Squeak-dev November 1999

squeak-dev@lists.squeakfoundation.org

164 participants
545 discussions

[Q, BUG?] perform:orSendTo problem?
by Jason McVay 01 Nov '99

01 Nov '99

hello again all-- i asked about this last week and only heard from one person (thanks Torsten). however, i would like to hear from some others on this on if this is a bug or not. from what i understand of the #perform: selector orSendTo: otherTarget method in both Object and Model, the receiever is supposed to be asked first if it can handle the selector; if not, then the selector is passed off to otherTarget. the comments in both classes says this is supposed to happen, but the implementation in both classes seems not to ask the receiever anything; instead, it just sends the selector onto otherTarget. other classes such as PluggableText, StringHolder, and TranscriptStream override this method to what looks to be the proper implementation. why is the message implemented like this in Object and Model? what am i missing? thanks for your consideration and replies, jason

2 1

[FIX] inset/raised borders for transparent morphs
by Doug Way 01 Nov '99

01 Nov '99

This changeset fixes a problem with drawing #inset or #raised borders around transparent morphs. Currently, if you set a BorderedMorph which is transparent to have an inset or raised border, the border ends up being a grayish color, no matter what the color behind the transparent morph is. My thought was that the border should be highlighted as if the morph was this "background" color (the color behind the transparent morph). (I suppose you could argue that a transparent morph shouldn't show any border at all, since it's transparent, but that seems less satisfying.) Mainly, this changes Morph>>colorForInsets so that if a morph is transparent, it recursively checks its owners until it finds a non-transparent owner. (This will be the background color.) It also tweaks BorderedMorph>>drawOn: so that #raised borders are fixed too. The changeset is at: http://www.mindspring.com/~dway/smalltalk/InsetBorderFix-dew.cs Anyway, this fix works well with the BobsUI simple widget framework, which is based on container morphs being transparent. (see http://www.charm.net/~arning/BobsUI.13Jun624pm.cs.txt) - Doug Way EAI/Transom Technogies, Ann Arbor, MI dway(a)eai.com, dway(a)mat.net http://www.transom.com

1 0

Mailing List Archives -- Re: Where's the source?
by JArchibald＠aol.com 01 Nov '99

01 Nov '99

=> 10/31/99 11:48:22 PM EDT, dway(a)mat.net => << Please, don't forget that there are two other mailing list archives (mentioned on the Swiki), which work perfectly fine and are up-to-date: >> Yes, this is true. However, they are second-source archives, and as such, are in no way official. Secondly, they only go back a finite period of time (although with the way the UIUC archive had been behaving, it has turned out to be even worse). Thirdly, and this turns out to be most important to me, there is no way to down-load the archive in its entirety--both of these resources (Thank God we at least have them!) allow you only to browse submission using their facilities, you must remain Internet connected to them to browse submissions. ~sigh~ Jerry Archibald systemObjectivesIncorporated

1 0

Re: Where's the source?
by JArchibald＠aol.com 01 Nov '99

01 Nov '99

Hello Ross, I hope I'm not going to tell you something you might already know about Squeak. Squeak, as do all Smalltalk variants (to a greater or lessor extent), provides all source for the entire application environment. First though you have to get within the environment to work. The home page 'squeak.org' will lead you to appropriate download information. Download a package for your platform which includes four files (VM, Image, Changes, Source). Run it. From the 'desktop' select open>>Browser. At that point you are looking at all the source (for the entire system, as well as the Smalltalk to C converter for creating the VM=virtual machine). The mailing list archives are broken (thus the Forbidden message); we have not been able to determine why at this point. Version 2.6 is the current version. But the download archives (=> Squeak.org) provide all versions for download. Usually facilities which have been previously available (such as mentioned for 2.4c, a very recent version), have been updated, or often will be updated if you request so from the submitter. Hope this helps, Jerry Archibald systemObjectivesIncorporated => 10/30/99 9:40:13 PM EDT, RossBoylan(a)stanfordalumni.org => << I'd like to get the source for the VM for the current squeak (2.6). At the moment I'm on Windows NT. I've looked around the download site and the squeak.org site, and I can't find it. I did see an indication that the Unix source was available, but I'd like something I can rebuild if I want to. I might want to link in some C++ code. I don't know if I need the VM source to do that, but I'd like to have it in any event. By the way, I tried going to the discussion archives from the main site, and I got a permission error. Also, I noticed that some items have particular versions of squeak attached. For example, on of the persistence frameworks says it's for 2.4c. Are these likely to work if I try them with 2.6? >>

2 1

"The Smalltalk-76 Programming System: Design and Implementation" paper now online....
by Dwight Hughes 01 Nov '99

01 Nov '99

I just put Dan Ingall's classic paper "The Smalltalk-76 Programming System: Design and Implementation" from the Fifth Annual ACM Symposium on Principles of Programming Languages, January 1978, online at my site: http://users.ipa.net/~dwighth/smalltalk/St76/Smalltalk76ProgrammingSystem.h… My thanks go to Dan and ACM for giving me their permission to publish this -- but especially to Dan since he even went to the trouble to send me scanned images of this paper to work from since I didn't have a copy of it available. Enjoy. -- Dwight

2 3

Jump to page: