September 2008 - Pharo-dev - lists.pharo.org

[Pharo-project] A .mcz of Pinesoft's accumulated fixes applicable to Pharo
by Simon Kirk Oct. 7, 2008

Oct. 7, 2008

Hi all. Gary and I decided we ought to give out an MCZ of our fixes that we've put together over the last two years. Obviously some of those we found have been pulled into Pharo already, but the attached MCZ represents those that we think are still relevant :) This is a rough list of the fixes: 1. Fixed flaky borwser comment text area sometime not showing. 2. Added getArgumentPermutation to support RefactoringEngine. 3. Fixes to CanvasCharacterScanner and MultiCanvasCharacterScanner to correctly align the first line of text when centered/justifed etc. 4. Added more constant names to Character class. 5. Modified ConnectionQueue to not accept connections when queue is full. 6. Changed GrafPort to answer a MultiDisplayScanner when appropriate (allows subclasses of MultiNewParagraph). 7. NaturalLanguageTranslator startup postion move to after PasteUpMorph since it will show a progress bar. 8. Fix to Random>>nextInt: to allow large numbers which would be out of range if using Floats (as mentioned recently on dev, use of Floats is dodgy anyway...) 9. Changed TranscriptStream>>clear to use a new buffer (was just reset which could leave a massive collection). 10. Changed RemoteString to handle concurrent processes safely. This one's particularly interesting: we were load-testing our Seaside app here and getting lots of syntax errors when hitting the image with around the 20 concurrent processes due to source file access. After this change we were hitting it with approximately 500 concurrent processes, and the failure was in the Java load tester (it timed out). Hope just submitting a .mcz is OK! :) Cheers, Simon and Gary ****************************************************************************************************************************************** This email is from Pinesoft Limited. Its contents are confidential to the intended recipient(s) at the email address(es) to which it has been addressed. It may not be disclosed to or used by anyone other than the addressee(s), nor may it be copied in anyway. If received in error, please contact the sender, then delete it from your system. Although this email and attachments are believed to be free of virus, or any other defect which might affect any computer or IT system into which they are received and opened, it is the responsibility of the recipient to ensure that they are virus free and no responsibility is accepted by Pinesoft for any loss or damage arising in any way from receipt or use thereof. ******************************************************************************************************************************************* Pinesoft Limited are registered in England, Registered number: 2914825. Registered office: 266-268 High Street, Waltham Cross, Herts, EN8 7EA

4 17

[Pharo-project] [ANN] Demo with both Truetype + Polymorph
by Marcus Denker Oct. 6, 2008

Oct. 6, 2008

Here is a Demo with TrueType fonts and Polymorph installed (only tested on Mac). https://gforge.inria.fr/frs/download.php/7268/Pharo0.1-PolymorphTT.zip Marcus -- Marcus Denker -- denker(a)iam.unibe.ch http://www.iam.unibe.ch/~denker

16 41

[Pharo-project] Fonts related from OLPC / rome also
by StÃ©phane Ducasse Oct. 3, 2008

Oct. 3, 2008

I started to go over http://tinlizzie.org/updates/etoys/updates to see if some of the fixes of yoshiki for unicode and others were there. Here is a list of names that may be should be further investigated - 1622fontFix-sw.cs - 1626FontNameGlitch-y.. - 1711UTF8Clipboard.cs - 1712utf8toSqueakFix-. - 1713UnicodeInput1-yo.cs - 1714UnicodeInput2-yo.cs - 1715utf8toSqueakFix2..> - 1725SystemFont-yo.cs - 1726SystemFont2-yo.cs - 1731UpdBtnFontFix-KR.cs - 1813macPasteUTF8-bf.cs - 1815latin1Chars-yo.cs - 1821UnicodeFntLoadin..> - 1826TTCFontSetRework..> - 1826TTCFontSetRework..> - 1827sketchColor-sw.cs - 1832TTCFontSetFix-yo.cs -1834FontRegFix-yo.cs - 1862TTCFontObjStream..> - 1869latin1PathEncodi..> - 1884TTObjForStream-y.. - 1907crlfFix-bf.cs - 1918UTF8InputFix-ar.cs - 1925unixLatin1Input-..> - 1935unicodeFallback-. - 2143macEnc-bf.cs about the rome canvas 1932romepango-base.cs 1933romepango.cs 1934romepango2-yo.cs 24

2 3

[Pharo-project] SIF working - needs a small fix
by Bill Schwab Oct. 1, 2008

Oct. 1, 2008

Hello, I would like to tell you all the incredible things I did to get SIF going - truth is, it loaded and just worked. The only gripe I have is that the Pharo browsers are displaying a large number of LF or other non-printable characters as boxes in the Pharo browsers. I have seen Squeak do this many times, and either just ignored it or deleted them, as my need to view the code and tolerance level allowed. It will get old with the amount of code I plan to move. How would YOU fix this? Is it a SIF problem (dumping out things it should not from Dolphin), a SIF problem (not supressing characters as they are loaded in Pharo), or something missing in the browsers that is expected to be fixed in Pharo? Bill 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) 273-6785 FAX: (352) 392-7029

2 1

Re: [Pharo-project] OSProcess under Pharo
by Bill Schwab Oct. 1, 2008

Oct. 1, 2008

Bryce, Thanks for your reply. Actually, I mentioned FFI etc. because that appeared to be a way to get what OSProcess wanted in terms of dependencies; using Universes should handle that part for me. FFI would be of interest on its own, but appears to make people nervous. Is that for good reason? Put another way, if I want to make some calls into a dll/so, do I really need to mess around with creating/compiling/distributing a plugin? It seems like a lot of overhead, especially since Dolphin does such a nice job based on a mechanism very similar (in apperance at least) to FFI. That said, I am interested in candid opinions; maybe things really are different in Squeak/Pharo. Thanks! Bill 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) 273-6785 FAX: (352) 392-7029 >>> bryce(a)kampjes.demon.co.uk 9/30/2008 2:12 PM >>> Bill Schwab writes: > Before doing something I might regret, I wanted to check with you about > installing OSProcess, VMMaker(?), FFI, etc. Do you have anything to say > "try it and report any problems?" I load OSProcess and FFI from universes, it works loading into a recent pharo-dev image. There's an issue with CommandShell as it subclasses off some MVC classes, I think it's safe to ignore. I'd load VMMaker from SqueakSource directly, but it depends on what you want to do. I've merged the FreeType changes into the Exupery VM's and will add closure support in the next few days. Loading Exupery-dev from the universes will load all the packages you listed. (Though, there's an incompatibility between Exupery and the stricter instance variable checks in Pharo, will be fixed in the next release.) Bryce _______________________________________________ Pharo-project mailing list Pharo-project(a)lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project

2 1

[Pharo-project] OSProcess under Pharo
by Bill Schwab Sept. 30, 2008

Sept. 30, 2008

Before doing something I might regret, I wanted to check with you about installing OSProcess, VMMaker(?), FFI, etc. Do you have anything to say "try it and report any problems?" Bill 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) 273-6785 FAX: (352) 392-7029

2 1

[Pharo-project] OB's icon support improved
by David RÃ¶thlisberger Sept. 30, 2008

Sept. 30, 2008

Hi there, taking the suggestions from the "On icons" thread, I have now changed and extended the icons in columns like this: - Removed icons for extension methods ('+'), instead use italic font for the method name label if it's a extension method and additionally red color if it's an override (as Lukas suggested) - Use new icons for abstract method and super send (credits to Stef who did the hard work of drawing them). The new icons make clear whether an abstract method or a method sending super is redefined in subclasses or not (they get an additional triangle pointing down if they are redefined). For a complete list of supported icons see: http://smallwiki.unibe.ch/hermion/icons/ There are still some icons missing for rare situations, if someone volunteers to draw them, we'd be happy. ;) - Explaining labels for all icons (keep the mouse over them) - Most icons are clickable: Clicking on a super send icon brings you to the method in the superclass, clicking on a triangle down icon brings you to the overridding method in the subclass or gives you a list of candidates if there are several sub-implementations. This I consider as a very convenient extension. - Classes or Traits can implement the method #browserIcon (answering a symbol for the icon name) to define the icon they will get in the class column themselves If you have more suggestions or wishes concerning icons (or anything else), let me know. Current ideas include to introduce icons for trait classes and methods (trait methods are shown in italic right now, hence somehow conflicting with extension methods), or to have an icon for methods without any additional implementor. I would be interested to know whether you consider these ideas as valuable. To load the latest version of OB, you can use this Installer script: Installer wiresong project: 'ob'; install: 'OmniBrowser-dr.419'; install: 'OB-Standard-dr.355'; install: 'OB-Morphic-dr.57'; install: 'OB-Enhancements-dr.217'. Cheers, David

7 17

[Pharo-project] Fwd: [Seaside-dev] preliminary Seaside 2.9 alpha universe
by Adrian Lienhard Sept. 30, 2008

Sept. 30, 2008

The new and easy way to load Seaside takes up to 10 steps (see below). We should try to make this even less painful in Pharo. I assume, having Universes pre-loaded will already reduce half of the steps, but how can it be further simplified? Adrian Begin forwarded message: > From: "Philippe Marschall" <philippe.marschall(a)gmail.com> > Date: September 24, 2008 10:21:16 PM GMT+02:00 > To: "Seaside - developer list" <seaside- > dev(a)lists.squeakfoundation.org> > Subject: [Seaside-dev] preliminary Seaside 2.9 alpha universe > Reply-To: Seaside - developer list <seaside-dev(a)lists.squeakfoundation.org > > > > Hi > > Seaside 2.9 was always hard to load. We always said that we would fix > this and this is a preview of what we expect the solution to look > like. > > After some evaluation of what is available in Squeak we had only two > options, either start from scratch or use universes [1]. We decided to > give universes a shot. Universes is something like a debian > distribution universe for squeak. > > So how do you load this stuff: > 1. get Universes, if you don't already have it (3.10) > 1.1 get an XML parser [2] > 1.2 get universes [3] > 1.3 optionally the the OB universe browser [4] > 2. save the *.packages to your image directory > 3. load the attached .mcz > 4. evaluate UUniverse switchSystemToUniverse: UUniverse > seasideAndServers > 5. in the world menu open a "Universe Browser" either basic or > enhanced > 6. "update list from network" > 7. choose a distribution > 8. choose a server > > There are still some unsolved issues like the naming of alpha an beta > versions. We probably want website integration as well. One possible > change would be to move the tests to a separate tests category. > > [1] http://wiki.squeak.org/squeak/3785 > [2] http://www.squeaksource.com/universes.html > [3] http://www.squeaksource.com/XMLSupport.html > [4] http://www.squeaksource.com/OBUniverse.html > > Cheers > Philippe > _______________________________________________ > seaside-dev mailing list > seaside-dev(a)lists.squeakfoundation.org > http://lists.squeakfoundation.org/mailman/listinfo/seaside-dev

12 48

[Pharo-project] Fwd: Strange bug
by StÃ©phane Ducasse Sept. 30, 2008

Sept. 30, 2008

> > > > Hi > > ImageSegment discoverActiveClasses. > is cool and it lets you (by swapping method dict behind your back) > knowing which classes > have been involved in a certain action. > > as follow > Now ImageSegment discoverActiveClasses. > -- do something typical -- > ImageSegment activeClasses size printString , ' classes were active > out of ' , > Smalltalk allClasses size printString. > > Now if following > recoverFromMDFaultWithTrace > > you put > > Smalltalk at: #MDFaultDict put: Dictionary new. > > you get a strange error, tell you that there is syntax error > > Stef > > >

2 2

[Pharo-project] Senders and implementors list with package
by StÃ©phane Ducasse Sept. 30, 2008

Sept. 30, 2008

Hi david in the latest version of pharo when you browse the senders of a message you get the usual list but the items show in addition the package in the item after the category (which is useless to my eye). Could you add that to OB senders and implementors? Stef

1 0

[Pharo-project] Total image lockup when using middle/right click menu on world
by Simon Kirk Sept. 30, 2008

Sept. 30, 2008

Hi all. Not sure where I should report this. If I use the middle/right (depending on one's preferences) click World menu (the one with Project and Configuration sub-menus down towards the bottom third of the menu), when my mouse goes over the "Configuration" menu section, the entire image freezes irrevocably, at least on OSX (my VM). A few fast-hit alt-. efforts can bring up a debugger on the current process, but it too is frozen. Don't know if this was already something knew about. Sorry for the duplication if so. Cheers, Simon ****************************************************************************************************************************************** This email is from Pinesoft Limited. Its contents are confidential to the intended recipient(s) at the email address(es) to which it has been addressed. It may not be disclosed to or used by anyone other than the addressee(s), nor may it be copied in anyway. If received in error, please contact the sender, then delete it from your system. Although this email and attachments are believed to be free of virus, or any other defect which might affect any computer or IT system into which they are received and opened, it is the responsibility of the recipient to ensure that they are virus free and no responsibility is accepted by Pinesoft for any loss or damage arising in any way from receipt or use thereof. ******************************************************************************************************************************************* Pinesoft Limited are registered in England, Registered number: 2914825. Registered office: 266-268 High Street, Waltham Cross, Herts, EN8 7EA

3 6

[Pharo-project] [OB] Implementor
by Alexandre Bergel Sept. 30, 2008

Sept. 30, 2008

Hi! Pressing alt-m on a method that is implemented on a class-side prevents from opening a implementor method. I will open an issue. Cheers, Alexandre -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.

3 3

[Pharo-project] Integrate early and release often and tests was: Installer was: Fwd: [Seaside-dev] preliminary Seaside 2.9 alphauniverse
by Simon Kirk Sept. 29, 2008

Sept. 29, 2008

Hi all. Apropos of the conversation this subject line evolved from, I was thinking about this from Marcus: "Pharo will do it completely differently: release often, integrate as soon as possible, keep the bugtracker down to managable levels and do not be religiously backwards-compatible." This lead me to think various comments recently from Stef (I think - I apologise if I'm miss-paraphrasing, but I can't find the exact messages) about how to tell which tests have failed/passed etc as the versions go by. I submit that in order for early integration and regular release to work, it needs to be an "all tests green or nothing" situation. Of course, it's very difficult to make sure all the tests are passing, but I just ran *all* the tests in the system, and there's only one error and one failure. This is fantastic work! Not only that but the error is down to BookMorph (due for removal iirc?) and a #printString problem with Array, the latter of which *has* to be simple to fix. If this level of passing tests (or even better, all of them; not necessarily possible as various things like EToys are stripped out I imagine) is maintained, there's no need to tell what tests have changed state from version to version; moreover Pharo will be far the better for it, imho. Just my â¬0.02 :) Cheers, Simon ****************************************************************************************************************************************** This email is from Pinesoft Limited. Its contents are confidential to the intended recipient(s) at the email address(es) to which it has been addressed. It may not be disclosed to or used by anyone other than the addressee(s), nor may it be copied in anyway. If received in error, please contact the sender, then delete it from your system. Although this email and attachments are believed to be free of virus, or any other defect which might affect any computer or IT system into which they are received and opened, it is the responsibility of the recipient to ensure that they are virus free and no responsibility is accepted by Pinesoft for any loss or damage arising in any way from receipt or use thereof. ******************************************************************************************************************************************* Pinesoft Limited are registered in England, Registered number: 2914825. Registered office: 266-268 High Street, Waltham Cross, Herts, EN8 7EA

1 0

[Pharo-project] SqueakMap Packae Loader - nil object
by Dan Corneanu Sept. 29, 2008

Sept. 29, 2008

Hi, there is an exception when trying to open SqueakMap Package Loader in the latest pharo/pharo-dev images with the latests updates. The object instance variable of SMSqueakMap is nil. I'm not sure if I should have put an issue on the tracker or not. Best regards, Dan Corneanu.

3 2

[Pharo-project] How do we manage todos?
by StÃ©phane Ducasse Sept. 29, 2008

Sept. 29, 2008

I would prefer a bugtraker with a tag todo that are not bug fixes. Stef

4 4

Re: [Pharo-project] OB's icon support improved
by Bill Schwab Sept. 29, 2008

Sept. 29, 2008

I apologize for not paying more attention to this thread. Visual feedback in the browser is very much appreciated, and you seem to be doing a very nice job! Had I been paying attention, I'd have said things like the following: (1) Red is probably not the best choice to have lurking on the screen. Maroon/purple/clay rose or just about any "muted" or muddy red IMHO is a much better choice for most things. It might help to add a small amount of blue to the greens too for the same reason. I recommend trying for the psychological benefit with as little eye strain potential as possible. (2) For collections, consider something like a circle drawn a few times from right to left to give an overlapping effect, maybe ... thrown in there if there is room. (3) For exceptions, I like the icon, but two concerns: it might be a little faint/bland; it might work better if its implied movement were upward. Dolphin 5 uses an explosion icon in some situations. That might work well for exceptions and errors. (4) Do we have any color blind people among us? IIRC, something around 12-14% of the population is affected to some degree, so we will encounter it. The test icons will be lost on some of the more extreme red/green sufferers. An alternative would be a boxed check, X or ?. The box (or other border shape) being the suggestion of the test, and the symbol indicating the state. It is fine to add color too, but we should give the red/green crowd a way to see what is going on. Not meaning to jab anyone in the eye, the Installer script is just the kind of thing I'd hope to be able to exchange in situations like this. I do not care so much that it is Installer that we use, but _something_ that can cope with the various publication methods and cope with the details is much needed. Bill 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) 273-6785 FAX: (352) 392-7029

4 3

[Pharo-project] About Network-YM.44.mcz
by Yann Monclair Sept. 29, 2008

Sept. 29, 2008

Hi list, I figured I'd explain my commit here, since I don't know how you handle your issue tracker. While I was loading some code in my image (Universes iirc), I got a lot of warnings pop up, about Url>>#toText being deprecated. After taking a look at it, I saw that the deprecation comment recommended using instead Url>>#asText. Which I tried, but it broke most of the SUnits and also the expected behaviour. Looking at the code, I saw that Url>>#toText simply did ^self asString, when Url>>#asText does ^self asString asText. The semantic leap between Text and String is too big (imo) to force that change on users. I thought it would be better to replace the deprecation comment to recommend replacing the obsolote Url>>#toText by Url>>#asString (which is what this method actually does). I would rather have removed the method, since it has been marked for obsolete for more than 3 years now! But I don't know what the approach is to deleting methods, and how to communicate to people that this isn't a missing method but an intentional removal. I hope this helps understand this version. Cheers, Yann

3 4

[Pharo-project] Update 10082
by Norbert Hartl Sept. 28, 2008

Sept. 28, 2008

In 10082 the script is buggy. script174 contains a line with only .mcz and one line with an archive Pinesoft-Removals-EToys-stephane.ducasse.1.mcz that is not in the repository. Norbert

2 5

[Pharo-project] who is YM
by StÃ©phane Ducasse Sept. 28, 2008

Sept. 28, 2008

because there is a package Network YM 44 Please please avoid to use initials. Stef

3 3

[Pharo-project] [ANN] 10082 More etoy cleaning
by StÃ©phane Ducasse Sept. 28, 2008

Sept. 28, 2008

Hi all I'm slowing going over gary's fixes and cleaning etoy. Etoy cleaning from gary (continued) Removed UndefinedObjectTest>>testNewTileMorphRepresentative Removed BooleanTest>>testNewTileMorphRepresentative Removed BookMorph>>chooseAndRevertToVersion Removed BookMorph>>revertToCheckpoint: secsSince1901 BookMorph>>installRollBackButtons BookMorph>>methodHolderVersions BookMorph>>methodHolders BookMoprh>>goToPageMorph: transitionSpec: runTransitionScripts: BookMorph>>goToPageMorph: transitionSpec: TabbedPalette>>selectTab: cleaned Removed TabbedPalette>>becomeStandPalette Removed>>currentPlayerDo: and fixed sender EncodedCharSet fixed initialize stef

1 0

[Pharo-project] what about squeakPage
by StÃ©phane Ducasse Sept. 28, 2008

Sept. 28, 2008

Hi guys I was browsing some etoy cleans of gary and I found somewhere sqPage and SqueakPage. Apparently we can save on disc a bookmorph. Any ideas if this is working/useful/fun? Stef

2 2

Re: [Pharo-project] Fwd: [Seaside-dev] preliminary Seaside2.9alphauniverse
by Bill Schwab Sept. 28, 2008

Sept. 28, 2008

Damien, Sounds good. Bill 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) 273-6785 FAX: (352) 392-7029 >>> damien.cassou(a)gmail.com 09/28/08 9:16 AM >>> On Sun, Sep 28, 2008 at 3:07 PM, Bill Schwab <BSchwab(a)anest.ufl.edu> wrote: > Damien, > > Now I'm confused :) If installing Seaside is easy, then the web image > is not all that necessary, agreed. With PolyMorph in Pharo, every image > gets better. No worries so far. > > You said "we want one image for Pharo." That is probably a good goal. > Will the RB go into Pharo, or will you continue to produce the developer > images? Yes, RB will be part of the official Pharo release and I won't generate specific pharo-dev images anymore. The list of packages in Pharo is maintained at: http://code.google.com/p/pharo/wiki/Packages -- Damien Cassou Peter von der AhÃ©: Â«I'm beginning to see why Gilad wished us good luckÂ». (http://blogs.sun.com/ahe/entry/override_snafu)

1 0

Re: [Pharo-project] Fwd: [Seaside-dev] preliminary Seaside 2.9alphauniverse
by Bill Schwab Sept. 28, 2008

Sept. 28, 2008

Damien, Now I'm confused :) If installing Seaside is easy, then the web image is not all that necessary, agreed. With PolyMorph in Pharo, every image gets better. No worries so far. You said "we want one image for Pharo." That is probably a good goal. Will the RB go into Pharo, or will you continue to produce the developer images? Bill 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) 273-6785 FAX: (352) 392-7029 >>> damien.cassou(a)gmail.com 09/28/08 5:38 AM >>> Hi Bill, I'm not sure I understand your email completly. Please tell me if it is not the case. On Sat, Sep 27, 2008 at 5:44 PM, Bill Schwab <BSchwab(a)anest.ufl.edu> wrote: > Damien, I suspect you put a lot of work into preparing your images, so please do not view this as anything other than gratitude driving some suggestions. With proper selection of browser and (sorry<g>) deactivating eCompletion, I very much enjoy your web image as a starting point. Whether I continue using it will depend on how hard it is to install Polymorph and the RB vs. Seaside and friends. Polymorph is going to be included in all Pharo images so you don't have to worry. RB as been in all my images from the beginning and will continue to be. I think I will stop generating web images for Pharo because we want only one image for Pharo. I will try to find a way to make installing Seaside easy. > Call me demanding, I want it all :) Installer can probably (either now or with some refinement) reduce Polymorph (which appears to be on its way into the Pharo base anyway???) and Seaside to one-liners in a script. Ditto cryptography. Add some lines to load .cs or whatever files I create on my own, and I should be able to download an image, and then run one script that builds the new image. > > I will stop short of asking you to manually keep up with the small releases, as I probably would end up waiting for the October release before building a new image. It would however be great to either reduce what you do to a simple script, or have it aggressively track updates once Pharo is officially released. I'm sorry but I do not understand this paragraph. > >From a gradually decreasing distance at least, Installer looks like it is trying to fill a gap. As always, I would like to see Pharo sprout an ability to apply multiple categories per method; Dolphin has done that from (almost?) the beginning, and it is quite addictive. My point being that however clever, systems that take the one available category to package methods make the restriction all the more obvious. > > I hope that makes some sense. Thanks to all for what is shaping up to be a nice system! > > Bill > > >

2 1

Re: [Pharo-project] Fwd: [Seaside-dev] preliminary Seaside 2.9 alphauniverse
by Bill Schwab Sept. 28, 2008

Sept. 28, 2008

Damien, Stef, I _think_ what I would want for a starting point is something like Damien's web image, followed by a script that gets the latest of a few things, specific versions of others, and then loads a bunch of .cs or other files for my code. Damien, I suspect you put a lot of work into preparing your images, so please do not view this as anything other than gratitude driving some suggestions. With proper selection of browser and (sorry<g>) deactivating eCompletion, I very much enjoy your web image as a starting point. Whether I continue using it will depend on how hard it is to install Polymorph and the RB vs. Seaside and friends. Call me demanding, I want it all :) Installer can probably (either now or with some refinement) reduce Polymorph (which appears to be on its way into the Pharo base anyway???) and Seaside to one-liners in a script. Ditto cryptography. Add some lines to load .cs or whatever files I create on my own, and I should be able to download an image, and then run one script that builds the new image. I will stop short of asking you to manually keep up with the small releases, as I probably would end up waiting for the October release before building a new image. It would however be great to either reduce what you do to a simple script, or have it aggressively track updates once Pharo is officially released. >From a gradually decreasing distance at least, Installer looks like it is trying to fill a gap. As always, I would like to see Pharo sprout an ability to apply multiple categories per method; Dolphin has done that from (almost?) the beginning, and it is quite addictive. My point being that however clever, systems that take the one available category to package methods make the restriction all the more obvious. I hope that makes some sense. Thanks to all for what is shaping up to be a nice system! Bill 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) 273-6785 FAX: (352) 392-7029 >>> damien.cassou(a)gmail.com 09/26/08 11:03 AM >>> On Fri, Sep 26, 2008 at 8:26 AM, StÃ©phane Ducasse <stephane.ducasse(a)inria.fr> wrote: > - For example I would like to have a real look at MC1.6 but the > message around it is blurry MC1.6 is not the topic here. > I tried LFP and it broke on me. It worked fine for me. You may want to try again and send bug report to Keith who answered all my emails. > - I like installer but when I look at the code I would like to refactor it. The code may not be perfect but it works fine. I don't think Keith would ignore our enhancements to his code. > - There is no document in a nice format that explain why and how (i'm > fixing that for installer and I was waiting because I will refactor it else I do not publish the chapter but I did not want to insult keith since this is not the point since I like installer). This is not fair. There is "no document in a nice format that explains" most of the tools you use everyday. But you still use them. The more people use a tool, the best chance you have to get documentation. Moreover, Keith and Matthew wrote a lot of things in pbwiki, squeak wiki and mailing lists. > - unknown is strange :) Should not be for us Squeakers and Pharoers. > - So if you want to help me I would love to have a > - Installer code refactored > - Installer chapter > - Sake (have a look at the code) > - Sake Chapter for the book I would love that too -- Damien Cassou Peter von der AhÃ©: Â«I'm beginning to see why Gilad wished us good luckÂ». (http://blogs.sun.com/ahe/entry/override_snafu)

3 3

[Pharo-project] Do you still need a Linux VM with FreeType?
by bryce＠kampjes.demon.co.uk Sept. 28, 2008

Sept. 28, 2008

I've stabilised the changes I was working on so now's a reasonable time for me to look at this if it's still required. Is it? Exupery already needs VMs for all major platforms, so I'm thinking about merging that requirement with what Pharo needs. I can build Linux VMs easily as it's my main development platform. Other platforms depend on time, so far development has been more important than maintaining ports for me. Bryce

4 7

[Pharo-project] Compression question
by Bill Schwab Sept. 28, 2008

Sept. 28, 2008

I am working on adapting some basic tools of mine to Pharo, and ran into a trick that use often: specifically, I use ZLib and Dolphin's ability to package globals to store compressed text for scripts, "empty" document files, etc. Then a small wrapper method uses ZLib to uncompress the content of the global w/o having to store all of data in the method source. There are other solutions, but I do this frequently and need to pick a way to handle it. At the most basic level, I found the compression stream and archives categories, but am unclear on where to look. If I want to have a byte array of data and "zip" or "unzip" it memory, is there a good entry point? Would I use a stream or an archive? Please do not feel pressed to do my browsing for me, but if somebody knows the answer, a class or method name would be appreciated. BTW, as further evidence of the accumulated effort, to check the jargon above, I just fired up an early 3.9 image that was handy, and the differences are striking! Maybe I do want to lean on Damien for an updated web image :) Bill 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) 273-6785 FAX: (352) 392-7029

3 2

[Pharo-project] possible TODOs...
by Marcus Denker Sept. 28, 2008

Sept. 28, 2008

Hi, I was already multiple times asked for a list of specific things that people could work on. Here are some ideas: -> World menu (after applying the simplification from bugtracker #166) -> Fullscreen of/on should be rendered like soundof/on, that is just one entry that changes according to state of display. -> in debug the start rendering entries should only be rendered when flag is set. (see code it calls) -> Preferences contains the code of the old preferences-browser. Remove all code related to this. (see Preferences class>>preferencesControlPanel) -> Browser has "printOut" menu entries that just do fileOut. Remove those entries. -> Browser has entry "browse all" in system category menu. This makes no sense. -> SqueakMap stores 1.5MB of crap in the image. make changeset that removes SqueakMap from the image. (check SqueakMap default purge) -> registerInFlapsRegistry... called in the class side initalize, please remove this and the infrastructure of flaps-registering. -> clean up Preferences. There are many preferences that nobody ever changes because the default makes a lot of sense. These need to be removed. -> CurvierMorph should be merged with PolygonMorph -> FileList2 should be merged with FileList. -> TranslatedMethod is a leftover from the original Jitter of 1998. Needs to be removed. Please note, it's in the specialObjectArray! -> Remove PluggableDictionary, PluggableSet (these are workarounds for the broken hashing in Squeak. We will fix hashing later for real) -> ColorTheme and subclasses should be removed. Refactoring needed for all users. (maybe best done after polymorph is added) -> remove BOBTransformationMorph. Refactor, merge with TransformationMorph? (maybe wait till etoys is gone) -> AlignmentMorphBob1 should be removed. Refactoring needed. (maybe wait till etoys is gone) -> Remove world-in-world support. start by removing classes EmbeddedWorldBorderMorph, NewWorldWindow after fixing all uses. -> TwoWayScrollPane shoud be removed (ScrollPane can do everything). Fix all useres of TwoWayScrollPane and delete class. -> CurrentProjectRefactoring is deprecated, yet it is referenced. Please fix all references (mostly just use Project current) and remove CurrentProjectRefactoring. -> Use Sophie's code to render a real pulldown menu entries on windows and Mac -- Marcus Denker -- denker(a)iam.unibe.ch http://www.iam.unibe.ch/~denker

7 19

[Pharo-project] Linux VM building - Where's Eliots closure code?
by bryce＠kampjes.demon.co.uk Sept. 27, 2008

Sept. 27, 2008

I've made decent progress with merging and building. Just got a working TrueType image. There's still work to be done on automating the build. Makefiles still require manual editing, and I haven't automated the image build fully. Where's Eliot's closure bytecode code? Preferably as an mcz. Hopefully, I'll be ready to look at that this weekend. Bryce

2 2

[Pharo-project] Regex11 -> regex 13. in VW
by StÃ©phane Ducasse Sept. 27, 2008

Sept. 27, 2008

Begin forwarded message: > From: "Steven Kelly" <stevek(a)metacase.com> > Date: September 17, 2008 1:24:16 PM CEDT > To: "Vassili Bykov" <smalltalkbigot(a)gmail.com>, Steffen MÃ¤rcker <merkste(a)web.de > > > Cc: vwnc mailing list <vwnc(a)cs.uiuc.edu> > Subject: Re: [vwnc] Fwd: Regex11 > > OK! I've updated the code, tests and documentation and published as > 1.3. > > I second the vote for a Character predicate for this in the base - > it would replace longer code in 7 places in a current standard > image. I think the name should be #isWordConstituent, since > "identifier" generally implies the first character is not numeric. > > Steve > >> -----Original Message----- >> From: vwnc-bounces(a)cs.uiuc.edu [mailto:vwnc-bounces@cs.uiuc.edu] On >> Behalf Of Vassili Bykov >> Sent: 17 September 2008 00:16 >> To: Steffen MÃ¤rcker >> Cc: vwnc mailing list >> Subject: Re: [vwnc] Fwd: Regex11 >> >> I think \w should include underscore to match what seems to be the >> expected behavior out there. Going a step further, perhaps a >> Character >> predicate named something like #isIdentifierConstituent to include $_ >> in addition to pure alphanumerics would be useful. >> >> Cheers, >> >> --Vassili >> _______________________________________________ >> vwnc mailing list >> vwnc(a)cs.uiuc.edu >> http://lists.cs.uiuc.edu/mailman/listinfo/vwnc > > _______________________________________________ > vwnc mailing list > vwnc(a)cs.uiuc.edu > http://lists.cs.uiuc.edu/mailman/listinfo/vwnc > > ------- Weitergeleitete Nachricht ------- > An: "Vassili Bykov" <smalltalkbigot(a)gmail.com> > Kopie: > Betreff: Regex11 > Datum: Mon, 15 Sep 2008 09:33:13 +0200 > > Dear Vassili Bykov, > > as far as I can see you've written the Regex11 package and I think > I've > found a bug. The documentation states: > > \w any word constituent character (same as [a-zA-Z0-9_]) > > but the actual implementation fails with underscores: > > '_' matchesRegex: '\w' "evaluates to false" > '_' matchesRegex: '[a-zA-Z0-9_]' "evaluates to true" > > Should the bug be fixed because we rely on the definition of \w or > is it > the better way to adjust the documentation in order not to break > existing > applications? How can I help?

1 0

[Pharo-project] Fwd: Published a fix for WideString fromPacked: 'TEXT' asPacked
by StÃ©phane Ducasse Sept. 27, 2008

Sept. 27, 2008

Begin forwarded message: > From: pharo-project-owner(a)lists.gforge.inria.fr > Date: September 27, 2008 9:04:14 AM CEDT > To: stephane.ducasse(a)free.fr > Subject: Published a fix for WideString fromPacked: 'TEXT' asPacked > > You are not allowed to post to this mailing list, and your message has > been automatically rejected. If you think that your messages are > being rejected in error, contact the mailing list owner at > pharo-project-owner(a)lists.gforge.inria.fr. > > > From: stephane ducasse <stephane.ducasse(a)free.fr> > Date: September 27, 2008 9:04:12 AM CEDT > To: An open mailing list to discuss any topics related to an open- > source Smalltalk <Pharo-project(a)lists.gforge.inria.fr> > Subject: Published a fix for WideString fromPacked: 'TEXT' asPacked > > > in multilangual in inbox > > Stef > > >

1 0

Re: [Pharo-project] About Flap
by Bill Schwab Sept. 27, 2008

Sept. 27, 2008

Hello all, Flaps will be easier to remove than to replace, and toolbars and menus come to mind, of course. "The" world menu is a concept that always bothers me a little: while I like having the ability to run Pharo with what amounts to MDI (child shells in a single main window), it would be nice to have each shell/SystemWindow be a host window. In the latter situation, flaps and world menu are (correct me if I am wrong) largely homeless. Ok, a flap could be a host window of sorts (perhaps hidden when Pharo does not have focus), but Pharo no longer owns the desktop, so the world menu is gone, right?? Maybe it simply _should_ be gone in that case. In terms of flaps, they no doubt add to drawing overhead. Would that not improve if the polling were converted to handling events? Another helpful change would be to either replace "tearing off" a new tool to clicking/releasing a command button that simply opens it at default location and size, or changing the dragging to respect the fast window drag setting. I invariably set that even on new machines (not that I am quick to replace old ones). Bottom line: faster GUI code never hurts. A pop or a fast drag off of a flap should help. Bill 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) 273-6785 FAX: (352) 392-7029 >>> stephane.ducasse(a)inria.fr 09/27/08 1:59 AM >>> OK we will go slowly there. One of the problem is that flaps are taking also a cpu On Sep 27, 2008, at 12:27 AM, Gary Chambers wrote: > I quite like the tools flap, when configured for auto-popout and > with "solid" style... > Just my 2c > > Gary. > > > ----- Original Message ----- From: "Igor Stasenko" > <siguctua(a)gmail.com> > To: "An open mailing list to discuss any topics related to an open- > sourceSmalltalk" <pharo-project(a)lists.gforge.inria.fr> > Sent: Wednesday, September 24, 2008 8:46 PM > Subject: Re: [Pharo-project] About Flap > > >> 2008/9/24 Marcus Denker <denker(a)iam.unibe.ch>: >>> >>> On 24.09.2008, at 20:14, Bill Schwab wrote: >>> >>>> Lukas, >>>> >>>> That's fine if my hands are on the keyboard, but if I have one on >>>> the >>>> mouse (common when browsing), then (in Dolphin at least), I end up >>>> either right clicking on a class in an open browser, or hitting a >>>> toolbar button. For the more obscure tools, the shell menus >>>> (Tools) do >>>> the trick. Shortcuts are great, but we should also provide mouse- >>>> based >>>> options. >>>> >>> >>> remember: there will be a menu, with a "tools" entry. And there will >>> always be the world-menu (click on the backround). >>> >>> the flap is not faster in the number of clicks than a pulldown >>> menu... >>> >> >> well, with flaps is more convenient because you can drag a tool >> window >> and place it where you want to. >> While from menu - first, it placing a window automatically, and then >> often you need to arrange it again. >> How about putting a window in hand, when you opening it from menu? >> I find it very convenient and useful feature - it gives you chance to >> arrange window with no extra clicks. >> >> >>> Marcus >>> >>> -- >>> Marcus Denker -- denker(a)iam.unibe.ch >>> http://www.iam.unibe.ch/~denker >>> >>> >>> _______________________________________________ >>> Pharo-project mailing list >>> Pharo-project(a)lists.gforge.inria.fr >>> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project >>> >> >> >> >> -- >> Best regards, >> Igor Stasenko AKA sig. >> >> _______________________________________________ >> Pharo-project mailing list >> Pharo-project(a)lists.gforge.inria.fr >> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project > > > _______________________________________________ > Pharo-project mailing list > Pharo-project(a)lists.gforge.inria.fr > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project > _______________________________________________ Pharo-project mailing list Pharo-project(a)lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project

1 0

[Pharo-project] Regression testing results within package manifesto
by StÃ©phane Ducasse Sept. 27, 2008

Sept. 27, 2008

Hi guys For Moose we started to add package manifesto using DrDoc. We will use the manifesto to document our package with license, documentation, class diagram... Now what I would really like to have is the results of SmallLint rules and the tests because when I run the tests if some of them were broken I do not know whether I made progress or not. I think that the manifesto could be a nice place to keep such information. After we could have a simple manisfesto dropping script Object allSubclasses do: .... what do you think about that, I would like to have a bit more self documentation and software engineering meta data within the system. Stef

3 4

Re: [Pharo-project] About Flap
by Bill Schwab Sept. 27, 2008

Sept. 27, 2008

Lukas, That's fine if my hands are on the keyboard, but if I have one on the mouse (common when browsing), then (in Dolphin at least), I end up either right clicking on a class in an open browser, or hitting a toolbar button. For the more obscure tools, the shell menus (Tools) do the trick. Shortcuts are great, but we should also provide mouse-based options. Bill 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) 273-6785 FAX: (352) 392-7029 >>> renggli(a)gmail.com 09/24/08 1:51 PM >>> > I like the way they are in dev images. You move the mouse to the right > side of the image and the flap opens. From there, you can drag&drop > your favorite tool. Do you have any faster way to open a browser or a > transcript? Ctrl+B and Ctrl+T -- Lukas Renggli http://www.lukas-renggli.ch _______________________________________________ Pharo-project mailing list Pharo-project(a)lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project

5 4

[Pharo-project] Bug? I cannot see the diff anymore doing merge or browse
by StÃ©phane Ducasse Sept. 26, 2008

Sept. 26, 2008

Hi I'm tracking a problem I have and which is quite annoying I cannot see the diff when I do a merge or browse with MC. I'm with 10074 and I woudl like to know if you have the same behavior. Stef

1 0

[Pharo-project] Interesting news
by Bill Schwab Sept. 26, 2008

Sept. 26, 2008

Hello all, Have a look at: http://www.object-arts.com/content/news/excitingNews.html Object Arts and Lesser Software are collaborating on a new version of Dolphin. They are still sticking to the Windows-only angle, which I find disappointing. However, even if they hold to that, we are better off keeping and enhancing what I consider to be the best development environment I have ever used. Bill 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) 273-6785 FAX: (352) 392-7029

1 0

[Pharo-project] Move to package from sender and implementors list
by StÃ©phane Ducasse Sept. 26, 2008

Sept. 26, 2008

Hi guys it would be great to have the senders and implementors tools offering move to package in their menus. Stef

4 3

[Pharo-project] On Icons
by Torsten Bergmann Sept. 25, 2008

Sept. 25, 2008

Matthew suggested >Perhaps OB could use composable icons Eclipse uses a similar concept, see http://www.eclipse.org/articles/Article-Decorators/decorators.html Typical icon sizes are 16x16 and 32x32 (at least on Win32 and in most icon collections on the web). The ones in OB currently look smaller and may be too small for decorators... Bye Torsten -- Ist Ihr Browser Vista-kompatibel? Jetzt die neuesten Browser-Versionen downloaden: http://www.gmx.net/de/go/browser

1 0

[Pharo-project] [ANN] some cleans
by StÃ©phane Ducasse Sept. 25, 2008

Sept. 25, 2008

I remove isUniClass: and clean the inspector following gary's clean. plus some leftOver related to uniclass. Stef

1 0

[Pharo-project] On Icons
by Lukas Renggli Sept. 25, 2008

Sept. 25, 2008

> OB-Enhancements version 0.203 I am horrified with all these colored icons on methods. That's worse than in the old browser with the button that had 8 different colors depending on the state of the method. - I don't want to see part of the "contents" of a method as an icon. The super-icon is useless, it only hides valuable information (for example that there is a sub-implementation). If I want to know if a method calls super (and when) I need to look at the method anyway. - The same is for abstract method icon. When browsing the protocol/interface I don't need to know that. I am more interested if any of the subclasses override it. - The same for extension methods. I see that, because it is in a different protocol and because its name is in italic (the same as with the extension classes). I don't need yet another icon. The only valuable thing for me as a developer is to know if there is a super/sub implementation of a method. And maybe if there are halts inside. The rest I can find out with one click or less, if necessary. This would greatly improve the interface. The browser starts to be really messy. Furthermore I would like to define my own icons for classes, for example Seaside component should have its own icon. So instead of hardcoding this into OB itself why not asking the class for an icon, as this is done in VW for example. Cheers, Lukas -- Lukas Renggli http://www.lukas-renggli.ch

6 27

[Pharo-project] Removing class category
by Alexandre Bergel Sept. 25, 2008

Sept. 25, 2008

Hi! I just created a package MorphicBalloon that contains the 3 classes defined in Morphic-Balloon. Here is the comment: -=-=-=-=-=-=-=-=-=-=-=-= This package contains the 3 classes that define Balloon (BalloonCanvas, BalloonRectangleMorph, MatrixTranformMorph). These three classes, are only referenced by classes defined in the class category Morphic-Support. I am not sure whether this is a good change, but it is a first step to remove class category -=-=-=-=-=-=-=-=-=-=-=-= This is an example that should raise the question about where we want to go. I feel that having class category as sub-monticellopackage is a bad thing. We should get rid of class category. Cheers, Alexandre -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.

3 2

[Pharo-project] Just for fun...
by StÃ©phane Ducasse Sept. 25, 2008

Sept. 25, 2008

Hi guys guess what... The old preferencePanelWindow is defined and created within the class Preferences of course preferencesControlPanel I will kill it. Stef

1 0

Re: [Pharo-project] about preparing the first release
by Bill Schwab Sept. 25, 2008

Sept. 25, 2008

Stability takes at least two forms: the system works AND is not going to experience violent change. Given that more of the latter is possible/expected, it might be better to start at 0.1. Bill 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) 273-6785 FAX: (352) 392-7029 >>> adi(a)netstyle.ch 09/25/08 5:21 AM >>> On Sep 25, 2008, at 11:12 , Damien Cassou wrote: > On Wed, Sep 24, 2008 at 10:49 PM, Adrian Lienhard <adi(a)netstyle.ch> > wrote: >> I'd suggest the common major.minor[.maintenance] scheme starting >> with "1.0". > > I would not start with "1.0" because it means this is already stable > enough for anyone to rely upon. isn't this exactly the goal of the release?! Only if the shipped version is reasonably stable the people will use it... Adrian _______________________________________________ Pharo-project mailing list Pharo-project(a)lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project

1 0

[Pharo-project] About traits support in OB
by David RÃ¶thlisberger Sept. 25, 2008

Sept. 25, 2008

Hi, I'm not terribly happy with the current traits integration in OB. To browse traits used in a given class, we have this fourth switch next to instance / ? / class, called 'traits'. Clicking on it then lists in the next column all traits used in this class and the class itself. This additional column duplicates the switches instance / ? / class. For classes using traits you can see methods coming from traits in italic, methods defined but not implemented in a trait in normal font when browsing the class and in blue when browsing the trait. You also see the hierarchy of the used traits in this 'trait column'. But it's also possible to open this 'trait column' for classes that do not use any traits, this is very confusing. Also it's not nice to suddenly get five columns instead of four. My suggestions to better integrate traits is to show them below the classes using them directly in the class column, hierarchially sorted and marked either with an icon or with a colored label (eg. in blue). Like this we can get rid of this additional column and the additional 'trait' switch. The only problem I see with this change is that it might confuse people as traits are displayed along with all classes of a package/class cat in the class column, although they might not be part of neither the package/class cat nor the class hierarchy. What do you think, should I change it like this or not? Cheers, David

2 2

[Pharo-project] On Icons
by Torsten Bergmann Sept. 25, 2008

Sept. 25, 2008

David, this crashes the browser for traits. I think #browserIcon returning nil is also required in Trait/TraitDescription class. At least this works for me then. Bye Torsten -- Der GMX SmartSurfer hilft bis zu 70% Ihrer Onlinekosten zu sparen! Ideal fÃ¼r Modem und ISDN: http://www.gmx.net/de/go/smartsurfer

4 5

[Pharo-project] about preparing the first release
by StÃ©phane Ducasse Sept. 25, 2008

Sept. 25, 2008

Hi guys I will work using the pinesoft etoy removal package. But it can take time to do that carefully since etoy put its fingers everywhere. Now I would really like that we define a date for the first release. I would really like to have something for maximum end of october. What about you? Stef

3 4

Re: [Pharo-project] about preparing the first release
by Bill Schwab Sept. 24, 2008

Sept. 24, 2008

Stef, Given a willingness to slip (either date or feature list) if the system is unstable, a deadline is great. Bill 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) 273-6785 FAX: (352) 392-7029 >>> stephane.ducasse(a)inria.fr 09/24/08 4:08 PM >>> Hi guys I will work using the pinesoft etoy removal package. But it can take time to do that carefully since etoy put its fingers everywhere. Now I would really like that we define a date for the first release. I would really like to have something for maximum end of october. What about you? Stef _______________________________________________ Pharo-project mailing list Pharo-project(a)lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project

1 0

Re: [Pharo-project] Re: Re: Re: Re: [ANN] Demo with both Truetype+Polymorph
by Bill Schwab Sept. 24, 2008

Sept. 24, 2008

It includes a Linux VM with the shutdown prompt fix - MUCH appreciated, not to mention the packages that were the point of demo :) Bill 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) 273-6785 FAX: (352) 392-7029 >>> stephane.ducasse(a)inria.fr 09/24/08 3:57 PM >>> thanks for the report. I guess that marcus will look into that. I do not know how he made the one click precisely. Stef

1 0

[Pharo-project] [ANN] 10075
by StÃ©phane Ducasse Sept. 24, 2008

Sept. 24, 2008

I removed ComponentLikeModel. Tomorrow morning we will applied our tools on pharo and start to fix package cycles. Stef going to bed :)

1 0

[Pharo-project] Other Morphic question: allKnownNames
by StÃ©phane Ducasse Sept. 24, 2008

Sept. 24, 2008

Hi I'm reading innocuousName allKnownNames allSubmorphNamesDo: and friends It seems to me that Morph submorph elements could be given names from the user interface and this would define automatically instance variables or something like that with the given name. My gut feeling is shouting inside me to tell me that we should remove that. I would really like to know your point of view. I checked in Squeak2.8 allKnownNames "Answer a list of all the names of all named objects borne in my instance variable values" ^ self instanceVariableValues select: [:e | (e ~~ nil) and: [e knownName ~~ nil]] thenCollect: [:e | e knownName] Stef

2 3

[Pharo-project] About Flap
by StÃ©phane Ducasse Sept. 24, 2008

Sept. 24, 2008

Hi guys we are applying our tools on pharo :) and I would like to know what do we decided for Flaps? I thought that we will remove this class. Stef

4 5

Re: [Pharo-project] Burning the disk packs: Who puts equality into Object?
by Bill Schwab Sept. 24, 2008

Sept. 24, 2008

I have always seen defaulting equality to identity to be a useful confusion, and I fear it would simply be copied down (often in frustration) into many classes vs. done once. It also removes a compatibility with other Smalltalks, which I think we should be working to maintain and enhance where it is not destructive to do so. Just my 2 asCents. Bill 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) 273-6785 FAX: (352) 392-7029 >>> diegogomezdeck(a)consultar.com 09/24/08 9:30 AM >>> Hi guys, I recently discovered Pharo project and I'd like to fire a proposal for cleaning Object class. The "simple" idea is to remove equality concept (implemented with #= and #hash messages) from Object. Equality means that different instances can represent the same value (like Numbers, Strings, Dates, etc). Identity means that any object is a different object from any other from the world. All objects have identity (#== and #basicHash messages), but not all object share the equality concept. To make the things worst, the default implementation of equality is based on the identity, producing a big confusion to newbies (and experts too!). One implementation option is to implement the Equality concept in a super-class of Magnitude (Comparable?), and reused all over the class hierarchy with traits. Of course such factorization will impact also on the Collections that (most of them) also confuses equality with identity. Probably some collections will need an Strategy-pattern to use identity or equality based behavior. What do you think? Cheers, -- Diego _______________________________________________ Pharo-project mailing list Pharo-project(a)lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project

2 1

Re: [Pharo-project] About Flap
by Bill Schwab Sept. 24, 2008

Sept. 24, 2008

Now THAT'S a reason for us to remove them!!!!!!! :) 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) 273-6785 FAX: (352) 392-7029 >>> damien.cassou(a)gmail.com 09/24/08 9:07 AM >>> On Wed, Sep 24, 2008 at 3:05 PM, Damien Cassou <damien.cassou(a)gmail.com> wrote: > On Wed, Sep 24, 2008 at 1:55 PM, StÃ©phane Ducasse > <stephane.ducasse(a)inria.fr> wrote: >> I would like to know what do we decided for Flaps? > > I like the way they are in dev images. You move the mouse to the right > side of the image and the flap opens. From there, you can drag&drop > your favorite tool. Do you have any faster way to open a browser or a > transcript? Note that latest versions of Microsoft Visual Studio now contains flaps too. -- Damien Cassou Peter von der AhÃ©: Â«I'm beginning to see why Gilad wished us good luckÂ». (http://blogs.sun.com/ahe/entry/override_snafu)

1 0

[Pharo-project] Burning the disk packs: Who puts equality into Object?
by Diego Gomez Deck Sept. 24, 2008

Sept. 24, 2008

Hi guys, I recently discovered Pharo project and I'd like to fire a proposal for cleaning Object class. The "simple" idea is to remove equality concept (implemented with #= and #hash messages) from Object. Equality means that different instances can represent the same value (like Numbers, Strings, Dates, etc). Identity means that any object is a different object from any other from the world. All objects have identity (#== and #basicHash messages), but not all object share the equality concept. To make the things worst, the default implementation of equality is based on the identity, producing a big confusion to newbies (and experts too!). One implementation option is to implement the Equality concept in a super-class of Magnitude (Comparable?), and reused all over the class hierarchy with traits. Of course such factorization will impact also on the Collections that (most of them) also confuses equality with identity. Probably some collections will need an Strategy-pattern to use identity or equality based behavior. What do you think? Cheers, -- Diego

1 0

[Pharo-project] About Flap
by Torsten Bergmann Sept. 24, 2008

Sept. 24, 2008

Damien wrote >Do you have any faster way to open a browser or a transcript? Hit Cmd+b for browser and Cmd+t for transcript. ;) Bye Torsten -- Psssst! Schon vom neuen GMX MultiMessenger gehÃ¶rt? Der kann`s mit allen: http://www.gmx.net/de/go/multimessenger

1 0

[Pharo-project] Please use your name and not initial
by Torsten Bergmann Sept. 24, 2008

Sept. 24, 2008

Hi Stef, that's me. It was just a basic test for committing to pharo income (just categorized an uncategorized method). You can drop the version and categorize the method yourself with a different name meanwhile (so actually it's not a license issue). Sorry for the confusion. Signed an agreement for Squeak already but it look like another one is required for Pharo. Where is the text and who should get a signed copy? Bye Torsten -- GMX Kostenlose Spiele: Einfach online spielen und SpaÃ haben mit Pastry Passion! http://games.entertainment.gmx.net/de/entertainment/games/free/puzzle/61691…

2 1

[Pharo-project] Please use your name and not initial
by StÃ©phane Ducasse Sept. 24, 2008

Sept. 24, 2008

Hi guys there is a Kernel-tbn in pharoInbox and this is important for us to know who is the person since we want him to sign the agreement (even if pubishing in the repository means that you license the code under MIT). Stef

1 0

[Pharo-project] First pharo-dev with limited package set
by Damien Cassou Sept. 24, 2008

Sept. 24, 2008

Hi, I've just released the first pharo-dev which contains a limited package set. See http://code.google.com/p/pharo/wiki/Packages for the list of packages we want in Pharo. You can download the image at https://gforge.inria.fr/frs/download.php/7404/pharo0.1-10074dev08.09.2.zip. Here is the list of packages included in this release: AST version 160 eCompletion version 0.90 OB-Enhancements version 0.203 OmniBrowser version 0.416 OmniBrowser-Morphic version 0.52 OmniBrowser-Refactory version 1.0 OmniBrowser-Regex version 0.9.1 OmniBrowser-Standard version 0.346 Pharo packages version 0.3 Refactoring Core version 5 RoelTyper version 0.60 Shout version 3.15-tween.72 ShoutWorkspace version 1-tween.2 VBRegex version 1.9 -- Damien Cassou Peter von der AhÃ©: Â«I'm beginning to see why Gilad wished us good luckÂ». (http://blogs.sun.com/ahe/entry/override_snafu)

7 21

[Pharo-project] [OT] Sophie in Java
by Michael Rueger Sept. 24, 2008

Sept. 24, 2008

Hi all, the Mellon foundation just approved the grant to Astea in Bulgaria to redevelop Sophie in Java! Astea (http://asteasolutions.com) are the nice guys that we trained and who then turned around, stating on their web site that they developed Sophie, took our Sophie documents and artwork for their proposal and managed to persuade Mellon to actually give them money... We moved the current Sophie project to http://opensophie.org, hoping to keep Sophie alive and making Sophie better than what they can come up with. And why I'm sending this to the Pharo list? Pharo already has so much promise as a base system, hopefully providing in the core already what we had to re-invent in Sophie because it didn't work in the main Squeak version. The Freetype stuff already turned out really well! Hopefully more to come. And once I finish these new etoys installers I can do some Pharo work too, hopefully ;-) Michael

6 7

[Pharo-project] Re: about http://bugs.squeak.org/view.php?id=6336
by StÃ©phane Ducasse Sept. 24, 2008

Sept. 24, 2008

Thanks. Another problem with real atomic loading. Stef > Hi Stef, > > there is nothing special with the two change sets except their > fileIn order, since instances of Character are used while both of > them are loaded and compiled ... > > The 3.10 release team attempted to make .mcz out of them but failed > miserably, so I suggest you just file them in. > > Let me know how it worked for you; perhaps also see if before/after > the .image indeed gets smaller by ca. 300 KB. > > Ah, and before I forget: because of Association #hash #= (since 3.8 > or so, Marcus knows the problem), compilation of Unicode methods > take looong, since one of its class variables has a populated > SparseLarge* thing which is #hash'ed and #='ed when compiling access > to each class variable :( But this is a genaral problem, which just > manifests, when compiling Unicode, with pain. > > Cheers, > Klaus > > On Tue, 23 Sep 2008 21:52:35 +0200, StÃ©phane Ducasse wrote: > >> Hi klaus >> >> I would like to include the following in pharo >> about http://bugs.squeak.org/view.php?id=6336 >> >> can you tell me if I should pay attention to something in particular. >> >> Stef >> > >

2 1

[Pharo-project] about http://bugs.squeak.org/view.php?id=6336
by StÃ©phane Ducasse Sept. 23, 2008

Sept. 23, 2008

Hi klaus I would like to include the following in pharo about http://bugs.squeak.org/view.php?id=6336 can you tell me if I should pay attention to something in particular. Stef

1 0

Re: [Pharo-project] First pharo-dev with limited package set
by Bill Schwab Sept. 23, 2008

Sept. 23, 2008

Stef, How about a preference for both? I generally find the color annoying in workspaces, and _greatly_ appreciate shout in browsers and the debugger. There is a preference to disable automatic definition of workspace variables??? Great idea!!! I'm not a fan of them. Bill 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) 273-6785 FAX: (352) 392-7029 >>> stephane.ducasse(a)inria.fr 9/23/2008 11:08:17 AM >>> Damien I would go for the workspace with color. and we should set the pref to have to type | | to declare varaibles. stef On Sep 23, 2008, at 4:30 PM, Damien Cassou wrote: > On Tue, Sep 23, 2008 at 2:17 PM, StÃ©phane Ducasse > <stephane.ducasse(a)inria.fr> wrote: >> uh? does not work for me no idea what is a Smalltalk workspace. > > a workspace to write Smalltalk code as opposed to a workspace to > write text. > > At the beginning, the ShoutWorkspace replaced the Workspace but some > people complained. That's why there are two workspaces now. I've never > used the non-colored workspace: when I want to write text, I enclose > it in quotes. > > -- > Damien Cassou > Peter von der AhÃ©: Â«I'm beginning to see why Gilad wished us good > luckÂ». (http://blogs.sun.com/ahe/entry/override_snafu) > _______________________________________________ > Pharo-project mailing list > Pharo-project(a)lists.gforge.inria.fr > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project _______________________________________________ Pharo-project mailing list Pharo-project(a)lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project

2 1

Re: [Pharo-project] First pharo-dev with limited package set
by Bill Schwab Sept. 23, 2008

Sept. 23, 2008

Damien, I thought about that, but it all get colored the way I LIKE having my method comments colored - I do NOT<g> like that for my notes. Shout is very helpful, and the rendering speed and stability are nothing short of amazing. However, I want a choice in what I use for a workspace, and it is almost always going to be plain text. Having the two menu options is fine/preferred, but a preference would work too. Bill 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) 273-6785 FAX: (352) 392-7029 >>> damien.cassou(a)gmail.com 9/23/2008 9:30:05 AM >>> On Tue, Sep 23, 2008 at 2:17 PM, StÃ©phane Ducasse <stephane.ducasse(a)inria.fr> wrote: > uh? does not work for me no idea what is a Smalltalk workspace. a workspace to write Smalltalk code as opposed to a workspace to write text. At the beginning, the ShoutWorkspace replaced the Workspace but some people complained. That's why there are two workspaces now. I've never used the non-colored workspace: when I want to write text, I enclose it in quotes. -- Damien Cassou Peter von der AhÃ©: Â«I'm beginning to see why Gilad wished us good luckÂ». (http://blogs.sun.com/ahe/entry/override_snafu)

2 1

[Pharo-project] Tree View in OmniBrowser
by David RÃ¶thlisberger Sept. 23, 2008

Sept. 23, 2008

Hello, It was very hard work, but finally and eventually I managed to integrate the tree view of Juraj in a reasonable way into OmniBrowser, so that we can use it for instance to browse packages. The package browser of OB-Enhancements is now using this tree in the first column to render packages on the first level and class categories on the second level. As this is a rather drastic change affecting also the underlying meta-model representing navigation in OB, I'd appreciate if some people give this a try and report any problems they might find to me. I extensively tested it, but you never know. I'm still working on some little things concerning this tree integration, mostly internal stuff, so nothing that should stand in way of using it already now. Just drop me a note if you encounter any problems, something that can still occur right now. Other recent enhancements include: - OB-TraitsIntegration is now part of OB-Enhancements as some people wanted to have it integrated in this package - mouse-over for icons in columns to get an explanation label - some icons are clickable, eg. to directly browse methods in superclasses - improvements in smart groups You need to load the following packages: OmniBrowser-dr.417 OB-Morphic-dr.54 OB-Standard-dr.351 OB-Enhancements-dr.207 ToDos: - refactor some of the code for the tree view - address several minor issues (eg. performance) - write more tests for tree view - integrate OmniBrowserFixes in OB-Enhancements Cheers, David

3 3

[Pharo-project] SharedQueue Fixes from OLPC distro
by StÃ©phane Ducasse Sept. 23, 2008

Sept. 23, 2008

http://tinlizzie.org/updates/etoys/updates 2009fixSharedQueue-b.. 2010fixSharedQueueAg..>

1 0

Re: [Pharo-project] Product Brand Pharo
by Bill Schwab Sept. 23, 2008

Sept. 23, 2008

Gary, There are two topics here: (1) allowing ourselves and others to easily retrofit tools into the various menus; (2) building a fast and elegant GUI framework. BookMorphs are a (not terribly good) implementation attempts at the latter. For the newbie (or even us trying to get something done), it can be very helpful to have tools offered in the right places. Cleaner is better of course, but the "user" in the case is more interested in whether the tool works than in how the tool works. Set out to write your own helpful wizard, and elegance is appreciated. Bill 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) 273-6785 FAX: (352) 392-7029 >>> gazzaguru2(a)btinternet.com 09/23/08 7:16 AM >>> I remember doing a "Wizard" framework in VisualAge... shudders... Still, nice to have standardised support for such things rather than ad-hoc BookMorphs (for example). Regards, Gary. ----- Original Message ----- From: "Bill Schwab" <BSchwab(a)anest.ufl.edu> To: <pharo-project(a)lists.gforge.inria.fr> Sent: Tuesday, September 23, 2008 12:03 PM Subject: Re: [Pharo-project] Product Brand Pharo > Markus, > > Based on the discussion, I started noting how long it takes my images to > load; a splash screen might be in order; I also fear that by the time VM > loads and the GUI is able to draw itself, it is no longer needed :( > Still, I'd add it to the milestones, but probably late in the process. > > About help, I am not very sympathetic on the "no documentation" front. > Someone seeking to join us will quickly find tutorials, and can post > questions. Dolphin's MVP framework almost invariably had the newbie > expressing pain and confusion, followed by "why was this so hard last > week???" Conversely, the MS Office IDL for Automation has LOTS of > "help" - almost all of it useless. The message (pun intended) needs to > be "this is not trivial to learn, but very worth your time." Ensuring > that good tutorials are linked from our Wiki would be a nice gesture. > > Wizards in the IDE - that's a very good idea, and Dolphin helps it along > through the IDE extension framework, which is little more than events > triggered at the right time, off of each tool's class and giving the > needed instances for the code to make the desired changes. Dolphin's > view resources, and our lack of them, would require some adaptation. > However we do it, it would be a nice addition. The idea is to do > _something_ that allows each of us to provide hooks into the IDE w/o > having to alter the code. That way, building a new image and updating > an existing one will not destroy the extensions. > > Bill > > > > > 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) 273-6785 > FAX: (352) 392-7029 > > > _______________________________________________ > Pharo-project mailing list > Pharo-project(a)lists.gforge.inria.fr > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project _______________________________________________ Pharo-project mailing list Pharo-project(a)lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project

2 1

Re: [Pharo-project] First pharo-dev with limited package set
by Bill Schwab Sept. 23, 2008

Sept. 23, 2008

Ok, but only if you allow a choice of which one to use. I am a HUGE fan of Shout, but I generally do not want a workspace trying to colorize my notes to myself. If I end up with enough code that the syntax is not clear, it probably belongs in a class and in a package or change set - at least I try to do that. Bill 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) 273-6785 FAX: (352) 392-7029 >>> adi(a)netstyle.ch 09/23/08 5:42 AM >>> Hi Damien, Thanks, looks good! Question: Does this include Universes? It's not listed below but I see the menu item (which does not work because of a deprecation warning). I think we need Universes mainly because Seaside will be packaged with Universes soon, and being able to easily load Seaside is a key requirement for Pharo. Another thing we need to look at (and I planned to do this) is the open menu. With all the additional packages in the dev version its even worse than before. For example there are 3(!) items for opening a workspace: Workspace Workspace (without colourisation) Shout Workspace There ought to be just one, and in the right place. Cheers, Adrian On Sep 23, 2008, at 09:52 , StÃ©phane Ducasse wrote: > This sounds good. > Now we should be able to see the overrides of ecompletion. > > Stef > > On Sep 23, 2008, at 8:20 AM, Damien Cassou wrote: > >> Hi, >> >> I've just released the first pharo-dev which contains a limited >> package set. See http://code.google.com/p/pharo/wiki/Packages for the >> list of packages we want in Pharo. >> >> You can download the image at >> https://gforge.inria.fr/frs/download.php/7404/pharo0.1-10074dev08.09.2.zip >> . >> >> Here is the list of packages included in this release: >> >> AST version 160 >> eCompletion version 0.90 >> OB-Enhancements version 0.203 >> OmniBrowser version 0.416 >> OmniBrowser-Morphic version 0.52 >> OmniBrowser-Refactory version 1.0 >> OmniBrowser-Regex version 0.9.1 >> OmniBrowser-Standard version 0.346 >> Pharo packages version 0.3 >> Refactoring Core version 5 >> RoelTyper version 0.60 >> Shout version 3.15-tween.72 >> ShoutWorkspace version 1-tween.2 >> VBRegex version 1.9 >> >> -- >> Damien Cassou >> Peter von der AhÃ©: Â«I'm beginning to see why Gilad wished us good >> luckÂ». (http://blogs.sun.com/ahe/entry/override_snafu) >> _______________________________________________ >> Pharo-project mailing list >> Pharo-project(a)lists.gforge.inria.fr >> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project > > > _______________________________________________ > Pharo-project mailing list > Pharo-project(a)lists.gforge.inria.fr > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project _______________________________________________ Pharo-project mailing list Pharo-project(a)lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project

1 0

Re: [Pharo-project] Product Brand Pharo
by Bill Schwab Sept. 23, 2008

Sept. 23, 2008

Markus, Based on the discussion, I started noting how long it takes my images to load; a splash screen might be in order; I also fear that by the time VM loads and the GUI is able to draw itself, it is no longer needed :( Still, I'd add it to the milestones, but probably late in the process. About help, I am not very sympathetic on the "no documentation" front. Someone seeking to join us will quickly find tutorials, and can post questions. Dolphin's MVP framework almost invariably had the newbie expressing pain and confusion, followed by "why was this so hard last week???" Conversely, the MS Office IDL for Automation has LOTS of "help" - almost all of it useless. The message (pun intended) needs to be "this is not trivial to learn, but very worth your time." Ensuring that good tutorials are linked from our Wiki would be a nice gesture. Wizards in the IDE - that's a very good idea, and Dolphin helps it along through the IDE extension framework, which is little more than events triggered at the right time, off of each tool's class and giving the needed instances for the code to make the desired changes. Dolphin's view resources, and our lack of them, would require some adaptation. However we do it, it would be a nice addition. The idea is to do _something_ that allows each of us to provide hooks into the IDE w/o having to alter the code. That way, building a new image and updating an existing one will not destroy the extensions. Bill 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) 273-6785 FAX: (352) 392-7029

2 1

Re: [Pharo-project] Halt once; workspace(?) menu
by Bill Schwab Sept. 23, 2008

Sept. 23, 2008

Stef, The class can be more flexible than what is in Pharo, and FWIW, is not built into the system. Since it does not need to be built-in, that's a plus in my thinking. Not much of a difference, I must admit. Bill 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) 273-6785 FAX: (352) 392-7029

1 0

[Pharo-project] UTF support in Linux
by StÃ©phane Ducasse Sept. 23, 2008

Sept. 23, 2008

Begin forwarded message: > From: Bert Freudenberg <bert(a)freudenbergs.de> > Date: July 9, 2008 12:20:52 AM CEDT > To: "A friendly place to get answers to even the most basic > questions about Squeak." <beginners(a)lists.squeakfoundation.org> > Subject: Re: [Newbies] UTF support in Linux? > Reply-To: "A friendly place to get answers to even the most basic > questions about Squeak." <beginners(a)lists.squeakfoundation.org> > > > Am 08.07.2008 um 23:09 schrieb smith02243: > >> >> Hi all, >> >> In the Linux VM, when I try to input, via paste or the right alt >> combination, non-ascii characters, like 'Ã¶' or 'Ã©' for example, >> Squeak >> merely shows 'o?' or 'e?' I researched into this and accordingly >> chose a >> font that supports UTF-8 and set the locale and VM command line >> options >> appropriately: >> >> export LC_CTYPE="en_IE.UTF-8" >> locale-gen en_IE.UTF-8 >> (...which claims that locale is upto date...) >> sudo dpkg-reconfigure -plow locales >> (...which outputs en_IE.UTF-8... done among other things....) >> squeak -encoding UTF-8 -textenc UTF-8 -pathenc UTF-8 >> Squeak3.10.2-7179-basic.image >> >> Now when I try to input the aforementioned characters I encounter a >> debug >> message: "Error subscript is out of bound." >> >> However, I have set Squeak's language to Spanish and then German, >> and the >> characters display perfectly well in the menus. Although, the same >> cannot be >> said for the any files with non-ascii characters in the file >> dialogue. So, >> it seems it's an input problem. >> >> Anyone with any ideas? Thanks. > > This has only been addressed recently, and may not have found its > way into a release yet. > > You need the very latest Squeak VM (3.10-1 at least). It supports > Unicode keyboard events (which are passed to the image as UTF32). > > To make use of these events your image needs to be using > UTF32InputInterpreter. Evaluate > > ActiveHand keyboardInterpreter > > to see the currently used one. It can be changed by patching > Latin1Environment class>>inputInterpreterClass. > > I don't know if this is in any official release yet, or scheduled > for one. > > Some relevant change sets from the Etoys image seem to be > > http://tinlizzie.org/updates/etoys/updates/1713UnicodeInput1-yo.cs > http://tinlizzie.org/updates/etoys/updates/1714UnicodeInput2-yo.cs > http://tinlizzie.org/updates/etoys/updates/1925unixLatin1Input-yo.cs > http://tinlizzie.org/updates/etoys/updates/1935unicodeFallback-bf.cs > > - Bert - > > > _______________________________________________ > Beginners mailing list > Beginners(a)lists.squeakfoundation.org > http://lists.squeakfoundation.org/mailman/listinfo/beginners >

4 8

[Pharo-project] Cannot load PolyMorph/UIEnhancements
by Norbert Hartl Sept. 23, 2008

Sept. 23, 2008

I tried to load the Polymorph package in a fresh pharo image and I didn't succeed. I tried pharo 10044, 10070 and 10074 image. And I tried loading from universes as well as via monticello. The result is the same: When the Loading progress bar appears it freezes after 10% of the bar is full. It doesn't matter if I load it with 64bit or 32bit (binary) vm. I had no trouble loading the same into a 3.10.2 squeak image. How do I debug the action while it is showing the progress bar? thanks in advance, Norbert

2 1

Sept. 23, 2008

http://interstices.info/jcms/c_39747/programmer-un-jeu-d-enfant There is also a podcast http://interstices.info/jcms/c_34709/podcast-interstices Stef

1 0

[Pharo-project] Product Brand Pharo
by Torsten Bergmann Sept. 23, 2008

Sept. 23, 2008

Markus wrote >One question: is the splash-screen really needed? Adrien wrote: >I don't see the point of a splash screen. I completely agree that it is useless in the sense of loading screen showing the progress for the standard pharo image. Maybe it makes more sense for the "slower" TrueType image where a progress bar currently shows the font updates. But this is not possible since the Squeak VM does not (yet) support an updating splash progress in the native splash window driven by Smalltalk code. But a splash can also be done "image side only" anyway. But that's not the point. I suggested to provide a splash screen more from the branding/marketing point of view - even if it is technically useless. At least I want to know what others think about it. A splash screen typically shows the Logo and some version information and people rember/recognize/identify the software through that. Some applications have an option to disable the splash or show it only on the first startup. Eclipse for instance uses a slightly modified splash screen theme for each yearly full release (Europa, Ganymede, ...) - nice idea since the user knows which one is started when the disk is full with different versions. Sure we should drop the splash screen from the technical point of view - but it would be a nice idea to have the welcome screen of each Pharo release with a slightly modified theme to see if it is a core, developer image, ... We dont have to clone native apps/Java IDE's - but at least we can learn a little bit about branding/marketing and how good design can attract more people. Take another example - will you find a Smalltalk IDE with good looking Wizards these days? >From the Smalltalkers point of view (knowing about #subclass: messages) it's technically useless and will bloat the image - but I've seen many newbies blaming Smalltalk to be that complicated the first time they just click around and search for the Class/Package wizard. Squeak is also not very supportive if one needs help - there is a help menu but no real help/documentation. That does not mean we should have the splash screen or spend the whole day implementing Wizards - but at least we should discuss and think about what the first impression for others/non Smalltalkers will be and how we can help them to take first hurdles and enjoy the power of Smalltalk... Bye Torsten -- GMX startet ShortView.de. Hier findest Du Leute mit Deinen Interessen! Jetzt dabei sein: http://www.shortview.de/wasistshortview.php?mc=sv_ext_mf@gmx

1 0

[Pharo-project] Halt once; workspace(?) menu
by Bill Schwab Sept. 23, 2008

Sept. 23, 2008

A couple of quick thoughts based on the Polymorph/TT image: (1) Halt-once looks like it is built into the image as a mode. It is a VERY useful concept, but Dolphin has long had a goodie that provides the feature as a class, used something along the lines of the following: abandonHopeAllYeWhoRaiseEndlessWalkbacksHere self someUnfortunateConditionDetected ifTrue:[ Once do:[ self halt. ]. " Once reset. " ]. Any takers? I did not write it, but I think we will find it licensed in a friendly way, or it could be easily reconstructed, if not improved by having registered instances, allowing independent conditional breakpoints, and all in code, or as I like to say "done in the language, not TO the language." (2) The extended search sub-menu. At the really picky level (ignore if there is ANY problem with this), it might be nicer to pop out a submenu vs. jumping to another menu. Bill 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) 273-6785 FAX: (352) 392-7029

2 1

Re: [Pharo-project] Re: Re: [ANN] Demo with both Truetype +Polymorph
by Bill Schwab Sept. 22, 2008

Sept. 22, 2008

Marcus, I'm fairly certain the problem is in Redmond. It would not be the first thing they botched. I'm just annoyed with myself for not fingering them sooner. By all means though, newbies will encounter it, so protecting Windows from itself is probably a good move. Is Pharo going to adopt the one-click structure? On Windows, I do not much care. The VM "installs" trivially, and shortcuts are cheap. Shortcuts are not much more expensive on Linux, or maybe I haven't learned enough about how to skip steps on Linux. The VM is a little more trouble on Linux, but nothing I can't handle, and I'll gain familiarity as I go. All that said, since there is something that is brain-dead simple on the main platforms, I'm not too proud to use it ;) I do not much care whether we use it or not. If the plan is to do so, I can save myself some hassle sooner rather than later, but whatever you guys want to do is fine by me. Thanks! Bill 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) 273-6785 FAX: (352) 392-7029 >>> denker(a)iam.unibe.ch 09/22/08 3:56 PM >>> On 22.09.2008, at 20:57, Andrew Tween wrote: > > "Bill Schwab" <BSchwab(a)anest.ufl.edu> wrote in message news:s8d7aeb8.028@AN9.ANEST.UFL.EDU > ... >> Got it. My basic algorithm for finding problems usually starts with >> "what could Microsoft have done to screw this up?" I faltered. In >> this >> case, MS could have made a mockery of the file names in the >> archive, and >> that is apparently just what they did. I do not care for 7-Zip's >> user >> interface, but the engine works. It magically understood the archive >> (what's that word again, standards???<g>), unzipped all the files, >> and >> now everything works. Very impressive BTW. > > That rings a bell. > I had the same problem with the early Sophie releases. > The Windows unzipper tried to create a folder and a file with the > same name. > > I reported the problem to the Sophie team, and with the most recent > releases of Sophie, Windows can unzip it ok. > Might be worth finding out how they got around the problem. > I think the problem is the Mac: the zip creates hidden files (starting with a dot, I think) for storing the resource fork of the old macos filesystem. I should create the zip differently... Marcus -- Marcus Denker -- denker(a)iam.unibe.ch http://www.iam.unibe.ch/~denker _______________________________________________ Pharo-project mailing list Pharo-project(a)lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project

2 1

Re: [Pharo-project] Re: Re: [ANN] Demo with both Truetype +Polymorph
by Bill Schwab Sept. 22, 2008

Sept. 22, 2008

Marcus, I'm fairly certain the problem is in Redmond. It would not be the first thing they botched. I'm just annoyed with myself for not fingering them sooner. By all means though, newbies will encounter it, so protecting Windows from itself is probably a good move. Is Pharo going to adopt the one-click structure? On Windows, I do not much care. The VM "installs" trivially, and shortcuts are cheap. Shortcuts are not much more expensive on Linux, or maybe I haven't learned enough about how to skip steps on Linux. The VM is a little more trouble on Linux, but nothing I can't handle, and I'll gain familiarity as I go. All that said, since there is something that is brain-dead simple on the main platforms, I'm not too proud to use it ;) I do not much care whether we use it or not. If the plan is to do so, I can save myself some hassle sooner rather than later, but whatever you guys want to do is fine by me. Thanks! Bill 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) 273-6785 FAX: (352) 392-7029 >>> denker(a)iam.unibe.ch 09/22/08 3:56 PM >>> On 22.09.2008, at 20:57, Andrew Tween wrote: > > "Bill Schwab" <BSchwab(a)anest.ufl.edu> wrote in message news:s8d7aeb8.028@AN9.ANEST.UFL.EDU > ... >> Got it. My basic algorithm for finding problems usually starts with >> "what could Microsoft have done to screw this up?" I faltered. In >> this >> case, MS could have made a mockery of the file names in the >> archive, and >> that is apparently just what they did. I do not care for 7-Zip's >> user >> interface, but the engine works. It magically understood the archive >> (what's that word again, standards???<g>), unzipped all the files, >> and >> now everything works. Very impressive BTW. > > That rings a bell. > I had the same problem with the early Sophie releases. > The Windows unzipper tried to create a folder and a file with the > same name. > > I reported the problem to the Sophie team, and with the most recent > releases of Sophie, Windows can unzip it ok. > Might be worth finding out how they got around the problem. > I think the problem is the Mac: the zip creates hidden files (starting with a dot, I think) for storing the resource fork of the old macos filesystem. I should create the zip differently... Marcus -- Marcus Denker -- denker(a)iam.unibe.ch http://www.iam.unibe.ch/~denker _______________________________________________ Pharo-project mailing list Pharo-project(a)lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project

1 0

Re: [Pharo-project] Re: [ANN] Demo with both Truetype + Polymorph
by Bill Schwab Sept. 22, 2008

Sept. 22, 2008

Got it. My basic algorithm for finding problems usually starts with "what could Microsoft have done to screw this up?" I faltered. In this case, MS could have made a mockery of the file names in the archive, and that is apparently just what they did. I do not care for 7-Zip's user interface, but the engine works. It magically understood the archive (what's that word again, standards???<g>), unzipped all the files, and now everything works. Very impressive BTW. Dumb question: to copy an image/changes between machines, would I simply copy the image and change files (a few levels down in Resources), and leave the rest of the "one click image" alone? Bill 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) 273-6785 FAX: (352) 392-7029

4 3

Sept. 22, 2008

Hi I would love if we could slowly deprecate caseOf: Stef

5 9

[Pharo-project] Product Brand Pharo
by Torsten Bergmann Sept. 22, 2008

Sept. 22, 2008

Beginners always lament on the default look of Squeak. Hopefully this will change with Pharo and Truetype/Polymorph. Some comments on the pharo one click image: I dont know who builds the custom VM for Pharo - but the icon is the one from Seaside and should be replaced. There are various parameters in the *.ini file to product brand a squeak VM based application therefore I would also recommend to add the folling lines into the Pharo.ini file: WindowTitle=Pharo This sets the window title to Pharo instead of Squeak and QuitDialogMessage="Quit Pharo without saving?" QuitDialogLabel=Pharo adapts the quit message box of the VM. At least on Win32. See http://www.squeakvm.org/win32/settings.html It would also be good to have a splash screen with the logo or some decent looking graphics/logo combination. The splash screen typically is a *.bmp file located in the same directory as pharo.exe and the default name is splash.bmp. You can set up another name again in the pharo.ini file: SplashScreen="pharo.bmp" SplashTitle=Pharo SplashTime=2000 Attached is a modified Pharo.ini and a sample splash file. Just put it in the same directory as the EXE file for the one click install. Do we have access to professinal designer for a good looking splash/website/icons? Are there any plans regarding this? Thanks Torsten -- Psssst! Schon vom neuen GMX MultiMessenger gehÃ¶rt? Der kann`s mit allen: http://www.gmx.net/de/go/multimessenger

5 4

Sept. 22, 2008

Hi I was checking the clean up of the menus. This is good but the changes...items are really missing to me. Especially the recover from .... We certainly do not need all of the changes related items but some are really important. Then from the windows menu item I would keep and remove Delete both of the above. Stef

3 4

Re: [Pharo-project] Re: [ANN] Demo with both Truetype + Polymorph
by Bill Schwab Sept. 22, 2008

Sept. 22, 2008

Igor, Ah, but you miss the point, wise and kind sir :) Dumbing down users makes them less able to pirate software. That aside, assuming I unzip the files into a folder of my chosing[*] and run Pharo.exe, what should happen? I will unzip it elsewhere to see what I find. One of my attempts lead to "file not found." It looks too correct to be corrupt download, so I'm thinking MS is tripping up over file names. Bill [*] Seriously, each update cycle, this particular XP box gets weirder and weirder. 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) 273-6785 FAX: (352) 392-7029 >>> siguctua(a)gmail.com 9/22/2008 12:28:05 PM >>> 2008/9/22 Bill Schwab <BSchwab(a)anest.ufl.edu>: > I appeared to get a valid download, but something isn't right. Is the > idea to download, unzip (MS seems to make this flakier with every > update???), and then launch Pharo.exe? Are there separate image and > change files? I do not see them, just a LOT of folders/files that are > giving my Java flashbacks. > MS people invented term of 'installing software' , while in most cases it can be done by simply copying files. And one of the results of such 'invention' that now people simply can't make simplest things like unpacking acrhive to choosen folder and run the program by themselves - they think in a terms of 'icon on desktop' or 'icon in the launch menu'.. A progress. What i can say. >From this POV, the distribution is still far away from the 'one-click', because , in my vision, a single click should be in browser, and after this click you should have everything up and running on your PC. :) > My assumption was that this is a prepared image with Polymorph and TT > loaded, and on seeing the structure, I figured it was a Sophie-like > distribution. Are my expectations skewed? Please let me know if I am > missing something. Otherwise, I will try this again later. > > Bill > > > > 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) 273-6785 > FAX: (352) 392-7029 > > > _______________________________________________ > Pharo-project mailing list > Pharo-project(a)lists.gforge.inria.fr > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project > -- Best regards, Igor Stasenko AKA sig. _______________________________________________ Pharo-project mailing list Pharo-project(a)lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project

3 2

Re: [Pharo-project] Re: [ANN] Demo with both Truetype + Polymorph
by Bill Schwab Sept. 22, 2008

Sept. 22, 2008

I appeared to get a valid download, but something isn't right. Is the idea to download, unzip (MS seems to make this flakier with every update???), and then launch Pharo.exe? Are there separate image and change files? I do not see them, just a LOT of folders/files that are giving my Java flashbacks. My assumption was that this is a prepared image with Polymorph and TT loaded, and on seeing the structure, I figured it was a Sophie-like distribution. Are my expectations skewed? Please let me know if I am missing something. Otherwise, I will try this again later. Bill 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) 273-6785 FAX: (352) 392-7029

3 2

[Pharo-project] Default windows sizes
by Adrian Lienhard Sept. 22, 2008

Sept. 22, 2008

The following changeset adjusts the default size of windows, e.g., it makes code browser, the inspector and the process browser larger. The senders/implementors windows are increased in their height and the ratio between the list height and code pane height is changed. Let me know if you prefer other settings. Adrian ___________________ http://www.adrian-lienhard.ch/

2 1

[Pharo-project] [update] 10074
by Marcus Denker Sept. 22, 2008

Sept. 22, 2008

update10074 ------------ - Issue 180: 10073 still needs some cleanup - Issue 166: Cleaning up the menus - Issue 182: Cleaning leftover from nebraska - Issue 184: delete PseudoPoolVariable - Issue 188: start of removing Vocabulary - Issue 187: delete Button, OneOnSwitch, ST80MenusTest, Switch - Issue 183: delete ClickExerciser, TestClickExerciser, DumberMenuMorph, PluggableFileList - Issue 181: Remove codepane+tilescript from browser - Issue 178: unsent messages - remove hashMappedBy: and friends (never called). - remove even more methods that have no senders (mostly etoys related) - Classes removed: RemoteHandMorph PseudoPoolVariable ElementTranslation MethodCall UnknownType Button OneOnSwitch ST80MenusTest Switch DumberMenuMorph, ClickExerciser, ClickExerciserTest, PluggableFileList - Postscript: SMSqueakMap default purge. Smalltalk cleanOutUndeclared. CustomEventsRegistry at: #scrolledIntoView put: IdentityDictionary new. CustomEventsRegistry at: #scrolledOutOfView put: IdentityDictionary new. Set rehashAllSets. -- Marcus Denker -- denker(a)iam.unibe.ch http://www.iam.unibe.ch/~denker

4 6

Sept. 22, 2008

Hi all I was ready to remove the novice mode and I think that we should remove DockingBar and TheWorldMainDockingBar Now gary it may introduce conflicts with etoy removal. What do you think? When will you send us the etoy removal? Stef

4 15

[Pharo-project] [ANN] prebuilt 10074
by Marcus Denker Sept. 21, 2008

Sept. 21, 2008

Hi, A prebuilt 10074 is availabe: Pharo-Core: https://gforge.inria.fr/frs/download.php/7265/Pharo0.1-10074-2008-09-21.zip Pharo: https://gforge.inria.fr/frs/download.php/7266/Pharo0.1-10074-oneclick.zip Marcus -- Marcus Denker -- denker(a)iam.unibe.ch http://www.iam.unibe.ch/~denker

1 0

Sept. 21, 2008

marcus should we close http://code.google.com/p/pharo/issues/detail?id=134&colspec=ID%20Type%20Sta…

2 1

[Pharo-project] Interesting post about Smalltalk and the effect of success (sorry for those who already saw this on Squeak Dev)
by Simon Kirk Sept. 21, 2008

Sept. 21, 2008

Hi all - as I said in my subject, I've already posted this on Squeak- Dev, but I thought those on here who don't subscribe to that list might be interested, too: http://www.robvens.nl/en/blog/The-Smalltalk-Trap.html Cheers, Simon ****************************************************************************************************************************************** This email is from Pinesoft Limited. Its contents are confidential to the intended recipient(s) at the email address(es) to which it has been addressed. It may not be disclosed to or used by anyone other than the addressee(s), nor may it be copied in anyway. If received in error, please contact the sender, then delete it from your system. Although this email and attachments are believed to be free of virus, or any other defect which might affect any computer or IT system into which they are received and opened, it is the responsibility of the recipient to ensure that they are virus free and no responsibility is accepted by Pinesoft for any loss or damage arising in any way from receipt or use thereof. ******************************************************************************************************************************************* Pinesoft Limited are registered in England, Registered number: 2914825. Registered office: 266-268 High Street, Waltham Cross, Herts, EN8 7EA

2 5

[Pharo-project] [ANN] first one-click image
by Marcus Denker Sept. 21, 2008

Sept. 21, 2008

Hi, There is now a first one-click image. This uses the packaging invented by Impara for Plopp, which later was used by Sophie and then even Seaside: https://gforge.inria.fr/frs/download.php/7078/Pharo0.1-10072-oneclick.zip on the Mac, this is an executable. On windows, this is a directory with an .exe, for linux, there is a pharo.sh. This has only been tested on Mac and Linux right now. Notes: => COPYRIGHT needs to be fixed => needs to be tested on Windows => more testing on Linux => fix icon for windows => Icon is not yet final => we need to add the truetype plugins / recompiled VM The idea behing this is that it should be very simple to get it running for new people. Those with lots of knowledge then can take the package appart and use it like they want. The idea is that the one-click will have more tools later (syntax highlighting, Refactoring, things like that). In addtion there will be a .zip with a "core" image that has the bare minimum for deployment. The "download" link of the pharo website will point to the one-click destribution. Marcus -- Marcus Denker -- denker(a)iam.unibe.ch http://www.iam.unibe.ch/~denker

7 19

Sept. 21, 2008

it says Distribution based on Sophie http://www.sophieproject.org/ Seaside http://www.seaside.st Stef

6 7

Sept. 20, 2008

looking at issue http://code.google.com/p/pharo/issues/detail?id=178&colspec=ID%20Type%20Sta… I saw that allUnsentMessages implementation is wrong it should be. Behavior>> allUnsentMessages "Answer an array of all the messages defined by the receiver that are not sent anywhere in the system." ^ SystemNavigation default allUnSentMessagesIn: self selectors Stef

1 0

[Pharo-project] More menu cleanup
by Adrian Lienhard Sept. 20, 2008

Sept. 20, 2008

Here is a changeset that cleans up and restructures the code context menu. In the process of removing stuff I also removed around 30 methods, some of which were huge (ever seen the implementation of the "explain" functionality?). Cheers, Adrian BTW: After loading the changeset save and restart to get all icons in the menu. ___________________ http://www.adrian-lienhard.ch/

2 2

Fwd: [Pharo-project] [ANN] first one-click image
by David Mitchell Sept. 19, 2008

Sept. 19, 2008

(Forgot to copy the list) Yep, that fixed them all after I did a Load code updates. Did have one new error in testPrinting: self assert: nonSEarray2 printString = 'an Array(#Array)' But I have to say, I have never experienced a Squeak release where there was only one test error. (Of course, before the test runner was integrated, there were no errors, but that doesn't count!) Great work everyone. On Fri, Sep 19, 2008 at 11:06 AM, Marcus Denker <denker(a)iam.unibe.ch> wrote: > > On 19.09.2008, at 17:31, Norbert Hartl wrote: >>> >>> >> No need to. None of these seem to fail in 10073. The one-click is >> based on 10072. > > Sorry for that... I will do a new release soon. > > Marcus > > > -- > Marcus Denker -- denker(a)iam.unibe.ch > http://www.iam.unibe.ch/~denker > >

2 1

[Pharo-project] Image size
by Norbert Hartl Sept. 19, 2008

Sept. 19, 2008

Hi, I have problems in understanding the image size of pharo. I monitored it since we started and I did not recognize a notably change. So I took a pharo 10073 and squeak 3.10.2 for comparsion. Amount classes, amount methods and lines of code are all 10% less in pharo. But the image is approx. 0,5% - 1% smaller than the squeak one. To clarify this. I closed all windows and did condenseChanges and garbageCollect, flushed the monticello cache... What could be the reason for this? thanks, Norbert

4 10

[Pharo-project] [ANN] update 10073
by Marcus Denker Sept. 19, 2008

Sept. 19, 2008

update 10073 ------------ - Issue 44: 7094 FixClassCompatibilityForClassVar (M6472) - Issue 86: 7138 TransformationMorph-isSticky - Issue 89: 7141 PrintingArraysWithMetaclasses - Issue 42: 7092 TTCFontDNUfix (M5309) - Issue 176:failing tests in 70072 - delete ParagraphEditor>>#initializeTextEditorMenus (never called, has undeclared) - revert Issue 95: 7149 ChangeHooksTest-tearDown-M6702-rej (broke tests) - MorphExtension now stores properties in SmallIdentityDictionary. - in loadscript: Smalltalk associations do: [:each | each beReadWriteBindingAnnouncing: true] Because it recompiles all methods multiple times, the update takes a while to load. (this is to be able to remove ReadOnlyVariableBinding, which has never seen any use) Marcus -- Marcus Denker -- denker(a)iam.unibe.ch http://www.iam.unibe.ch/~denker

3 7

Sept. 19, 2008

> We have two important fixes: compiler fixes and trait that are in > Changesets or cannot be loaded (in the case of the compiler > with MC). > I suggest for now to add them as changesets > Then we should keep the Compiler package to evaluate if MC1.6 can > be used. > I asked matthew to see if he can load the package. > > I still believe that having atomic loading in pharo is important. > > > > Stef > > >> Issue 167: Please check the traits fixes >> http://code.google.com/p/pharo/issues/detail?id=167 >> >> Comment #2 by adrian.lienhard: >> additionally remove caching of marker methods and make condensing >> work >> looks good now >> >> how do we integrate this since it is a changeset? >> >> Attachments: >> TraitsSharingCM-3.cs 14.6 KB > > > > > > >

3 6

Sept. 19, 2008

Hi I took 10072 and pressed update. After 10 or 15 min I got DoIt ^ server aborted early So may be the server stopped in the middle. Stef

2 3

[Pharo-project] Non-pharo reports in pharo bugtracker...
by Marcus Denker Sept. 18, 2008

Sept. 18, 2008

Hi, There are reports in the pharo tracker that are not solvable because the software the bug is reported about is not part of pharo. I wonder if that makes sense... from the bug-fixing of Squeak, I have found that it's very important to keep the numner of open issues as small as possible. (the nearly 1000 open reports they gatheres when doing 3.10 in Squeak is completely unmanagable, for example). I would suggest to put these reports on a wiki page, instead of the bugstracker. Or report them to the developers of the package. Marcus -- Marcus Denker -- denker(a)iam.unibe.ch http://www.iam.unibe.ch/~denker

4 6

[Pharo-project] Bugs on bugs.squeak.org
by Marcus Denker Sept. 18, 2008

Sept. 18, 2008

Hi, bugs.squeak.org is a mess. (and people wonder why I always stress that doing harvesting was a lot of work... we got the tracker down to 250 entries with 3.9, if I remember correctly). Even for 3.11, they will have some real work ahead to get this in some form of managable shape. Now for Pharo, I don't see that we can benefit from the bug-tracker at all anymore: there is so much stuff, no way to do here anything in a strategic, "all covering" way, as we did for example when moving from BFAV to Mantis. So if someone has reported a bug against Squeak, the only way I see this change ending up in pharo is that the bug is reported again on the pharo tracker. So if you want something to be fixed: http://code.google.com/p/pharo/issues/list (down to 79 open issues, we already fixed and closed 96 reports) Marcus -- Marcus Denker -- denker(a)iam.unibe.ch http://www.iam.unibe.ch/~denker

2 1

[Pharo-project] Re: Monticello 1.5
by Hilaire Fernandes Sept. 17, 2008

Sept. 17, 2008

Oh, Chris I am just a dumb developer and I don't know much about the tools, I just know it is a pain to get MC 1.5 working. I am cc to Pharo firends with a small recall of the situation, may be they have some tips Chris is working on new beta version of Magma which I am testing with him. As far as I can understand, Chris use Monticello configuration to be able to load in one click several Magma mcz packages. This requires Monticello 1.5, but this ones does not work equally on the different version of Squeak nor with Pharo. Which tools could be suggested to Chris ? (I can give more details if needed but really this is becoming insane --and unproductive--, and I really hope Pharo people will do it right) Hilaire 2008/6/27 Chris Muller <ma.chris.m(a)gmail.com>: > Well of course I am always open to suggestions. Everyone knows what > the requirement is: one-click loading of an entire (multi-package) > system. Do you know another way I can accomplish this other than > Monticello? What is everyone else doing? > > > > > > On Thu, Jun 26, 2008 at 3:09 PM, Hilaire Fernandes <hilaire(a)ofset.org> wrote: >> Chris, >> >> As MC seems to be a source of frustration, I am wondering why do you >> need it for the beta release of Magma? >> >> Now I really think I want to move in the future to Pharo as the >> development plans are clearer (https://gforge.inria.fr/projects/pharo) >> Yet Seasider are participating to Pharo so I am confident. >> >> I will first try Magma beta and my code as well. >> >> Hilaire >> >> >> -- >> http://blog.ofset.org/hilaire >> > -- http://blog.ofset.org/hilaire

10 27

[Pharo-project] Call for ideas for a Continuous Integration server
by Alexandre Bergel Sept. 16, 2008

Sept. 16, 2008

Dear List, Here are the notes I took on AmbientTalk. In Ambientalk: Table: >def table[5] { z := z + 1 } >>[1, 2, 3, 4, 5] Table splicing: >[1,@[2,3],4] >>[1, 2, 3, 4] >def [first, @rest] := [1,2,3,4] >>[1, 2, 3, 4] >rest >>[2, 3, 4] Function >def square (x) { x*x } >>nil >square(5) >>25 There is no check about fulfilling required methods. There is a way to declare a method as required. Required method are part of a structural type. Reflection is then used to ensure that a composition makes sense at composition time. Aliasing changes the name on the surface (it is not a rename), and does not create a new entry. The import clause (which is used to use a trait), may accept any expression, and not only a trait name. There is no precedence of the locally defined methods. Note that traits are composed when object composer are instantiated. Ambientalk may use the same trick than in Smalltalk to declare a method as required (self requiredMethod), but the lack of precedency is problematic. The fact that you do not have precedency, implies that a method used in a trait needs to be specify as required (using the reflective trick mentioned above) in case of the composer is likely to not provide an implementation for this method. Else, better is not mark the used method as required since the composer will have to exclude it when using the trait. Two methods that are acquired using different import path are not conflicting since they are the same compiled method. In Ambientalk, this will always cause a conflict since methods are âduplicatedâ between composers. In Ambientalk, deep equality (i.e., comparing bytecode instructions) may not lead to what we want since methods can be nested. The Morphic framework. Inspired from the Self version and has been influenced by John Maloneyâs work. Requirement between traits is specified with for exemple require: [Trait1, Trait2], in that case, all public methods defined in Trait1 and Trait2 will be set as required method for the trait that made the require. This framework is composed of 18 traits -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.

2 1