Squeak-dev June 2010

squeak-dev@lists.squeakfoundation.org

96 participants
269 discussions

Unicode strings, benchmarks
by Janko Mivšek 01 Jan '11

01 Jan '11

Hi Squeakers, I already extended String with TwoByteString and did a "scaling" with auto conversion to wider string when a wider character is put into a string. So far so good and this already works in Aida/Web. I also did a bit better UTF8 conversion but is only 25-80% faster that existing one in UTF8TextConverter. To prepare for even better results, I made a benchmark, which measure conversion time for English, French, Slovenian, Russian and Chinese 2500 characters long text. It measure 100 conversions which accumulates to 250K characters of text. Here are results in VW, Squeak with old UTF8 converter and a new one: VW old new english 30 313 248 ByteString, pure ASCII french 32 323 251 ByteString, ISO8859-1 (Latin 1) slovenian 48 578 480 TwoByteString Latin 2 russian 112 1306 720 TwoByteString Cyrillic chinese 107 1544 3825 TwoByteString Notice an exceptional 10x VW performance comparing to Squeak, and they do all encodings in plain Smalltalk! No primitives! So how come that Squeak is so slow here? Above benchmark was done on Squeak 3.9 on Suse Linux 10.1, P3.2GHz. Best regards Janko -- Janko Mivšek AIDA/Web Smalltalk Web Application Server http://www.aidaweb.si

9 17

Open Cobalt on Squeak 4.1 is available
by Matthew Fulmer 12 Dec '10

12 Dec '10

Hi. Open Cobalt now runs mostly as it should on Squeak 4.1. It is not a finished port, but it is definitely enough to start playing with, testing out OpenGL, packaging it up for installation into current images, or whatever. Up to date instructions are kept at http://cobalt.cs.duke.edu/pub/INSTALLING.txt Reply with any questions. Current copy of installation instructions reproduced below. ----------------------------------------------------------------- Welcome to the beta version of Cobalt ported to Squeak 4.1 To run me, you will need: - the latest cobalt-trunk.*.zip file from http://cobalt.cs.duke.edu/pub/ - the latest opencobalt-1.0alpha*.zip from http://www.cs.duke.edu/~jdougan/downloads/1.0alpha1/ - A recent VM from http://squeakvm.org - http://ftp.squeak.org/sources_files/SqueakV39.sources.gz 1. Unzip the opencobalt-1.0alpha*.zip folder somewhere 2. Unzip the cobalt-trunk.*.zip file inside the opencobalt-1.0alpha* folder. It contains 3 files: cobalt-trunk.*.image cobalt-trunk.*.changes config/cobalt-browser.conf 3. unzip SqueakV39.sources.gz and put SqueakV39.sources into the opencobalt-1.0alpha* folder 4. run cobalt-trunk.*.image using your VM 5. Do this: CProjectMorph open: CobaltModularPanel 6. Click the menu item Cobalt > Start Cobalt == About this image == This image is a trunk image from November 2009, updated to the latest code in the trunk repository, except where that conflicts with Cobalt. Feel free to make a better image based on the current, 4.1 images. Known issues are listed at http://cobalt.cs.duke.edu/pub/cobalt41todo.txt -- Matthew Fulmer (a.k.a. Tapple)

6 7

quick handling of graphics files
by Ross Boylan 01 Dec '10

01 Dec '10

I have PNG files with document scans, and would like to view them as part of an application. I will also need to scale and/or pan them. One file took over a minute to open when I selected "open graphic in window" from the file browser. (ImageReadWriter formFromFileNamed: 'myfile.png') asMorph openInHand had about the same speed. I'm looking for advice on how to speed that up, or the information that is not possible. I found a similar question on the newbie list, but no one answered it. I'm guessing that it is the initial read and translation of the file that's taking the time, since resizing the image is much faster than loading it. This is really too slow to be practical, and much slower than imagemagick's display command, which opens in under 5 seconds. The image is a 400dpi black and white scan of an 8.5x11" page; the file is 144k bytes. BTW is there any particular reason the file browser opens a SketchMorph while asMorph on the ImageReadWrite result produces an ImageMorph? The former is closer to what I want since it is resizeable and not transluscent; on the other hand, the comment in SketchMorph says they are produced by the PaintBox, which is not true in this case. Squeak 3.10.2 on Linux with 32 bit color in X. I think squeak is using 32 bit color as well, but couldn't find the setting on a quick look (not in preferences browser).

11 24

WideString hash is way slower than ByteString hash.
by christophe.jalady＠free.fr 30 Sep '10

30 Sep '10

Just print this two scripts: |s| s := WideString with: (Character value: 16r55E4). [100000 timesRepeat: [s hash]] timeToRun. |s| s := ByteString with: (Character value: 16rE4) with: (Character value: 16rE4) with: (Character value: 16rE4) with: (Character value: 16rE4). [100000 timesRepeat: [s hash]] timeToRun. I've seen that "ByteString>>stringHash:initialHash" use a primitive. Could we have such optimisation for WideString ? Christophe.

5 25

Etoys developer chat today
by Bert Freudenberg 27 Sep '10

27 Sep '10

Please join our weekly Etoys developer chat today, as every Monday at 12 noon Squeakland Time. Follow this link to see the time in your local time zone: http://tinyurl.com/368xxgn To participate, enter your name as nickname on http://chat.squeakland.org/ and press "Login". Then say hi :-) Alternatively, use an IRC client of your choice to join #etoys on irc.freenode.net Everyone is welcome! - Bert -

1 14

[Documentation] What about package comments?
by Hannes Hirzel 13 Sep '10

13 Sep '10

Hello I found the proposal by Bernhard Pieber and Ralph Johnson to introduce package comments? It was in the thread HttpView2 where I think some people interested in documentation will miss it. An application could be to automatically generate documentation, i.e. for example at Squeaksource. --Hannes >From the HttpView2 thread On 4/23/10, Ralph Johnson <johnson(a)cs.uiuc.edu> wrote: > On Fri, Apr 23, 2010 at 5:39 AM, Bernhard Pieber <bernhard(a)pieber.com>wrote: > >> - IMHO, there should be first class package comments along the lines of >> class and method comments, i.e. version controlled in PackageInfo. This >> would greatly improve the Monticello browser. And SqueakSource could use >> them as project descriptions. (I once argued [1] that the easiest >> implementation of this would be a PackageInfo subclass for each package. >> Its >> class comment would be the package comment. Pre and post load code could >> be >> put there as well. You could reuse all the development tools like that.) >> >> [1] >> http://lists.squeakfoundation.org/pipermail/squeak-dev/2005-February/088181… >> >> > This seems like a good idea. What tools would have to change to make this > work, and could we make this change backwards compatible? For example, if > you move the preload and postload code to the PackageInfo subclass then you > could leave forwarding code behind. Tools wouldn't bother displaying the > forwarding code if they saw the PackageInfo subclass, but would keep it so > that older versions of the tools would still work. > > -Ralph >

10 19

Etoys developer chat log
by Bert Freudenberg 08 Sep '10

08 Sep '10

Hi all, almost missed the chat yesterday. And when I logged in I got distracted by #olpc-devel talk. A minute later David came back, but I didn't see that. Sorry. But: If I'm not there, that's no reason not to chat :) Anybody can lead the discussion if it gets too disorganized, but usually we're pretty civilized, so it should rarely be necessary. So here is a rule: Do not wait more than 15 minutes. Just start. I'll see it in the logs later so I can still post a summary (though it would be awesome if someone beat me to it). Anyway, next chat is next Monday, same time as always: http://tinyurl.com/3ytdm3d Hope you'll be there (and I'll try to attend too). - Bert - [21:02:37] --> Richo2 (~c9d56472(a)skylab.immuexa.com) has joined #etoys [21:05:08] --> dcorking (~dcorking(a)82.152.173.174) has joined #etoys [21:05:13] <Richo2> hi [21:06:06] <dcorking> hi Richo2 - when does the meeting start? [21:07:36] <dcorking> is storybot down again? [21:11:53] <Richo2> I think it should have started already [21:13:09] <dcorking> I have only lurked in a couple of these. bertf normally leads them doesn't he? Is randall travelling at the moment? [21:13:31] --> scottwal (~scottwal(a)99-40-5-235.lightspeed.sntcca.sbcglobal.net) has joined #etoys [21:17:00] <-- Richo2 (~c9d56472(a)skylab.immuexa.com) has quit (Quit: CGI:IRC (Ping timeout)) [21:21:13] <bertf> hi folks - sorry I'm late [21:21:26] <bertf> you could've started already ;) [21:24:59] <bertf> dcorking: storybot has been down for months, and it's unlikely to come back [21:25:12] <bertf> jecel was looking onto other logging options [21:25:19] <bertf> for now I just email the logs manually [21:26:18] <dcorking> shame: I guess a decision should be made and the link removed from http://squeakland.org/discuss/ [21:36:37] <dcorking> is the meeting starting? if not, any thoughts on jecel's idea on porting Cog VM to the XO-1? [21:38:46] <jecel> I was looking at the Cog sources to get an idea of how complex this would be. It seems doable, but not easy [21:40:05] <dcorking> hmm. Sounds like something that would distract the squeak experts from doing the next Sugar release on time, and not something for average smalltalkers. [21:40:25] <jecel> About the logging, I didn't remember we had agreed that I would check the options and so haven't yet done so [21:40:47] <dcorking> however, I wonder if qemu could be used to emulate the missing instructions: at least as a proof of concept [22:07:44] <dcorking> is the meeting over? [22:10:43] <dcorking> does the showcase still give trouble when a user uploads a new version of an existing project? [23:47:14] <-- dcorking (~dcorking(a)82.152.173.174) has quit (Quit: Leaving)

1 5

Documentation Team
by Casey Ransberger 25 Aug '10

25 Aug '10

Looking at http://squeak.org/Community/Teams/ I don't see a documentation team listed. Has it been disbanded? If the documentation team has been disbanded, I would like to offer to lead a new documentation team (in which case I would draft up a proposal with expected deliverables, etc.) If the documentation team has not been disbanded, I would like to join it.

7 16

Colstme in June
by Alexander Lazarević 03 Aug '10

03 Aug '10

Hi! Just a short reminder that the next Cologne Smalltalk Meetup will be happening this Thursday at the usual time and location. For any details have a look at [1] and [2]. Cheers, Alex [1] http://colstme.selfish.org/events/stammtisch03 [2] http://colstme.selfish.org/blog/2010-06-02

1 2

Nuking OldSocket, OldSimpleClientSocket and HTTPSocket
by Andreas Raab 16 Jul '10

16 Jul '10

Hi - I'm interested in doing a little assault on the aforementioned classes. The only reason for the existence of Old[SimpleClient]Socket is that they're superclasses of HTTPSocket which is a beast to get rid of since we need it for updating etc. However, with WebClient [1] at hand, we can convert HTTPSocket into a simple facade that just delegates everything to WebClient and in addition to having a better HTTP client implementation we could also get rid of these Old* classes. What do people think? Is there any value in preserving the Old* classes and HTTPSocket or should we try to take an end-run on the issue? Obviously, there will be some tricky updating involved so I'd like to get opinions beforehand. Cheers, - Andreas [1] http://www.squeaksource.com/WebClient.html

8 21

Jump to page: