Squeak-dev December 2001

squeak-dev@lists.squeakfoundation.org

137 participants
354 discussions

Query: moving stuff between projects to hide from publication
by Ross Boylan 30 Dec '01

30 Dec '01

I have a project I'm working on that I intermittently publish to BSS (see the Activity project if you're curious, but it's sparsely documented). It's a priority planning tool, and I now have a fair amount of data in it. I want to publish it without the data to keep a little privacy, and also to save space. How can I do that? I figure if I temporarily move things to another project I can avoid publishing them. I have a number of morphs I'd like to move. Is there a reasonable way to move them? I looked quickly in the Project and Morph protocols, and didn't see anything obvious. Perhaps I somehow get the Project's world and put the morph in there? There is data separate from the Morphs. I've put my data in the Smalltalk dictionary, and one of my classes accesses it. Is that enough to keep it out of being published? I assume if I just left it as a class variable it would get published (since the class itself is published).

3 6

Losing my latin
by ducasse 30 Dec '01

30 Dec '01

Hi I have the following strange behavior with the parse tree nodes. Context: I cannot find where a comment is stored in a returnNode or its child like the comment at the end of this method. testCommentAtTheEnd ^ 1 + 2 "a comment after a return" I looked into the resulting tree and all the comments are nil. Problem: I just took my swiss knife and put a self halt in the method that pretty print a return node. visitReturn: aNode value: val self stream nextPutAll: '^ '. self visitNode: val. self halt. self printComment: val. and I did not get at all the comment pretty printed. The only change I did was this halt. When I run it and press proceed. I lost the comment. When I remove the halt from the code I got back the comment! You can imagine that I retried several times. But apparently I'm redoing the same error. Can somebody tell me where I'm wrong??? Thanks Stef

1 0

Naive questions on cascade syntaxe
by ducasse 30 Dec '01

30 Dec '01

Hi I'm trying to come to the simplest set of rules to put () around Smalltalk expressions. I'm on the cascade. is it correct to say that a cascade is - either without () I mean outside of the cascade this happens when there is no message send to the cascaded expressions itself like in x := Set new add: 5 ; yourself - either there is one message not in the cascade and the cascade should be parenthesised. like in (x := Set new add: 5 ; yourself) class or (Set new add: 5 ; yourself) + 2 or (Set new add: 5 ; yourself) includes: 5 I read the ANSI standard but could not really found a clear description. Then inside the cascade you need () only if the receiver expression selector is a keyword and that you have keyword in the messages examples: Set new add: 5 ; add: 5 ; yourself (the formatter of VisualWorks put () around Set new but I guess that this is just to stress the receiver). Set new reset; reset ; reset with binary I do not see what example would not be silly (Set with: 5) add: 6 ; yourself

2 2

Re: [GOODIE] TestMaker 0.2 [CSOTD]
by danielv＠netvision.net.il 30 Dec '01

30 Dec '01

Sorry I was rude. I like your refactoring, it's cleaner than my code was, and creating the class is nice. Didn't understand default class and method name code. What would be a good example to use? Daniel "David T. Lewis" <lewis(a)mail.msen.com> wrote: > > --Boundary_(ID_ZZbbHv6BpKgt6aGqeqChuw) > Content-type: text/plain; charset=us-ascii > Content-transfer-encoding: 7BIT > Content-disposition: inline > > On Sun, Dec 23, 2001 at 12:23:14AM +0300, Daniel Vainsencher wrote: > > "David T. Lewis" <lewis(a)mail.msen.com> wrote: > > > On Fri, Dec 21, 2001 at 10:27:08PM +0300, Daniel Vainsencher wrote: > > > This is a very good idea, I like it. > > Hey David. Thanks. > > > > > As long as it is OK to propose things > > > that create work for other people, I have some suggestions ;) > > > > Hey - no fair - you want to make suggestions, you're supposed to also > > cough up some helpful code :-) In the spirit of Goran's brave commitment > > (Go CSOTD!), give us a test, or implement at least one of the > > suggestions... > > OK, you shamed me into it. Here is a CSOTD to add to TestMaker 0.2. It > attempts to provide reasonable defaults for test class names and selectors. > This implementation expects to see a class name as part of the expression > being printed, which would be a typical case for an expression in a > workspace, although not necessarily for an expression in a debugger > pane. > > Dave > > > --Boundary_(ID_ZZbbHv6BpKgt6aGqeqChuw) > Content-type: application/x-gunzip; NAME=TestMaker-0-2-changes-dtl.cs.gz > Content-transfer-encoding: base64 > Content-disposition: attachment; filename=TestMaker-0-2-changes-dtl.cs.gz > > H4sICPdGJjwAA1Rlc3RNYWtlci0wLTItY2hhbmdlcy1kdGwuY3MArVhbb9tGFn7mA//D2F5A > FGCrsuQkWC7chWvHQLFJXNRq98FQgBE5sqamhupwZFkL//h+Z4aXISUlKLyBgZCccz/fuYx6 > tzpfsvs/14I/jQfnPFstOMvnbPSB3YuVEcuZ0Gw0HJ6zh4wbURi2XqV4iNnJxfjiw5Tlio3G > 7EYkHik37CIeXcTjD2y17B2Fx9cLrh4FJJo4CCaQ8pk/CX02PBudKbFxp8VZarLwhmQHQdCV > GV6tzSLXdHTDn2XKJgP2SWxkEYaThSwY/ji7vr+b3DCTs5lgPE1FSs+1OjYcjMLwZwPz4NfK > FHS60jmkCaYFL3LFZ5lgqZjzdYbjea6ZdTnJeFEwxZcCWlTKCpGJxOS6GDCnfLnKxFIow41E > PMTLCsdWfCFgicfPeMFWXBsKsVkIItWiKIhrJqR6hD1SGZEOwvBz/gx1uRJsKeB6yuY2U0YT > 2YwnTwyUvnt44c85QpNka3ho6ZxmxzQIj4/CMDxqOJzg4hZxZT3raakqQTjIlR4rDF+ucIrc > sPPRD6PxDzbD5+N49B6ZtYSCJFoh/FdRIHSeW/im7mZ/IB4sDF5tOEsXnKovFBT6ek2W2rdk > kRdC2fcqF/blSqX3ZeDZaxh4ouJLSsm8op7UJ981ahAGh1R0hFbnn2uzCyqd2HMJwjyn4ssw > CKJbmWU/q8lC/JRx9YQYBIEWqLbCdEJeYapHJFJJI3l2pYqNgAeHTMS76be1yuIjoL1lcj7R > a1Tpw1dyYjroxstjOa1Alejm2fAZXlq+tbNk3fu+dw30/45jc6kLs6PysG/R/ZJnmeHZk1NI > 5GncxlUf/F9kBisqfjLoIbJJTnIFnchmzyGa9U7b3KesR7WLbsWK9cx5hRomqF3zQvy717fi > IPuWZ0VjXPmxUQiVPsArlDWFROLuSxUdF6bT0NfQkdPEgJsOI0y/3y5nebYLhD0p9HJYtR7k > ZE9EfvwRHxskOV/bGd5BUKl5N5dKWut8p5J8uZKZ8KW4/Mq5pPye0PcjdvR/amrDf8bnYzS1 > v9FGrvNUlIZRe0uaV3Qo761KMy9+4RqDYiEK+T+RutOWHISg0nTJ5pRpL0a9UkghNGUHFdro > mLY4DTh2GFOhtvvYoopPFv+RKr2bx+zjS4LxDyf7npQ21zeEMiqXS/uxEg0E2qNvpgvT7xll > cHjsnMdDGjsY7hMtxCfBMSMBDzzfzR3CaezagGbUS+BAGBzfZjTyFSYxUVLdVrSztYRtsy2O > 7hOulNCD4xBAbQlMIfAhFpTSIBJekK605tt+6FX4g0tQ2zzRNWkatjqFd0KsYJja+Posbw3b > aERhO4C+iXgxd9q9Ul4jlHXrY509tlarUoSRCfrNFpnu995m3Tg+H7Z2iXYLrDs6Mnkj5hIb > keu/tIplOXqJws5gV0gMjkxQAr9eu9ahmXjm2dourNE3Gm0ViloXZVXlRs639ruygwPKHu1c > Qa7DN/v8/qDPO9aQR72Kqh5AaIBIVOI15BDtFxpVIn7nWtI2axcV6Mc/NCQi3XuyyvPsRlqk > 4djjgHWPuaYCt/rPflPlJwIQkbwtDO/ii3dNxz24Y/EKm9gg8yehCrchuCmhxZJj6qjHiTvK > 5JPIttUuYccfMs82oMRIQbwKuz46YjRmtq9mo7IduImPZye8NqVva/pOp0KL1KtfkKN0Peug > oFSVCgMiaiWGvXrTesGLa39pmaLjfsHeTx3Xdhi2kWZBGagfqD103YaiKPpVcCprwZeM5pPT > 3MedbQJrG6v+xbb5WpPf7uyjwo0j2BO4am61j2xyo/a+0Qjv25uKteqTpC2iYyk0UT4geo/C > ci/wtrQHO8R77dXHfTvdJ8Eagob122oldEL1UjWvveTw4ae1uSUmiurGBvK/WhrhxbFsfi65 > 2pJ0Ig2PcPSAM24QzSnbLNB+nLkh7XzEo2mBfKHFNogSdnnJ/hG3p0fFzhpPN47nlzUCGZUC > fO/608408RkSO0gI8dDeZAidomomvSYbNtFrJbH7VRGKqWqwG2N4uvuwD1SIpSnlQ5Q0uScI > e2NreB+fU2vYgzzuKquDMod1dMrjRiMV7x4BdafbdoWcRCd04z6xiM5zdrKiMJ7MuMayucn7 > NFtem/s/hANr+B9hEs9Cb+8wGLWVR/aQTIbSRPtIJf1qYhH3aiNeCkDcS3eaHxWoEsFYdw4C > lv1SFgclvA0vC7xdUBITge0Aql3mHKgDD7t2j+8owLkPLA/JdF/YhW1JYPEXBN3YwBgPWJak > CheOOtTOuurci8E+I9X36zfwtNWrHaWp9O4qy2gFKynbuYNs5VVEda0qaaMmr1Il2TqlSdIS > 0HeEgbdT+8fTjnkOAvbR9alP3LWpoIZI1hzWTYzuUm8ezOMLVF+3HXDnb7cVOAi7M1shSRvS > 9ja0/57qs9bxBJCuZrgG0rT8ypura6tukp2yiUZ2Kr8bDnFrUX/YaYupSb+T2SYcF+v5XL6w > WK2Xbpvfl7CoVHjKHHm/Nqvf7rRfD1PaHJGWavaU1s+Z0NrdTAFvQeDAlslWOS6Us2xLP1ya > BceVlast843DBaVu2S61fm61yPUjV9jJjsLowKW3/VOdf4vde9/1CPqQWQOmvSfZi83+W8XB > df6bK+/3NsHdbt7p4B3EtoHaDxG4shlgK8mfPWij8Sei5clR+Bd6Axl4HhcAAA== > > --Boundary_(ID_ZZbbHv6BpKgt6aGqeqChuw)--

2 1

Seeking advice on demos/documentation
by Ross Boylan 29 Dec '01

29 Dec '01

If I can get past a few more bugs, I'm hoping to put a little application out. The documentation is rather sparse, and it seemed to me a walk through a typical usage cycle might help. I was just going to write a narrative, but it struck me that having a real demo might help. Ideally, it would be good to have something that the user could follow along with, and that would display text messages as it went (my ability to record audio is doubtful, and anyway that would be huge). In this ideal world, the playback could be paused or its pace changed. I've seen some references to EventRecorder for doing this, and wonder if that is a reasonable place to start. Also, if people have ideas about other approaches to conveying the information, I'd love to hear them. I know that personally I usually find demos annoying, so I'm not quite sure why it seems appealing to me now. I guess one factor is I think I can produce it faster than a really full-featured documentation set. Since I'm playing around with the application, it's in nothing like a final state, so the complete documentation would be extravagant even if I had tons of time. Thanks in advance.

2 1

Debugging Layout problems
by Ross Boylan 29 Dec '01

29 Dec '01

I have a bug in some layout code which is bringing up the emergency evaluator. Because of this, I can not bring up an ordinary debugger. I realize there may be no silver bullets, but does anyone have any suggestions for getting to the bottom of this? I'm in a morphic problem, and the problem is at least a bit subtle because the object displays properly the first time. Here are some strategies that have occurred to me. Which horse to bet on? 1) Debug by inspection. Unfortunately, I've looked at the code and the instance variables, and nothing has jumped out. I do know where the problem is occurring because of the emergency evaluator stack trace. 2) Put a halt in the code. But this causes the same loop that brings up the emergency evaluator. 3) Put a halt in the code but start it with the mouse off the problem morph, in hopes this will avoid the display loop. It didn't. 4) Learn how to use the emergency evaluator well enough to go up the stack and inspect the instance variables. 5) Do old fashioned debugging with trace statements to a file. Aside from the other problems, I'm not sure if buffering would prevent the interesting stuff from just before the problem from getting out. I'm currently at 3.2alpha #4441. Many months ago I remember that the system would sometimes display a red box with a yellow X across it in somewhat similar circumstances. If there's a way to get that behavior, that would certainly make debugging easier.

3 3

[ENH] Re: Request for Explorer improvements (3)
by Joern Eyrich 29 Dec '01

29 Dec '01

> Joern Eyrich wrote: > > > > Then does somebody introduced shortcut for the > explorer? > > > > apparently yes, use > > (Ctrl|Alt|Meta|Cmd|Opt|whatever-your-platform-uses)-I to get the > > explorer instead of -i for the inspector > > True, but once your inside the explorer you don't have the shortcuts > to open other explorers, inspectors etc. > > Karl Ah, I see. Try this hack. you can use 'e' and 'i' now. I don't think this is how it's supposed to be done, but I don't know Morphic well enough... How could I use Cmd-e instead, for example? Joern

1 0

[ENH]NorwegianVocabulary
by Karl Ramberg 29 Dec '01

29 Dec '01

A quick and dirty adaptation to Norwegian of the SwedishVocabulary for Etoy scripting. Karl

1 0

[ENH] Re: Request for Explorer improvements
by Joern Eyrich 29 Dec '01

29 Dec '01

> does somebody made an enhancement so that the explorer window acts as the > window label of the inspector: showing the class of the inspected object. > It would be cool if the class of root of the explorer or the selected items > would be displayed in the window label. attached change set displays the same title for the root object as the inspector does > Then does somebody introduced shortcut for the explorer? apparently yes, use (Ctrl|Alt|Meta|Cmd|Opt|whatever-your-platform-uses)-I to get the explorer instead of -i for the inspector

3 3

[FIX][VM][UNIX] vmPath (was: "Re: [BUGS] Looking for plugins (Linux)")
by Joern Eyrich 29 Dec '01

29 Dec '01

On Fri, Dec 28, 2001 at 03:10:03PM -0800, Tim Rowledge wrote: > > On a related note, "Smalltalk vmPath" incorrectly answers the > > directory that the image resides in. > So it does. And it does it with deliberate care - see > platforms/unix/vm/sqXWindow.c>RecordFullPathForImageName(). Looks like a > fix is needed from someone that understands how to get path info for the excutable. fix (context diff) is attached. I don't like that the result has a slash appended (plus, it's inconsistent with SystemDictionary>>imagePath, which doesn't the trailing slash), but the source file loading mechanism seems to rely on that (SystemDictionary>>sourcesName); If I would leave out the slash and change the image accordingly, other platforms would probably break. There seems to be a bit of confusion about the source file name anyway, FileDirectory class>>openSources:forImage: seems to expect just the file name, but is passed the absolute pathname. anybody knows what's *supposed* to be going on here? btw, what do I have to do to be able to check in files into the SF CVS? Joern

2 1

← Newer
1
2
3
4
5
6
7
...
36
Older →

Jump to page: