Hi all, I have posted an article on SquekPeople: I have read the "rant" and "unrant" subject with a mix of uneasiness (embarrassment) and fear. The "Rant" thread began on the 15th of december 2004, and produced a lot of discussion. I will not summarize the thread, because a lot of people read it. Instead, I will go forward and try to find a pragmatic and constructive of helping the community (and yes, I am also very bold and handosome :) << ironic comment).
See the rest on http://people.squeakfoundation.org/article/41.html ....and happy new year :)
Quoting Giovanni Giorgi giovanni.giorgi@siforge.org:
Hi all, I have posted an article on SquekPeople: I have read the "rant" and "unrant" subject with a mix of uneasiness (embarrassment) and fear. The "Rant" thread began on the 15th of december 2004, and produced a lot of discussion. I will not summarize the thread, because a lot of people read it. Instead, I will go forward and try to find a pragmatic and constructive of helping the community (and yes, I am also very bold and handosome :) << ironic comment).
See the rest on http://people.squeakfoundation.org/article/41.html ....and happy new year :)
-- Giovanni Giorgi http://www.squeak.org http://www.siforge.org .............................................................
Thank you for your posting and beeing prepared to volunteer to advance things. I perceive it as an honour that the GNU smalltalk representative is willing to do more for Squeak than he already does ;-)
May I ask you two questions of clarification and add a note.
How do you perceive the problem you want contribute to solve? To put it more bluntly - I think we need a more concise problem discription.
You are mentioning that additional SUnit tests are needed. I agree with this. We have about 1400 and still could need another 4000 ..8000 probably. (Eclipse BTW has about 20'000 AFAIK). This means again we need to do set priorities. Where do you see these priorities? And it is in fact a hard job to write tests after the code has been written as Stephane D, Roel Wuyts and confirmed in earlier threads in this list.
As a note I would like to propose that a useful thing in any case is to look for another subsystem for which it is relatively easy to write a removal and adding script. This should be followed by the proposal to remove the said subsystem from 3.8g before releasing. For the full image of course it should be added again. Even if the proposal is not actually carried out at least we have the benefit that poeple who do not want that subsystem can easily remove it from their base image before loading other packages. Removing another subsystem would follow the recent request by Andreas Raab that the base image shouldn't grow anymore in number of classes.
This is actually the same thing as has you are already doing with Celeste and thank you for maintaining it. BTW what is it's status for 3.8g?
To summarize: More quality means we need ways to assure quality and this currently means more tests. But which ones is the question. There is probably no easy solution.
Hannes
Giovanni,
after thinking again about my previous post. I think I can summarize it with one short question:
Are you voluteering to develop the so-called 'minimal image' as it has been discussed in this list during this year?
(Note: We have full image <------------- built at release time by adding 'official packages' to the base image. base image <------------- the current Squeak 3.8g image) minimal image <-------- not yet done; equals base image minus certain things .... (to be defined).
Hannes
Hi--
Hannes asks Giovanni:
Are you volunteering to develop the so-called 'minimal image' as it has been discussed in this list during this year?
Lest people think there's no one currently working on this, note that I am, as part of Spoon[*]. Comments and suggestions are welcome, as always.
thanks,
-C
-- Craig Latta improvisational musical informaticist craig@netjam.org www.netjam.org [|] Proceed for Truth!
On 31/12/04 02:06, "Craig Latta" craig@netjam.org wrote:
Lest people think there's no one currently working on this, note that I am, as part of Spoon[*]. Comments and suggestions are welcome, as always.
thanks,
Craig:
You has done a terrific job with squat/spoon. We are waiting one article on how Spoon could take Squeak code.
Edgar
Hi, first of all excuse me for forgetting to cite "Spoon" on my article. I am subscribed to the mailing list, and I have tried it a lot of time ago. The Spoon approach is very interesting because it is very original and can compete with the so called "network language" (read: .NET and Java). We can assume Craig's Spoon a very smart way to achieve the minimal image.
Answering to Hannes Hirzel, my primary goal is the opposite: to mantain the first "Rock Solid" image, to keep the fears about instability to the minimum, and to have a measure of the abandoned code/packages/methods in the Squeak Major subsystems.
By the way, after some discussion on my article, I will send an email to the guide to give my help to the Rock Solid project, adding more details to this idea.
....Happy new year :)
On 31/dic/04, at 06:06, Craig Latta wrote:
Hi--
Hannes asks Giovanni:
Are you volunteering to develop the so-called 'minimal image' as it has been discussed in this list during this year?
Lest people think there's no one currently working on this, note that I am, as part of Spoon[*]. Comments and suggestions are welcome, as always.
squeak-dev@lists.squeakfoundation.org