Squeak-dev October 2003

squeak-dev@lists.squeakfoundation.org

178 participants
791 discussions

[BUG] BFAV requires net access to write a response
by Daniel Vainsencher 16 Dec '03

16 Dec '03

I was trying to open as a mail a response I'd written, so that it get put in my Celeste queue for later sending. I think this shouldn't require net access - being able to work offline is an important feature. The following lines show the culprit. BugFixArchiveMailClient class>>mailMessageFromUrl: ArchivePost>>asMailMessage CommentNotePad>>setThreadingHeadersIn:withToAddress: CommentNotePad>>mailCommentsTo: CommentNotePad>>openCommentsAsMailToList 12 August 2003 2:38:24 am VM: unix - Squeak3.4 of 1 March 2003 [latest update: #5170] Image: Squeak3.6beta [latest update: #5387] NetNameResolver class(Object)>>error: Receiver: NetNameResolver Arguments and temporary variables: aString: 'Could not resolve the server named: swiki.gsug.org' Receiver's instance variables: superclass: Object methodDict: a MethodDictionary() format: 2 instanceVariables: nil organization: ('as yet unclassified') subclasses: nil name: #NetNameResolver classPool: a Dictionary(#DefaultHostName->'' #HaveNetwork->true #ResolverBusy->...etc... sharedPools: nil environment: a SystemDictionary(lots of globals) category: nil [] in NetNameResolver class>>addressForName:timeout: Arguments and temporary variables: hostName: 'swiki.gsug.org' secs: 20 deadline: 345293765 result: nil msg: 'Could not resolve the server named: swiki.gsug.org' rcvr: NetNameResolver Semaphore>>critical:ifError: Receiver: a Semaphore() Arguments and temporary variables: mutuallyExcludedBlock: [] in NetNameResolver class>>addressForName:timeout: errorBlock: [] in NetNameResolver class>>addressForName:timeout: blockValue: NetNameResolver hasError: true errMsg: 'Could not resolve the server named: swiki.gsug.org' errRcvr: NetNameResolver msg: 'Could not resolve the server named: swiki.gsug.org' rcvr: NetNameResolver Receiver's instance variables: firstLink: nil lastLink: nil excessSignals: 1 NetNameResolver class>>addressForName:timeout: Receiver: NetNameResolver Arguments and temporary variables: hostName: 'swiki.gsug.org' secs: 20 deadline: 345293765 result: nil msg: 'Could not resolve the server named: swiki.gsug.org' rcvr: NetNameResolver Receiver's instance variables: superclass: Object methodDict: a MethodDictionary() format: 2 instanceVariables: nil organization: ('as yet unclassified') subclasses: nil name: #NetNameResolver classPool: a Dictionary(#DefaultHostName->'' #HaveNetwork->true #ResolverBusy->...etc... sharedPools: nil environment: a SystemDictionary(lots of globals) category: nil --- The full stack --- NetNameResolver class(Object)>>error: [] in NetNameResolver class>>addressForName:timeout: Semaphore>>critical:ifError: NetNameResolver class>>addressForName:timeout: - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - HTTPSocket class>>httpGetDocument:args:accept:request: HttpUrl>>retrieveContentsArgs:accept: HttpUrl>>retrieveContentsArgs: HttpUrl>>retrieveContents BugFixArchiveMailClient class>>mimeDocumentFromUrl: BugFixArchiveMailClient class>>mailMessageFromUrl: ArchivePost>>asMailMessage CommentNotePad>>setThreadingHeadersIn:withToAddress: CommentNotePad>>mailCommentsTo: CommentNotePad>>openCommentsAsMailToList CommentNotePad>>perform:orSendTo: [] in MenuItemMorph>>invokeWithEvent: BlockContext>>ensure: CursorWithMask(Cursor)>>showWhile: MenuItemMorph>>invokeWithEvent: MenuItemMorph>>mouseUp: MenuItemMorph>>handleMouseUp: MouseButtonEvent>>sentTo: MenuItemMorph(Morph)>>handleEvent: MorphicEventDispatcher>>dispatchDefault:with: MorphicEventDispatcher>>dispatchEvent:with: MenuItemMorph(Morph)>>processEvent:using: MorphicEventDispatcher>>dispatchDefault:with: MorphicEventDispatcher>>dispatchEvent:with: MenuMorph(Morph)>>processEvent:using: MenuMorph(Morph)>>processEvent: MenuMorph>>handleFocusEvent: [] in HandMorph>>sendFocusEvent:to:clear: PasteUpMorph>>becomeActiveDuring: HandMorph>>sendFocusEvent:to:clear: HandMorph>>sendEvent:focus:clear: HandMorph>>sendMouseEvent: HandMorph>>handleEvent: HandMorph>>processEvents [] in WorldState>>doOneCycleNowFor: Array(SequenceableCollection)>>do: WorldState>>handsDo: WorldState>>doOneCycleNowFor: WorldState>>doOneCycleFor: WorldState>>doOneSubCycleFor: PasteUpMorph>>doOneSubCycle MenuMorph>>invokeModalAt:in:allowKeyboard: MenuMorph>>invokeModal: PluggableTextMorph>>yellowButtonActivity: PluggableTextMorph(ScrollPane)>>mouseDown: PluggableTextMorph(Morph)>>handleMouseDown: MouseButtonEvent>>sentTo: PluggableTextMorph(Morph)>>handleEvent: ...etc...

2 1

Squeak on Athlon 64?
by Aaron J Reichow 16 Dec '03

16 Dec '03

Hello folks- Anyone here done any benchmarking on a new Athlon 64? I'm going to be in the market for an upgrade for my ancient K6-2 350 MHz, and the machine will be as exclusively Squeak as I can pull off. The idea is to find a machine that runs Squeak the fastest for the least amount of cash. :) Regards, Aaron -- "if i don't stay true to live and hate, how do i differentiate between chasing cream and chasing dreams" :: atmosphere

4 3

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

[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

Re: (fwd) [Test-Runner Results] Squeak3.6beta latest update: #5411
by Daniel Vainsencher 21 Nov '03

21 Nov '03

Thanks, didn't know the version name was part of the updates.list. Is that definitely linked to the ftp directory name? My scsh script ended up looking like this - *************** #!/usr/bin/scsh -s !# (define versionname (string-drop (last (filter (lambda (s) (string-prefix? "#" s)) (run/strings (| (curl -s "http://update.squeakfoundation.org/external/updates/updates.list") (tr "\r" "\n"))))) 7)) (define ftpdir (string-append "ftp://ftp.squeakfoundation.org/official-dist/" versionname "/")) (define imagename (string-drop-right (last (string-tokenize (last (run/strings (curl -s ,ftpdir))))) 4)) (run (wget ,(string-append ftpdir imagename ".zip"))) (run (unzip -o ,(string-append imagename ".zip"))) (run (squeak -memory 64M ,(string-append imagename ".image") runTests.st)) ****************** where runTests.st is a short script that loads and runs Markus' code. scsh is currently my favorite shell language (which isn't hard, I hate most of them ;-). Daniel Ned Konz <ned(a)bike-nomad.com> wrote: > On Tuesday 09 September 2003 12:21 pm, Daniel Vainsencher wrote: > > * I don't know how to automate getting the image from the correct > > most-bleeding-edge directory. I've hard coded 3.6beta for now and > > will change it manually. Can we have a currentBleedingEdge symbolic > > link in the ftp servers? > > "Download the latest image if necessary; extract the zip to the > default directory." > > list _ (HTTPClient httpGet: > 'http://update.squeakfoundation.org/external/updates/updates.list') > contents. > versions _ (list findTokens: String crlf) select: [ :line | line > beginsWith: '#' ]. > bleedingEdge _ versions last copyFrom: 8 to: versions last size. > ftpDirectory _ ServerDirectory on: > ('ftp://st.cs.uiuc.edu/pub/Smalltalk/Squeak/', bleedingEdge). > latestImage _ (ftpDirectory getFileList contents findTokens: String > crlf) asSortedCollection last. > (FileDirectory default fileExists: latestImage) ifFalse: [ | zipStream > client | > zipStream _ FileDirectory default newFileNamed: latestImage. > client _ ftpDirectory openFTPClient. > client binary. > client getFileNamed: latestImage into: zipStream. > zipStream close. > ]. > zip _ ZipArchive new readFrom: latestImage. > zip members do: [ :m | zip extractMember: m ]. > > -- > Ned Konz > http://bike-nomad.com > GPG key ID: BEEA7EFE

20 51

Re: [ENH]browserNagIfNoClassComment (v5)
by Daniel Vainsencher 12 Nov '03

12 Nov '03

I agree with Doug that the ideal solution would - have the comment visible - not change the class creation messages (have you considered file outs? compatibility with older images? other dialects?) - not have a notifier on every uncommented class. The idea of ask for a comment when you've changed a class sounds nice... Lets start with something that is more informative but has no bad effects. We can always decide to be more pushy later. Daniel Karl Ramberg <karl.ramberg(a)chello.se> wrote: > > > Markus Gaelli wrote: > > > > Hi Doug, > > > > > This is looking better, although I think I prefer having the comment > > > show up in double-quotes below the template as in your previous > > > changeset, rather than extending the template method to > > > #subclass:instVarNames:...comment:, for a couple of reasons: > > > > > > - It looks a bit better visually to have a multi-line comment start on > > > a new line (with an extra cr in between). > > So why not > > > > Class >> > > instanceVariableNames: someVarNames > > classVariableNames: someClassNames > > poolDictionaries: somePoolDics > > category: categoryName > > comment: > > aCommentText > This is easy to fix. It's just to ass a extra cr. > > > > This shows in the creation method of a class that > > it is _essential_ for it to have comments. > > > > - You have to worry about escaping or handling extra single quotes > > > when accepting a new comment in the definition pane. Hm, it looks > > > like this is handled with some sort of magic in your changeset, but > > > this causes some weirdnesses, such as if you put a space after > > > comment: (to make it consistent with the other parameters), it inserts > > > a new single-quote. > > I think this is a solvable technical problem, just keep the text > > as long as Text as possible, it was not needed until > > now and can be done in a more elegant and possibly > > more error-free way than I did in my little hack. > > > Also, I think the inform: pop-up isn't as necessary if we have the > > > comment in the definition pane, as long as there is some warning text > > > if the comment is missing. The pop-up message unfairly punishes > > > innocent newbies who simply want to browse code, and will probably > > > scare people away. :) > > I think that Karl wants to put this to "default on" only in > > "images in-between" so that would be o.k. for me. > Exactly. Running in alpha and beta is only for daredevils:-) > > > > > > So, I'd personally be happy to approve a changeset which: > > > > > > - Has the inform pop-up and preference removed. > > > - Shows the class comment (in read-only mode) > > I believe that Karl and me had both the idea to make it writable > > in the class definition pane independent from each other... > > Still think that this would lower the barrier to write comments, > > and this is what this change-sets are all about, no? > Yes, I think authoring should always be on. > > > > > - If there is no class comment, the quoted part in the definition pane > > > would say something like "This class has no comment!" > > Yep, that should be the empty/ default comment. > Fixed. > > > - Escape any double-quotes in the comment by doubling them in the > > > definition pane. (otherwise a class comment containing double-quotes > > > will mess things up when you add instvars, etc. > > Don't know about this one right now. > I'll look into it. > > Karl? > > > > Markus > > > > >>>> The snag with both is that the comments are in read only > > >>>> mode but in 'comment mode'. > > >>>> I'm not sure how to fix that. > > >>>> > > >>> > > >>> I did a little hack, which should do the trick... > > >>> though not as I naively thought by enhancing > > >>> Class with something like >> subclass: '' instVarNames: '' (...) > > >>> comment: '' > > >>> > > >>> The reason is, that the Class >> classComment: '' expects a text and > > >>> not > > >>> a string. (' in strings are ugly to write and read...) > > >>> > > >>> So I hacked > > >>> Browser >> defineClass: defString notifying: aController > > >>> > > >>> which is already to long....:-( > > >>> > > >>> I still reformated it like it could be a method sent to class, > > >>> maybe this is missleading? > > >>> > > >>> Attention: I did not look into timestamps issues of the comment, so > > >>> this might be wrong. > > >>> Also the comment does not update while on the class side... > > >>> Also no tests. > > >>> > > >>> So there is room for improvement... ;-) > > >>> > > >> Thanks, this is starting to look good. > > >> > > >> I fixed a issue with a single quote being added to the end of each > > >> comment. > > >> > > >> Karl > > >> > > >> <NagIfNoClassComment.9.cs.gz> > > > > > >

2 1

X11 Server for Squeak!
by David Faught 11 Nov '03

11 Nov '03

Aaron J Reichow wrote: >I suppose the point of this post is to let anyone else interested in >"Living out of Squeak" that this project exists *and* to let those >working on the project (Ian, and perhaps others?) that I'm very >excited about this! > > A functional X server for Squeak would allow many things- in my >case, it would allow me to make Squeak my entire GUI environment, >running it directly on the framebuffer on Linux, and running Mozilla >Firebird within Squeak using this X Server. After all, I can do >pretty much everything with the tools we have now in Squeak- >excepting a web browser, for which I rely on the host OS. This does sound like an interesting idea, but I have a couple of questions. 1. First of all, how do you make Squeak run directly on the frame buffer on Linux? I know I have seen some discussion of this before on the list, but I couldn't find it by searching the list on GMane or searching the swiki. 2. Although having the X Server in Squeak would be cool and allow you to run Mozilla Firebird from "inside" Squeak, wouldn't it also be fun to use a plugin or FFI interface to Mozilla directly? __________________________________ Do you Yahoo!? The New Yahoo! Shopping - with improved product search http://shopping.yahoo.com

4 5

Re: Erlang (was: Re: Generics)
by Blaine Buxton 10 Nov '03

10 Nov '03

Hey, I've been reading up on Erlang myself. They have an approach where everything is a process and you send messages to these processes. Just imagine having a separate process for each Squeak object. They argue that processes should be as easy to create as objects. You would think this would kill performance, but actually, the systems they have written in it scale very well (the web server they wrote in Erlang can handle 10x the capacity of a Apache server, if I remember correctly). They also boast numbers of 99.99999% availablity for the Ericsson switch that was written in mostly Erlang. Now, those boasts are what got me interested. It is a functional language and they are big on no mutable state like most functional languages. They argue that Erlang makes multi-processing easy and in fact easier than single process systems. They seem very anti-OO on their list, but I have found that the whole process thing maps very well to objects. But, I haven't gotten that far. It just seems they have been scarred by the {} crowd and static typed OO systems. I haven't had a whole lot of time to really get much deeper than that. I bought the book and have been working through some of the examples. They do a lot with pattern matching and it's made me think about certain programming topics differently. I know I wasn't the person you were asking for the answer, but I thought I'd chime in. ------------------------------------------------ Blaine Buxton My Amps: Smalltalk, Lisp, and Ruby http://www.blainebuxton.com >From: Daniel Vainsencher <danielv(a)netvision.net.il> >Reply-To: The general-purpose Squeak developers >list<squeak-dev(a)lists.squeakfoundation.org> >To: The general-purpose Squeak developers list ><squeak-dev(a)lists.squeakfoundation.org> >CC: squeak-dev(a)lists.squeakfoundation.org >Subject: Erlang (was: Re: Generics) >Date: Mon, 29 Sep 2003 13:38:37 +0300 > >Hi Richard. > >"Richard A. O'Keefe" <ok(a)cs.otago.ac.nz> wrote: > > I like Smalltalk (and Erlang and Prolog) a lot >[snip] > >Sorry about the off topic interjection, but can you write a paragraph or >two about Erlang? I've tried to get the flavor of it from >site/documentation, but haven't quite made it yet. > >Daniel > _________________________________________________________________ Get McAfee virus scanning and cleaning of incoming attachments. Get Hotmail Extra Storage! http://join.msn.com/?PAGE=features/es

5 6

[FIX] InternalThreadNavigationMorph
by Norberto Manzanos 09 Nov '03

09 Nov '03

A very small fix for InternalThreadNavigationMorph. When you have more than one ITNM, and one of them has more projects than the other, if you switch from the bigger to the other and the bigger is at an index greater than the number of pages of the smaller, the morph becomes empty. Perhaps this is the expected behaviour. I prefer, having in mind a final user, to lead to the last project instead of an empty navigator. Cheers Norberto Manzanos http://www.amanza.com.ar

4 5

← Newer
1
2
3
4
5
6
7
8
9
...
80
Older →

Jump to page: