Community dev

[Benson Margulies <bi...@gmail.com>]

Folks,

I just had a look at the commit history, and there are very few people
making very many commits. In fact, the #1 committer is rather far
ahead even of the #2 committer.

This is not a recipe for a successful escape from the incubator.
Writing code is wonderful and all that, but if you want to be a TLP, I
would advise you to put some effort into marketing and attracting more
participants.

--benson

Re: Community dev

[James Kosin <ja...@gmail.com>]

Hi Henry,

No, on the contrary... I just think it may be a while before we can get
contributors.  NLP takes time to absorb the real power of the technology.
We had one develop an HTML finder for names and addresses for a Rolodex
type application.  I'm in the process of developing a different kind of
document comparison application.  Others just use the library as is to
parse text and report on the tagged sentences for trying to determine
context and meaning of sentences.  The possibilities are endless.

I really don't think there are bad contributors; because anyone willing
to contribute is worth the effort they put into the project.  I just
don't think this is the kind of project anyone can jump in and start
contributing to in less than 8 hours.  True we need to beat the drums a bit.

James

On 5/31/2011 11:00 PM, Henry Saputra wrote:
> HI James,
>
> Are you worried about contributions that would not help the project to
> be better?
>
> I think as incubator project, it needs to attract more interest and
> good contributions to grow the committers and community of the
> project.
>
>
> - Henry
>
> On Tue, May 31, 2011 at 7:35 PM, James Kosin <ja...@gmail.com> wrote:
>> On 5/31/2011 1:22 PM, Benson Margulies wrote:
>>> Folks,
>>>
>>> I just had a look at the commit history, and there are very few people
>>> making very many commits. In fact, the #1 committer is rather far
>>> ahead even of the #2 committer.
>>>
>>> This is not a recipe for a successful escape from the incubator.
>>> Writing code is wonderful and all that, but if you want to be a TLP, I
>>> would advise you to put some effort into marketing and attracting more
>>> participants.
>>>
>>> --benson
>> Benson,
>>
>> The landscape is a bit daunting currently.  Hopefully it will change as
>> we add support for the open model architecture and allow easy
>> integration of different neural training networks into the landscape.
>> More than just a perception and maxent, not that there is anything wrong
>> with them.
>>
>> I currently, am working on my thesis; so, don't have too much time to
>> spend on the project at current.  I have been only a recent addition
>> with some of the items I've found in the architecture ... OpenNLP is
>> still growing as far as projects go and it may take us a while to catch
>> up.
>>
>> Jorn currently has been the large pusher of changes recently; however,
>> Jason is also a large contributor.  I'm sure the list is a lot bigger in
>> the large scale; however, the area of NLP is a bit new/old and requires
>> a lot of groundwork.  I think, most miss the mark what the real power of
>> NLP actually is other than finding text in a large document, or telling
>> what form of speech is being used in a paragraph.
>> Most of the libraries here are useful to many; however, the real power
>> comes when integrating with a real Human-computer interface using
>> natural language understanding.  In theory, with the right tools, the
>> computer will be able to read a book and understand the content and
>> extract ideas and thoughts from the document easily and be able to
>> understand them.
>>
>> I'm all for recruiting more help; but, the real question is are there
>> people willing to take on the challenge of the real landscape?
>>
>> James
>>
>
>

Re: Community dev

[Henry Saputra <he...@gmail.com>]

HI James,

Are you worried about contributions that would not help the project to
be better?

I think as incubator project, it needs to attract more interest and
good contributions to grow the committers and community of the
project.


- Henry

On Tue, May 31, 2011 at 7:35 PM, James Kosin <ja...@gmail.com> wrote:
> On 5/31/2011 1:22 PM, Benson Margulies wrote:
>> Folks,
>>
>> I just had a look at the commit history, and there are very few people
>> making very many commits. In fact, the #1 committer is rather far
>> ahead even of the #2 committer.
>>
>> This is not a recipe for a successful escape from the incubator.
>> Writing code is wonderful and all that, but if you want to be a TLP, I
>> would advise you to put some effort into marketing and attracting more
>> participants.
>>
>> --benson
> Benson,
>
> The landscape is a bit daunting currently.  Hopefully it will change as
> we add support for the open model architecture and allow easy
> integration of different neural training networks into the landscape.
> More than just a perception and maxent, not that there is anything wrong
> with them.
>
> I currently, am working on my thesis; so, don't have too much time to
> spend on the project at current.  I have been only a recent addition
> with some of the items I've found in the architecture ... OpenNLP is
> still growing as far as projects go and it may take us a while to catch
> up.
>
> Jorn currently has been the large pusher of changes recently; however,
> Jason is also a large contributor.  I'm sure the list is a lot bigger in
> the large scale; however, the area of NLP is a bit new/old and requires
> a lot of groundwork.  I think, most miss the mark what the real power of
> NLP actually is other than finding text in a large document, or telling
> what form of speech is being used in a paragraph.
> Most of the libraries here are useful to many; however, the real power
> comes when integrating with a real Human-computer interface using
> natural language understanding.  In theory, with the right tools, the
> computer will be able to read a book and understand the content and
> extract ideas and thoughts from the document easily and be able to
> understand them.
>
> I'm all for recruiting more help; but, the real question is are there
> people willing to take on the challenge of the real landscape?
>
> James
>



-- 
Thanks,
Henry

Re: Community dev

[James Kosin <ja...@gmail.com>]

On 5/31/2011 1:22 PM, Benson Margulies wrote:
> Folks,
>
> I just had a look at the commit history, and there are very few people
> making very many commits. In fact, the #1 committer is rather far
> ahead even of the #2 committer.
>
> This is not a recipe for a successful escape from the incubator.
> Writing code is wonderful and all that, but if you want to be a TLP, I
> would advise you to put some effort into marketing and attracting more
> participants.
>
> --benson
Benson,

The landscape is a bit daunting currently.  Hopefully it will change as
we add support for the open model architecture and allow easy
integration of different neural training networks into the landscape. 
More than just a perception and maxent, not that there is anything wrong
with them.

I currently, am working on my thesis; so, don't have too much time to
spend on the project at current.  I have been only a recent addition
with some of the items I've found in the architecture ... OpenNLP is
still growing as far as projects go and it may take us a while to catch
up. 

Jorn currently has been the large pusher of changes recently; however,
Jason is also a large contributor.  I'm sure the list is a lot bigger in
the large scale; however, the area of NLP is a bit new/old and requires
a lot of groundwork.  I think, most miss the mark what the real power of
NLP actually is other than finding text in a large document, or telling
what form of speech is being used in a paragraph.
Most of the libraries here are useful to many; however, the real power
comes when integrating with a real Human-computer interface using
natural language understanding.  In theory, with the right tools, the
computer will be able to read a book and understand the content and
extract ideas and thoughts from the document easily and be able to
understand them.

I'm all for recruiting more help; but, the real question is are there
people willing to take on the challenge of the real landscape?

James

Re: Community dev

[Jörn Kottmann <ko...@gmail.com>]

Anyway we should talk a little about what we can do to improve
the situation.

Our website is still missing a "Getting Involved" section, that should be
fixed, so that interested potential contributes can see how they
can contribute.

It would be nice to add an overview section to the website so potential 
interested users
can get a better feeling of what OpenNLP can provide, this page should
also include an online demo. I actually had a short discussion with that
off list with James.

Further we should create more jira issues to train OpenNLP on new 
languages and
corpora, and point out that we are looking for contributions. Actually 
that worked
quite well for the documentation where we received two patches. I often have
the feeling that people would like to contribute, but never really 
figure out what they
can do. Maybe it is a good idea to make a "Contributions Wanted" section on
the website somewhere.

I believe quite a few people are interested to use the name finder but 
are disappointed
by the performance of the models we provide, the only good fix I see for 
this is to
get a community labeling project started. Then these people might start 
to use OpenNLP
and some of them might start to contribute.

Any other suggestions?

Jörn


On 5/31/11 7:40 PM, Benson Margulies wrote:
> I'm just providing a distant early warning.
>
> On Tue, May 31, 2011 at 1:33 PM, Jörn Kottmann<ko...@gmail.com>  wrote:
>> On 5/31/11 7:22 PM, Benson Margulies wrote:
>>> This is not a recipe for a successful escape from the incubator.
>>> Writing code is wonderful and all that, but if you want to be a TLP, I
>>> would advise you to put some effort into marketing and attracting more
>>> participants.
>> I believe we are on a good way compared to the sourceforge days, we worked
>> a lot on the website, made it much easier with the documentation to use
>> OpenNLP and it is
>> easier now to work on the code itself,  sure we still have a long way to go,
>> but at least we are moving.
>>
>> Jörn
>>
>>

Re: Community dev

[Benson Margulies <bi...@gmail.com>]

I'm just providing a distant early warning.

On Tue, May 31, 2011 at 1:33 PM, Jörn Kottmann <ko...@gmail.com> wrote:
> On 5/31/11 7:22 PM, Benson Margulies wrote:
>>
>> This is not a recipe for a successful escape from the incubator.
>> Writing code is wonderful and all that, but if you want to be a TLP, I
>> would advise you to put some effort into marketing and attracting more
>> participants.
>
> I believe we are on a good way compared to the sourceforge days, we worked
> a lot on the website, made it much easier with the documentation to use
> OpenNLP and it is
> easier now to work on the code itself,  sure we still have a long way to go,
> but at least we are moving.
>
> Jörn
>
>

Re: Community dev

[Jörn Kottmann <ko...@gmail.com>]

On 5/31/11 7:22 PM, Benson Margulies wrote:
> This is not a recipe for a successful escape from the incubator.
> Writing code is wonderful and all that, but if you want to be a TLP, I
> would advise you to put some effort into marketing and attracting more
> participants.

I believe we are on a good way compared to the sourceforge days, we worked
a lot on the website, made it much easier with the documentation to use 
OpenNLP and it is
easier now to work on the code itself,  sure we still have a long way to go,
but at least we are moving.

Jörn

Re: Community dev

[Jason Baldridge <ja...@gmail.com>]

This is all good. I must admit that I have hesitated to contribute because I
far prefer working with Scala and start grating my teeth when I go back to
Java. Sorry about that, but it is just how it is for me.

On Wed, Jun 1, 2011 at 5:51 AM, Jörn Kottmann <ko...@gmail.com> wrote:

> On 6/1/11 4:30 AM, Jason Baldridge wrote:
>
>> I've also implemented a spell checker before, would be happy to help with
>> tips and pointers.
>>
>
> Maybe we can work this into a jira issue, and then advertise it that we are
> looking
> for a spellchecker contribution.
>
> I think there are quite a few things where it would be really nice to get
> contributions
> fore, e.g. lemma detection,  support for more languages,
> dependency parser, stemmers, CRF, topic modeling,  etc.
>
> What do you think?
>
> Jörn
>



-- 
Jason Baldridge
Assistant Professor, Department of Linguistics
The University of Texas at Austin
http://www.jasonbaldridge.com
http://twitter.com/jasonbaldridge

Re: Community dev

[Jörn Kottmann <ko...@gmail.com>]

On 6/1/11 4:30 AM, Jason Baldridge wrote:
> I've also implemented a spell checker before, would be happy to help with
> tips and pointers.

Maybe we can work this into a jira issue, and then advertise it that we 
are looking
for a spellchecker contribution.

I think there are quite a few things where it would be really nice to 
get contributions
fore, e.g. lemma detection,  support for more languages,
dependency parser, stemmers, CRF, topic modeling,  etc.

What do you think?

Jörn

Re: Community dev

[Jason Baldridge <ja...@gmail.com>]

I've also implemented a spell checker before, would be happy to help with
tips and pointers.

I'm now encouraging my students to use/modify OpenNLP in their work, so hope
to stir some more developers up that way.

-Jason

On Tue, May 31, 2011 at 6:28 PM, Henry Saputra <he...@gmail.com>wrote:

> Hi Jörn,
>
> Not the one I have worked on, its copyrighted by my prev employer =(
>
> Still poking around the code but looking forward to add contribution
> to the project.
>
> -  Henry
>
> On Tue, May 31, 2011 at 4:07 PM, Jörn Kottmann <ko...@gmail.com> wrote:
> > On 6/1/11 12:52 AM, Henry Saputra wrote:
> >>
> >> Hi OpenNLP devs,
> >>
> >> I am really interested in this project. Having had to build a decision
> >> tree and spellchecker from scratch, having a project like this is
> >> awesome.
> >> Still playing around with it so I havent had chance to add patches or
> >> new features.
> >>
> >
> > Just let us know if you would like to work on something, so we can
> support
> > you.
> >
> > Actually we do not have a spellchecker yet,
> > would you be interested to contribute yours?
> >
> >> If I may add some comments, I think the project needs a quick demo
> >> project to showcase how the existing features work.
> >>
> > +1, I believe we should make a webapp to demonstrate our components
> combined
> > with an overview section to explain what OpenNLP can do.
> >
> > Jörn
> >
> >
>
>
>
> --
> Thanks,
> Henry
>



-- 
Jason Baldridge
Assistant Professor, Department of Linguistics
The University of Texas at Austin
http://www.jasonbaldridge.com
http://twitter.com/jasonbaldridge

Re: Community dev

[Henry Saputra <he...@gmail.com>]

Hi Jörn,

Not the one I have worked on, its copyrighted by my prev employer =(

Still poking around the code but looking forward to add contribution
to the project.

-  Henry

On Tue, May 31, 2011 at 4:07 PM, Jörn Kottmann <ko...@gmail.com> wrote:
> On 6/1/11 12:52 AM, Henry Saputra wrote:
>>
>> Hi OpenNLP devs,
>>
>> I am really interested in this project. Having had to build a decision
>> tree and spellchecker from scratch, having a project like this is
>> awesome.
>> Still playing around with it so I havent had chance to add patches or
>> new features.
>>
>
> Just let us know if you would like to work on something, so we can support
> you.
>
> Actually we do not have a spellchecker yet,
> would you be interested to contribute yours?
>
>> If I may add some comments, I think the project needs a quick demo
>> project to showcase how the existing features work.
>>
> +1, I believe we should make a webapp to demonstrate our components combined
> with an overview section to explain what OpenNLP can do.
>
> Jörn
>
>



-- 
Thanks,
Henry

Re: Community dev

[Jörn Kottmann <ko...@gmail.com>]

On 6/1/11 12:52 AM, Henry Saputra wrote:
> Hi OpenNLP devs,
>
> I am really interested in this project. Having had to build a decision
> tree and spellchecker from scratch, having a project like this is
> awesome.
> Still playing around with it so I havent had chance to add patches or
> new features.
>

Just let us know if you would like to work on something, so we can 
support you.

Actually we do not have a spellchecker yet,
would you be interested to contribute yours?

> If I may add some comments, I think the project needs a quick demo
> project to showcase how the existing features work.
>
+1, I believe we should make a webapp to demonstrate our components combined
with an overview section to explain what OpenNLP can do.

Jörn

Re: Community dev

[Henry Saputra <he...@gmail.com>]

Hi OpenNLP devs,

I am really interested in this project. Having had to build a decision
tree and spellchecker from scratch, having a project like this is
awesome.
Still playing around with it so I havent had chance to add patches or
new features.

If I may add some comments, I think the project needs a quick demo
project to showcase how the existing features work.

Just my 2-cents. Looking forward to help for this project to be top level ASF.

- Henry

On Tue, May 31, 2011 at 10:22 AM, Benson Margulies
<bi...@gmail.com> wrote:
> Folks,
>
> I just had a look at the commit history, and there are very few people
> making very many commits. In fact, the #1 committer is rather far
> ahead even of the #2 committer.
>
> This is not a recipe for a successful escape from the incubator.
> Writing code is wonderful and all that, but if you want to be a TLP, I
> would advise you to put some effort into marketing and attracting more
> participants.
>
> --benson
>



-- 
Thanks,
Henry

Re: Community dev

[Benson Margulies <bi...@gmail.com>]

That's pretty much the same as the numbers I got from svnsearch that
started me thinking about all of this.

On Tue, May 31, 2011 at 2:08 PM, Jörn Kottmann <ko...@gmail.com> wrote:
> On 5/31/11 7:22 PM, Benson Margulies wrote:
>>
>> I just had a look at the commit history, and there are very few people
>> making very many commits. In fact, the #1 committer is rather far
>> ahead even of the #2 committer.
>
> Here are ohloh statistics about that:
> http://www.ohloh.net/p/apache-opennlp/contributors
>
> The numbers only reflect the commits of the last 6 month,
> there are also statistics for the time before at sourceforge:
> http://www.ohloh.net/p/opennlp-sourceforge/contributors
>
> Jörn
>

Re: Community dev

[Jörn Kottmann <ko...@gmail.com>]

On 5/31/11 7:22 PM, Benson Margulies wrote:
> I just had a look at the commit history, and there are very few people
> making very many commits. In fact, the #1 committer is rather far
> ahead even of the #2 committer.

Here are ohloh statistics about that:
http://www.ohloh.net/p/apache-opennlp/contributors

The numbers only reflect the commits of the last 6 month,
there are also statistics for the time before at sourceforge:
http://www.ohloh.net/p/opennlp-sourceforge/contributors

Jörn