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

Re: [Pharo-project] about ensuring file closing
by Bill Schwab July 20, 2008

July 20, 2008

Stef, The streams are a mess, and the more I work with them, the more I understand the desire to replace them with Nile. For now, I am simply inserting some methods that behave well (work when they should, raise errors when they should) to allow me to begin doing some real work. My view is that I can always add helper classes and methods as needed. Whether that gets around fragile design (#nextOne) or is a one-place fix for the myriad times I open a file, it can always be rewritten to work with whatever we create. URIs make sense. I have never been happy with the ways I split and combine paths, so those things will have to be rewritten, especially since they are Windows-specific. 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) 846-1285 FAX: (352) 392-7029 >>> stephane.ducasse(a)inria.fr 07/19/08 5:49 AM >>> On Jul 18, 2008, at 5:01 PM, Bill Schwab wrote: > Stef, > > I do not doubt that the image is littered with such problems. On > several occasions, I lost 10-15 minutes tracking down errors that > resulted from unopened files. It is all a continuation of my streams > "rant" (a well-intentioned one, of course). > > Meaning no disrespect to the milestones (in particular, streams being > slated for the future), I have undertaken an effort similar to what > you > propose: putting an end to silent failures and lack of protection. > As a > streams addict, I need stable semantics, and have thus embarked on my > #nextOne, #nextMany:, etc. solution to reaching that goal. It also > became clear that FileStream was quite unfriendly, so I started adding > Dolphin-compatible methods, at least to a point. My primary goal is > to > have errors arise when appropriate vs. waiting for > potentially/apparently disconnected collateral damage to occur. I have the impression that FileStream should be patched but also just rewritten from scratch. I was discussing with mike about the use of URI in sophie. He told me that they got a really nice solution which was offering cross platform real portability > > Dolphin's semantics will not be completely appropriate for Squeak, > because there are OS (in)dependencies to consider. I want that to > work > well so I can treat my Redmond allergy. > > I urge Smalltalkers to avoid #next and #next: like the plague, but one > could always use my changes and ignore the replacements. A very > incomplete but gradually evolving change set is available on request. ok good to know. _______________________________________________ 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] Improving the GUI
by Bill Schwab July 20, 2008

July 20, 2008

Agreed. As an example, I would like to see the browse-it command grouped with do/inspect/explore. It also is not quite as flexible as Dolphin's verion - it should first evaluate the selected expression or current line and the browse the instance class of the result. I am a bit strange on fonts. For programming, I like them mono-spaced and a bit small. No guessing between letters and numbers, punctuation does not get kerned out of existence, etc. With those improvements, I can tolerate a slightly smaller font size than with kerned fonts. Regardless, options are good, especially when they are easy to change to one's personal preference. I would like to see better control over background colors. Editing a couple of methods makes me fairly happy with the results, but it should be preference based. I am happy to use code to set things, as that allows me to build a script to get what I want. 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) 846-1285 FAX: (352) 392-7029 >>> adi(a)netstyle.ch 07/20/08 5:22 AM >>> Hi, I'd like to get your opinion about how we could improve the GUI. I believe that even a few small changes will make a significant difference, especially for new people that come to Pharo. Just the cleaned up world menu is such a relief... Here a three suggestions: Better fonts ---------------- I think the current code font is not a good one and it is too small (for today's screen resolutions). I like for example the font used in TextMate, which is Monaco 12, a monotype font. The page http://www.lowing.org/fonts/ gives some samples for good code fonts. I've tried out the other fonts in the Pharo image like the BitstreamVeraSansMono but they are all very blurry. Just not good enough. I'm not a font expert, so I don't know what the options for a better font are. Better menus ------------------------------- Current problems: - bad organization menus only work when you have learned where something is located - duplicated entries e.g., Appearance is both in the top level world menu and in the submenu Help - useless entries e.g., set font, set style etc. in the context menu are useless since code is never styled - defective entries: e.g., "Make project link" in the context menu does not work Larger windows ------------------------------- The default size of windows is usually too small. Often, after opening a window I have to resize it. For example the hierarchy browser shows only the 5 first classes. Any other ideas what could be improved in the short term? I'm looking for stuff that we can do now without having to rewrite the whole system... Cheers, Adrian ___________________ http://www.adrian-lienhard.ch/ _______________________________________________ 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] [ANN] New release is out 10050
by StÃ©phane Ducasse July 19, 2008

July 19, 2008

Hello Pharoers [ANN] New release is out 10050 update10047 Fixing the release. Adrian made a few cosmetic changes and tagged testUnimplementedNonPrimitiveCalls as expected failure Original description: - added allClassesWithUnimplementedCalls to SystemNavigation to return structured classes with unimplemented calls - added convenience method PackageOrganizer>>allPackagesContainingUnimplementedCalls - adjusted ReleaseTest>>testUnimplemented... to exclude TestCase classes ------------------------------------------------------ update10048 Fix the Scriptloader logStream + Collection tests http://bugs.squeak.org/view.php?id=6355 + Pharo Flaps ------------------------------------------------------ update10049 7073 CollectionAndCollectionTest * [http://bugs.squeak.org/view.php?id=6367 M6367] * seems good [http://bugs.squeak.org/view.php?id=5700 M5700] Add 2 tests for RunArray class + at:put: See : http://bugs.squeak.org/view.php?id=6070 ------------------------------------------------------ update10050 This update includes a set of Mantis fixes (that also went into Squeak 3.10): http://code.google.com/p/pharo/issues/detail?id=27 http://code.google.com/p/pharo/issues/detail?id=31 http://code.google.com/p/pharo/issues/detail?id=32 http://code.google.com/p/pharo/issues/detail?id=30 mark testUnwindDebuggerWithStep as expected failure

1 0

[Pharo-project] Enough changes to release?
by Damien Cassou July 19, 2008

July 19, 2008

Hi, I don't know when you decide to release Pharo versions, but it might be interesting to do that now. The previous version is 2 week old and the changes take a long time to load. Bye -- 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] About HTMLFileStream
by StÃ©phane Ducasse July 19, 2008

July 19, 2008

Hello I was browsing and browsing code and I found HTMLFileStream Apparently positionableStream introduces command: just for this class. I have the impression that if we want a good htmlwriter we could extract it from seaside. But HTMLFileStream looks outdated. So I would propose to remove it. Any thoughts on that one? Stef

2 2

Re: [Pharo-project] about ensuring file closing
by Bill Schwab July 19, 2008

July 19, 2008

Stef, I do not doubt that the image is littered with such problems. On several occasions, I lost 10-15 minutes tracking down errors that resulted from unopened files. It is all a continuation of my streams "rant" (a well-intentioned one, of course). Meaning no disrespect to the milestones (in particular, streams being slated for the future), I have undertaken an effort similar to what you propose: putting an end to silent failures and lack of protection. As a streams addict, I need stable semantics, and have thus embarked on my #nextOne, #nextMany:, etc. solution to reaching that goal. It also became clear that FileStream was quite unfriendly, so I started adding Dolphin-compatible methods, at least to a point. My primary goal is to have errors arise when appropriate vs. waiting for potentially/apparently disconnected collateral damage to occur. Dolphin's semantics will not be completely appropriate for Squeak, because there are OS (in)dependencies to consider. I want that to work well so I can treat my Redmond allergy. I urge Smalltalkers to avoid #next and #next: like the plague, but one could always use my changes and ignore the replacements. A very incomplete but gradually evolving change set is available on request. 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) 846-1285 FAX: (352) 392-7029 >>> stephane.ducasse(a)inria.fr 07/18/08 3:40 AM >>> I was looking at the ensure: usage in the image. And I found this one (besides the third line :) and I was wondering if it would not make sense to check them all (ensure:, forceNewFileNamed:.....) because if forceNewFileNamed: returns nil the ensure will not work. writeEncoded: bytes | file | fileName ifNil:[^self]. false ifTrue:[^self]. file := FileStream forceNewFileNamed: fileName. [file nextPutAll: bytes] ensure:[file close]. I was also looking at readOnlyFileNamed: and it does not return nil readOnlyFileNamed: fileName "Open an existing file with the given name for reading." | fullName f | fullName := self fullName: fileName. f := self new open: fullName forWrite: false. ^ f isNil ifFalse: [f] ifTrue: ["File does not exist..." ((FileDoesNotExistException fileName: fullName) readOnly: true) signal]. "StandardFileStream readOnlyFileNamed: 'kjsd.txt' " I found some inconsistencies: For example BMPReadWriter>>nextPutImage: aForm as a stream close as the last messages while all the other methods do not and should be protected by an ensure: may be we could have a task for going over the implementation of fileStream and also the ensure: uses. Stef

2 1

[Pharo-project] A board for pharo
by StÃ©phane Ducasse July 19, 2008

July 19, 2008

Hello everybody We are happy to announce a board for Pharo! So far, Pharo has mainly been "directed" by Stef who initiated the project. Now that the community grows and Pharo makes progress, the board will bring some more structure into the decision making process. We don't have bylaws or any rules written down yet but the idea is to run it similar to the Apache board. That is, new board members are elected by existing board members. The rational is that we want to assure continuity to be able to steadily push Pharo forward. The current board members are: - Marcus Denker - Stefane Ducasse - Adrian Lienhard - Michael Rueger Stef

7 12

July 18, 2008

This Warning is signalled by methods which are deprecated. The use of Object>>#deprecatedExplanation: aString and Object>>#deprecated: aBlock explanation: aString is recommended. Idiom: Imagine I want to deprecate the message #foo. foo ^ 'foo' I can replace it with: foo self deprecatedExplanation: 'The method #foo was not good. Use Bar>>newFoo instead.' ^ 'foo' .... but deprecatedExplanation: aString "This method is OBSOLETE. Use #deprecated: instead." self deprecated: 'Use Object>>deprecated: instead of deprecatedExplanation:.'. Preferences showDeprecationWarnings ifTrue: [Deprecation signal: ('{1} has been deprecated. {2}' translated format: {thisContext sender printString. aString})] and deprecated: aBlock explanation: aString "This method is OBSOLETE. Use #deprecated:block: instead." self deprecated: 'Use Object>>deprecated:block: instead of deprecated:explanation:.'. Preferences showDeprecationWarnings ifTrue: [Deprecation signal: ('{1} has been deprecated. {2}' translated format: {thisContext sender printString. aString})]. ^ aBlock value.

1 0

July 18, 2008

Alex if you want to run the FixUnderscore and check if we can reload the package then we would integrate them. Stef

1 0

[Pharo-project] Classes in more than one package
by Alexandre Bergel July 18, 2008

July 18, 2008

I also noticed that there are classe that are contained in more than one package. For example, InstallerTest is contained in the package Installer and Installer-Test. Same thing for the class Installer. It is contained in the package Installer and Installer-core. I guess that packages Installer-Test and Installer-Core must be removed, since there are categories in the package Installer. Cheers, Alexandre -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.

6 12