Squeak-dev January 2001

squeak-dev@lists.squeakfoundation.org

112 participants
318 discussions

Squeak VM for Zaurus PDA
by Yoshiki Ohshima 29 Jan '01

29 Jan '01

Hello, I updated the VM for Zaurus PDA series. Supported hardware include the latest MI-E1. The VM is based on 2.9 alpha source, but it has many hardware and OS specific VM for ... a project. See http://www.is.titech.ac.jp/~ohshima/squeak/squeak-zaurus-e.html for more detail. -- Yoshiki

1 0

Changes acceptance and polling issues
by Lex Spoon 29 Jan '01

29 Jan '01

"Peter Schuller" <peter.schuller(a)infidyne.com> wrote:> > Similarly, the IRC implementation is a polling one (you have to invoke > processIO repeatedly, instead of the I/O running in as eparate thread). > As a result, I'll probably write an alternative implementation for personal > use. > In this case, it's a deliberate design issue. It's certainly open for debate, but let me explain the reasoning. First, using a blocking interface means adding a thread, and adding a thread add a lot of complexity. This includes design work (where do the locks go?), complexity (implementing and using the locks), and added difficulty in testing (let's run this code a million more times and try to make the bug appear again). IMNSHO, many programmers severely underistemate what they are getting into with threads. To contrast, the added complexity for polling is something like "how does my polling routine get invoked"; once that is worked out, the focus is on the function of the system. Second, there are cases where polling doesn't have the ills that are usually ascribed to it. In particular, polling works well when failed polls are cheap, and where polls can happen infrequently. If a problem meets both of these criterea, then polling will work fine. CPU usage will not be abnormally high, and responsiveness will be good. IRC meets both criteria, while HTTP serving probably does not. > And finally, there is the GUI which isn't entirly event driven (eg, moving > a window consumes 100% CPU In Morphic, dragging a large window easily uses up all of my 366 MHz CPU, and it could easily use up 3-5 times more, as well. It's not wasted CPU in this case -- there are a lot of bits being shoved around! In MVC, or with "fastDrag", well, the CPU is wasted. See below. ; and in MVC (but not in morphic), pop-up menu > navigation also consumes 100% CPU; That's a deep design point in MVC: the currently active controller polls the mouse and keyboard. This dates back to machines where Smalltalk was the OS, or at least the shell, of single-user ("personal") computers. Morphic makes more sense for Unixy systems where there might be multiple users, and where CPU should be shared. Incidentally, I went on a CPU-squashing hunt in MVC a while back, and those squashes are in Squeak nowadays. I didn't try to squash all the instances, but instead only worried about cases where the user is idle. The idea was, the important case is when the user really is idle. If a user is busily clicking around, then using 100% CPU doesn't seem like a very big deal. Furthermore, there are simply a *lot* of polling loops for various kinds of interactive code in MVC. Fixing them would mean adding a lot of code, and I'm not sure it would be worth it. Morphic is a different story. In Morphic, the system can simply limit it self to 50-100 frames per second (whatveer you set it at). If the framerate excedes this, Morphic slows down. It only requires a little extra code in the main Morphic loop, and it's all finished. This difference in MVC versus Morphic is a classic case of having the right abstractions for a job being explicit in the code. -Lex

2 1

RE: RE: RE: RE: newbie. morphic is sluggish on my PC
by Pennell, David 28 Jan '01

28 Jan '01

[snip] > > so, the next questions are: > > - What does the morphic world contain, and > - Where was the mouse, relative to the stuff in the world, > while the MessageTally was running? For that tally, the world contained a browser and a workspace. The mouse was clear of both. I just ran it again with a completely empty world and the results are similar. BTW - if I collapse Squeak completely, the Windows CPU monitor still indicates the same 30+% activity. -david

1 0

Re: RE: RE: RE: newbie. morphic is sluggish on my PC
by Bob Arning 28 Jan '01

28 Jan '01

On Sun, 28 Jan 2001 12:32:55 -0500 "Pennell, David" <DPennell(a)quallaby.com> wrote: >Ah, very nice... > >Looks like a lot of time spent in an expensive primitive... David, In your MessageTally, - 741 tallies, 18247 msec. **Tree** 99.9% {18229ms} DisplayScreen>>doOneCycleMorphic 99.5% {18156ms} PasteUpMorph>>doOneCycle 99.3% {18119ms} WorldState>>doOneCycleFor: 66.7% {12171ms} WorldState>>interCyclePause: |66.3% {12098ms} Delay>>wait | 66.1% {12061ms} primitives 28.6% {5219ms} WorldState>>doOneCycleNowFor: |19.3% {3522ms} HandMorph>>processEvents | |18.1% {3303ms} MouseOverHandler>>processMouseOver: ... the 66.7% {12171ms} WorldState>>interCyclePause: |66.3% {12098ms} Delay>>wait | 66.1% {12061ms} primitives indicates you are giving up control to the OS about 66% of the time (or Squeak is using 34%, conversely). Using 2.9a on a Mac, I see - 669 tallies, 13004 msec. **Tree** 100.0% {13004ms} DisplayScreen>>doOneCycleMorphic 100.0% {13004ms} PasteUpMorph>>doOneCycle 100.0% {13004ms} WorldState>>doOneCycleFor: 99.7% {12965ms} WorldState>>interCyclePause: 99.7% {12965ms} Delay>>wait **Leaves** 99.7% {12965ms} Delay>>wait which indicates Squeak is essentially giving up complete control to the OS. The difference between mine and yours is that yours has 28.6% {5219ms} WorldState>>doOneCycleNowFor: |19.3% {3522ms} HandMorph>>processEvents | |18.1% {3303ms} MouseOverHandler>>processMouseOver: so, the next questions are: - What does the morphic world contain, and - Where was the mouse, relative to the stuff in the world, while the MessageTally was running? Cheers, Bob

1 0

Re: newbie. morphic is sluggish on my PC
by Ned Konz 28 Jan '01

28 Jan '01

On Sunday 28 January 2001 07:28, Pennell, David wrote: > Thanks, but I'm not sure how I helped. I think I'm still > having a serious problem with 2.9 performance. > > I'm running a 2.9a image with updates through 3193. > If I open a fresh Morphic project and enter it, it uses > 30-35% of my CPU. If I do the same with a MVC project, > it uses 100% of the CPU. I see the same results with > the 2915 and 3125 updates that Bruce published. On Linux, I see a fresh MVC project using between 3% and 95% of a CPU. If I open a single browser window in that MVC project, usage goes down to 6 or 7 percent. There seems to be an inefficiency if there are no MVC windows. A fresh Morphic project uses about 10% of a CPU; if I kill off the project navigator it goes to about 6%. -- Ned Konz currently: Stanwood, WA email: ned(a)bike-nomad.com homepage: http://bike-nomad.com

1 0

Re: Comanche on Linux eating CPU?
by Mark A. Schwenk 28 Jan '01

28 Jan '01

Cees de Groot wrote: > > I've seen Comanche eating CPU on Linux boxen twice now - the first time > I thought it was because I had lots of changesets in the image that could > be the cause, but my second installation was from a clean image and had > the same behavior. It's not much, maybe ~20%, but on an Internet server > it's a waste of cycles (especially when said server is doing some unimportant > background processing, like hosting 8,000 domains ;-)). > > Both installations where Squeak 2.8, Linux 2.2.x on a multi-processor box, > with the VM attached to an Xvnc server. Anything that I'm doing wrong > with this setup? > > -- > Cees de Groot http://www.cdegroot.com <cg(a)cdegroot.com> > GnuPG 1024D/E0989E8B 0016 F679 F38D 5946 4ECD 1986 F303 937F E098 9E8B I'm using Xvnc also on Linux 2.2.x (some multi-processor and some uniprocessor systems). With my swikis, I've mainly had CPU hogging when a morphic project is left open. Leaving an MVC project open instead has dramatically lowered the CPU usage. What type of project is open in your image? -Mark Schwenk WellThot Inc.

1 0

RE: Squeak as an NT service
by Raab, Andreas 27 Jan '01

27 Jan '01

Peter, It may certainly be useful, but the fact of the matter is that I have no idea of how to make it happen :-( Any Windows hackers out there who can help?! - Andreas > -----Original Message----- > From: Peter Crowther [mailto:Peter.Crowther@melandra.com] > Sent: Friday, January 26, 2001 1:22 AM > To: 'squeak(a)cs.uiuc.edu' > Cc: recipient list not shown > Subject: RE: Squeak as an NT service > > > > From: Raab, Andreas [mailto:Andreas.Raab@disney.com] > > As I said in my previous post you cannot interact > > (graphically) with Squeak > > when it's running as a service. While this may be > > theoretically possible it > > would also imply that everyone who is logged on could do > > whatever she wants > > with Squeak and that is most certainly not what you want. > > The same's true for anything set up that way, but it's still useful on > standalone server in order to start the service as the > machine boots. If > you don't do this, you need to configure auto-login on an NT > box, which has > a nasty side-effect of leaving an unencrypted password in a > public area of > the registry that can be read remotely. > > Just make sure you put the server in a safe place :-). > > - Peter >

1 0

Re: [OT] Frame and Word [Was: Squeak Books!]
by Alan Knight 27 Jan '01

27 Jan '01

Interesting. I just did write a book using Word (Mastering ENVY/Developer), and it wasn't a particularly unpleasant experience. Wait, let me rephrase that. The word processing software was the least of my worries. I think it depends a lot on how close to the final product you have to get. Writing something in Word where you want to get exact formatting is fairly difficult, but doing something that is in the right general ballpark can be quite easy. Since all I had to do was define styles (which were not supposed to look particularly like the final output) and explain to the typesetting people what a particular style meant (and they had a pretty broad range of predefines ones to start with) it was quite straightforward. It also depends on size. Administrative staff tend to write things a few pages long, and for administrative documents precise look is not all that important. At 04:50 PM 1/26/01 -0800, Andrew P. Black wrote: >I have an explanation for this. I think of writing like programming. I >think: I would like more space after this kind of paragraph, so I change >the definition of "this kind of paragraph". I think: "I need a new >subsection here", or, "I need a list here", and I "send the message" >(i.e., use the style) "subsection" or "list". Initially, I start with >one of the sample documents that come with Frame, because they have all of >these styles defined, and they look good. If I want to change the >appearance later, I know I can do that -- just import a new batch of >paragraph definitions. Another example: the way that numbering schemes >for paragraphs and figures work will be immediately intuitive to any >programmer: it's based on named variables. > >My administrative staff thought of writing like typing: what you see is >all you've got. If you want a bigger paragraph break, type two <CR>s. If >you want a subsection, type some text in bold, type <CR>, and presto, it >_looks_ like a subsection heading. You want to make paragraph breaks 1.5 >lines insead of 2 lines? You want to change those subsection headings so >that they are in small caps in the left margin? Sorry, you can't get >there from here. > >Hence, I love Frame, and they hate it. The only thing that I don't like >in Frame is the UI for the equation editor; too much mousing, and I can >never seem to select what I want. > >I refuse to use word, except that I have it so that when people send me >Word documents, I can read them. Or, at least try to. A word document >looks different on MacOS from the way that it looks on Windows; the >pagination changes, as Mark mentioned. One thing that I notice is that >the Word documents that I'm sent almost always use visual formatting, and >rarely use styles. > > Andrew -- Alan Knight [|], Cincom Systems knight(a)acm.org aknight(a)cincom.com http://www.cincom.com/smalltalk/downloads.html

1 0

Re: Squeak V3 (was: Final call for 2.9 updates)
by Tim Rowledge 27 Jan '01

27 Jan '01

Henrik Gedenryd wrote: > > Bert Freudenberg wrote: > > > There were a lot of modifications proposed that would justify a new major > > version when finished: Morphic clean up, Network streaming support, new > > method format, actually using Exceptions, block closures, you name it. > > Eh, that sounds more like 7.0 to me ;-) Anyway, having version 2.11 etc. is > at least as bad imo. How does someone new know that this is not 2.1.1? > > There should be the new CompiledMethod format, as a major version number has > been said to signify image incompatibility. It's probably a couple of weeks work to bring this up to date; I don't even know if the Systemtracer still works in 2.9. I can do it, but I'm certainly not going to spend time on it unless I know it will be used. tim

1 0

Squeak development screenshot
by Raymond Asselin 27 Jan '01

27 Jan '01

This one case of 'Skeak Development Environment'... You have in the same envionment: - A transcript - A browser (Whisker) - A CRC cards Book - Object Diagram build simply with rectangleMorph and connectorMorph So you can go from CRC cards to Object Diagram (or Class Diagram) to code... Nice. I tried to upload that screenshot to the Swiki where are the Squeak's ScreenShots, but it happen that until now I never been able to upload something to a minnow swiki I don't know why (my browser is ICAB but I tried also with Netscape without success... So I send it to the list...

1 0

← Newer
1
2
3
4
5
6
7
8
...
32
Older →

Jump to page: