I'm back... (sort of)

[Raphaël Luta <ra...@networks.groupvu.com>]

Hi all,

I'm back on this list :)

Status update:

I could practivaly not work on jetspeed over the last 2 weeks so I did not make
a lot of progress on the new layout.
However, I did commit the current configuration OM I have on my laptop so that
I can start getting feedback on it.
In this version, I don't use any Castor generated API anymore but rather a 
Javaben
like implementation which can be mapped to XML files using Castor mapping 
files.

Tonight, I plan to commit the other package I'm working on (aggregation engine)
but I'm not very pleased with its current status and architecture.

Also, I'm flying to NY tomorrow and then make a short trip to LA before 
going to
ApacheCon next week, so while I should have complete net access during these
2 weeks, I can't guarantee I'll be very active during this time... :(


--
Raphaël Luta - raphael.luta@networks.groupvu.com
Vivendi Universal Networks - Services Manager / Paris


---------------------------------------------------------------------
To unsubscribe, e-mail: jetspeed-dev-unsubscribe@jakarta.apache.org
For additional commands, e-mail: jetspeed-dev-help@jakarta.apache.org

Re: I'm back... (sort of)

[Jon Stevens <jo...@latchkey.com>]

on 3/26/01 2:56 PM, "Santiago Gala" <sg...@hisitech.com> wrote:

> Now I see the point. I thought commits were moderated because of
> potential harm of some commits to the repository (for instance, the one
> I did of the jetspeed docs with all the tables badly translated due to a
> version clash).

That is why the word "version" is in CVS.

> It could make sense to have moderated commits in some cases, but it
> looks very tricky to implement. It would be simpler to have a rollback
> facility. Both features would require to serialize the changes to the
> repository very carefully.

Uh. CVS does have rollback. That is why the word "version" is in CVS.

> Now I see there was again a misunderstanding. I was thinking there were
> some kind of "rules", moderation, younameit about these entries, while
> you were thinking I was trying to get it done by free (as in beer, not
> speech :)

The only rule is that you get to do the work.

> Now, I'll do it for whoever is commiter and send me either the entry as
> such or a diff. I imagine the size should be reasonable, like the
> current entries.
> 
> (A friend told me that I was not Apache commiter, since i was not in
> this page. That is how I noticed it.)

You aren't an "Apache commiter". You are a "Jetspeed and Jakarta site2
commiter".

No one is an "Apache commiter" as that doesn't really have any real meaning.

-jon


---------------------------------------------------------------------
To unsubscribe, e-mail: jetspeed-dev-unsubscribe@jakarta.apache.org
For additional commands, e-mail: jetspeed-dev-help@jakarta.apache.org

Re: I'm back... (sort of)

[Santiago Gala <sg...@hisitech.com>]

Jon Stevens wrote:

> on 3/26/01 2:13 PM, "Santiago Gala" <sg...@hisitech.com> wrote:
> 
> 
>> First thought, it looked natural to moderate the whole commit process...
>> 
>> But then, my hackish mind came to play, and I questioned myself:
>> 
>> What happens if the moderated mails arrive out of order to the cvs
>> engine? Potentially a disaster. Or if the moderator accepts part of a
>> commit, but not the rest.
>> 
>> So I am in doubt, and I ask. I still don't see it. It looks too complex
>> for me to work this way.
> 
> 
> Ok. CVS Admin 101:
> 
> When a commit happens. cvs reads a file called: loginfo that is stored in
> the CVSROOT directory.
> 
> the file looks something like this:
> 
> DEFAULT $CVSROOT/CVSROOT/log_accum.pl %s
> 
> That causes a perl script to be run. %s is the username and the data for the
> commit is passed into stdin.
> 
> That script then sends the commit email. That commit email has the From:
> header set to be the userid that does the commit (passed in via %s). The To:
> header is set to be the name of the repo + cvs. So, a commit would be:
> 
> From: sgala@apache.org
> To: jakarta-jetspeed-cvs@apache.org
> 
> That message is then sent to the list by way of an alias that points:
> 
> jakarta-jetspeed-cvs -> jetspeed-dev
> 
> Look in ~apmail for the aliases...(ls -la ~apmail)
> 
> So, because I'm the moderator of the list, I get the message because you are
> not subscribe to the jetspeed-dev list with your apache.org address, you are
> subscribed with your sgala@hisitech.com address.
> 
> When I moderate the message, I do it in such a way to "accept" all further
> messages from that address automatically. So, the moderation only has to
> happen once.
>


> Now, why would the actual cvs commits get moderated?
> 


Now I see the point. I thought commits were moderated because of 
potential harm of some commits to the repository (for instance, the one 
I did of the jetspeed docs with all the tables badly translated due to a 
version clash).

It could make sense to have moderated commits in some cases, but it 
looks very tricky to implement. It would be simpler to have a rollback 
facility. Both features would require to serialize the changes to the 
repository very carefully.

But the "real" problem is about posting to the list without having 
subscribed to it.

As you can see, I'm completely hopeless, getting a cold, and will go to 
bed right now :)


> 
>>>> so I can do it for me, and for other jetspeed commiters that send me
>>>> their patches.
>>> 
>>> oh come on...think about it for like 2 seconds...
>> 
>> Here again. I feel thick tonight. I don't see. Does it mean yes, no,
>> maybe, or completely hopeless?
> 
> 
> This is a volunteer organization. You do the work, otherwise it doesn't get
> done until someone else volunteers to do it. A while back, Ceki volunteered
> to do it and sent an email out to all the committers. Maybe you missed it.
> 

Possibly I was not yet in jakarta.

Now I see there was again a misunderstanding. I was thinking there were 
some kind of "rules", moderation, younameit about these entries, while 
you were thinking I was trying to get it done by free (as in beer, not 
speech :)

Now, I'll do it for whoever is commiter and send me either the entry as 
such or a diff. I imagine the size should be reasonable, like the 
current entries.

(A friend told me that I was not Apache commiter, since i was not in 
this page. That is how I noticed it.)


---------------------------------------------------------------------
To unsubscribe, e-mail: jetspeed-dev-unsubscribe@jakarta.apache.org
For additional commands, e-mail: jetspeed-dev-help@jakarta.apache.org

Re: I'm back... (sort of)

[Jon Stevens <jo...@latchkey.com>]

on 3/26/01 2:13 PM, "Santiago Gala" <sg...@hisitech.com> wrote:

> First thought, it looked natural to moderate the whole commit process...
> 
> But then, my hackish mind came to play, and I questioned myself:
> 
> What happens if the moderated mails arrive out of order to the cvs
> engine? Potentially a disaster. Or if the moderator accepts part of a
> commit, but not the rest.
> 
> So I am in doubt, and I ask. I still don't see it. It looks too complex
> for me to work this way.

Ok. CVS Admin 101:

When a commit happens. cvs reads a file called: loginfo that is stored in
the CVSROOT directory.

the file looks something like this:

DEFAULT $CVSROOT/CVSROOT/log_accum.pl %s

That causes a perl script to be run. %s is the username and the data for the
commit is passed into stdin.

That script then sends the commit email. That commit email has the From:
header set to be the userid that does the commit (passed in via %s). The To:
header is set to be the name of the repo + cvs. So, a commit would be:

From: sgala@apache.org
To: jakarta-jetspeed-cvs@apache.org

That message is then sent to the list by way of an alias that points:

jakarta-jetspeed-cvs -> jetspeed-dev

Look in ~apmail for the aliases...(ls -la ~apmail)

So, because I'm the moderator of the list, I get the message because you are
not subscribe to the jetspeed-dev list with your apache.org address, you are
subscribed with your sgala@hisitech.com address.

When I moderate the message, I do it in such a way to "accept" all further
messages from that address automatically. So, the moderation only has to
happen once.

Now, why would the actual cvs commits get moderated?

>>> so I can do it for me, and for other jetspeed commiters that send me
>>> their patches.
>> 
>> oh come on...think about it for like 2 seconds...
> 
> Here again. I feel thick tonight. I don't see. Does it mean yes, no,
> maybe, or completely hopeless?

This is a volunteer organization. You do the work, otherwise it doesn't get
done until someone else volunteers to do it. A while back, Ceki volunteered
to do it and sent an email out to all the committers. Maybe you missed it.

-jon


---------------------------------------------------------------------
To unsubscribe, e-mail: jetspeed-dev-unsubscribe@jakarta.apache.org
For additional commands, e-mail: jetspeed-dev-help@jakarta.apache.org

Re: I'm back... (sort of)

[Santiago Gala <sg...@hisitech.com>]

Jon Stevens wrote:

> on 3/26/01 1:25 PM, "Santiago Gala" <sg...@hisitech.com> wrote:
> 
> 
>> I didn't know. (Now I do :)
> 
> 
> i posted about it when we first moved over...i said that if there is a delay
> in mail getting through (especially for commits) that it was because i'm the
> moderator for this list...
> 
> 
>> A question: the commits themselves are moderated, or just the mails?
> 
> 
> oh come on...think about it for like 2 seconds...
> 

First thought, it looked natural to moderate the whole commit process...

But then, my hackish mind came to play, and I questioned myself:

What happens if the moderated mails arrive out of order to the cvs 
engine? Potentially a disaster. Or if the moderator accepts part of a 
commit, but not the rest.

So I am in doubt, and I ask. I still don't see it. It looks too complex 
for me to work this way.


> 
>> A second question: are we supposed to add ourselves to the
>> "Contributors" page in site2 (jakarta.apache.org)? I have the karma now,

It is really committers, under whoweare.html


>> so I can do it for me, and for other jetspeed commiters that send me
>> their patches.
> 
> 
> oh come on...think about it for like 2 seconds...
> 

Here again. I feel thick tonight. I don't see. Does it mean yes, no, 
maybe, or completely hopeless?


---------------------------------------------------------------------
To unsubscribe, e-mail: jetspeed-dev-unsubscribe@jakarta.apache.org
For additional commands, e-mail: jetspeed-dev-help@jakarta.apache.org

Re: I'm back... (sort of)

[Jon Stevens <jo...@latchkey.com>]

on 3/26/01 1:25 PM, "Santiago Gala" <sg...@hisitech.com> wrote:

> I didn't know. (Now I do :)

i posted about it when we first moved over...i said that if there is a delay
in mail getting through (especially for commits) that it was because i'm the
moderator for this list...

> A question: the commits themselves are moderated, or just the mails?

oh come on...think about it for like 2 seconds...

> A second question: are we supposed to add ourselves to the
> "Contributors" page in site2 (jakarta.apache.org)? I have the karma now,
> so I can do it for me, and for other jetspeed commiters that send me
> their patches.

oh come on...think about it for like 2 seconds...

-jon


---------------------------------------------------------------------
To unsubscribe, e-mail: jetspeed-dev-unsubscribe@jakarta.apache.org
For additional commands, e-mail: jetspeed-dev-help@jakarta.apache.org

Re: I'm back... (sort of)

[Santiago Gala <sg...@hisitech.com>]

Jon Stevens wrote:

> on 3/26/01 8:45 AM, "Santiago Gala" <sg...@hisitech.com> wrote:
> 
> 
>> ??? The are no entries in the jetspeed-cvs list.
>> I thought they would appear there, even if done in a branch.
> 
> 
> I had to moderate approve them.
> 

I didn't know. (Now I do :)

A question: the commits themselves are moderated, or just the mails?

A second question: are we supposed to add ourselves to the 
"Contributors" page in site2 (jakarta.apache.org)? I have the karma now, 
so I can do it for me, and for other jetspeed commiters that send me 
their patches.


---------------------------------------------------------------------
To unsubscribe, e-mail: jetspeed-dev-unsubscribe@jakarta.apache.org
For additional commands, e-mail: jetspeed-dev-help@jakarta.apache.org

Re: I'm back... (sort of)

[Jon Stevens <jo...@latchkey.com>]

on 3/26/01 8:45 AM, "Santiago Gala" <sg...@hisitech.com> wrote:

> ??? The are no entries in the jetspeed-cvs list.
> I thought they would appear there, even if done in a branch.

I had to moderate approve them.

-jon


-- 
If you come from a Perl or PHP background, JSP is a way to take
your pain to new levels. --Anonymous
<http://jakarta.apache.org/velocity/ymtd/ymtd.html>


---------------------------------------------------------------------
To unsubscribe, e-mail: jetspeed-dev-unsubscribe@jakarta.apache.org
For additional commands, e-mail: jetspeed-dev-help@jakarta.apache.org

Re: I'm back... (sort of)

[Santiago Gala <sg...@hisitech.com>]

Raphaël Luta wrote:

> Hi all,
> 
> I'm back on this list :)
> 
> Status update:
> 
> I could practivaly not work on jetspeed over the last 2 weeks so I did 
> not make
> a lot of progress on the new layout.
> However, I did commit the current configuration OM I have on my laptop 
> so that
> I can start getting feedback on it.

??? The are no entries in the jetspeed-cvs list.
I thought they would appear there, even if done in a branch.


> In this version, I don't use any Castor generated API anymore but rather 
> a Javaben
> like implementation which can be mapped to XML files using Castor 
> mapping files.
> 

I will look at this carefully. I was about to remove Castor API from the 
feed daemon implementation, since it builds potentially large disposable 
objects. Instead, I prefer having a SAX parse of the ocs file, building 
one entry at a time, and instantiating it in the registry.


> Tonight, I plan to commit the other package I'm working on (aggregation 
> engine)
> but I'm not very pleased with its current status and architecture.
> 

> Also, I'm flying to NY tomorrow and then make a short trip to LA before 
> going to
> ApacheCon next week, so while I should have complete net access during 
> these
> 2 weeks, I can't guarantee I'll be very active during this time... :(
> 

See you there!


---------------------------------------------------------------------
To unsubscribe, e-mail: jetspeed-dev-unsubscribe@jakarta.apache.org
For additional commands, e-mail: jetspeed-dev-help@jakarta.apache.org