Squeak-dev May 2008

squeak-dev@lists.squeakfoundation.org

141 participants
174 discussions

Bogus Class Category in 3.10.1-7175 - Mantis #7070
by Ken Causey 29 May '08

29 May '08

Note that in the released 3.10.1-7175 you will find a category called #Morphic-CandidatesForGo. Do NOT click any of the classes listed in that category as they are not really there but removed in 3.10-7159. The result is an annoying, seemingly unstoppable, series of debug prompts. I made a mistake when building 3.10.1-7175, you can look at http://bugs.squeak.org/view.php?id=7070 For more of the gorey details. My point with this email, other than to apologize, is to ask what I should do. First note that this issue is related to my pre-release cleanup and is not due to any changeset in the update stream. Take a 3.10-7159 and update it to 3.10.1-7175 and you will not have this problem. The first solution might be to simply build a replacement 3.10.1-7175-basic.zip without the problem and replace the copy on the FTP site. However I don't think that would be a good idea. Having spent a lot of time over the years helping users with issues I know how confusing it can be to try to replicate an issue and one of the keys is being able to start with the same image. Imagine a user has downloaded my faulty release and is experiencing this problem and is confused. Imagine also a helpful person who wants to assist them with this problem. The helper finds out they are using 3.10.1-7175 and downloads it from the FTP site. Unfortunately this helper has not heard of this issue and tries the newer 3.10.1-7175 image and is unable to replicate the problem. Substitute appropriately modified "Who's on First" variant; hilarity ensues at least if you are not a participant. So, I started a Known Issues list at http://ftp.squeak.org/3.10/KNOWNISSUES that prominently features this issue. I would like to provide a workaround of some kind there and I would appreciate any insights into a code snippet to address this issue in 3.10.1-7175 that I can provide as a workaround. Otherwise the only other solution I can think of is to yet again start on a quick fix release and call it 3.10.2. Thoughts? I apologize for this mistake, I can only hope (unrealistically) that was my only one. Clearly having someone jump into a release branch as I have done here, without the proper knowledge, has consequences. Ken

3 5

Re: [V3dot10] Patch should be applied: SystemDictionary>>recreateSpecialObjectsArray
by Keith Hodges 29 May '08

29 May '08

Torsten Bergmann wrote: > The latest Squeak3.10.1-7175-basic.zip image from http://ftp.squeak.org/current_stable/ > also left out the patch to make FFI usable in 3.10: > > http://bugs.squeak.org/view.php?id=6980 > > Thx > Torsten > I added the above patch to http://installer.pbwiki.com/MinorFixesUnstable-Squeak3%3A10 All of the listed patches can be loaded using Installer install: 'LatestUnstable'. We still need to establish a "gate" process for moving patches from Unstable to Stable. Those patches listed on the Stable page would be the ones that are loaded by: Installer install: 'Latest'. These are considered to be the fixes that will be included in 3.11 We could have a "vote for approval" approach, a "vote for disapproval" approach or a "vote for suspicion" approach. Therefore I invite readers to review the list of fixes and at first glance just email me with any that they feel suspicious about, or have reason to dissapprove. In two weeks I will move the rest to "Stable". regards Keith

2 1

Anyone attending RailsConf2008
by Jason Rogers 29 May '08

29 May '08

I'm here in Portland attending RailsConf 2008 and wouldn't mind catching up with some Squeakers. -- Jason Rogers "I am crucified with Christ: nevertheless I live; yet not I, but Christ liveth in me: and the life which I now live in the flesh I live by the faith of the Son of God, who loved me, and gave himself for me." Galatians 2:20

2 2

Funding (was: What Constitutes a Complete and Final Release?)
by Stephen Pair 29 May '08

29 May '08

On Sun, Apr 6, 2008 at 4:11 AM, Hilaire Fernandes <hilaire(a)ofset.org> wrote: > tim Rowledge a écrit : > > > > On 4-Apr-08, at 8:49 PM, Klaus D. Witzel wrote: > > > > > > It seems to me that the board has a problem here, with both 3.9 and > > > 3.10 :( > > > > Well, no. It's not the board that has any sort of problem; the board is > > a bunch of people that meet every couple of weeks to discuss an agenda that > > is primarily organisation based (ie getting to a state where we can join the > > SFLC foudation thingy). The board is not a team that can be asked or > > expected to solve all *our* problems. *You lot* get to do the actual work. > > Until and unless you provide funds to pay a team.... > > I found this statement pretty irresponsible :( > And reading this statement one may think what is the Sqf good for (beside > the never ending registration in SFLC...) and outsiders may even fell that > Squeak community is not a good place to go (no clear roadmap). > > Hilaire > I think you're overstating things a bit. The point is that unless you are willing to jump in and help, or provide other means of help (i.e. funding), one shouldn't complain about the state of affairs. The people doing this work are doing it as a free service to others and for that, we should be grateful. At the same time, maybe we should be thinking about alternative means of funding. The way I see it, funding is a means of enabling more people to spend more of their time working on squeak (presumably something they enjoy). I think the bounty model doesn't work all that well, and I think being dependent on benevolent commercial ventures to develop and donate substantial code is also sub-optimal. With commercial ventures, their attention is always going to be focused on their product. Innovating and developing the core squeak system will consequently only get their attention when it is necessary to further the commercial product. And then, they will only develop it as far as it needs to go to serve their interests and no further. I find it a real shame that the economy around core platforms and tools seems to have evaporated with the open source movement. It seems that there is precious little funding available to those that want to develop new languages, new operating systems, new development tools and the like. There is a modest amount of innovation going on in spite of this reality, but progress is frustratingly slow. Here's an idea for a funding model. Instead of always releasing new stuff immediately under MIT, how about releasing some stuff under a time limited proprietary license with a specific date at which it reverts to MIT (this would be baked into the original license such that the "freedom date" is irrevocable). The date could be 3 months, 6 months or even a year or two into the future. In the interim, anyone wishing to use that software would need to pay some fee to the author(s). This is the essentially the same idea that is behind copyright laws, but on a dramatically shorter timeframe that is more appropriate for software (the life of copy rights in the US extends for life+70 years I believe, which certainly doesn't work for software). If we could make something like this work successfully (and I'd say that success means that more people are able to spend more of their time working on squeak (instead of other endeavors to pay the bills)), I think people would quickly take notice. Imagine if this came to be an expectation of all software licenses. Imagine if all commercial software were sold like this. To me, this is more in line with basic economics. The cost to produce software is substantial, but the marginal cost to distribute software is effectively zero. So, over time the price of software (or any intellectual property) must approach zero. It's unnatural for software to be locked up under proprietary licenses for what is effectively an indefinite period of time. It's also unnatural in my opinion to rely solely on indirect means of funding (relying on people deriving inherent fun in the work or commercial benevolence or services associated with the free software, etc). Musicians will deal with this by deriving money from live performances or associated merchandising and hence be able to offer recordings for free right from the start. I doubt too many people will pay to watch many of us as we write code, so we have to think of alternative means. Maybe if enough people were willing to pay $10 for a high quality 3.10, it would be enough to entice a few people into spending a larger chunk of their time working on that problem. The compromise we make is that for the first 6 months (or whatever the author decides), it's not free software...you must pay $10. The small size of the squeak community won't make people rich in so doing, but over time, a better squeak means a larger community and a larger community means that more people will have an opportunity to participate as a supplier in such an economy. - Stephen

36 78

Algernon issues with sq3.10-7159dev08.05.2
by Torsten Bergmann 29 May '08

29 May '08

Hi Igor, I already reported this offlist. See comments from the developers below. Bye Torsten -------- Original-Nachricht -------- Datum: Tue, 20 May 2008 10:09:41 -0600 Von: Erik Hinterbichler <xxx(a)xxx.xxx> An: Damien Cassou <xxx.xxx(a)xxx.xxx> CC: "Torsten Bergmann" <xxx(a)xxx.xxx>, xxx(a)xxx.xxx, xxx(a)xxx.xxx Betreff: Re: Algernon in Dev-Images It looks like this happens because refreshItems is run every time a new class is added, and so it's being run many times when you load a new package. We did that to make Algernon automatically stay up to date when new items were added, but if it slows down loading of packages we may need to remove that feature. Either that or we would need to figure out a way to prevent refreshItems from being run until the very end of a package being loaded. We'll experiment and try to have a new version out before the new image is created. Torsten, in the meantime if Algernon is slowing you down you can remove it by bringing it up with shift-return and then bringing up a halo on it and deleting it. Thanks, Erik On May 20, 2008, at 9:23 AM, Damien Cassou wrote: > Hi Torsten, > > thank you for the information. I will generate a new image in 10 days > approximately. If I have no answer from the developers, I will remove > it. However, I would really appreciate if this could be fixed as I > think Algernon can really help developers. > > > Bye > > On Tue, May 20, 2008 at 8:36 PM, Torsten Bergmann <xxx(a)xxx.xx> > wrote: >> Hi Damien, >> >> sq3.10-7159dev08.05.2.zip from https://gforge.inria.fr/frs/?group_id=1299 >> now also includes "Algernon" but this extremely slows down loading >> of other packages. >> >> Try loading KomHttpServer 7.0.4 using the universe browser >> in this image. A lot of time is spent in Algernon>>refreshItems >> during the load. >> >> Maybe someone fixes this or you should make this an optional package. >> >> Thx >> Torsten >> > > > > -- > Damien Cassou > Peter von der Ahé: «I'm beginning to see why Gilad wished us good > luck». (http://blogs.sun.com/ahe/entry/override_snafu) -- Ist Ihr Browser Vista-kompatibel? Jetzt die neuesten Browser-Versionen downloaden: http://www.gmx.net/de/go/browser

2 1

Algernon issues with sq3.10-7159dev08.05.2
by Igor Stasenko 28 May '08

28 May '08

- it takes ages to load anything new into image (algernon updates its huge tables with each system change notification) - unable to unload it from image w/o errors -- Best regards, Igor Stasenko AKA sig.

1 2

path for volatile files
by K. K. Subramaniam 28 May '08

28 May '08

Hi, How can I access a 'temporary directory' in Squeak? When I run Squeak from USB flash memory, I don't want temporary files created in the (FileDirectory default) folder. I wanted a VM-unique volatile path where I can create temporary files. e.g. /tmp/squeakvm-$VMPID/ (Unix) C:\WINDOWS\TEMP (Windows) Do I hack FileDirectory to create a unique temp directory on startup and clean the cruft on quit. Is there a better way? Thanks in advance, Subbu

2 1

Mantis bug 215, JPEG plugin problems on Sparc
by tim Rowledge 28 May '08

28 May '08

Does anyone care enough about this to spend time on it? Are there any running Sparc systems left? See http://bugs.squeak.org/view.php?id=215 tim -- tim Rowledge; tim(a)rowledge.org; http://www.rowledge.org/tim Strange OpCodes: PSM: Print and SMear

3 2

Image Building Tool
by Keith Hodges 28 May '08

28 May '08

This is in response to requests over IRC, an overview of some of the new tools. Installer ====== A single class implementing a "Domain Specific Language" for loading packages into the image. The class comment includes extensive documentation. Example: Installer sake addPackage: 'Pier'; install. LevelPlayingField ============= Level playing field is a single script available from http://installer.pbwiki.com/f/LPF.st that can be passed as a startup document when launching an image. It loads the image with the following tools in order to give everyone in all versions of squeak a similar starting point. Once the script has completed it hands over to "Launcher". Installer Launcher Monticello1.5 MonticelloConfigurations PackageInfo Launcher ======= Launcher is installed by LevelPlayingField and it processes subsequent command line parameters Example: #> squeak starting.image http://installer.pbwiki.com/f/LPF.st Installer do="MinorFixes;Packages" SmalltalkImage save=new.image +quit installer.pbwiki.com ============== Is intended for publishing installation scripts that are used in early stages of building a squeak image. i.e scripts which may run prior to more advanced tools, such as Sake/Packages, being loaded into the image. This wiki is on Installers webscript search path. You can publish scripts for any and all versions of squeak and Installer will run the appropriate script for the version of squeak that is running. Sake ==== Based upon concepts in "Make" and "Rake", implements a scheme for defining tasks, with dependencies and for running tasks that are needed. Tasks are typically defined as class methods, and may, unlike Rake, take parameters. An example task: Packages class >> #taskGenerateAll ^ SakeTask define: [ :task | task dependsOn: { "a list of other tasks upon which this task is dependent" self taskGenerateAllUniverses }. task if: [ self squeakmap notNil ]. "the action is run if the needed block returns true" "the action itself is a block or a list of tasks" task action: { self squeakmap taskGenerateSqueakMapPackageTasks } ] An example task that takes a parameter is referenced as a dependency like so: task dependsOn: { SakeClassTask class: #ClassA ensureSubclass: #ClassB. }. The basics of Sake are fully implemented, in SakeTask, SakeClassTask, and SakeFileTask which uses Rio to reference files. We have not as yet implemented bulk multiple file handling as is the traditional staple diet of "Make". Sake/Packages =========== A Universes like package manager built using Sake. Package definitions are methods defining tasks, similar to the example above but with some more meta information. Each package definition is a method in a class, enabling the complete set of package definitions for a version of squeak to be managed in a single monticello package. This also allows all package definitions for all squeak versions to be loaded simultaneously, and managed using a standard class browser. Currently most of the package definitions are generated automatically from the Universes server, but there is the ability to subclass and specialise them. Sake/Packages definitions define a load: action and an unload action. The actions themselves can be any task or script, facilitating more flexibility than the model constrained Universes. Sake/Tasks ======== A package for collecting miscellaneous image housekeeping tasks, according to squeak version. thats about as brief an overview as I can manage, all the best Keith

1 0

Squeak and i18n
by muyuubyou 27 May '08

27 May '08

I haven't posted to the list in a few years but I think this is a very important topic. First of all, thanks to all developers and contributors for all the effort. The improvement in the last 2 years has been nothing short of astonishing. One thing that I still find lacking is squeak's internationalization support. How is it there are no official images in other languages? How about font packages for non-romanized alphabets? Given the amount of international involvement there should at least be a version with some sort of Unicode font (I realize that full Unicode fonts are huge). This is a show stopper for a lot of people these days. How is Squeak-jp working? because the Japanese kids will struggle to play with etoys in English. I've automated the localization of several software packages and I could give a hand with this. We need some sort of extractor-injector of text strings. A propose using a CSV file in UTF-8 with the following information, for every language: TARGET_LANGUAGE_CODE.csv (for instance, en-us, en-uk, en-ca, fr-fr, fr-ca, es-es, es-arg, es-uy, es-co, zh-tw, ru-ru, ru-ua, ua, jp, de-aut, de-ger, etc...) -------------------------------------------------- contents of en-us.csv: string code(placeholder), string, context STR_WORLD_MENU_TITLE,World,WORLD_MENU STR_INSPECT_MENU_INSPECT,Inspect,INSPECT_MENU [...] contents of es-es.csv: string code(placeholder), string, context STR_WORLD_MENU_TITLE,Mundo,WORLD_MENU [...] contents of jp.csv: string code(placeholder), string, context STR_WORLD_MENU_TITLE,世界,WORLD_MENU [...] contents of ru-ru.csv: string code(placeholder), string, context STR_WORLD_MENU_TITLE,мир,WORLD_MENU [...] ... and so on. Or whatever structure you like, just so you can sort by "context", "string" or "placeholder" in OpenOffice, Excel or whatever sorts CSV files. Sorry for the boring post.

2 2

← Newer
1
2
3
4
5
6
7
8
...
18
Older →

Jump to page: