Squeak-dev December 2005

squeak-dev@lists.squeakfoundation.org

129 participants
187 discussions

veryDeepCopy, ImageSegments, and actionMaps
by Ned Konz 17 Aug '23

17 Aug '23

Any Object can have dependents whether or not it has an instance variable to keep track of them (these are kept in the DependentsFields class variable of Object). The only class that keeps its own dependents is Model. Any Object can have an actionMap. These are supplied by EventManager class by default, and are stored in the ActionMaps class variable of EventManager. The only class that keeps its own action maps is Morph. When we veryDeepCopy an object we also copy the global dependents (see DeepCopier>>fixDependents). However, we don't copy the actionMap in those cases where it's held by the EventManager class. When we file out an object that uses DependentsFields out into an ImageSegment, we also file out the dependents (see ImageSegment>>dependentsSave:) if it's in a project. We do this by adding a project parameter called #GlobalDependentsInProject. However, we don't file out the actionMap in those cases where it's held by the EventManager class. My question is this: should we fix these behaviors (copying and serialization) to include these "global action maps" just like we already include the "global dependents"? Or should we discourage the use of the "global action maps"? -- Ned Konz http://bike-nomad.com GPG key ID: BEEA7EFE

3 2

Trying to build SerialPlugin for IrDA under MPW (MacOS 8.6)
by Hans-Martin Mosner 24 May '12

24 May '12

Hi, I recently got a serial IrDA adapter and wanted to play around with it under Squeak. This adapter has a baudrate switching mechanism which depends on toggling the DTR and RTS lines in the correct sequence. To support this, I wanted to build a new SerialPlugin with an additional primitive. However, the plugin as built with MPW does not seem to be accepted by the VM, which still uses the internal plugin in all cases. I've tried placing the plugin in a 'Plugins' subfolder of the VM's folder, and then directly into the VM folder, without any success. I have considered placing debug outputs in appropriate places in the VM, but building a complete VM with MPW is not easy - I would have to create an MPW makefile for the whole shebang, which is more work than I can invest at the moment. Does any of you Mac aficionados have an idea where the problem could be? I have appended the plugin and corresponding source files in a ZIP file (it's not big) so if you want to try it out you could do it right now :-) BTW, would it make sense to incorporate support for switching serial port parameters of an open port into the standard plugin? Closing and opening a port just to change the baud rate seems a little overkill... Cheers, Hans-Martin

6 14

hash method
by Houssam Fakih 31 Jan '07

31 Jan '07

Hi, When I redefine the = operation for a class I should redefine its hash method. My question is about the definition of the hash method. I read that the method must return a SmallInteger whose value is related to the receiver. I would like to know if there is a general pattern adopted in Squeak for that. I get a look in the Squeak classes but it doesnt help. Thanks, Houssam ____ "Le paradis touche les pieds des mères..." >>> http://csl.ensm-douai.fr/fakih

4 4

RE: Progrmaming in Bytecode?
by Swan, Dean 24 Nov '06

24 Nov '06

>> It's not uncommon to find x86 code padded with a lot of NOPs >> that weren't in the original source. > >Cowardly assembler! ;-) > >> [NOPs...] Sometimes they're there to pad the execution pipeline, and >> sometimes it's just because the code generator allocates a whole large >> enough for the largest form of an instruction to simplify the branch >> target calculations. > >We must be talking about M$oft assemblers here. ;^p) Yes. This has been pretty characteristic of MS assemblers as far back as I can remember (i.e. about MASM 2.0 - I remember being quite excited when MASM 4.0 came out - '186 and '286 code!). >The real problem is that the solution in general might require N >iterations for a piece of code with N branches -- if promoting one of them >from short to long causes the next (or, much more problematically, a >preceding) one to become long too. You're such an optimist. O(n) is close to best case for this problem. >This is a weak form of what I suppose one might call an "oscillating >constraints" problem. (Hands up anyone who has ever seen [La]TeX get Hence, stochastic optimization. Even MS uses stochastic optimizers now, even in vb. (didn't want to say that too loud here. ;) ) >The usual cause for "padding" in (capable) assemblers (other than for >prefetch alignment, as you point out) is due to dynamic linking where the >size of call sites isn't known until link time. (On modern systems the Not only prefetch alignment. I have heard, although I don't know the details that sometimes it's desirable to force a dead cycle in the execution pipeline. I can only guess that this would be to influence the scheduling of the execution units for complex instructions, but I consider myself to be only a "casual observer" with respect to this kind of thing. >In Smalltalk you've also got a bigger problem: for very long methods the >range of even unconditional branches might not be enough, at which point >I know this has nothing whatsoever to do with Smalltalk...) Well, to kind of tie this back into something relevant to Squeak - I know a few people have talked about "new" VM instruction sets. It seems that it might help to think about a VM where all instructions are the same size. This could simplify code generation (i.e. compiling methods), and possibly improve performance and *reduce* code size (at least in the case of long methods). The anticipated performance improvement would come from forcing aligned fetches more often and potentially simpler instruction decoding. It's really nice to work with an architecture where all instructions take up one word of memory, take one cyle to execute, and sometimes can cause five or six different useful things to occur during that one cycle. (ok, so I write a lot of DSP code...) The single cycle thing isn't necessarily relevant for Smalltalk, but those "extended" bytecodes are really distasteful to me. -Dean

10 15

[ANN] Audio Video Library
by Martin Kuball 07 Oct '06

07 Oct '06

Hi I would like to annonce my project to build a Library for processing audio and video material with squeak. I put a 1st version on the SqueakSource at kilana.unibe.ch. It's not much yet. But you can load and analyze avi container with mpeg 4 video and mp2 and mp3 audio content. I did this because I needed a tool to repair some broken avi files. Other tools I tried just failed badly. Try this: RIFFArchive openOn: 'test.avi' "use explore it to get a nice tree view of the chunks in the avi" "open a transcript and execute the following to get some information about the contents" (RIFFArchive openOn: 'test.avi') analyzeVideoStream "To extract the streams inside the avi container do:" (RIFFArchive openOn: 'some.avi') extractAllStreamsToFile: 'base_fileName'. "If you have raw streams you can create an avi this way:" | avi | avi := AVIFile with: (M4VStream on: 'video.m4v') with: (MPAStream on: 'audio.mp3'). avi runtime: 100. avi frameRate: 24000 / 1001. avi createIndex. "not working yet" avi writeTo: (StandardFileStream newFileNamed: 'test.avi') Here is my tentative todo list. - write more documentation and tests - cleanup and speedup - add more container formats like mp4, matroska - add mpeg2 handling (-> dvd authoring???) - connect with the other multimedia bits in squeak (sound, video playing, ...) The project is of course open for everyone to join. Enjoy, Martin

5 6

squeak & support for video?
by polishook＠optonline.net 24 Feb '06

24 Feb '06

what's the [current] status of squeak and video support? - input, output, display, etc .... r there future plans for squeak and video?

10 11

[ANN] SM 2.1
by goran＠krampe.se 12 Jan '06

12 Jan '06

Hi people! I think I missed the announcement of the SM upgrade... :) Anyway, I posted about the improvements in this release in an earlier post: http://lists.squeakfoundation.org/pipermail/squeak-dev/2005-November/09 8168.html ...but essentially this is what you need to know: - When you do "update map from the net" your client will notice that the server (map.squeak.org) is running SM2.1 and it will ask you to upgrade your image. If you don't it can't update the map. - While upgrading it will start blabbering about Monticello vs MCInstaller for upgrading the loader, pick Monticello. Using MCInstaller might work too, but I haven't looked carefully. - Eventually it will reopen the loader and all should be fine. The loader doesn't contain much any change at all so you can't really see it is version 2.1. :) If you are running multiple images in the same directory (as most of us probably do) and you upgrade one of them to SM 2.1 then you will end up with a map snapshot version 2.1 in the "/sm" dir. If you now start *another* non updated image and do "upgrade from net" in the SqueakMap Package Loader it will end up asking about what class to convert to yaddayadda... This is a "problem" I will try to fix in the next SM release (it doesn't detect that a new snapshot is already on disk in the new format) but right now you need to press alt-. and then manually execute in a workspace: SMSqueakMap bootStrap ...to trigger the upgrade and then it should work as above. regards, Göran PS. The main difference in this release from the client perspective is that it now has a fallback on a server cache so the problem with packages not being online should hopefully be a problem of the past.

5 6

Say Hello to Stanley
by Frank Caggiano 10 Jan '06

10 Jan '06

Stanford's souped-up Volkswagen blasted through the Mojave Desert, blew away the competition, and won Darpa's $2 million Grand Challenge. Buckle up, human - the driverless car of the future is gaining on you. http://www.wired.com/wired/archive/14.01/stanley.html ------------------------ Frank Caggiano frankcag at crystal-objects dot com http://www.crystal-objects.com The best education for the best is the best education for all. Robert Maynard Hutchins

8 11

Caché Database System Integration With Squeak
by Ron Teitelbaum 05 Jan '06

05 Jan '06

Hello All, Has anyone worked with the Caché Database system: http://www.intersystems.com/ . Im interested in possible integration with squeak. Ron Teitelbaum President / Principal Software Engineer US Medical Record Specialists www.USMedRec.com <http://www.usmedrec.com/> Ron(a)USMedRec.com

5 8

Steward teams
by Cees De Groot 03 Jan '06

03 Jan '06

This is just a quick ramble before I go to sleep - interpret it as such :) I've been a bit worried about the lack of manpower in the Stewards teams. Not about the current absolute numbers, but the fact that with the current granularity, we'll need a lot more teams and thus people and I'm not sure where we are going to get them from. At the same time, I think we all agree that we should move away from the monolithic development mode where one team maintains everything. Therefore, I would like you to think about an intermediate way: cut the image in rough chunks, maybe four our five, and assign these to teams. "Core Class Library" (Kernel, Objects, Collections), "I/O" (Files, Network), "UI" (Morphic, MVC), "Development" (Tools and assorted stuff) is more or less along the lines I'm thinking of (again, I'm just throwing an idea at the list so I haven't checked whether this way of cutting covers everthing). Four teams are easier to staff than the 10-15 we'd need if we keep the package-level granularity. And when the teams grow big, they can always divide work internally, and maybe later on split up when it makes sense. Anyway, sounds truer to "divide and conquer" than what we're currently doing (which a pessimist might interpret as "chop and destroy" ;-)). G'night, Cees

9 12

Jump to page: