[squeak-dev] Re: Summer(s) of Code 2007/2008

Giovanni Corriga giovanni at corriga.net
Fri Mar 7 22:08:26 UTC 2008


Klaus D. Witzel ha scritto:
> On Fri, 07 Mar 2008 02:22:07 +0100, Giovanni Corriga wrote:
> 
>> Andreas Raab ha scritto:
>>> Klaus D. Witzel wrote:
>>>> I bet one (1.-) CHF that GTK support will not be included in the 
>>>> next Squeak release, not counting 3.10, not talking about an 
>>>> optional package.
>>>  Why does everyone seem so exclusively focused on doing something 
>>> that ends up "in the next Squeak release"? Is this a requirement for 
>>> GSOC? I certainly don't see the HydraTools ever getting into a Squeak 
>>> release but that doesn't make them any less useful for Squeak.
>>
>> As I said in my reply to Klaus, there's no need for the projects to be 
>> geared toward inclusion in the next Squeak release. It's just that I'd 
>> like for our students to be able to deliver something at the end of 
>> the Summer.
> 
> This *is* a good point and I would really like to be in favor of such 
> policy. But I cannot since reality dictates something else. Let me 
> mention the Delta Stream project as an example, its sheer complexity 
> makes such an enterprise look like it has several ends [pun not 
> intended] but it is nevertheless possible for a summer coder to make a 
> significant contribution to DS without also making DS usable for the 
> masses on the next day, no?
> 
> Same goes for Cedrick's project suggestion, please don't expect that he 
> will create *the* AI by that; all he wants is a port to Squeak Seaside 
> for multi users (thereby enabling any form of persistency that users can 
> afford BTW), instead of the current plain Java desktop. Thousands of 
> other people will be able to do the "remainder" with Cedrick's base 
> work, perhaps in the next SoC? This kind of policy *must* be possible, 
> unless we want only gray-bearded experts to apply to 
> autumn-of-life-long-coding projects.
> 
> Think small, keep it open (= always unfinshed *and* always [re-]usable, 
> as in Smalltalk) and do great things?

That's a nice idea, generally speaking, but it doesn't really agree with 
what Google expects. To use your example, "finish DeltaStreams" isn't a 
good project, since the scope is too wide and vague; "adding this four 
kinds of deltas" to the system is much better in this regard.

	Giovanni




More information about the Squeak-dev mailing list