Squeak-dev July 2000

squeak-dev@lists.squeakfoundation.org

121 participants
566 discussions

Re: Squeak Swiki out of order again
by Mark Guzdial 05 Jul '00

05 Jul '00

At 10:46 AM +0200 7/5/00, Bert Freudenberg wrote: >There could be a read-only mirror (for example at the european gsug.org >site): Each time a page is modified in the master swiki, a rendered >version of it could be uploaded to a special directory, an email >notification about the changed page would be sent to the mirror, it would >fetch the page. While rendering, the "special links" like edit, upload, >history etc. should point to the master swiki, while all others should be >relative. So if the master goes down, reading would still be possible. That's already in there, as it was in PWS. The problem is that rendering is more complicated in ComSwiki since page names can be renamed and thus impact many other pages. We could render at a regular time each night, though... Mark -------------------------- Mark Guzdial : Georgia Tech : College of Computing : Atlanta, GA 30332-0280 Associate Professor - Learning Sciences & Technologies. Collaborative Software Lab - http://coweb.cc.gatech.edu/csl/ (404) 894-5618 : Fax (404) 894-0673 : guzdial(a)cc.gatech.edu http://www.cc.gatech.edu/gvu/people/Faculty/Mark.Guzdial.html

1 0

SVG in Squeak?
by Peter Munro 05 Jul '00

05 Jul '00

Hi, Is anyone working on SVG (Scalable Vector Graphics) support in Squeak and/or Morphic? I was thinking of a few applications, like visualisation, navigating (perhaps remote) object repositories (eg for a web server's contents), visual simulation, model building/viewing. Besides, I'm sure the Disney folk could put it to some good use. I don't know how best to integrate it with Squeak, but an SVG-capable morph might be a nice idea. Of course, we'd also need form & canvas drawing ability. If you want more info on SVG, look at: http://graphicssoft.about.com/compute/graphicssoft/msubformatssvg.htm http://www.w3.org/Graphics/SVG/ http://www.adobe.com/svg/ Cheers, Pete

1 0

Re: Squeak Swiki out of order again
by squeak＠sysa.barnett.sk 05 Jul '00

05 Jul '00

hehe, When you will broke one leg, you will not fix it but you will take polaroid snapshoot to have picture of all of them, in case you will broke the second one :) ... Jan Barger http://www.barnett.sk http://www.napri.sk "Those who know do not say; Those who say do not know." Lao-tzu's http://www.ccnet.com/~elsajoy/spiritus3.html On Wed, Jul 05, 2000 at 10:46:26AM +0200, Bert Freudenberg wrote: > On Tue, 4 Jul 2000, Andrew C. Greenberg wrote: > > > Not surprising during a holiday. I'm sure Mark and company will have > > it up soon enough. > > It's still dead. > > > Next step: can we build a redundant slave server that can mirror the > > Swiki and take over the IP address if a heartbeat from the primary > > server fails -- now THAT would be something. Indeed, perhaps we can > > arrange for it to do reasonable load-balancing in the interim? > > There could be a read-only mirror (for example at the european gsug.org > site): Each time a page is modified in the master swiki, a rendered > version of it could be uploaded to a special directory, an email > notification about the changed page would be sent to the mirror, it would > fetch the page. While rendering, the "special links" like edit, upload, > history etc. should point to the master swiki, while all others should be > relative. So if the master goes down, reading would still be possible. > > -- Bert >

1 0

Re: [ENH][ANSI] Collection>>allSatisfy:
by Marcel Weiher 05 Jul '00

05 Jul '00

> From: Michael Lucas-Smith <ich(a)driftwood.darktech.org> > I could have _sworn_ this was called #conform: aBlock ...? Maybe thats > just an IBM Smalltalk thing? Well, the ANSI spec, or at least the draft that I have, calls it #allSatisfy:, and Squeak does have #anySatisfy:, but neither #conform: nor #allSatisfy: . So it may actually be just an IBM Smalltalk thing, but I don't really know that much about standard protocols. Incidentally, I found the ANSI spec an excellent overview of standard protocols, with everything collected in neat summaries. That does rekindle the desire for formal protocols, with protocols managing the message name, a separate protocol hierarchy and browser and even [gasp!] optional means of specifying protocols in messages. The common syntactic means of specifying protocol (types) in Smalltalk extensions seem to be angle brackets, <Integer> myInt. Wouldn't it be a good idea to modify the standard compiler so that it ignores such constructs? That way, experiments in that direction could produce code that's still compilable on everybody else's system. Marcel

1 0

Re: [Question] SUnit 2.7 and Squeak: how do you structure test
by Robert Hirschfeld 05 Jul '00

05 Jul '00

Piero, In the 2.8 port of SUnit you can limit the test cases executed by providing selection patterns for the subclasses of TestCase and their test methods. For instance, if one of your projects uses a classname prefix, say 'Xxx', to simulate a namespace, you can use this prefix to select those classes via [TestRunner new runTests: 'Xxx*'] To narrow down the set of tests further, you can add another selection pattern for the names of actual test methods to be run like all test methods that end in 'Yyy': TestRunner new runTests: 'Xxx* test*Yyy' More convenient though is the use of the TestModel (open it with TestModel openAsMorph in a morphic environment, or with TestModel open in a MVC environment. The mechanism to select specific tests is the same. -Robert Piero Campanelli wrote: > > Hi, > > I have used JUnit with VisualAge Java and i have founded the integration > with the IDE very useful. Now i am working with Squeak but I see there is no > integration: > > - TESTRUNNER search for all TestCases classes and fill a list widget. It > isn't useful if my project has 750 TestCases :). Isn't it ? > > - If I want to run all test cases at once i have to do this: > > t := TestSuite new. > t addTest: ClassOneTest suite. > t addTest: ClassTwoTest suite. > t addTest: ClassThreeTest suite. > ............ > (t run) inspect > > My question is: > could I extend TestRunner so it runs a predefined list of TestCases or is > there someone who is just working on it ? > > Sorry, probably this is the right mailing list... > Tnx > Piero ___________________________________________ Robert Hirschfeld mailto:hirschfeld@acm.org http://www.prakinf.tu-ilmenau.de/~hirsch

1 0

Base64MimeConverter
by Andres Valloud 05 Jul '00

05 Jul '00

Hi. Why is is that Base64MimeConverter>>mimeDecodeToBytes: returns an RWBinaryOrTextStream (and this class keeps things internally as text!), while mimeDecodeToChars: returns a ReadWriteStream? Shouldn't both methods answer a ReadWriteStream? From the senders of mimeDecodeToBytes: it seems to be that a ReadWriteStream would do... Andres.

1 0

Re: Base64MimeConverter
by Bob Arning 05 Jul '00

05 Jul '00

On Tue, 04 Jul 2000 16:58:31 -0700 Andres Valloud <avalloud(a)exobox.com> wrote: >Why is is that Base64MimeConverter>>mimeDecodeToBytes: returns an >RWBinaryOrTextStream (and this class keeps things internally as text!), >while mimeDecodeToChars: returns a ReadWriteStream? Shouldn't both >methods answer a ReadWriteStream? From the senders of mimeDecodeToBytes: >it seems to be that a ReadWriteStream would do... Andres, My guess would be that some senders (not necessarily the ones you see in the image) of #mimeDecodeToBytes: might wish to toggle between #binary and #ascii modes, which a ReadWriteStream doesn't support. Cheers, Bob

1 0

PIC/Smalltalk Release
by Jon Hylands 05 Jul '00

05 Jul '00

Okay, I've done the port, it works, it's out there. You can grab it at http://www.huv.com/squeak/PICSmalltalk.cs or http://www.huv.com/squeak/PICSmalltalk.zip The change set doesn't file in completely correctly for some reason -- the pool dictionaries in PicObject and its subclasses are getting messed up somehow. If someone can have a look at that and let me know what is wrong, I'll fix it and update them. There's an example of interfacing to a servo included. To open a browser, evaluate: PicSmalltalkBrowser openBrowser in a Morphic project. This gives you a regular class browser with a new sub-menu in the class and message list views. The 'Program Chip' menu item doesn't launch the programmer, since I didn't bother to figure out how to launch a DOS executable from Squeak yet. I suspect a lot of you aren't using Windows anyway, so you'll need to fix that method to do the right thing depending on your OS and PIC programmer. This thing is designed to work with PIC 16F84 chips. It may work for others, but I haven't tried. It includes a full-blown Smalltalk parser, since the version of Smalltalk I wrote it in originally doesn't include one. If you have problems, for now I guess report them to the Squeak list. If it becomes a problem there, we can start a new egroups group or something for it. I haven't extensively tested this in Squeak yet, so user beware. I did just port it a couple months ago from Visual Smalltalk to Dolphin, so I worked out most of the porting issues when I did that. There are no real classes or instances in this system. instance variables are global, so name them as such. The linker does a full closure on the methods called from the reset method and the isr method, so you can include lots of other support code that you don't use, and it doesn't end up costing space in the end assembler code. I don't guarantee that the assembler produced from the included macros is 100% correct -- check the assembler output yourself carefully. If you don't know PIC assembler, you shouldn't be programming PICs, or you should buy a commercial C compiler. If you look at macro methods in PicObject, you'll see that they basically just define a block. The code in those blocks is (of course) real Smalltalk code that is evaluated by the code generator (and thus the Squeak Compiler) when the PIC/Smalltalk code gets generated. You can take that as far as you need it in terms of optimization of code, or doing weird stuff. Anyway, comments are welcome, preferrably to the list. I really wasn't planning on releasing this to the public, but you asked for it, so now you get it. If you really like it, feel free to send me nice stuff, or job offers :-) Later, Jon -------------------------------------------------------------- Jon Hylands Jon(a)huv.com http://www.huv.com/jon Project: Micro Seeker (Micro Autonomous Underwater Vehicle) http://www.huv.com

1 1

Re: Squeak Swiki out of order again
by guzdial＠cc.gatech.edu 04 Jul '00

04 Jul '00

I did check it yesterday and it was running fine. Perhaps a network/DNS glitch? If there is a problem, please write to swiki-bugs(a)cc.gatech.edu and we will look into it. We've said this several times on the list -- we do monitor that list, but we don't always check all our (ten) servers daily. Mark > > The Squeak Swiki > http://minnow.cc.gatech.edu/squeak/1 > can not be accessed since two days or so. > > It is also not the first time that the site is down > for more than a few hours. > > I do not like to complain about a voluntary service > provided by the people who are operating that site, > but it does not serve Squeak if the main documentation > (besides the source and changes files) is offline. > > Is there anything which can be done to improve this > situation? > > Andreas > > >

1 0

Re: [swiki] How to recover a password
by Bert Freudenberg 04 Jul '00

04 Jul '00

On Mon, 3 Jul 2000, Francisco Garau wrote: > We have a running swiki in a government institution. Unfortunately, one > user lock a page and forgot the password. How can this page be > recovered? In the swiki's pages folder, locate the page file. It's named "pagenumber.xml". In the settings section, remove the lockPassword entry and set lock to false. Or, since the password is stored as is, just bring up a webbrowser and unlock the page using this password. -- Bert

1 0

← Newer
1
...
49
50
51
52
53
54
55
56
57
Older →

Jump to page: