From asf at boinkor.net Tue Oct 10 10:13:02 2006 From: asf at boinkor.net (Andreas Fuchs) Date: Tue, 10 Oct 2006 12:13:02 +0200 Subject: [mcclim-devel] braindump: TODOs before a new mcclim release Message-ID: <452B722E.9080902@boinkor.net> Okay, I'd like to do a new release, but with the beginning of the semester, I could use a bit of assistance with a few things. Here are the release-critical things that I won't be able to do alone (ordered by severity): * Fix clisp build failure: message "pointer-butt missing (mcclim 0.9.2)" by Eric Y. Kow, 2006-06-09: I lack a clisp setup. * Test current CVS on non-SBCL and non-ACL. CMUCL is a good candidate (: (Of course, testing on SBCL and ACL is always welcome, too!) The rest looks pretty ok, AFAICT. But if you want to help with any of the following, you're welcome: * Fix the X server memory/resource leak when using double-buffering and mcclim-freetype. * Find a way to make mcclim asdf-installable that makes all contribs/Apps/Experimental systems available right away. * Improve performance in incremental redisplay (; Thanks!, -- Andreas Fuchs, (http://|im:asf@|mailto:asf@)boinkor.net, antifuchs From ahefner at gmail.com Sat Oct 14 18:38:31 2006 From: ahefner at gmail.com (Andy Hefner) Date: Sat, 14 Oct 2006 14:38:31 -0400 Subject: [mcclim-devel] draw-arrow with transformation, freetype with fix(ed) fonts In-Reply-To: <87r6zcea20.fsf@matley.muppetslab.org> References: <87r6zcea20.fsf@matley.muppetslab.org> Message-ID: <31ffd3c40610141138q7c676cc1i8856811a9fe006d7@mail.gmail.com> On 8/19/06, Luigi Panzeri wrote: > 1) draw-arrow* use the line-style-thickness to calculate distances in > order to draw the head, but that thickness is not transformed with the > current user transformation. Thanks for pointing this out. However, your patch is not quite correct. First, transform-distance returns two values, so your use of it would only be correct where the arrow was aligned with the x-axis. Also, I believe the right thing is to adjust for scaling on the variable 'thickness', rather than scaling only 'offset'. You get the same result for 'offset', but thickness is also used in the edge case where the arrow heads would overlap, and the idea is that we just get the thickness (in some unspecified units which we'll pretend are stream coordinate units) into the current coordinate system. I've adapted your patch as such and applied it to McCLIM. Hopefully this resolves the issue. Thanks for the patch. Sorry for the delayed response. From dtc at scieneer.com Wed Oct 18 00:37:45 2006 From: dtc at scieneer.com (Douglas Crosher) Date: Wed, 18 Oct 2006 10:37:45 +1000 Subject: [mcclim-devel] Patch: Fix system pathnames. Message-ID: <45357759.7080603@scieneer.com> * defsystem :clouseau: avoid pathname directory namestrings in file component names; reworking to be more portable. -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: patch-pathnames URL: From dtc at scieneer.com Wed Oct 18 00:40:43 2006 From: dtc at scieneer.com (Douglas Crosher) Date: Wed, 18 Oct 2006 10:40:43 +1000 Subject: [mcclim-devel] Patch: add the stopwatch example. Message-ID: <4535780B.1040803@scieneer.com> * defsystem :clim-examples: add the stopwatch example. -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: patch-stopwatch URL: From dtc at scieneer.com Wed Oct 18 00:42:02 2006 From: dtc at scieneer.com (Douglas Crosher) Date: Wed, 18 Oct 2006 10:42:02 +1000 Subject: [mcclim-devel] Patch: Update support for the Scieneer CL. Message-ID: <4535785A.6010700@scieneer.com> * Update support for the Scieneer CL. -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: patch-scl URL: From dtc at scieneer.com Wed Oct 18 01:04:23 2006 From: dtc at scieneer.com (Douglas Crosher) Date: Wed, 18 Oct 2006 11:04:23 +1000 Subject: [mcclim-devel] Patch: Color name symbol case. Message-ID: <45357D97.1050602@scieneer.com> * Adjust some of the color names to support lower case CL variants. Note that the names used for colors could use some review. Many colors are named using C style conventions such as +RoyalBlue3+, but this is exported as +royalblue3+ which is no different under standard CL but this does make a difference on a case sensitive CL variant. Perhaps renaming these all to use CL conventions such as +royal-blue3+ would be appropriate? The attached patch makes minimal additions to the color names to get the code and examples working. Regards Douglas Crosher -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: patch-color-case URL: From dtc at scieneer.com Wed Oct 18 01:11:52 2006 From: dtc at scieneer.com (Douglas Crosher) Date: Wed, 18 Oct 2006 11:11:52 +1000 Subject: [mcclim-devel] Patch: Improve support for lower case mode CL variants. Message-ID: <45357F58.802@scieneer.com> * Improve support for lower case mode CL variants. -------------- next part -------------- A non-text attachment was scrubbed... Name: patch-case1.bz2 Type: application/x-bzip Size: 20525 bytes Desc: not available URL: From dtc at scieneer.com Wed Oct 18 01:13:35 2006 From: dtc at scieneer.com (Douglas Crosher) Date: Wed, 18 Oct 2006 11:13:35 +1000 Subject: [mcclim-devel] Patch: Rename key name symbols for support of lower case mode CL variants. Message-ID: <45357FBF.4020804@scieneer.com> * Rename key name symbols for support of lower case mode CL variants. -------------- next part -------------- A non-text attachment was scrubbed... Name: patch-case2.bz2 Type: application/x-bzip Size: 21841 bytes Desc: not available URL: From csr21 at cantab.net Wed Oct 18 07:25:40 2006 From: csr21 at cantab.net (Christophe Rhodes) Date: Wed, 18 Oct 2006 08:25:40 +0100 Subject: [mcclim-devel] Patch: Rename key name symbols for support of lower case mode CL variants. In-Reply-To: <45357FBF.4020804@scieneer.com> (Douglas Crosher's message of "Wed, 18 Oct 2006 11:13:35 +1000") References: <45357FBF.4020804@scieneer.com> Message-ID: Douglas Crosher writes: > * Rename key name symbols for support of lower case mode CL variants. Instead of these huge diffs, would it be possible instead to bind a variable (did I see ext:*case-mode*?) so that SCL is in ANSI mode while compiling and loading McCLIM? (Probably this ought to be done for plain old vanilla ANSI-conforming treatment of reader case, too, as I suspect that McCLIM wouldn't build if the lisp has an exotic readtable-case such as :invert) The disadvantage that I see in merging support for non-ANSI treatment of case is that it is terribly easy to break by writing regular ANSI-conforming Lisp; it doesn't feel robust to me. (I would also like to encourage people who work on cut-and-pasted copies of files to find some way of unifying the information in those files; CLX/keysymdefs.lisp and gtkairo/keysymdefs.lisp are identical apart from the package header.) Cheers, Christophe From dtc at scieneer.com Wed Oct 18 08:10:17 2006 From: dtc at scieneer.com (Douglas Crosher) Date: Wed, 18 Oct 2006 18:10:17 +1000 Subject: [mcclim-devel] Patch: Rename key name symbols for support of lower case mode CL variants. In-Reply-To: References: <45357FBF.4020804@scieneer.com> Message-ID: <4535E169.2050405@scieneer.com> Christophe Rhodes wrote: > Douglas Crosher writes: > >> * Rename key name symbols for support of lower case mode CL variants. > > Instead of these huge diffs, would it be possible instead to bind a > variable (did I see ext:*case-mode*?) so that SCL is in ANSI mode > while compiling and loading McCLIM? (Probably this ought to be done > for plain old vanilla ANSI-conforming treatment of reader case, too, > as I suspect that McCLIM wouldn't build if the lisp has an exotic > readtable-case such as :invert) This is not really practical. While it is possible to compile in one mode and load in another, there are limits. Further many of the symbols affected are exported so the requirement would be contagious. I understand your position w.r.t the maintainability of the code. Regards Douglas Crosher From langstefan at gmx.at Sat Oct 28 12:57:27 2006 From: langstefan at gmx.at (Stefan Lang) Date: Sat, 28 Oct 2006 14:57:27 +0200 Subject: [mcclim-devel] Question about source tarball Message-ID: <200610281457.27819.langstefan@gmx.at> Hello all, The source tarball linked from http://common-lisp.net/project/mcclim/ contains only the directory "mcclim/cvsroot" including CVS files and subdirectories. I.e. mcclim/ - cvsroot/ - CVSROOT/ - Emptydir/ - checkoutlist - checkoutlist,v . . . - mcclim - Apps - Debugger - clim-debugger.lisp,v . . . . . . Is that intended? Or is the tarball build broken? Or am I missing something? Thanks, Stefan From langstefan at gmx.at Sat Oct 28 13:01:57 2006 From: langstefan at gmx.at (Stefan Lang) Date: Sat, 28 Oct 2006 15:01:57 +0200 Subject: [mcclim-devel] Question about source tarball In-Reply-To: <200610281457.27819.langstefan@gmx.at> References: <200610281457.27819.langstefan@gmx.at> Message-ID: <200610281501.57369.langstefan@gmx.at> On Saturday 28 October 2006 14:57, Stefan Lang wrote: > Hello all, > The source tarball linked from > http://common-lisp.net/project/mcclim/ > contains only the directory "mcclim/cvsroot" including > CVS files and subdirectories. I.e. Let me clarify that a bit: 1. I'm not familiar with CVS. 2. I expected the get something like "svn export" would give me for an svn repository. Is that expectation wrong and the tarball (intentionally) contains the whole CVS repository?! -- Stefan From athas at sigkill.dk Sat Oct 28 14:45:16 2006 From: athas at sigkill.dk (Troels Henriksen) Date: Sat, 28 Oct 2006 16:45:16 +0200 Subject: [mcclim-devel] Question about source tarball In-Reply-To: <200610281501.57369.langstefan@gmx.at> (Stefan Lang's message of "Sat, 28 Oct 2006 15:01:57 +0200") References: <200610281457.27819.langstefan@gmx.at> <200610281501.57369.langstefan@gmx.at> Message-ID: <87odrwjy83.fsf@sigkill.dk> Stefan Lang writes: > 2. I expected the get something like "svn export" would > give me for an svn repository. Is that expectation wrong > and the tarball (intentionally) contains the whole CVS > repository?! You have downloaded a CVS snapshot, I guess it contains the CVS control files on purpose. I believe that the CVS snapshots are autocreated by common-lisp.net, so I doubt we'd be able to do much about it. -- \ Troels "Athas" /\ Henriksen From asf at boinkor.net Sat Oct 28 17:49:51 2006 From: asf at boinkor.net (Andreas Fuchs) Date: Sat, 28 Oct 2006 19:49:51 +0200 Subject: [mcclim-devel] Patch: Update support for the Scieneer CL. In-Reply-To: <4535785A.6010700@scieneer.com> References: <4535785A.6010700@scieneer.com> Message-ID: <4543983F.1060008@boinkor.net> Douglas Crosher wrote: > * Update support for the Scieneer CL. Thanks! I committed this patch and the "Fix system pathnames" and "add the stopwatch example" patches. One comment from David Lichteblau on #lisp (see http://www.ircbrowse.com/channel/lisp/20061028?utime=3371045114#utime_requested for context): Setting floating point traps in def-cairo-fun is probably not right; it might be better to do it in with-cairo-floats. I left it as it the way your patch does it and added a small explanation for people trying to understand the code. I'm not certain about the symbol name case patches (on the basis of hugeness and maintainability), so I will leave them out for this release. I hope leaving them out doesn't discourage you from sending more of your high-quality patches. I would like to find a solution that makes all users (including the ones of "modern mode" lisps) happy, but right now I prefer to get that release done before we fall back to 1-year time slots. (-: Cheers, -- Andreas Fuchs, (http://|im:asf@|mailto:asf@)boinkor.net, antifuchs From asf at boinkor.net Sat Oct 28 17:59:21 2006 From: asf at boinkor.net (Andreas Fuchs) Date: Sat, 28 Oct 2006 19:59:21 +0200 Subject: [mcclim-devel] Ok, release plans. Freezing... Message-ID: <45439A79.6060801@boinkor.net> Hi there, time for a new release. To give myself less wiggle room, we freeze now, and release on November the 2nd. The release code name will be "All Souls' Day". You read that one correctly. Freeze period starts now. Release on 2006-11-02. (~20:00) 120 hours to go. I think we all have some testing to do now. (-8 Cheers, -- Andreas Fuchs, (http://|im:asf@|mailto:asf@)boinkor.net, antifuchs From asf at boinkor.net Sat Oct 28 18:13:52 2006 From: asf at boinkor.net (Andreas Fuchs) Date: Sat, 28 Oct 2006 20:13:52 +0200 Subject: [mcclim-devel] Platform support / testing status and reports Message-ID: <45439DE0.1000002@boinkor.net> I'd like to try out an experiment this release: There's a mcclim cliki page at http://mcclim.cliki.net/0.9.3%20Status, where I would like you, if you test, to put your results. The format is very simple for now; I hope it will work out for the gigantic mcclim community (: This should allow me to write more accurate release notes and allow volunteers to better coordinate their efforts. Everybody wins! (That's the theory. Let's see how well it works) Cheers, -- Andreas Fuchs, (http://|im:asf@|mailto:asf@)boinkor.net, antifuchs From metawilm at gmail.com Sat Oct 28 23:52:56 2006 From: metawilm at gmail.com (Willem Broekema) Date: Sun, 29 Oct 2006 01:52:56 +0200 Subject: [mcclim-devel] small Allegro issues Message-ID: Hi, There are some issues with current CVS version that make it not work with Allegro 8 (Ansi mode) yet: - (load "..") inside (eval-when (compile load eval) ..) fails, if current directory is not the McClim root directory (as is usually happens if McClim is loaded using asdf). This can be fixed by binding *default-pathname-defaults* as in the patch below, althogh I'm not sure this is the right solution. - In port.lisp there is (in a #+allegro piece) reference to *x-tcp-port* where probably xlib::*x-tcp-port* is meant (which has value 6000). With these changes McClim compiles (Allegro 8, Ansi Mode, Mac OS X). As I have yet to configure X11 on this Mac, I can't check whether the demos work. - Willem - - - fix loading of files in eval-when - - - banaan:~/dev/mcclim willem$ cvs diff -c *.lisp Index: commands.lisp =================================================================== RCS file: /project/mcclim/cvsroot/mcclim/commands.lisp,v retrieving revision 1.64 diff -c -r1.64 commands.lisp *** commands.lisp 17 Sep 2006 20:34:39 -0000 1.64 --- commands.lisp 28 Oct 2006 22:46:30 -0000 *************** *** 21,26 **** --- 21,36 ---- (in-package :clim-internals) + (eval-when (compile load eval) + #+allegro + (let ((*default-pathname-defaults* (or *compile-file-truename* *load-truename*))) + (load "presentations") + (load "presentation-defs")) + #-allegro + (progn + (load "presentations") + (load "presentation-defs"))) + ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;; ;;; ;;; Command tables Index: package.lisp =================================================================== RCS file: /project/mcclim/cvsroot/mcclim/package.lisp,v retrieving revision 1.55 diff -c -r1.55 package.lisp *** package.lisp 29 Jun 2006 08:16:02 -0000 1.55 --- package.lisp 28 Oct 2006 22:46:32 -0000 *************** *** 2,7 **** --- 2,14 ---- (in-package :common-lisp-user) + (eval-when (compile load eval) + #+allegro + (let ((*default-pathname-defaults* (or *compile-file-truename* *load-truename*))) + (load "patch")) + #-allegro + (load "patch")) + ;;; ;;; CLIM-LISP ;;; - - - fix socket port number - - - banaan:~/dev/mcclim/Backends/CLX willem$ cvs diff -c cvs diff: Diffing . Index: port.lisp =================================================================== RCS file: /project/mcclim/cvsroot/mcclim/Backends/CLX/port.lisp,v retrieving revision 1.122 diff -c -r1.122 port.lisp *** port.lisp 23 Mar 2006 08:45:26 -0000 1.122 --- port.lisp 28 Oct 2006 23:02:36 -0000 *************** *** 1428,1434 **** :remote-filename (format nil "/tmp/.X11-unix/X~D" display) :format :binary) (socket:make-socket :remote-host (string host) ! :remote-port (+ *x-tcp-port* display) :format :binary)))) (if (streamp stream) stream --- 1428,1434 ---- :remote-filename (format nil "/tmp/.X11-unix/X~D" display) :format :binary) (socket:make-socket :remote-host (string host) ! :remote-port (+ xlib::*x-tcp-port* display) :format :binary)))) (if (streamp stream) stream From dtc at scieneer.com Sun Oct 29 00:19:29 2006 From: dtc at scieneer.com (Douglas Crosher) Date: Sun, 29 Oct 2006 11:19:29 +1100 Subject: [mcclim-devel] Patch: Update support for the Scieneer CL. In-Reply-To: <4543983F.1060008@boinkor.net> References: <4535785A.6010700@scieneer.com> <4543983F.1060008@boinkor.net> Message-ID: <4543F391.9000708@scieneer.com> Andreas Fuchs wrote: > Douglas Crosher wrote: >> * Update support for the Scieneer CL. > > Thanks! I committed this patch and the "Fix system pathnames" and "add > the stopwatch example" patches. Thank you. > One comment from David Lichteblau on #lisp (see > http://www.ircbrowse.com/channel/lisp/20061028?utime=3371045114#utime_requested > for context): Setting floating point traps in def-cairo-fun is probably > not right; it might be better to do it in with-cairo-floats. I left it > as it the way your patch does it and added a small explanation for > people trying to understand the code. Thank you for pointing this out. The attached patch implements 'with-cairo-floats for the SCL and also CMUCL. > I'm not certain about the symbol name case patches (on the basis of > hugeness and maintainability), so I will leave them out for this > release. I hope leaving them out doesn't discourage you from sending > more of your high-quality patches. I would like to find a solution that > makes all users (including the ones of "modern mode" lisps) happy, but > right now I prefer to get that release done before we fall back to > 1-year time slots. (-: 'Modern mode' is a Franz term and may be trademark. Some customers prefer the ANSI CL symbol names in lower case, and I'll just call it a non-standard lower case mode, and the SCL supports this as an option. Since the changes are all neutral in ANSI upper case mode it would seem harmless to integrate them, and they have been developed and continue to be supported by Scieneer. Further I believe Franz offer ACL for free evaluation and this could be used to maintain McCLIM support for these changes. The modified source code is available from the SCL freeware collection, and ACL modern mode patches are also welcome, see: http://www.scieneer.com/s/product.html?code=56150 Regards Douglas Crosher -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: patch-with-cairo-floats URL: From asf at boinkor.net Sun Oct 29 09:56:34 2006 From: asf at boinkor.net (Andreas Fuchs) Date: Sun, 29 Oct 2006 10:56:34 +0100 Subject: [mcclim-devel] small Allegro issues In-Reply-To: References: Message-ID: <45447AD2.7030409@boinkor.net> Willem Broekema wrote: > Hi, > > There are some issues with current CVS version that make it not work > with Allegro 8 (Ansi mode) yet: > > - (load "..") inside (eval-when (compile load eval) ..) fails, if > current directory is not the McClim root directory (as is usually > happens if McClim is loaded using asdf). This can be fixed by binding > *default-pathname-defaults* as in the patch below, althogh I'm not > sure this is the right solution. I don't think this is the right way to go about it. Are you using mcclim.asd to load mcclim? If you are, inter-file dependencies should be recorded there. At any rate, I committed a dependency from package.lisp to patch.lisp there. > - In port.lisp there is (in a #+allegro piece) reference to > *x-tcp-port* where probably xlib::*x-tcp-port* is meant (which has > value 6000). committed. > With these changes McClim compiles (Allegro 8, Ansi Mode, Mac OS X). > As I have yet to configure X11 on this Mac, I can't check whether the > demos work. Thanks for testing & the fixes! I got it to compile & run demos on acl 8.0/linux. Cheers, -- Andreas Fuchs, (http://|im:asf@|mailto:asf@)boinkor.net, antifuchs From langstefan at gmx.at Sun Oct 29 12:52:58 2006 From: langstefan at gmx.at (Stefan Lang) Date: Sun, 29 Oct 2006 13:52:58 +0100 Subject: [mcclim-devel] Question about source tarball In-Reply-To: <87odrwjy83.fsf@sigkill.dk> References: <200610281457.27819.langstefan@gmx.at> <200610281501.57369.langstefan@gmx.at> <87odrwjy83.fsf@sigkill.dk> Message-ID: <200610291352.58746.langstefan@gmx.at> On Saturday 28 October 2006 16:45, you wrote: > Stefan Lang writes: > > 2. I expected the get something like "svn export" would > > give me for an svn repository. Is that expectation wrong > > and the tarball (intentionally) contains the whole CVS > > repository?! > > You have downloaded a CVS snapshot, I guess it contains the CVS > control files on purpose. I believe that the CVS snapshots are > autocreated by common-lisp.net, so I doubt we'd be able to do much > about it. IMO, then the sentence "A compressed tar file of the sources is made nightly." on the webpage is misleading and should instead be something like: "Here is a backup of the entire CVS repository." I'll just checkout via cvs. Thanks, Stefan From mikemac at mikemac.com Sun Oct 29 15:27:46 2006 From: mikemac at mikemac.com (Mike McDonald) Date: Sun, 29 Oct 2006 08:27:46 -0700 Subject: [mcclim-devel] Question about source tarball In-Reply-To: Your message of "Sun, 29 Oct 2006 13:52:58 +0100." <200610291352.58746.langstefan@gmx.at> Message-ID: <200610291527.k9TFRku27665@saturn.mikemac.com> >To: mcclim-devel at common-lisp.net >Date: Sun, 29 Oct 2006 13:52:58 +0100 >From: Stefan Lang > >On Saturday 28 October 2006 16:45, you wrote: >> Stefan Lang writes: >> > 2. I expected the get something like "svn export" would >> > give me for an svn repository. Is that expectation wrong >> > and the tarball (intentionally) contains the whole CVS >> > repository?! >> >> You have downloaded a CVS snapshot, I guess it contains the CVS >> control files on purpose. I believe that the CVS snapshots are >> autocreated by common-lisp.net, so I doubt we'd be able to do much >> about it. > >IMO, then the sentence >"A compressed tar file of the sources is made nightly." >on the webpage is misleading and should instead be >something like: "Here is a backup of the entire CVS repository." > >I'll just checkout via cvs. The web page is from when I was hosting McCLIM. I was making a nightly tarball of just the sources. I guess the common-lisp.net guys got their shell script wrong. Mike McDonald mikemac at mikemac.com From asf at boinkor.net Mon Oct 30 22:23:57 2006 From: asf at boinkor.net (Andreas Fuchs) Date: Mon, 30 Oct 2006 23:23:57 +0100 Subject: [mcclim-devel] Question about source tarball In-Reply-To: <200610291527.k9TFRku27665@saturn.mikemac.com> References: <200610291527.k9TFRku27665@saturn.mikemac.com> Message-ID: <45467B7D.2040003@boinkor.net> Mike McDonald wrote: >> "A compressed tar file of the sources is made nightly." >> on the webpage is misleading and should instead be >> something like: "Here is a backup of the entire CVS repository." >> >> I'll just checkout via cvs. > > The web page is from when I was hosting McCLIM. I was making a > nightly tarball of just the sources. I guess the common-lisp.net guys > got their shell script wrong. I think that's deliberate - encouraging us developers to keep offshore backups of the sources. I have now corrected the text linking to the CVS repository snapshot. Cheers, -- Andreas Fuchs, (http://|im:asf@|mailto:asf@)boinkor.net, antifuchs