Design decisions for new Protocol: Adium vs Libpurple

[Evan Schoenberg, M.D.]

On Jun 26, 2010, at 9:04 PM, Peter Hosey wrote:

> On Jun 26, 2010, at 13:35:34, David Smith wrote:
>> … the upside is just the additional front ends …
> 
> Which is a mixed upside, since you then have to test your plug-in on those additional front ends, on any platforms you support.
> 
> I'd say don't worry about Pidgin/Finch support (or, conversely, Adium support) until you have an additional developer willing to test those platforms for you.
> 
> So it comes down to whether you'll ever want to support Pidgin or Finch. If not, you might as well write it directly as an Adium plug-in. If you might, you can write it as a Libpurple plug-in, and only release Adium builds until you have another developer to test Pidgin/Finch.
> 
> Of course, if you use Pidgin or Finch as well as Adium, you can disregard all the foregoing and just go for it. You'll still be testing for two, but you'll have self-interest motivating you.
> 
> 

On the other hand, supporting the additional platforms also grants the possibility of having a wider userbase; in the case of supporting Pidgin on Linux, that allows the possibility of a userbase that may be more likely to contribute back to your project.

-Evan
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://adium.im/pipermail/devel_adium.im/attachments/20100628/05faadd7/attachment-0002.html>