Continue our own Twitter interface library?

[Dave Briccetti]

Considering that the Twitter API grows and changes (search, lists,
cursors, etc.) and that we TalkingPuffin developers have limited time,
perhaps we should discuss the pros and cons of continuing to enhance
and maintain our own Scala Twitter interface library. How good are the
Java libraries? Are we duplicating effort working on a Scala version
of libraries written in Java? In the beginning, having our Twitter
access code be pure Scala appealed to me, but it seems less important
now.

[Mikael Sundberg]

I think this depends on your goal with the projekt.

If your goal is to write a GUI in scala, then its no point in having a twitterAPI in scala.
if your goal is to write a full program in scala, then you should keep the api.
in the end it boils down to what you enjoy most, since its just to face the fact, your the one doind most of the coding. And since you ask the question, that shows abit that you prefer the gui part over the api part :-)
Yes there are some pretty nice twitterAPI's in java, havent tried them myself but i have heard good things.

I think the API is the most fun part. GUI's are best made with a gui-builder that scala lacks. but thats me and I have way to little time to do anything i want nowdays :-(


/Micke

2009/10/5 Dave Briccetti <da...@davebsoft.com>

[Dave Briccetti]

Since you wrote that, Mikael, I have renewed my interest in our
Twitter API library. In fact I’ve added some new features, including
partial support for Twitter lists. I hope Mark McBride remains a
contributor, but even if he can’t do much I think I’ll keep it part of
TalkingPuffin for the time being.

The most fun for me has been in trying to parallelize batch operations
against the Twitter Web services, such as for finding people with
lists, and adding people to lists. (See recent commits if you’re
curious about that.) Serialized, it took several minutes to export
tagged users to a Twitter list. Now the same thing takes just several
seconds.

Dave