GSoc 2009

[M Breugelmans]

On 1/18/09, David Nolden <zwabel at googlemail.com> wrote:
> Am Sunday 18 January 2009 12:33:58 schrieb Milian Wolff:
>> - PHP (basic support is there)
>> - JavaScript
>> - CSS
>> - HTML
>> - Python (there is a plugin, not sure of the state)
>> - Ruby (there is a plugin, not sure of the state)
>> - Java
>>
>> This should increase the userbase of KDevelop considerably. And that otoh
>> should get you more developers, right? ;-)
>>
>> And the first four languages are a requirement for Quanta.
>
> The problem is that implementing a good language-support takes a lot more
> effort than what can be done by a student during a SOC project. We already
> had several language-support GSOC projects, and all of them failed, because
> the student didn't finish what he started. Probably it just isn't satisfying
> enough if you don't have a _real_ will to finish the thing.
>
> Also we want to initially only really support C++ in the main application,
> so
> we should focus on projects that make KDevelop better within that context.
>
> Examples:
> - Working-sets
> - Distribution integration
> - Making the debugger rock
> - Documentation integration
>

+1 for a kick-ass debugger.

more ideas:
- Superb set of C++ refactorings
- Smart static code inspections (~Krazy++)

Both are realistic and sure to deliver at least some useful and
finished stuff. In my book these kind of features make the difference
between a good and an awesome IDE. Also this would increase DUChain's
bus factor, which is good.

IntelliJ's documentation could serve as a source of inspiration:
http://www.jetbrains.com/idea/features/refactoring.html
http://www.jetbrains.com/idea/features/code_inspection.html


regards,
Manuel