Squeak-dev December 2001

squeak-dev@lists.squeakfoundation.org

137 participants
354 discussions

[GOODIE] TestMaker (0.2 - was: Re: [GOODIE] TestMaker)
by Daniel Vainsencher 23 Dec '01

23 Dec '01

"David T. Lewis" <lewis(a)mail.msen.com> wrote: > On Fri, Dec 21, 2001 at 10:27:08PM +0300, Daniel Vainsencher wrote: > This is a very good idea, I like it. Hey David. Thanks. > As long as it is OK to propose things > that create work for other people, I have some suggestions ;) Hey - no fair - you want to make suggestions, you're supposed to also cough up some helpful code :-) In the spirit of Goran's brave commitment (Go CSOTD!), give us a test, or implement at least one of the suggestions... > - It would be good if ParagraphEditor>>createTestFor: could suggest the > right class name and selector to use. This would encourage consistency > in naming, and would help folks who are not familiar with the naming > conventions. Attached a modified version that gives some helpful hints and offers to create the class if it doesn't exist. Maybe someone can make it show a list of subclasses? > - This deserves its own 'make test case' menu choice. Sure, I just really don't like doing UI integration under the current scheme with ParagraphEditor and friends. Never quite got the hang of the MVC/Morphic compatibility scheme. I'm hoping someone who does will take it up. For my part, I factored most of the functionality into a small class, so it's not too messy, and easy to call from wherever is the right place. > - Perhaps some creative person could do a UI that generates the test > cases. I could imagine one pane in the UI which would contain a (possibly > empty) list of expressions which should succeed, and another pane > containing expressions which should fail. The end result might look > like a "TestCaseBrowser" which could be used for browsing classes and > interactively generating test cases. If it could also interactively run your tests and show you how they fail, that'd be very nice. A guy called Peter demoed an interesting test-based development tool at ESUG last year, for one of the other Smalltalk platforms. I think SUnit is just begging to have a smarter tool suite built around it - something to avoid bogging down in irrelevant, long running tests that are certain not to succeed when your basic ParserTest isn't even running, let alone passing, for example... > This is very nice, thanks! Daniel from preamble: "Change Set: TestMaker Date: 21 December 2001 Author: Daniel Vainsencher This changeset replaces print-it with a test creation utility. Select something that raises an exception or returns the wrong answer, answer a few questions, and voila, a new test. If someone posts tests and it helps them do it, I'll polish this scrap into an ENH..."!

2 1

RE: [rant] open source and support (was owner of squeak)
by G.J.Tielemans＠dinkel.utwente.nl 23 Dec '01

23 Dec '01

> -----Original Message----- > From: John Hinsley [mailto:jhinsley@telinco.co.uk] > Sent: vrijdag 21 december 2001 18:18 > To: squeak-dev(a)lists.squeakfoundation.org > Subject: Re: [rant] open source and support (was owner of squeak) > > > G.J.Tielemans(a)dinkel.utwente.nl wrote: > > John: OK, rises to bait (I'm getting so predictable) although I agree with > most of this ;-) > > > Suppose you go to your boss and tell him all about Linux > > and Staroffice: > > > > - free stuff, so no one is earning money on that stuff? > > - giving away for free?!? you are kidding... > > - then there is no money for supporting the product. > > But that's _precisely_ how people make money out of Open Source. > You take your choice, everything from free support via > newsgroups to a team of dedicated IBM folk available 24/7. > > > > ... Of course we then have to admit that in the Squeak > > community we only do what we like, > > You cannot build a business on our fun: > > - Problems with Squeak under Windows? buy a Mac! > > - Browser problems with the Squeak plugin? change the > browser(version)! > > Mmmmm. Or sue Microsoft. > Certainly, if I was hiring someone to do Squeak Ger: hmm, close your eyes and the world does no longer exist, didn't we do that when we were (very) young? As I said often on this list: IF you want to live outside the Squeak-zoo THEN you have to accept that people play dirty in the real world. > support for a M$ based business, these are some of the issues > I'd expect them to look at. But instant solutions are no more likely to be > available at Squeak support com than at Microsoft. IIRC, it > took Cincom > 2 or 3 weeks to isolate and cure the "tiny print" problem people were > experiencing with VisualWorks (eventually tracked down to an > undocumented change Microsoft had made). > No one took them to court, and seems to me that that was a pretty fine effort. > Ger: Yes I know this kind of games: Mid 1997 we selected Oracle InterOffice as the new foundation for our groupware-support in education. Jan 1998 Oracle took it back from the market, as a spokesman said: "Because Microsoft everytime did change the (Ger: undocumented?) functions under the hood of windows. May 1998 we had to deliver.... so we took our refuge too Lotus Domino: - in these days a great product with a terrible marketing.... - Now it is in the hands of IBM, better marketing, yes... for websphere... > > (Although I have to admit that in other communities like > SourceForce the community looks much more proficient Apache, PHP, MySQL... > and so many packages, right-out-the-box (PHPDEV for example)) > > I think Squeak support is pretty proficient. But the larger the > community (and the Apache and Linux communities are huge, with a much > higher developer to user ratio than M$, for obvious reasons) > the quicker and more wide ranging support is likely to be. > Ger: yes you are right... how many Squeakers are there really? (Not only platforms but real people > > > > Another point is that normal people have habits and want to > keep these: If you offer a product like StarOffice, that is so different > in look and feel, then you have to invest a lot of extra time before you can > go back to your normal work... > > Is Star/Open Office _so_ different? I'd think that after 2 hours group > training anyone would be able to make the switch, although > it'd probably take around a week to become equally proficient. Ger: do you remember your first switch to your second wordprocessor? (or platform? I believe that lots of people are still on their first) > I hear that Ford are looking at switching.... > Just think: simply by ditching M$ Office you save £150, > ditch the OS as well and you save around £200 per user. Ger: Yes, isn't that a fascinating calculation? Or are to many people using illegal software? (Trying to bring a faculty on the legal software side, a university professor flamed me for blocking Univwersity research (1995) > Chances are that you can leave your 2 yearly upgrades another year or > two. Hell, I've seen it argued that it's cheaper to build a > Linux based system for 500 users (not 500 terminals) around a new IBM > mainframe than it is to do it around Windows 2000. > Ger: You look at the details, so -- are you a manager?.... > > > And last but not least: You do not need the best product as > > a businessman, you just need the best advertisement and other tricks to > > grab the market. > > The desktop market was always going to be the place Microsoft circled > the wagons. It's also the only place where the user base is hopelessly > uneducated and naive. Elsewhere, Apache has pretty much gobbled up the > webserver market simply because, while it's _not_ the best > webserver, it > offers a huge range of features and the best > speed/cost/reliability/configurability/multi-platform compromise. > Ger: Yes, in the past I did look at GNU & Emacs, always a promise for the future, but now looking at SourceForce and the useful examples they deliver, I believe that free becomes professional and also useful in busimess environment. Squeak has potential to survive ther too: if it can create the spin-off I did expect from Squeakland when I first saw it: solving the plug-in problem is for me the ultimate test... (For myself: repair the upload problems in BigTimeSwiki and try to communicate with SourceForce initiativs like the Apache XML project, it would be fascinating..) > Cheers > > John > -- > Reputed to be the reason Windows 2000 was nearly a year late, (paid in > shares M$Ds needed the cash and kudos) Netproject's Eddie > Bleasdale has, > renewed his challenge to virus writers. The first person to infect his > Linux box wins 10,000 pounds. > > http://www.silicon.com/bin/bladerunner?30REQEVENT=&REQAUTH=210 46&14001REQSUB=REQINT1=48211

4 3

[FIX] assert: LargeIntegers
by Marcus Denker 23 Dec '01

23 Dec '01

"Change Set: assertFIX-md Date: 22 December 2001 Author: Marcus Denker fixes a problem building the LargeIntegers Plugin on MacOS X by renaming assert: to sqAssert:" -- Marcus Denker marcus(a)ira.uka.de -- Squeak! http://squeakland.org

1 0

[ENH] QueueSoundEnh
by Eddie Cottongim 23 Dec '01

23 Dec '01

from preamble: "Change Set: QueueSoundEnh Date: 23 December 2001 Author: Eddie Cottongim Adds a class comment to QueueSound with particular attention to the done: method so that people won't leave QueueSounds playing indefinitely. Modifies QueueSound's nextSound method to trigger #soundFinished each time it is called so that you can keep tabs on a QueueSound using events instead of polling. "!

1 0

[VM][Simulator] InterpreterSimulator broken? (repost)
by Stephan Rudlof 23 Dec '01

23 Dec '01

Dear Squeakers, since there hasn't been any answer, I repost this, though I know that it's really hard stuff for Christmas! I'd really like to know if InterpreterSimulator>>testBecome works for anybody! This check shouldn't take much time for someone already using the simulator ;-) BTW: my modified VM works now, but this bug appears as for it as for standard VMs. Greetings, Stephan ----- Dear Squeakers, I have problems with the InterpreterSimulator. The InterpreterSimulator>>testBecome hasn't been successful yet. #become:with:twoWay: works for the created arrays, but *not* for the values on the stack. I have played with a modified VM, but since it doesn't work, I have tried the simulator. After faced with the problem described here with my modified VM I tried the normal one, but there is the same problem :-(( I'd like to know, which are my and which are other errors... Does InterpreterSimulator>>testBecome work for somebody? Does anybody know this problem? Greetings, Stephan Look for "!!!" below: all values on the stack haven't been affected by the #become:with:twoWay: operation there (before the #popStack in this line). Same result as with Squeak3.2alpha#4567 as with (last) Squeak3.1beta#4478. InterpreterSimulator>> testBecome "Become some young things. AA testBecome " | array list1 list2 p1 p2 p3 p4 | array _ self splObj: ClassArray. list1 _ self instantiateClass: array indexableSize: 2. list2 _ self instantiateClass: array indexableSize: 2. p1 _ self instantiateClass: (self splObj: ClassPoint) indexableSize: 0. self push: p1. self storePointer: 0 ofObject: list1 withValue: p1. p2 _ self instantiateClass: (self splObj: ClassPoint) indexableSize: 0. self push: p2. self storePointer: 1 ofObject: list1 withValue: p2. p3 _ self instantiateClass: (self splObj: ClassMessage) indexableSize: 0. self push: p3. self storePointer: 0 ofObject: list2 withValue: p3. p4 _ self instantiateClass: (self splObj: ClassMessage) indexableSize: 0. self push: p4. self storePointer: 1 ofObject: list2 withValue: p4. (self become: list1 with: list2 twoWay: true) ifFalse: [self error: 'failed']. "!!!" self popStack = p2 ifFalse: ["stops here!" self halt]. self popStack = p1 ifFalse: [self halt]. self popStack = p4 ifFalse: [self halt]. self popStack = p3 ifFalse: [self halt]. (self fetchPointer: 0 ofObject: list1) = p3 ifFalse: [self halt]. (self fetchPointer: 1 ofObject: list1) = p4 ifFalse: [self halt]. (self fetchPointer: 0 ofObject: list2) = p1 ifFalse: [self halt]. (self fetchPointer: 1 ofObject: list2) = p2 ifFalse: [self halt]. BTW: Not using #pushRemappableOop: here doesn't seem to be clean to me, but this hasn't been the problem here for me (at least with enough simulator mem reserved). -- Stephan Rudlof (sr(a)evolgo.de) "Genius doesn't work on an assembly line basis. You can't simply say, 'Today I will be brilliant.'" -- Kirk, "The Ultimate Computer", stardate 4731.3

1 0

Type Inferencer to the Rescue
by ducasse 22 Dec '01

22 Dec '01

Hi I **tried** to fix the unexpected behavior of SortedCollection accepting addLast: (this was easy and dirty like in VW). add: should call super addLast:. Then I tried to verify if clients of SortedCollection would not break so I looked for all the senders of addLast: and tried to see what was the type of the receiver......endless. I tried to check all the references to SortedCollection and crosscut this with the previous set and taking into inheritance.....nearly impossible. So I would like to know what is the status of the type inferencers? Is Lucid still developed? I know that Franscico would like to make his type inferencer operational for crazy code containing primitives. I was wondering if somebody is still working on the topic and also if other alternative like spying the run-time of squeak and collecting types have been explored. Does somebody worked on something similar as the CoverageBrowser developed by michel tilman? Stef By the way how harvesters of bug fixes operate?

1 0

Leandro's Traffic Light Morph
by John Hinsley 22 Dec '01

22 Dec '01

I'm (slowly) translating LC's tutorial into English, re-doing the gifs and methods to match 3.2 as I go along. Slowly because Morphic has changed quite a lot in the meantime and my Spanish is dreadful (usually I make as much sense as I can out of the Spanish by pretending it's Italian and then grab the dictionary!): but I recon that if I don't do it, it'll never get done. At some point I decided to get it running and got ahead of the translation. Anyway, I rewrote the initialization method, thus: | green orange red | super initialize. self listDirection: #topToBottom. self wrapCentering: #center; cellPositioning: #topCenter. self extent: 66@152. self color: (Color r: 0.0 g: 0.355 b: 0.258). self borderWidth: 8. self borderColor: (Color r: 0.0 g: 0.581 b: 0.613). green _ LampMorph new. green color: (Color green). orange _ LampMorph new. orange color: (Color orange). red _ LampMorph new. red color: (Color red). self addMorph: green; addMorph: orange; addMorph: red I guess I could improve on the indenting, but this works just fine. Is it acceptable? Can I improve on it? BTW, some of the methods in the Morphic version seem to be missing altogether. Is this intentional (as in Squeakland's "hard learning" or whatever?)? Cheers John -- Reputed to be the reason Windows 2000 was nearly a year late, (paid in shares M$Ds needed the cash and kudos) Netproject's Eddie Bleasdale has renewed his challenge to virus writers. The first person to infect his Linux box wins 10,000 pounds. http://www.silicon.com/bin/bladerunner?30REQEVENT=&REQAUTH=21046&14001REQSU…

2 1

An Idea: A bit more support
by Stephane Ducasse 22 Dec '01

22 Dec '01

Hi I was wondering (may be in the same way as the Squeaks news) that it would be good to have a CD release every six months or yearly of Squeak containing the stable versions well organized and self contained (meaning including the source in the zip) so that normal people do not have to look around to find squeakV3.sources.... plus the actual code, goodies.... in synch with the version. It would be easier for some people to just pay for having a complete CD than to download stuff over the web. I'm thinking about teachers in particular. The SqueakLand site is good for people connected. Then as a university we could buy this cd to support Squeak community. Tell me what you think. Stef PS: this makes me wondering the status of the squeak foundation. Any information?

7 8

[GOODIE] TestMaker
by Daniel Vainsencher 22 Dec '01

22 Dec '01

A few blessed voices have been heard about adding more tests to the image. Unfortunately, our system is a little biased against this - * There's no really trivial way to post a Test. * SUnit is new - we're not that well versed making them. * If you're coding an ENH or Goodie, you're somewhat likely not to feel very militant about finding bugs in it right away, you're too happy the darned things works. In the tradition of proposing things that create work for other people, but maybe people will like it: * Add a small, simple tool for adding tests, and a button for mailing them to the list. The hack attached below is an initial implementation. * Add a fast track in the harvesting process for tests. There are various reasons to do this - - A test is usually trivial to integrate, so it should be pretty easy. - Then people can easily submit this sort of report in an effective way that'll stick a red light in everyone's eyes until it's fixed. - Then the Fix category of mails can have a very simple definition (Fixes CelesteTestCase>>mumbleMumble) which would make *them* easier to integrate too. - People would have a good expectation that a test will be put in the image and then promptly fixed. This will encourage reporting bugs in the best way known, which should help quality... * Simply submit Tests instead of complaining, like Stef did. If a few more people do this, others will have lots of examples to learn how to write good tests, which is a potential good causality loop. from preamble: "Change Set: TestMaker Date: 21 December 2001 Author: Daniel Vainsencher This changeset replaces print-it with a test creation utility. Select something that raises an exception or returns the wrong answer, answer a few questions, and voila, a new test. Doesn't do error checking, and it should offer to create a new test class or select from the existing ones, and offer to mail the test right off to the list. If someone posts tests and it helps them do it, I'll polish this scrap into an ENH..."!

3 2

Gutenberg is for beta testers
by ducasse 22 Dec '01

22 Dec '01

Hi I started to introduce a visitor for the ParseNode hierarchy. Based on this visitor I wrote a pretty printer called Gutenberg (I moved all the print related methods on visitor) so that we can create better and easy parse tree wlaking and decorating tools. Now I have a early beta pretty printer (just show code correctly not pretty printed ;)) and a SQ00 syntax pretty printer (not even in alpha). I would like to know what a selector only containing : ? Then if you are interested contact me. Stef

1 0

← Newer
1
...
6
7
8
9
10
11
12
...
36
Older →

Jump to page: