Squeak-dev November 2003

squeak-dev@lists.squeakfoundation.org

174 participants
566 discussions

Morph>>Delete
by Jeff Longland 05 Dec '03

05 Dec '03

I'm trying to create a button that will delete a morph named JeffsTestMorph (sublass of Morph). Through inheritance, JeffsTestMorph inherits the delete method does it not? On mouseDown for my button, I have the following: mouseDown: evt JeffsTestMorph delete. But when I click on the button I get MessageNotUnderstood: delete. Could someone tell this Squeak newbie what he's doing wrong? Thanks, Jeff

10 13

[ENH][VM] faster image startup
by John M McIntosh 03 Dec '03

03 Dec '03

Make image startup slightly faster. a) Don't rewrite all oops to clear the mark bit, only rewrite the ones that have the mark bit set. b) Don't rewrite all oops to change oops address if the offset is zero which is the case on the mac when the same vm reads a recently saved image because of how the memory map is located. Unknown if this is the case on wintel. c) Don't rescan the entire image a second time to null out compiled methods with an external primitive call, remember those oops on the first scan, then later null out the 320 or so contexts, versus rescaning 260 thousands oops Now one of the 'ugly' things I did here to avoid altering the Interpreter class definition and having 16K of memory sucked up by a global one-time use table is to allocate a local 4096 element Array in ObjectMemory>>initializeObjectMemory: which gets passed to an altered adjustAllOopsBy: which remembers the methods oops that have refer to external primitives as it's looking for Oops with the Mark Bit set. That array with the stored count is returned and at the end of initializeObjectMemory we decide to invoke the old flushExternalPrimitives if the count is >= 4095 or in my testing with the base image since we only have 320 oops that match it means just invoking flushExternalPrimitiveOf: on the remembered oops. At startup time we reduce the CPU time by 8% based on the changes to initializeObjectMemory:, and remove another 4% based on a no-longer required rescan by flushExternalPrimitives. Although these changes reduce powerpc startup time by just a little it should make some noticeable changes for 100MB images, or on machines that are *much* slower. Other changes go into the mac source code to reorganize how the menu bar is manipulated and not to start the 1/60 clock pthread until it's really required. That shaves a bunch of clock time off and now startup is much more snappy. -- ======================================================================== === John M. McIntosh <johnmci(a)smalltalkconsulting.com> 1-800-477-2659 Corporate Smalltalk Consulting Ltd. http://www.smalltalkconsulting.com ======================================================================== ===

4 6

deficience in Squeak
by ye juan 03 Dec '03

03 Dec '03

Hi, I have developed in visualworks 5i.4 from Cincom Corporation,= another Smalltalk platform. Now I will develop some applications= in Squeak, however, I find many basic methods disappearing, such= as "matchesPattern: ignoreCase: do:". So I want to know whether= this is deficient in Squeak, or there exist some better methods= to replace the old ones. What is more, if a new method contains= unknown methods, which will be defined in the future, then it= can not be accepted. I think it is not convinient for developing= application. __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com

24 37

[ENH] ParenBlinking ( [et] reposted as changeset )
by marcus＠ira.uka.de 02 Dec '03

02 Dec '03

from preamble: "Change Set: ParenBlinking Date: 8 October 2003 Author: Avi Bryant md: reposted as real changeset. Avi writes: This has been on SqueakMap forever, but I figured with the new BFAV and so on it might be worth posting it as an ENH, in the hopes that it'll actually make it into the image. This makes two changes to the ParagraphEditor: 1. upon typing ), }, or ], the matching (, {, or [ will be made bold until the next keystroke 2. the meanings of return and control-return are switched (ie, auto-indent is the default) I have had this changeset loaded in every image I've used for the last year and a half, with no ill effects. I think it makes the ParagraphEditor significantly more useful for writing code (which is, after all, what most of us use ParapraphEditor for most of the time). If people want change 1 without change 2, I'm happy to introduce a preference that controls that (although I personally think of auto-indent as something any good text editor does, and can think why you wouldn't want it). Separating them into two changesets is more difficult, however, since they both need to modify #dispatchOnCharacter:with:. Cheers, Avi "!

7 10

[ENH] yaxoRemoval
by marcus＠ira.uka.de 02 Dec '03

02 Dec '03

from preamble: "Change Set: yaxoRemoval-mir Date: 22 July 2003 Author: Michael Rueger md: this is from SqueakMap. We have currently a pretty old Yaxo version in the image, and actually it's not used by anything in the image. Added a postscript for removing the categories. Removes Yaxo, the XML parser from the image."!

5 4

[VM] pending change to Cross platform sq.h
by John M McIntosh 02 Dec '03

02 Dec '03

I'm proposing a change to sq.h such that /* image file and VM path names */ extern char imageName[] becomes char *imageName(void); The VM supplies an image name to the image open logic. This image name comes from the VM support routines once they decide what it should be as part of the startup process, or later when you do a save or save as operation. Having a function return a pointer to the characters as a result of calculations that determine issues like character encoding is 'better' than referring to an MacRoman ASCII byte string. This allows the os-x version to translate a unicode path name to the desired mapping when needed, since that mapping might change under Smalltalk application control. Yes, there will need to be a change set with this at some point to alter Interp.c to generate imageName() versus imageName -- ======================================================================== === John M. McIntosh <johnmci(a)smalltalkconsulting.com> 1-800-477-2659 Corporate Smalltalk Consulting Ltd. http://www.smalltalkconsulting.com ======================================================================== ===

4 5

An '*?%&$' mood
by raymondasselin＠sympatico.ca 01 Dec '03

01 Dec '03

1. Can somebody explains what to do (is there something to do?) when you install a package from SM and there is undefined classes lists on the Transcript window? Like : (SixxXmlUtil is Undeclared) (SixxXmlUtil is Undeclared) (SixxXmlUtil is Undeclared) (SixxXmlParseError is Undeclared) (SixxSetting is Undeclared) (SixxXmlUtil is Undeclared) (VWXMLXMLParser is Undeclared) (SixxMockClassForUnitTest is Undeclared) (SixxMockClassForUnitTest is Undeclared) (SixxMockClassForUnitTest is Undeclared) (SixxMockClassForUnitTest is Undeclared) (SixxMockBinaryData is Undeclared) (SixxMockBinaryData is Undeclared) (SixxMockClassForUnitTest is Undeclared) (SixxMockClassForUnitTest is Undeclared) (SixxMockClassForUnitTest is Undeclared) (SixxMockClassForUnitTest is Undeclared) (SixxMockClassForUnitTest is Undeclared) (SixxMockClassForUnitTest is Undeclared) (SixxMockClassForUnitTest is Undeclared) 2. And what to do with UndefinedObject when trying to install Refactoring Browser ! My image is 3.7 5566 and I want to try SmallWiki so I started with a clean 3.7 image... =========== 3. I must say that lots of things are harder to get together with Squeak from 3.6. It seems to me that this was'nt the case with the 3.5 release. Few examples: - At Job in WinNT4.0 I can't bring a 3.5 Squeak project (.pr) in 3.6 Squeak when trying to save it and load it in a 3.6 image there are problems. - It is the same when trying to save a 'drawing' in Connectors to be able to load it in a 3.6 image, or 3.7 one. - Recently there was problems with the MacOSX 10.3 and the VM - There is or was problem whith filePath in Mac don't know if this is fix, nor in which vm or image it is fix, nor if I need to install a patch and which in which image. - When I tried to use new version of Seaside with 3.6 Squeak, Seaside had changed, Comanche had changed, some package on SM try to load a Seaside version not anymore served.... - Did you tried to load BFAV to understand it and see if you can help in a minimal way the Squeak community...I don't know how it is now but two weeks ago this was a mess... ...so......fewwwww...!!! So: - projects from image to image is'nt stable.. - 'drawings' from image to image is'nt stable...(may be it is the same problem...don't know). - using SM is'nt stable yet (I know Göran.. the SM2 will be...). No, -- backtracking --, SM is stable but the documentation which say explicitly which package is a prerequisite an what is the loading order is'nt, so for the end user this is the same: the experience is'nt stable. In fact I lost at least 30 hours to try few things in Squeak and nothing was working the first time and often I did'nt know how to solve the problem. I don't mind about these hours but i'm a bit 'lost', I don't think it is possible to master all aspect of Squeak to use it, and I don't think I should understand it all to use it. BTW, I know what have been done in Squeak from 2.2 image, and thats a lot. In fact what is frustrating is'nt new 'alpha' stuff, it is the old one that didn't function any more in new version of Squeak. A part of the problem is mine as I should have been signaling every bug as a bug to the community, but this is time consuming when english is your first langage, and even more when it is not...this is why I tried to load and use BFAV. ========= Conclusion: - the effort to identify 'maintainers' for differents parts of Squeak will be a big improvment. I think that I now, master Smalltalk sufficiently to help but not to be a lead maintainer. - the SM2 is an absolute necessity (I know SM2 alpha is there but did'nt check it yet). - Sorry...I was in a '*?%&$' mood. Let come back to Squeaking. Raymond

5 7

VM file name character encoding support (pending)
by John M McIntosh 01 Dec '03

01 Dec '03

I believe that to support different file name character sets we're going to need to alter sqVirtualMachine.c and take VM_PROXY_MINOR to 6 by adding three new calls to support information about what the file name encoding information is. getVMEncodings returns Smalltalk OOPS Array of supported encodings which the VM supports. getCurrentVMEncoding returns magic integer signifying the encoding being used. setCurrentVMEncoding takes magic integer, returns 1 if set, zero if not a supported encoding. where the magic integer is and I'll be platform independent and define them as: MacRoman = 0, WindowsLatin1 = 1, /* ANSI codepage 1252 */ ISOLatin1 = 2, /* ISO 8859-1 */ Unicode = 3 UTF8 = 4 This allows plugins, like the mpeg plugin to accept a file name from the VM and know what to do with it before issuing the file open by asking the VM what the encoding is. It also of course allows a programmer the ability to see and choose an encoding to use within the Image. I'm also considering using a reserved header word in the image so we can determine what the image encoding is at startup time because things like the VMPath are loaded early and we can better ensure file name encoding is done correctly versus relying on some startup task stamping the image with the encoding information. What is unclear, but could be defined is how to deal with existing path information, such as the vmpath once you switch encoding. I suppose you could leave that information in native platform format and return a converted value based on the currently chosen encoding. Thoughts on this are welcome. -- ======================================================================== === John M. McIntosh <johnmci(a)smalltalkconsulting.com> 1-800-477-2659 Corporate Smalltalk Consulting Ltd. http://www.smalltalkconsulting.com ======================================================================== ===

3 4

Squeak on the Sigmarion III
by Aaron J Reichow 01 Dec '03

01 Dec '03

Hello- especially Ned- I have aquired a Sigmarion III device as a review/demo unit. There is a lot of interest within some circles (HPC users, people who want an upgrade from their MobilePro or Jornada7xx), but being a japan-only device no good reviews have surfaced on the english mobile sites. Replying to some comments by Ned when Yoshiki-san brought up the Sig3 and my own impresions. The Sigmarion III is a newish Handheld PC-like device that runs WinCE 4.1. It is decently priced (as low as $360 in Japan) compared to anything else in its class. It has the usual specs for devices like this these days- 400 MHz PXA255, 64 MB RAM, SD and CF slots. That's about where it ends as far as the "usual" specs are concerned. In addition to the above, the Sigmarion III is special in a few ways- * A USB Host (1.X) *and* a seperate USB slave port: the former is for attaching USB devices, the latter for connecting the Sig3 to a desktop for syncing, usbnet, etc. I don't know of any other XScale PDAs that have this, although I think some Toshiba PocketPCs have at least some capability in this respect. You can get a CF card to a USB host port to a PocketPC or Zaurus, but they are pretty expensive (US$140), take up a CF slot (where does wifi go?!) and bulkier. Reading various forums, people have been able to hook up a number of devices to the USB port- keyboards, mice, USB hard drives scanners, and cd burners (!). * A 5" 800x480 touch screen. What is there more to say? A smaller DPI, so the screen is a bit taller on the vertical axis than the Zaurus is when they both have 480 pixels. Like the Zaurus C7x0 the screen is very sharp, packing a lot of pixels in a tiny space. Not as bright of a screen as the Zaurus C7x0, but easier to read for me for longer periods of time, though that is probably the DPI. * A touch-typable keyboard. Very much like the one on the Jornada 720. Which means that the keyboard feel like a laptop keyboard, a laptop keyboard with very small keys, but the same mechanism. As opposed to the click-click TV remote control-style of buttons on the Zaurus C7x0. I have pretty stubby hands and I have no problem typing on the Sig3's kb. There are all sorts of cellular data services you can get through NTT DoCoMo if you are in Japan. I am not and have no clue what options there are- but they do look sweet. As far as running Squeak, things are not perfect yet. I have spoken with Yoshiki-san some about it, but nothing has been resolved so far. Using the older, stable VM (which is pointed to on the WinCE Squeak website) there seems to be a limit on the size of image that works for me, even with 58 MB of RAM free, I can only load an image around 10 MB in size. I tried to use his beta VM that allow the user to set the size of the memory heap, but the slider control doesn't show up. I've had no problems running all sorts of standard image versions though. 0 tinyBenchmarks yields: 18017133 bc/sec and 633568 sends/sec. These numbers are very close to what the C7{5,6}0 get. They can be ordered through various Japan-US importers. The one with the best price (US$525) is Conics.net <http://conics.net/shp/pda/sigmarion3/> who also have some good photos up. Other places that carry them include Dynamism (US$599) and www.Japan-Direct.com (US$699). Regards, Aaron -- "Like the creators of sitcoms or junk food or package tours, Java's designers were consciously designing a product for people not as smart as them." :: paul graham

5 6

Re: [TFNR] Looking for maintainers!
by diegogomezdeck＠consultar.com 01 Dec '03

01 Dec '03

Hi Guys... [snip] > Volunteers? I want to postulate myself to be the maintainer (or one of the maintainers) of eToys and to be the "assembler" (or one of the assemblers) of the Full- Image. eToys: The initial set of goals I want to push in eToys are the followings: - Synchronize the Squeakland's image with the "big" one. The eToys in Squeakland's image evolved and these changes are not in the big image. - Make eToys translatable with Babel. It means to remove the current translation mechanism and replace it by Babel. - Coordinate the work with "Squeaklanders" to keep the 2 versions synchronized. - Expand eToys to cover more morphs and more functionality. -- Example: InternalThreadNavigatorMorph has to be scriptable from eToys to create active essays with complex navigation flows, etc. -- All the multimedia morphs has to be fully scriptable from eToys. MediaSqueak (aka Full-Image): For the Full-Image the goal is to make this distribution the "MediaSqueak". The full image has to include all the media stuff available in SM (Connectors, More graphics files support, Video Processing, TTF, etc). Also the full image has to include all the languages available for Babel and some good initial look&feel. Since SqC moves away we're working to have less in the image I think we need somebody that can take care of giving more in the full-image. What can I offer: Since January/2004 (or before) I'll have a full-time job with Squeak working with the folks from LinEx project. Just now we're using a customized 3.5 based image (The Small-Land's one). My plan is to combine these 2 projects and build the Small-Land image on top of the full one. I'd like to hear your opinions... Cheers, -- Diego Gomez Deck http://www.small-land.org

8 11

← Newer
1
...
4
5
6
7
8
9
10
...
57
Older →

Jump to page: