Squeak-dev March 2009

squeak-dev@lists.squeakfoundation.org

134 participants
173 discussions

Problem with cdelc and apicall
by Mariano Martinez Peck 13 May '09

13 May '09

People: I am having a problem with cdelc and apicall in windows with SqueakDBX. I have a abstract class OpenDBX with the subclasses: OpenDBXWin32 and OpenDBXUnix. SqueakDBX detects which OS you are using and instances one of them. They have these methods: OpenDBXWin32: apiBind: handle database: databaseName name: userName password: password method: method "int odbx_bind(odbx_t* handle, const char* database, const char* who, const char* cred,int method )" <apicall: long 'odbx_bind' (ulong char* char* char* ulong) module: 'libopendbx-1.dll'> ^self externalCallFailed OpenDBXUnix: apiBind: handle database: databaseName name: userName password: password method: method "int odbx_bind(odbx_t* handle, const char* database, const char* who, const char* cred,int method )" In linux, I don't have problems, but in windows, when I try to download it from MC, I have a "sintax errror" "OpenDBXUnix api calls apiBind: handle database: databaseName name: userName password: password method: method" apiBind: handle database: databaseName name: userName password: password method: method "int odbx_bind(odbx_t* handle, const char* database, const char* who, const char* cred,int method )" <> expected ->cdecl: long 'odbx_bind' (ulong char* char* char* ulong) module: 'opendbx'> ^self externalCallFailed <cdecl: long 'odbx_bind' (ulong char* char* char* ulong) module: 'opendbx'> ^self externalCallFailed Does anyone have an idea of what can be happening ? thanks in advance. Mariano

2 8

Contributing to 3.11: Step-by-step
by Andreas Raab 11 May '09

11 May '09

Folks - To summarize some of the discussion on IRC, here is a step-by-step guide for how to contribute to 3.11: 1. Report an issue at http://bugs.squeak.org/ 2. If you have a fix/patch attach it to the bug report 3. Add an installer script to the notes, like here: (for more options see http://installer.pbwiki.com/Installer#Mantis) "fix begin" Installer mantis bug: 1234 fix: 'FixTheBug.cs'. "fix end" 4. Test your fix: Installer mantis ensureFix: 1234 5. To propose this fix as a candidate for 3.11: a) Obtain "developer" status on Mantis. You can do this by mailing box-admins(a)lists.squeakfoundation.org. b) Update the issue and set the "Status" field to "testing" c) Update the issue and set "Fixed in Version" field to "3.11" This field is only accessible in the "advanced" view, so if you do not see it you need to change to the "advanced" view by clicking on "view advanced" or "update advanced" in the top right. 6. Load all the fixes with testing status: (Installer mantis bugsAll select:[:bug| ((bug status = 'testing' or:[bug status = 'resolved']) and:[bug fixedIn = '3.11']]) do:[:bug| bug ensureFix]. 7. Rinse and repeat. Cheers, - Andreas

3 4

More distributed version control
by Andreas Raab 28 Apr '09

28 Apr '09

Folks - Just as a side note to anyone who is interested in distributed version control systems, I thought it's interesting to see that they get more and more widely adopted. This just popped up on Python-dev: http://mail.python.org/pipermail/python-dev/2009-March/087931.html I'm wondering if this has any impact on our thinking with respect to Monticello, change sets etc? Cheers, - Andreas

9 13

Loading the OpenGL package (was: Games with Squeak)
by Diego Fernandez 17 Apr '09

17 Apr '09

On 10/13/06, Andreas Raab <andreas.raab(a)gmx.de> wrote: > Edgar J. De Cleene wrote: > > Andreas, as you see no package said " OpenGL" in the provided link. > > I think you should learn a little about using Squeaksource: "Versions" Ouch! I missed the package too, because in www.squeaksource.com usually there is one MC package per "Project". Anyway, I'm trying to load OpenGL-ar.36.mcz in Squeak 3.9 but I get: Syntax Error: glBindBufferARB(target, buffer) "This method was generated by OGLExtManager" <Argument expected ->apicall: void 'glBindBufferARB' (long ulong)> Any clue on how to solve this error? (It looks that it's created with a "special" compiler) And another question... As I wrote in the previous mail I've started to do some SDL bindings. But I think that maybe it's a waste of time, since today nearly all machines have OpenGL, and Squeak already has objects to deal with sound and joysticks.... so instead of using SDL for 2D graphics is better to use OpenGL for the same thing, what do yo thing about that? Cheers, Diego

10 19

Squeak Oversight Board Blog
by Ken Causey 01 Apr '09

01 Apr '09

It has been mentioned before but I want to make it clear that the current Squeak Oversight Board (the team previously known as either the SqueakFoundation Board or the Squeak Leadership Team) has started a blog where we will post our thoughts and plans and where we want to receive your comments. http://squeakboard.wordpress.com/ Please visit this site if you have not already and check back regularly. Thanks, Ken

1 1

Squak main window oddly transparent
by Martin Drautzburg 01 Apr '09

01 Apr '09

Since I am using compiz on linux, the mail squeak window looks somewhat transparent or actually just mangled. It is not really transparent but the stuff behind it affects the way things are displayed. Often it helps to set the color depth to 32, but even that is difficult to do, because I can hardly read the menus. I noticed the same phenomenon with qemu, but it was fixed in a recent version. Is there a way I can avoid this? -- Martin

2 2

[Ann] Cuis: A new Squeak distribution
by Juan Vuletich 31 Mar '09

31 Mar '09

Hi Folks, I'm pleased to announce that Cuis is available. Cuis is a Squeak distribution with emphasis on simplicity. You can download it from http://www.jvuletich.org/Cuis/Cuis1.0-0158.zip . The web is not done yet. Some of the main ideas and objectives for Cuis are: - Aimed at Smalltalk developers and people learning about Smalltalk - Close to Smalltalk-80 and Dan Ingalls' ideas - Compatible with the latest Squeak VMs - MIT license (based on the work by Yoshiki Ohshima, from VPRI) - Evolution by removing unnecessary complexity, not adding it - Reduced Morphic - No MVC or Etoys - No M17N, Traits, Monticello, Omnibrowser, TTFonts, FreeType - Includes support for building VM plugins, but not for building the VM - Includes a set of high quality, antialiased StrikeFonts - Backwards compatibility not important - Include code from the Squeak project (various versions, Mantis, squak-dev, etc) Cuis is under MIT license. It is the result of several years of cleanup, and it is currently being used in one commercial project. It is small and fast, and good for dev and learning. Besides PCs, it also runs great on older machines, PDAs and embedded devices. I hope you like it. Cheers, Juan Vuletich

16 52

Problems building the VM
by Hernán Morales Durand 31 Mar '09

31 Mar '09

Hello, I've followed the steps described here : http://wiki.squeak.org/squeak/6102 for compiling the VM, but it seems there's something wrong in my procedure or the page needs some review: C:\>cd C:\SqZ3950\platforms\win32\build\ C:\SqZ3950\platforms\win32\build>make Makefile:44: src/plugins.ext: No such file or directory Makefile:45: src/plugins.int: No such file or directory make: *** No rule to make target `src/plugins.int'. Stop. C:\SqZ3950\platforms\win32\build>cd .. C:\SqZ3950\platforms\win32>build C:\SqZ3950\platforms\win32>make -f Makefile.mingw32 Makefile.mingw32:6: plugins/plugins.int: No such file or directory make: *** No rule to make target `plugins/plugins.int'. Stop. Then I saw two pairs of them: plugins.int/plugins.ext C:\SqZ3950\winbuild\src\ (not modified) C:\SqZ3950\src32\ (modified but plugins.int list is empty : INTERNAL_PLUGINS = ) I've copied both modified files to C:\SqZ3950\platforms\win32\plugins\ and started the dancing again ('make' fails like above) with the build.bat: C:\SqZ3950\platforms\win32>make -f Makefile.mingw32 gcc -o ./release/sqWin32Alloc.o -g -mpentium -mwindows -Werror-implicit-function -declaration -fomit-frame-pointer -funroll-loops -fschedule-insns2 -O2 -I. -I. - I./vm -Ic:/dx7sdk/include -DWIN32_FILE_SUPPORT -DNO_SERVICE -DNO_STD_FILE_SUPPO RT -DNDEBUG -DLSB_FIRST -DVM_NAME=\"Croquet\" -DX86 -DSQUEAK_BUILTIN_PLUGIN -DCR OQUET -c ./vm/sqWin32Alloc.c ./vm/sqWin32Alloc.c:14: sq.h: No such file or directory make: *** [sqWin32Alloc.o] Error 1 I'm using the latest WMMaker from SqueakSource and Squeak 3.11.2 (alpha) Mar 27 2009. BTW I'm not building the Croquet VM :) Any help would be appreciated. Hernán

2 2

MC load order?
by Andreas Raab 31 Mar '09

31 Mar '09

Folks - Is there anyone out there who really understands how Monticello decides in which order to load definitions in a package? I'm having constant troubles with it and I want to fix a few things regarding load order but I don't really understand what MC uses to decide in which order to load things. Any explanations would be greatly appreciated. Cheers, - Andreas

1 0

[FFI] Endless C structs and unions
by Hernán Morales Durand 31 Mar '09

31 Mar '09

Dear list, I have a couple of questions, I started to play with FFI, so these are newbie ones, my intention is to port the ZOOM library. I was there happily writing ExternalStructure's until I found one that kills the VM. Then I've realized I have some basic doubts about FFI: 1) I have this C struct odr Struct Reference =============== int direction int error unsigned char * buf int top int size int pos const unsigned char * bp NMEM mem struct Odr_private * op which I wrote (with a lot of brilliant inspiration) like: ZOOMODR class>> fields " self defineFields " ^ #( (direction 'long') (error 'long') (buf 'char *') (top 'long') (size 'long') (pos 'long') (bp 'char *') (mem 'ZOOMNMem') (odrPrivate 'void *') " ========= HERE ========= " ) I don't know if this exists, the HERE means : I want to avoid to write an external structure for this... Odr_private Struct Reference ===================== struct odr_constack * stack_first struct odr_constack * stack_top const char ** tmp_names_buf int tmp_names_sz struct Odr_ber_tag odr_ber_tag yaz_iconv_t iconv_handle int error_id char element [80] void(* stream_write )(ODR o, void *handle, int type, const char *buf, int len) void(* stream_close )(void *handle) int can_grow int t_class int t_tag int enable_bias int choice_bias int lenlen FILE * print int indent After a look at the fields, I supose I do not care about the contents of that struct. It is ok the ... (odrPrivate 'void *') ... magic line to "stop" these C structs? If not, is there a way to say "I don't care what follows from here" (i.e. pass automagically nil to that pointer) ? 2) Z_NamePlusRecord Struct Reference ================================= Z_DatabaseName * databaseName int which union { Z_External * databaseRecord Z_DiagRec * surrogateDiagnostic Z_FragmentSyntax * startingFragment Z_FragmentSyntax * intermediateFragment Z_FragmentSyntax * finalFragment } u Is there something special to do about an union in the #fields method? Thanks in advance. Hernán

3 3

Jump to page: