so we could use a maven program to pull
>> them down and extract them to the target directory where we
>> can then do our little page generation.  A stub with
>> those ideas in comments:
>> 
>>   http://svn.apache.org/repos/asf/openejb/trunk/sandbox/tools/src/main/java/org/apache/openejb/tools/examples/GenerateIndex.java
>> 
>> # Retweeting
>> 
>> We should monitor this feed http://twitter.com/#!/OpenEJB/contributors and retweet
>> anything that mentions OpenEJB.
>> 
>> So if anyone in the contributors list tweeted about
>> OpenEJB, the OpenEJB twitter account would retweet it.
>> 
>> Two things will happen as a result:
>>   -  The more activity on the OpenEJB twitter
>> account the more followers it will get
>>   -  The more @joe and other contributors are
>> seen on the account, the more followers they will get
>> 
>> The OpenEJB twitter account has more followers than most
>> everyone else so getting it to retweet is a good way to
>> expose people to all our wonderful contributors and get them
>> some followers and help the project at the same time.
>> 
>> The result is we as a community will have more ability
>> overall to get the word out!
>> 
>> Twitterfeed.com was the obvious first idea, but it turns
>> out twitter does not allow you to post content from twitter
>> back onto twitter.  Not unless you use their API http://dev.twitter.com/doc .  So we could maybe
>> hack together some tool we run hourly and retweet things
>> that contributors tweet.  A little stub with comments
>> here:
>> 
>>   http://svn.apache.org/repos/asf/openejb/trunk/sandbox/tools/src/main/java/org/apache/openejb/tools/twitter/Retweet.java
>> 
>> 
>> Both are up for grabs!  If you're looking for
>> something to do, either one would be excellent ways to
>> improve the project!
>> 
>> 
>> 
>> -David
>> 
>> 
>>   
>> 
>> 
>> 
>> 
>> 
>> 

Re: Ideas for Getting the word out

[Mark Struberg <st...@yahoo.de>]

Hi David!

As this is true for a lot Apache projects, we might also discuss this on a foundation level?

I think there is already such a small incubator project which contains 'inhouse' tools as far as I remember.

LieGrue,
strub

--- On Sat, 5/28/11, David Blevins <da...@gmail.com> wrote:

> From: David Blevins <da...@gmail.com>
> Subject: Ideas for Getting the word out
> To: dev@openejb.apache.org
> Date: Saturday, May 28, 2011, 11:50 PM
> Some IRC chat resulted in some neat
> ideas on ways to dramatically improve how easy it is to
> consume and learn about OpenEJB.
> 
> # Examples
> 
> One thing that hit home is that we have now 46 examples in
> trunk!!!  Amazing.  However the benefit of that is
> dramatically reduced as most of them are only available in
> zip file form.  Links to svn aren't that
> effective.  And examples change too as the technology
> improves so keeping wiki pages up to date is hard -- which
> version of the example do you show?  What if you really
> want all versions available to see?
> 
> So the idea was to use README files that are formatted in
> Markdown and use that to generate a page for each
> example.  No more having part of the example in svn and
> part of it in confluence and then always breaking.  A
> stackoverflow inspired solution -- Markdown + Google's
> prettyprint.
> 
> All the examples are in a zip file that is published in the
> maven repository, so we could use a maven program to pull
> them down and extract them to the target directory where we
> can then do our little page generation.  A stub with
> those ideas in comments:
> 
>   http://svn.apache.org/repos/asf/openejb/trunk/sandbox/tools/src/main/java/org/apache/openejb/tools/examples/GenerateIndex.java
> 
> # Retweeting
> 
> We should monitor this feed http://twitter.com/#!/OpenEJB/contributors and retweet
> anything that mentions OpenEJB.
> 
> So if anyone in the contributors list tweeted about
> OpenEJB, the OpenEJB twitter account would retweet it.
> 
> Two things will happen as a result:
>   -  The more activity on the OpenEJB twitter
> account the more followers it will get
>   -  The more @joe and other contributors are
> seen on the account, the more followers they will get
> 
> The OpenEJB twitter account has more followers than most
> everyone else so getting it to retweet is a good way to
> expose people to all our wonderful contributors and get them
> some followers and help the project at the same time.
> 
> The result is we as a community will have more ability
> overall to get the word out!
> 
> Twitterfeed.com was the obvious first idea, but it turns
> out twitter does not allow you to post content from twitter
> back onto twitter.  Not unless you use their API http://dev.twitter.com/doc .  So we could maybe
> hack together some tool we run hourly and retweet things
> that contributors tweet.  A little stub with comments
> here:
> 
>   http://svn.apache.org/repos/asf/openejb/trunk/sandbox/tools/src/main/java/org/apache/openejb/tools/twitter/Retweet.java
> 
> 
> Both are up for grabs!  If you're looking for
> something to do, either one would be excellent ways to
> improve the project!
> 
> 
> 
> -David
> 
> 
>   
> 
> 
> 
> 
> 
>