Hi all,

I am very excited to see the many pull requests the project has
received lately. This is more activity than the project has seen in
quite some time. Thank you to all - the contributions and overall
activity in the project are both welcome and very good to see.

Since some newer contributors might not be familiar with certain
aspects of OpenNLP, I wanted to provide a few key points about the
project's history. I don't think these points are much different than
many other projects but they have had a special focus in OpenNLP.

The first point is that OpenNLP has traditionally avoided bringing in
external dependencies. It keeps size down, lowers threat surface, and
makes OpenNLP being a dependency in other projects easier (e.g.
Lucene/Solr). An example is if you need a function in commons-lang,
it's more favorable to provide an implementation instead of bringing
in the whole dependency just for one function. There have been
exceptions, but we do try to follow this in general.

A second point is over time the project has been driven by need.
Functionality has been added based on exhibited use-cases. This is a
low bar, but it helps the project focus on capabilities that are used.
OpenNLP has never had the large contributor base that some other
Apache projects have, so we must be careful about adding new code, as
it becomes more code that the same small group must maintain over
time. A short conversation on dev@ will help refine it and get things
moving.

The last point is the documentation. As features are added, let's put
a special emphasis on documenting the code added with example
use-cases of when that feature can/should be used. It helps the devs,
end-users, and AI assistants understand OpenNLP's inner workings. And
while writing this I recognize these points need to be in the
developer documentation, too!

OpenNLP has been around since 2009 (or so?) and I think these three
points are part of the reason why it is still a valued NLP framework
today and I don't want to lose sight of them.

Again, I am very excited to see the contributions and extend a warm
welcome to each new contributor.

Thanks,
Jeff

Reply via email to