- Pharo-dev - lists.pharo.org

Re: [Pharo-project] about creating a task for the next milestone
by Bill Schwab July 12, 2008

July 12, 2008

Stef, I am very interested in serializers, but I (right or wrong) do not have good feeling about ReferenceStream and friends. When I last tested them (a while back), the output was BIG, and I was not comfortable about the location[*] of the conversion methods, nor with the timing: as usual, if I goof, please raise an error, don't put browsers in front of my end users =:0 That said, I am expecting to port some of my code, and completely rewrite other parts using things like Magritte. A variation on ReferenceStream might turn out to be of use to me. However, I am also considering something, again along the lines of Magritte (if not part or an extension thereof) to make my persistence independent of the internal representation of objects. With extensive unit testing, I think I could make that work, and should at least investigate it. Dolphin's STB and in particular its handling of layout changes is pretty good. I also found some things to like about SIXX and might try to blend the two. However clunky, I want to get my BibTeX system up and running, and then I will probably turn attention to an out-filer that writes code that Pharo can read. Some gnashing of teeth later, I will start to port sub-systems, redesigning some as I go. Bill [*] IIRC, they end up in ReferenceStream itself, rather than in the class that "needs help." It struck me as asking for trouble. Wilhelm K. Schwab, Ph.D. University of Florida Department of Anesthesiology PO Box 100254 Gainesville, FL 32610-0254 Email: bschwab(a)anest.ufl.edu Tel: (352) 846-1285 FAX: (352) 392-7029 >>> stephane.ducasse(a)inria.fr 07/11/08 3:34 AM >>> Hello guys I was reading the class ReferenceStream and sisters/mother. The ones that are in System-Object Storage. I think that these classes are important. They have a good class comments, even if we could improve. I would really like to use them as a test-bed for a canonical documentation effort: - unit tests - simple how to use I would also like to see if they can be rebuild on top of Nile. So I would like to know if some of you are interested working on this task. stef _______________________________________________ Pharo-project mailing list Pharo-project(a)lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project

3 2

[Pharo-project] connexion sluggisness
by StÃ©phane Ducasse July 12, 2008

July 12, 2008

hi guys I just arrived to a place where the net does not work.... Yes it exists. :) So I'm trying to discuss with the provider arghhhh. So do not expect me to be reactive.... Right now I'm in public library which has a nice open wifi. Stef

1 0

[Pharo-project] about creating a task for the next milestone
by StÃ©phane Ducasse July 11, 2008

July 11, 2008

Hello guys I was reading the class ReferenceStream and sisters/mother. The ones that are in System-Object Storage. I think that these classes are important. They have a good class comments, even if we could improve. I would really like to use them as a test-bed for a canonical documentation effort: - unit tests - simple how to use I would also like to see if they can be rebuild on top of Nile. So I would like to know if some of you are interested working on this task. stef

1 0

[Pharo-project] testSubstrings fails in 10049
by Adrian Lienhard July 9, 2008

July 9, 2008

It seems that #testSubstrings was introduced by http://code.google.com/p/pharo/issues/detail?id=26 . This test fails in update 10049. Why was no fix applied? Adrian

3 3

[Pharo-project] Issue tracker shows closed issues when filtering for "Issues to verify"
by Adrian Lienhard July 9, 2008

July 9, 2008

The expected behavior would be to show only issues that are in state Fixed. Does somebody know how to change this? Adrian

2 1

[Pharo-project] Status of issue #1: Create a class Author out of Utilities?
by Adrian Lienhard July 9, 2008

July 9, 2008

Hi Damien You created http://code.google.com/p/pharo/issues/detail?id=1. Can you explain what you mean with this issue? It is in state fixed. Is there code pending that needs to be harvested? If yes, please indicated the slice or package. If no, the state should be updated. Thanks, Adrian

2 2

[Pharo-project] Keeping track of external issues
by Norbert Hartl July 9, 2008

July 9, 2008

I played a little with the squeak and the mantis tickets. First, I decided to use the usual way to indicate a mantis bug number: M1234. This way it is easily distinguishable from other version numbers. Then I changed the title of the mantis tickets to follow the number convention. M6355: [ENH] Lots of tests for Collections I started to change the squeak issues to include the mantis numbers in the title. It is in () at the end of the title. E.g. 7073 CollectionAndCollectionTest (M3574 M5700 M6367 M6777) So, if you are looking for a mantis number just search for Mantis and Squeak Tickets: http://code.google.com/p/pharo/issues/list?can=1&q=label%3AType-Mantis +OR+label%3AType-Squeak You get a page with all mantis numbers on it. I think this way it is not too hard to keep track. If you have a shell und you love nasty one liners you can do: wget -O - "http://code.google.com/p/pharo/issues/list?q=label:Type-Mantis%20OR% 20label:Type-Squeak&can=1" | perl -ne 'while(m/(M[0-9]+)/g) { print "$1 \n"; }' | sort which outputs: M2446 M3568 M3574 M4654 M5251 M5392 M5567 M5619 M5641 M5700 M5704 M6070 M6355 M6367 M6402 M6405 M6520 M6702 M6711 M6777 M7090 M7107 hope you like it, Norbert

3 6

July 8, 2008

marcus math could'nt you fix the compiler to read := _ and generate := ? Stef

2 4

July 8, 2008

>>> Couldn't you do this on Squeak? >> > > So, couldn't you continue on Squeak? Because - 1 we need freedom. We do not want to maintain: etoy, nebraska, alice and the like. We do not want to have other people putting stress and their dreams on us. We did that for 3.9 and people pissed on us (random refactorers...) complained that we pushed traits in Squeak..... read the archives of the mailing-list everything is there. - 2 people said that we did not have a process, that social engineering was the solution (which is true when you apply it). - 3 we proposed a 3.10 roadmap and people complained. - 4 we proposed our help to 3.10 but it was declined. - 6 after the sad adventure of 3.10 and the lack of reaction of people we decided that we did our best. in 3.9 we harvested 625 bugs fixed but apparently this was not valuable. We think that 3.9 was a positive experience even if marcus got sick and burnt out. - 7 The squeak model is flawed. We do not want a model were consumers can put exigences on producers. Even if good customers are important and this is why we want to have Aida/Seaside and a lot of cool projects to run on pharo. - 8 I'm sick of opening squeak and having to turn my eyes from problems. After a certain point we should learn and take decision even if they are not easy. > I can't understand why you forked the project instead of improving it. Because we do not want to have complain all the times. We want to be able to invent the future, rethink design decision, apply good practices but not maintain the cool but old features and experiments of others. Sorry but we did that and now this is over. We are also sorry but we do not believe in big bang changes and we waited too much after spoon. So may be we are wrong, probably but again we had to do something for ourselves. >> I do not know. >> FunSqueak ;D > > So you quit Squeak... Not really. We will work on a new Volume Two and we did not renamed yet our book Pharo by Example :) May be pharo will be a cool squeak distribution or something else. For now our focus is getting things done. > I still can't see a valid reason for this fork. There are plenty of them but may be you are not in the right places to understand them. Here is an extra one. I have no choice this is pharo or ruby and I'm not joking. I was terribly sad when I see people willing to go back to 3.8! We spent a lot of energy to make sure that the forks could reunite but now this is over. We tried to really make a better Squeak and got sick, burnt and more. So now if I do not want to destroy the little flame we have for Smalltalk we **must** do pharo, because this is like new dry wood that you put on the fire place. It brings exciting future. Stef

3 5

[Pharo-project] pharo website proposal
by Samuel MORELLO July 8, 2008

July 8, 2008

Hi all, Please find a proposal for the pharo website. The example text is from the iolanguage website because i like this site and the way they present the language and the content. (structure and clarity) I know that the stained texture can seems strange, may be at first, but this allow to bring colors and a nice texture like water or stars with imagniation ;-) http://www.ouvanous.com/images/pharo/pharo-ws-design-0.1.png and a new version of the logo http://www.ouvanous.com/images/pharo/ppharo-0.8.0.pdf Hope you will like it Regards, Samuel

8 21