Squeak-dev digest, Vol 1 #175 - 40 msgs

Jeroen Gruter jeroen at gruter.nl
Mon Oct 8 04:54:25 UTC 2001


Hello,
I am interested in Smalltalk en Squeak because of Smalltalk. Is this
newsbullitin for Smalltalk like 'what hotel do I have to book' or are you
going to do this at your own e-mail because I hope I now get only
interesting mails.
Greatings
J. Gruter, The Netherlands

----- Original Message -----
From: <squeak-dev-request at lists.squeakfoundation.org>
To: <squeak-dev at lists.squeakfoundation.org>
Sent: Monday, October 01, 2001 4:53 AM
Subject: Squeak-dev digest, Vol 1 #175 - 40 msgs


> Send Squeak-dev mailing list submissions to
> squeak-dev at lists.squeakfoundation.org
>
> To subscribe or unsubscribe via the World Wide Web, visit
> http://lists.squeakfoundation.org/listinfo/squeak-dev
> or, via email, send a message with subject or body 'help' to
> squeak-dev-request at lists.squeakfoundation.org
>
> You can reach the person managing the list at
> squeak-dev-admin at lists.squeakfoundation.org
>
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of Squeak-dev digest..."
>
>
> Today's Topics:
>
>    1. Help with sound (Daniel Joyce)
>    2. Re: Squeak BOF at OOPSLA (Henrik Gedenryd)
>    3. Re: Squeak BOF at OOPSLA (Henrik Gedenryd)
>    4. [ENH] isEqualToInAspects (gaelli at emergent.de)
>    5. Re: [ENH][Win32]RE: Font Support (Christian Langreiter)
>    6. invalid Win32Rectangle compiledSpec (field definition) (Gerald Leeb)
>    7. Skip List in #4308 (Randal L. Schwartz)
>    8. Re: [ENH][Win32]RE: Font Support (Thomas Kuehne)
>    9. Getting accents into Squeak (John Hinsley)
>   10. Re: Help with sound (Ned Konz)
>   11. Re: [modules] Ginsu in 3.1 (Hans-Martin Mosner)
>   12. Re: Skip List in #4308 (Karl Ramberg)
>   13. Re: Skip Lists?! (Lex Spoon)
>   14. Re: Skip List in #4308 (Scott A Crosby)
>   15. Re: Desired Archive behavior w/r/t absolute paths? (Karl Ramberg)
>   16. Re: Skip List in #4308 (Karl Ramberg)
>   17. Re: Skip Lists?! (Scott A Crosby)
>   18. Re: Computerchannel.de: Squeak 3.0 tested (Karl Ramberg)
>   19. Re: [modules] Ginsu in 3.1 (ducasse stephane)
>   20. [BUG] Relative directory strangeness? (Ned Konz)
>   21. Cache lines and branch prediction. (Scott A Crosby)
>   22. Re: Computerchannel.de: Squeak 3.0 tested (Fleeberz at aol.com)
>   23. [UNIX][VM] network audio support (Lex Spoon)
>   24. Re: Storing and Retrieving Point (raymondasselin at sympatico.ca)
>   25. Re: Why we should remove {} from Squeak (David N. Smith (IBM))
>   26. Re: Getting accents into Squeak (Torge Husfeldt)
>   27. Re: Computerchannel.de: Squeak 3.0 tested (Kevin Fisher)
>   28. Re: [modules] Ginsu in 3.1 (danielv at netvision.net.il)
>   29. [OT] aRts (KDE) and Squeak...one year later (Kevin Fisher)
>   30. Re: Help with sound (Daniel Joyce)
>   31. Re: Which Squeak is preferred for Mac OS X ? (Doug Way)
>   32. Re: Skip List in #4308 (danielv at netvision.net.il)
>   33. Re: Skip Lists?! (Tim Rowledge)
>   34. Re: [ENH][Win32]RE: Font Support (PhiHo Hoang)
>   35. Re: Localization in code (Richard A. O'Keefe)
>   36. Re: [ENH][Win32]RE: Font Support (PhiHo Hoang)
>   37. Re: Getting accents into Squeak (John Hinsley)
>   38. Caveats (Daniel Joyce)
>   39. Augh... (Daniel Joyce)
>   40. Re: Localization in code (Richard A. O'Keefe)
>
> --__--__--
>
> Message: 1
> From: Daniel Joyce <crusoe at gaia.theinternetone.net>
> To: squeak-dev at lists.squeakfoundation.org
> Subject: Help with sound
> Date: Sat, 29 Sep 2001 23:15:50 -0500
> Reply-To: squeak-dev at lists.squeakfoundation.org
>
> The Midi works great.
>
> But none of the sound playing demos work.
>
> The wave editor, etc.
>
> Is sound that bad on Unix?
>
> NB, normal window/open/close  sound effects and others work fine too.
>
> Where is that code handled? Perhaps I can look there? Or is it just that
all
> the sampled sound demos are broken?
>
> -Daniel
>
>
> --__--__--
>
> Message: 2
> Date: Sun, 30 Sep 2001 13:49:38 +0200
> Subject: Re: Squeak BOF at OOPSLA
> From: Henrik Gedenryd <Henrik.Gedenryd at lucs.lu.se>
> To: <squeak-dev at lists.squeakfoundation.org>
> Reply-To: squeak-dev at lists.squeakfoundation.org
>
> Doug Way wrote:
>
> > Looking forward to the BOF (Birds Of a Feather) session at OOPSLA.
> > Maybe I'll try to keep the informal Squeak "bug fixing party" going at
> > Camp Smalltalk there this year, too.
> >
> > Just a quick suggestion to anyone going to OOPSLA who hasn't booked a
> > hotel room yet:  I was originally going to try to find a somewhat
> > cheaper hotel downtown than the suggested ones (which had "special"
> > OOPSLA rates of $140-$150 per night), such as the Holiday Inn several
> > blocks away for about $100.  But then I happened to try priceline.com,
> > and set a price of $50/night for any 3-star hotel, and it actually
> > reserved the Marriott Waterside (adjacent to the conference, originally
> > $150)!  (This was for Sat.-Wed.)
> >
> > - Doug Way
> > dway at riskmetrics.com
>
> Doug,
>
> Before I book anything I thought I might ask if you would be interested in
> sharing a room. Of course we will be sharing the bill as well. According
to
> what I saw, the reservation you got was for a 2-person room.
>
> Let me know what you think and any reservations you may have.
>
> best regards,
> Henrik
>
>
>
>
> --__--__--
>
> Message: 3
> Date: Sun, 30 Sep 2001 13:51:02 +0200
> Subject: Re: Squeak BOF at OOPSLA
> From: Henrik Gedenryd <Henrik.Gedenryd at lucs.lu.se>
> To: Squeak list <squeak-dev at lists.squeakfoundation.org>
> Reply-To: squeak-dev at lists.squeakfoundation.org
>
> Henrik Gedenryd wrote:
>
> > Doug,
> >
> > Before I book anything I thought I might ask if you would be interested
in
> > sharing a room. Of course we will be sharing the bill as well. According
to
> > what I saw, the reservation you got was for a 2-person room.
> >
> > Let me know what you think and any reservations you may have.
> >
> > best regards,
> > Henrik
>
> Yup, that was supposed to be a personal mail. Oops(la). Sorry.
>
> Henrik
>
>
>
>
> --__--__--
>
> Message: 4
> Date: Sun, 30 Sep 2001 14:19:20 +0200
> Subject: [ENH] isEqualToInAspects
> To: squeak-dev at lists.squeakfoundation.org
> From: gaelli at emergent.de
> Reply-To: squeak-dev at lists.squeakfoundation.org
>
>
> --==CelesteAttachment25598==
>
> from preamble:
>
> "Change Set: isEqualToInAspects
> Date: 30 September 2001
> Author: Markus Gaelli
>
> Object >> isEqualTo: anObject inAspects: someSymbols
>
> Tests, whether I am equal to <anObject> in all aspects defined by
> <someSymbols>;
> gives falses as soon as I am not equal in one of the symbols.
> Gives true for an empty collection of <someSymbols>
> "!
> --==CelesteAttachment25598==
> Content-type: application/octet-stream
> Content-transfer-encoding: base64
> Content-disposition: attachment;filename="isEqualToInAspects.cs.gz"
>
> H4sIAAAAAAAAAN2RS08CQRCEz3Po/1BwARMjixgTVyQhPogH4wFuRpNZ6IXVeaw7g4bEH28v
> r2A8efU0SfXUV9M1zeuFdnPGmGOqVBFu35faTPy9G4aSpzHQjY4sE9VL5E4Z2WZc4TRJujRc
> xoWv6tmDrt6WASPNxhREj9mrWDEYYM9Lod1WLnboFMFbHq9s5k0gmnCI4RifC44LibiHtuDa
> jejR39mF6aCNgd5AMOO8cDxDtkL/gDe4JDUvPjgg1ybIoYPEeVefa7TzcYsXoHcMn0OCETb+
> E1KjtT1WS0buK1kAbMu4wtQbI9GFwMTzI5SaDSJqbDe1somfhTspCa2pt6WuCjdvIURtS5Hs
> XOOi00s6dZ3onqXJeatBf+hMNf9DaU0i9XIg1C8d61iEfJXiKdUbGV8gpdqBTY6SK4FbKWgz
> O8IV2vuyfk2fSTUgH/MN0Bl+8uwCAAA=
>
> --==CelesteAttachment25598==--
>
>
> --__--__--
>
> Message: 5
> From: "Christian Langreiter" <chris at langreiter.com>
> To: <squeak-dev at lists.squeakfoundation.org>
> Subject: Re: [ENH][Win32]RE: Font Support
> Date: Sun, 30 Sep 2001 16:10:12 +0200
> Reply-To: squeak-dev at lists.squeakfoundation.org
>
> > HostFont textStyleFromUser.
> >
>
> THANK YOU SO MUCH!
>
> Works like a charm, easy, fantastique!
>
> -- Chris
>
>
>
> --__--__--
>
> Message: 6
> From: "Gerald Leeb" <leeb at anecon.com>
> To: <squeak-dev at lists.squeakfoundation.org>
> Subject: invalid Win32Rectangle compiledSpec (field definition)
> Date: Sun, 30 Sep 2001 17:03:42 +0200
> Organization: ANECON
> Reply-To: squeak-dev at lists.squeakfoundation.org
>
> Hi squeakers!
>
> After loading a new image (3.0 and 3.1 alpha) the byteSize of the
> Win32Rectangle returns 0 but should return 16.
>
> To correct the compiledSpec (field definition) type
>
>     Win32Rectangle compileFields.
>
>
>
> keep squeaking
> Gerald.
>
>
>
> --__--__--
>
> Message: 7
> To: <squeak-dev at lists.squeakfoundation.org>
> Subject: Skip List in #4308
> From: merlyn at stonehenge.com (Randal L. Schwartz)
> Date: 30 Sep 2001 08:21:51 -0700
> Reply-To: squeak-dev at lists.squeakfoundation.org
>
>
> Skip lists got snuck in with changeset 4308 with *no documentation*
> except the source code.  Not even a class comment!
>
> Can SqueakCentral at least set the bar high enough that *new* classes
> into the core will not be accepted without at least having a class
> comment?  Please?
>
> Especially something that is in the Collections category!
>
> --
> Randal L. Schwartz - Stonehenge Consulting Services, Inc. - +1 503 777
0095
> <merlyn at stonehenge.com> <URL:http://www.stonehenge.com/merlyn/>
> Perl/Unix/security consulting, Technical writing, Comedy, etc. etc.
> See PerlTraining.Stonehenge.com for onsite and open-enrollment Perl
training!
>
>
> --__--__--
>
> Message: 8
> Date: Sun, 30 Sep 2001 17:35:52 +0200
> From: Thomas Kuehne <kuehne at informatik.uni-kl.de>
> Organization: University of Kaiserslautern
> To: squeak-dev at lists.squeakfoundation.org
> Subject: Re: [ENH][Win32]RE: Font Support
> Reply-To: squeak-dev at lists.squeakfoundation.org
>
> Andreas,
>
> what a great changeset. At last there is an easy way to import standard
fonts...
> Thanks many, many times.
>
> I noticed that the carriage return character (13) is displayed as a box so
I tried to redefine it (along with character "_" and "^").
>
> I set Arial to be the default Font and then did
>
> (TextStyle default fontAt: 2) edit: (13 asCharacter)
>
> Here the "2" selects a certain size of the font.
> That did not work in Squeak3.1a-4332 since the message
> pixelValueForDepth: aColor is not understood by Form instances.
>
> It seems that Form>>pixelValueFor: aColor
> does exactly what is required so I provided it under the above name as
well. Perhaps it is better to change the caller so that it calls the
existing method?
>
> Anyway, after fixing this the above code (e.g., "... edit: $_") allows one
to change any character one does not like.
>
> Thanks again,
>
>     Thomas
>
> --
> Dr. Thomas Kuehne
> +49 178 4314387, http://www-agce.informatik.uni-kl.de/~kuehne
> Experts are people who successfully calibrated their intuition. -- TK
>
>
>
> --__--__--
>
> Message: 9
> Date: Sun, 30 Sep 2001 16:45:46 +0100
> From: John Hinsley <jhinsley at telinco.co.uk>
> To: "squeak-dev at lists.squeakfoundation.org"
<squeak-dev at lists.squeakfoundation.org>
> Subject: Getting accents into Squeak
> Reply-To: squeak-dev at lists.squeakfoundation.org
>
> I've just noticed this (the localisation stuff may have some kind of
> bearing on it, but......).
>
> There I was trying to put a name with "ü" (that's u with a dieresis, I
> think) in my Rolodex. Of course, you can copy and paste into a workspace
> and it'll appear correctly, but you can't copy and paste into a Text or
> ShowEmptyText morph (maybe I should take a look at that -- I think it
> would be useful). But is there any way (and here I'm working in Linux --
> I guess the old number pad method works in Windows) of typing it in
> directly, given a UK keyboard? (Here, locales or foreign keyboards are
> not really a viable option: I'd need some bizzare combination of
> Italian, German and Spanish.
>
> Cheers
>
> John
>
>
> --
> Can't cope anymore? Desperate for help?
> Join the 12 step program for those who yearn to give up Microsoft:
> http://home.earthlink.net/~penguinrox/index.html
>
>
> --__--__--
>
> Message: 10
> From: Ned Konz <ned at bike-nomad.com>
> To: squeak-dev at lists.squeakfoundation.org
> Subject: Re: Help with sound
> Date: Sun, 30 Sep 2001 08:57:09 -0700
> Reply-To: squeak-dev at lists.squeakfoundation.org
>
> On Saturday 29 September 2001 09:15 pm, Daniel Joyce wrote:
> > The Midi works great.
>
> Under Unix? How? The MIDI code is just stubbed out. You must be on a
> different OS.
>
> > But none of the sound playing demos work.
> >
> > The wave editor, etc.
> >
> > Is sound that bad on Unix?
>
> Sound playing works under Unix.
> Sound recording doesn't (at least in the stock image).
> Lex Spoon has gotten both to work well.
>
> > NB, normal window/open/close  sound effects and others work fine too.
> >
> > Where is that code handled? Perhaps I can look there? Or is it just that
> > all the sampled sound demos are broken?
>
> What kind of operating system and sound architecture are you using?
>
> If you're using the Gnome or KDE desktop environments, you may have to use
> esddsp or artsdsp to start Squeak.
>
> --
> Ned Konz
> currently: Stanwood, WA
> email:     ned at bike-nomad.com
> homepage:  http://bike-nomad.com
>
>
> --__--__--
>
> Message: 11
> Date: Sun, 30 Sep 2001 14:48:28 +0200
> From: Hans-Martin Mosner <hm.mosner at cityweb.de>
> Organization: Home Sweet Home
> To: squeak-dev at lists.squeakfoundation.org
> Cc: Joseph Pelrine <jpelrine at acm.org>, Sarkela <sarkela at home.com>
> Subject: Re: [modules] Ginsu in 3.1
> Reply-To: squeak-dev at lists.squeakfoundation.org
>
> danielv at netvision.net.il wrote:
> >
> > Hey guys,
> > We need to be more organized about this - part of the changes you
> > enumerate are included in the ones I posted on the 23rd.
> >
>
> Just shows how ill-organized I am :-(
> I think I remember reading somewhere about at least one of the issues,
> but could not find the place quickly, so I just started off with the ZIP
file.
> Getting the current version of something out of n different sources, and
> combining things such that they work together is something I was never
> good at...
> Perhaps we need some kind of module and repository concept here :-)
>
> BTW, I'm now working on producing a complete first modularization of the
> 3.1 image, but this will take some more time... Running the module
> builder again and again is somewhat time consuming.
>
> Cheers,
> Hans-Martin
>
>
>
> --__--__--
>
> Message: 12
> Date: Sun, 30 Sep 2001 19:28:48 +0200
> From: Karl Ramberg <karl.ramberg at chello.se>
> To: squeak-dev at lists.squeakfoundation.org
> Subject: Re: Skip List in #4308
> Reply-To: squeak-dev at lists.squeakfoundation.org
>
>
>
> "Randal L. Schwartz" wrote:
> >
> > Skip lists got snuck in with changeset 4308 with *no documentation*
> > except the source code.  Not even a class comment!
> >
> > Can SqueakCentral at least set the bar high enough that *new* classes
> > into the core will not be accepted without at least having a class
> > comment?  Please?
> >
> > Especially something that is in the Collections category!
>
> This came up just a few day ago.
> Someone who understand it could make a comment enhancement:-)
> Karl
>
> On Tue, 25 Sep 2001, Andreas Raab wrote:
>
> > Folks,
> >
> > I noticed them coming into Squeak a while ago and always since I wanted
to
> > ask this question: What the heck are they good for?! ;-) I don't see
them
>
> Skiplists are an implementation of the sorted-list pattern, in that they
> are paramaterized by some sort of ordering relation, and support the
> following operations. Basically, this is the SortedCollection pattern:
>
>
> Create()
> Insert(item)
> Delete(item)
> Next(item)
> Prev(item)
> First()
> Last()
> Lookup(Item)  ;; return the item in the list that is equal (by the
>               ;; ordering relation) with the requested item.
>
> Items can be thought of as being key-value pairs, where the ordering
> relation in the items is the ordering relation on the keys in those items.
>
> The typical implementation of these uses balanced binary trees (splay
> trees, red black trees, AVL trees, treaps, etc). That way you get O(log n)
> for all operations except O(1) to create.
>
> Skiplists are a different way of doing this; they are simpler than trees
> and have tighter inner loops.
>
> BTW, squeak currently uses arrays to implement SortedCollection, thus
> Insert and Delete cost O(n) instead of O(log n). I was thinking of
> reimplementing SortedCollection myself if profiling indicated that this
> was a problem.
>
> As an interesting aside, assuming that you have no other intervening
> accesses or mutations to the splay tree, you can do enumerate
> through all of the elements of a splay tree from start to
> finish in O(1) amortized time, even though the individual operations are
> O(log n) time.
>
>
> > used anywhere and I've never heard of 'em before (yeah, well, my
> > introductory courses in algorithms and data structures were plenty of
years
> > back ;-) but somebody found them important enough to write and include
them.
> > Since I like data structures (in particular efficient ones) I'd like to
hear
> > more about the places in which SkipLists could and should be used
instead of
> > other approaches. Pointers to comparisons would be nice, examples in
Squeak
> > even nicer ;-)
>
> Basically, Skiplists (and/or balanced tree implementations of
> SortedCollection) should be used anywhere where you have a lot of
> inserts/removals from a sorted collection, and profiling shows that that
> is a problem.
>
> Now, if you integrated them so that they supported the same sort of
> operations as SortedCollection, I'd really thank you. (I'm not at the
> point where I am dealing with large collections of sorted data for the
> difference to be noticable in the profiling results. As such, I've not
> bothered with examining the code closely, but I think that they should be
> resuperclassed to inherit from OrderedCollection, not Collection.)
>
>
> Scott
>
>
> --__--__--
>
> Message: 13
> Date: Sun, 30 Sep 2001 14:11:01 -0500
> Subject: Re: Skip Lists?!
> To: squeak-dev at lists.squeakfoundation.org
> From: "Lex Spoon" <lex at cc.gatech.edu>
> Reply-To: squeak-dev at lists.squeakfoundation.org
>
>
>
> > As far as I can tell it is a very good alternative to balanced binary
> > trees (much easier implementation being prime advantage I believe) -
> > performance wise they are much the same. I read somewhere that database
> > implementations still might prefer Btrees because of some property that
> > make them more efficient when it comes to distribution on secondary
> > memory - but I don't know the details on that.
> >
>
> On disk, the exact number of dereferences makes a big difference.  With
> a binary tree, you divide your search space by 2 with each dereference;
> with a btree, you divide it by a larger number.  I'd guess it's similar
> for skip lists, though I've only heard about them, not studied them.
>
> Hmm, with "main" memory being so much slower than cache, btrees might
> even start making sense for in-memory structures.  But not necessarily
> for Squeak, where an "array of foo's" won't necessarily have all of the
> foo's next to each other, anyway, and thus where accessing items within
> a node still causes fetches to distant parts of memory.
>
> Anyway, I'd rather work on something higher-level.  This
> hyper-optimization stuff is only useful on small portions of most of the
> programs I ever work on.
>
>
> -Lex
>
>
> --__--__--
>
> Message: 14
> Date: Sun, 30 Sep 2001 14:24:36 -0400 (EDT)
> From: Scott A Crosby <crosby at qwes.math.cmu.edu>
> To: Karl Ramberg <karl.ramberg at chello.se>
> Cc: <squeak-dev at lists.squeakfoundation.org>
> Subject: Re: Skip List in #4308
> Reply-To: squeak-dev at lists.squeakfoundation.org
>
> On Sun, 30 Sep 2001, Karl Ramberg wrote:
>
> >
> > "Randal L. Schwartz" wrote:
> > >
> > > Skip lists got snuck in with changeset 4308 with *no documentation*
> > > except the source code.  Not even a class comment!
> > >
> > > Can SqueakCentral at least set the bar high enough that *new* classes
> > > into the core will not be accepted without at least having a class
> > > comment?  Please?
> > >
> > > Especially something that is in the Collections category!
> >
> > This came up just a few day ago.
>
> Can you include attributation next time you include another message?
>
> Ideally, skiplists should just be something else that has the same
> interface as SortedCollection and a reference to the skiplist paper.
>
> I may get to that point in a while, I'm going to need to do a lot of
> prefix-matching to lists that are constantly changing; the easiest
> way to do that is to store the list in sorted order. Search, then do a
> prefix-match on the Next() item.
>
> > Someone who understand it could make a comment enhancement:-)
>
> Or someone could cut&paste the comment. :)  Or better yet, integrate them
> so that they can drop-in replace SortedCollection.  *hint hint*
>
> Scott
>
>
>
> --__--__--
>
> Message: 15
> Date: Sun, 30 Sep 2001 20:41:15 +0200
> From: Karl Ramberg <karl.ramberg at chello.se>
> To: squeak-dev at lists.squeakfoundation.org
> Subject: Re: Desired Archive behavior w/r/t absolute paths?
> Reply-To: squeak-dev at lists.squeakfoundation.org
>
>
>
> Ned Konz wrote:
> >
> > On Thursday 27 September 2001 11:28 pm, Karl Ramberg wrote:
> >
> > > I personally like the non intrusive expansion so I see which files
folders
> > > etc that are involved, and from there move the files manually to where
> > > they are expected to be.
> >
> > Which one is non-intrusive? All I'm doing (at worst) is removing leading
> > slashes to make everything relative.
>
> Sorry, I ment relative.
>
> With absolute paths stuff ends up in
> places hard to find them. When/if a absolute path is used
> it should make a text document wich tells what's been
> installed and where.
>
> Karl
>
>
> --__--__--
>
> Message: 16
> Date: Sun, 30 Sep 2001 20:44:32 +0200
> From: Karl Ramberg <karl.ramberg at chello.se>
> To: Scott A Crosby <crosby at qwes.math.cmu.edu>
> Cc: squeak <squeak-dev at lists.squeakfoundation.org>
> Subject: Re: Skip List in #4308
> Reply-To: squeak-dev at lists.squeakfoundation.org
>
>
>
> Scott A Crosby wrote:
> >
> > On Sun, 30 Sep 2001, Karl Ramberg wrote:
> >
> > >
> > > "Randal L. Schwartz" wrote:
> > > >
> > > > Skip lists got snuck in with changeset 4308 with *no documentation*
> > > > except the source code.  Not even a class comment!
> > > >
> > > > Can SqueakCentral at least set the bar high enough that *new*
classes
> > > > into the core will not be accepted without at least having a class
> > > > comment?  Please?
> > > >
> > > > Especially something that is in the Collections category!
> > >
> > > This came up just a few day ago.
> >
> > Can you include attributation next time you include another message?
>
> Sorry, I just made a quick cut and paste.
> I'll be more carefull next time :-)
>
> Karl
>
>
> --__--__--
>
> Message: 17
> Date: Sun, 30 Sep 2001 14:53:38 -0400 (EDT)
> From: Scott A Crosby <crosby at qwes.math.cmu.edu>
> To: Lex Spoon <lex at cc.gatech.edu>
> Cc: <squeak-dev at lists.squeakfoundation.org>
> Subject: Re: Skip Lists?!
> Reply-To: squeak-dev at lists.squeakfoundation.org
>
> On Sun, 30 Sep 2001, Lex Spoon wrote:
>
> >
> > On disk, the exact number of dereferences makes a big difference.  With
>
> Its the number and the predictability of the dereference. If you can
> preload it, it doesn't matter. In this case, the cost of loading 16 bytes
> versus 16k is so little, that its faster to load the 16k, do 10 binary
> tree comparisons, than do 10 disk seeks.
>
> > a binary tree, you divide your search space by 2 with each dereference;
> > with a btree, you divide it by a larger number.  I'd guess it's similar
> > for skip lists, though I've only heard about them, not studied them.
> >
>
> > Hmm, with "main" memory being so much slower than cache, btrees might
> > even start making sense for in-memory structures.  But not necessarily
>
> Nope. Most cachelines are 16 bytes.. You pay a lot more with the extra
> comparisons. Nor do you save anything with teh ability to prefetch; you
> cannot predict which branches you'll need. You might get semi-close with
> splay trees though, because they can use locality of access, so the
> effective number of accesses is much better.
>
> I'll probably eventually implement them.
>
> > for Squeak, where an "array of foo's" won't necessarily have all of the
> > foo's next to each other, anyway, and thus where accessing items within
> > a node still causes fetches to distant parts of memory.
>
> Distant parts don't hurt; cache lines are 16 bytes. The closest they could
> get to hurting is if they hit the TLB too heavily. (TLB for x86 has 64
> entries for 4kb pages. Or, 256k. access outside that costs about 3 memory
> lookups to read in the new page table entry for that 4kb. PTE lookups can
> be cached.)
>
> >
> > Anyway, I'd rather work on something higher-level.  This
> > hyper-optimization stuff is only useful on small portions of most of the
> > programs I ever work on.
>
> Exactly.. The O(log n) will beat your O(n) every time you start getting a
> lot of data. If it does in profiling, then implement the O(log n)
> algorithm, and contribute it.
>
>
> Don't discount profiling.. Profiling has found out places in my network
> stream library where invoking #origionalContents instead of #contents sped
> it up by 3x,
>
>
> And this little beauty: fidding with one line of code has lead to speedups
> of 30% in the queueing code.
>
> I found out that:
>    ``readPosition  > (writeLimit/2)'' is about 30 times slower than
>    ``readPosition * 2 > writeLImit''
> (if writeLimit is randomly distributed. If writeLimit is always odd, its
> 60x.)
>
> Which was the absolute last thing I expected to be consuming 20% of my
> CPU time. Profiling is useful, use it. :)
>
> Scott
>
>
> --
> No DVD movie will ever enter the public domain, nor will any CD. The last
CD
> and the last DVD will have moldered away decades before they leave
copyright.
> This is not encouraging the creation of knowledge in the public domain.
>
>
>
> --__--__--
>
> Message: 18
> Date: Sun, 30 Sep 2001 21:00:36 +0200
> From: Karl Ramberg <karl.ramberg at chello.se>
> To: squeak-dev at lists.squeakfoundation.org
> Subject: Re: Computerchannel.de: Squeak 3.0 tested
> Reply-To: squeak-dev at lists.squeakfoundation.org
>
>
>
> Fleeberz at aol.com wrote:
>
> > I haven't gotten the hang of Smalltalk AT ALL yet, but that's not going
to
> > keep me from reading this list, getting all the goodies and updates and
> > following links to more information.
> > Someday, I'm sure I'll get the hang of it, and be grateful for "hanging
in
> > there." - Regardless, I know I'm not really such an idiot either, even
though
> > I strongly feel that Squeak is very powerful, easy to use (for the rest
of
> > you, obviously) and innovative.
>
> The way I learned was by finding a simple morphic demo or program,
> change a few parameters, open the morph and see what had happened...
> Do this over and over again until you know what to change to get
> what you want. Then expand by cutting and pasting in code from other
> places, testing and debugging a few thousen times you will suddenly
> just know how to do it the next time.
>
> I allways browse with optinal buttons on and use the senders of and
> implementors of all the time to see how stuff is used elsewere in
> the image.
>
> Karl
>
>
> --__--__--
>
> Message: 19
> Date: Sun, 30 Sep 2001 21:10:42 +0200
> Subject: Re: [modules] Ginsu in 3.1
> From: ducasse stephane <ducasse at iam.unibe.ch>
> To: <squeak-dev at lists.squeakfoundation.org>
> Cc: Henrik Gedenryd <Henrik.Gedenryd at lucs.lu.se>
> Reply-To: squeak-dev at lists.squeakfoundation.org
>
>
> Henrik this would be really interesting to see if the concept of
visibility
> used in Ginsu is applicable to your modules. Because we could use Ginsu as
a
> tools to modularise the image, this way we could add functionality to
Ginsu
> and use it as a slicing tools.
>
>
> > BTW, I'm now working on producing a complete first modularization of the
> > 3.1 image, but this will take some more time... Running the module
> > builder again and again is somewhat time consuming.
>
> Hans-Martin just continue. This would be great if we could identify a
gross
> architecture of the system and all the place to fix (where method have to
be
> moved around...I should find some times to extract the fileList. I will
try
> this week.
>
> Stef
>
>
>
> --__--__--
>
> Message: 20
> From: Ned Konz <ned at bike-nomad.com>
> To: squeak-dev at lists.squeakfoundation.org
> Subject: [BUG] Relative directory strangeness?
> Date: Sun, 30 Sep 2001 12:45:48 -0700
> Reply-To: squeak-dev at lists.squeakfoundation.org
>
> /home/ned/tmp is an empty directory.
> /tmp/xx.txt exists and is a file.
>
> d _ FileDirectory on: '/home/ned/tmp'.
> d fileExists: 'tmp/xx.txt'          "true, but shouldn't be"
> d fileExists: './tmp/xx.txt'          "true, but shouldn't be"
> d fileExists: '/tmp/xx.txt'        "true"
> d fileExists: '/home/ned/tmp/tmp/xx.txt'          "false"
>
> It seems to me that there should be some way to test for the existence of
> arbitrary files even in subdirectories. I think that
>
> d fileExists: 'tmp/xx.txt'
>
> and
>
> d fileExists: './tmp/xx.txt'
>
> should both return false.
>
> Anyone else got an opinion on this?
>
> --
> Ned Konz
> currently: Stanwood, WA
> email:     ned at bike-nomad.com
> homepage:  http://bike-nomad.com
>
>
> --__--__--
>
> Message: 21
> Date: Sun, 30 Sep 2001 16:00:47 -0400 (EDT)
> From: Scott A Crosby <crosby at qwes.math.cmu.edu>
> To: Squeak List <squeak-dev at lists.squeakfoundation.org>
> Subject: Cache lines and branch prediction.
> Reply-To: squeak-dev at lists.squeakfoundation.org
>
> Whoops. Looks like I am obsolete with my memory of CPU's.. Oh, for the
> days when cache lines were 16 bytes. I feel old already.
>
> It looks like you can now assume them to be 32 (Pentium 2) or even 64
> bytes.  (AMD Athalon)
>
> So, given the bigger cache line, which we can expect to be big enough to
> contain a full entry, my idea of using two tables is no longer
> useful; it will lead to using more cache lines because entries will be
> split up between lines. (And the other portions of those cache lines will
> most likely be wasted.)
>
> Given this, my 'cache conscious' design is no longer relevant, and a
> variant of the current scheme will work well:
>
> 1. Make the cache hold 2k-16k entries.
> 2. Upon a cache hit on the first slot, use it as-is.
> 3. Upon a cache miss on the first probe, but a hit upon a later probe
> Replace the first slot probed with the entry of the hit.
> 4. If all probes miss, lookup the correct method. If there is an
> empty slot use it, otherwise replace either the 2nd or 3rd (chosen
> randomly) probe with the lookedup information.
>
> Advantages of this: It is still pretty similar to the current scheme and
> only requires very minimal changes.  Also, we never add to the second
> slot.
>
> Reason:
>
>  #1: Morphic is very heavyweight and uses a lot of selectors, classes, and
> methods.
>
>  #3: If it is used a second time before being removed, put it in the first
> slot so we notice it quickly.
>
>  #4: We do not want to ever erase an entry.
>
> Note that this scheme still wastes 1/3 of the slots, but it will probably
> increase the hit rate, of the full cache (because its a larger cache) and
> increase the hit rate of the first probe... Until Jitter comes along and
> makes this all irrelevant.
>
>
> --
>
> Now, for branch prediction:
>
>
http://developer.intel.com/design/intarch/techinfo/Pentium/operatn.htm#10047
86
>
> A branch prediction hit on the intel has a cost of one clock cycle, and
> may be run in parallel with any other integer instruction, even a
> comparison!!!!
>
> A miss costs a full pipeline flush. (I do not know the cost of this, but
> I'd assume at least 6 ticks, maybe over 10 clock cycles.) and another 3 to
> 4 clock cycles.
>
> The example, taken from the above:
>
>    for(k=i+prime;k<=SIZE;k+=prime) flags[k]=FALSE;
>
> Can run in 2 clocks/iteration. (plus any cache miss overhead) It runs the
> store&add in one tick, and the compare&branch in the second tick.
>
> So a branch prediction miss costs what? About 10x that of a hit.
>
>
> So, you can see why a branch prediction hit rate of 60% at that critical
> if-then is so ugly. Increasing it to (say) 95% theoretically (which is our
> hit rate after 3 probes) would be a 3x improvement.
>
> And having a method cache that cannot fit in the CPU cache is not much of
> a problem either; the CPU cache will figure out what entries are important
> and hold them.
>
>
> Scott
>
>
>
>
> --__--__--
>
> Message: 22
> From: Fleeberz at aol.com
> Date: Sun, 30 Sep 2001 16:28:33 EDT
> Subject: Re: Computerchannel.de: Squeak 3.0 tested
> To: squeak-dev at lists.squeakfoundation.org
> Reply-To: squeak-dev at lists.squeakfoundation.org
>
>
> In a message dated 2001-09-30 1:14:18 PM, karl.ramberg at chello.se writes:
>
> Fleeberz at aol.com wrote:
> >> I haven't gotten the hang of Smalltalk AT ALL yet, ...
>
> >The way I learned was by finding a simple morphic demo or program,
> >change a few parameters, open the morph and see what had happened...
> >Do this over and over again until you know what to change to get
> >what you want. Then expand by cutting and pasting in code from other
> >places, testing and debugging a few thousen times you will suddenly
> >just know how to do it the next time.
>
> A few thousand times? Oh weary me. I know that is not as much of a joke as
it
> sounds like (although I can appreciate some humor from it as well), and is
> really likely to be a low estimate for many people, er, me. ;^)
>
> >I allways browse with optinal buttons on and use the senders of and
> >implementors of all the time to see how stuff is used elsewere in
> >the image.
>
> Surely I've seen references to these things, but can't seem to recall
where;
> looks like it's time to start looking. With luck, I'll trip over them
early
> in the "search."  Will be looking for preferences, and perhaps browser
menu
> items such as "show implementors of..." and so on. See? It's easy! (?)
>
> Thanks, Karl.  I'd ask Cees or Bijan about it, but considering how I've
> treated those wonderful people in the past, I'm better off having my tail
> between my legs than nothing at all. :-/ (C, B, I'm sorry. Me bad.)
>
> --Flee
>
>
> --__--__--
>
> Message: 23
> Date: Sun, 30 Sep 2001 16:42:42 -0500
> Subject: [UNIX][VM] network audio support
> To: <squeak-dev at lists.squeakfoundation.org>
> From: "Lex Spoon" <lex at cc.gatech.edu>
> Reply-To: squeak-dev at lists.squeakfoundation.org
>
> I've dug into "NAS" (Network Audio System) and added some basic support
> to Squeak on Unix.  The reason this is useful is that multiple programs
> using NAS can share the audio device, plus they can access audio across
> the network.  This matters if you run programs other than Squeak that
> use sound, which as you might guess I do.  :)
>
> http://minnow.cc.gatech.edu/squeak/uploads/298/Squeak-lex.6.tgz
>
> or for the individual patches:
>
> http://www.cc.gatech.edu/~lex/squeak  (wait until tomorrow morning for
> this to update)
>
>
> This driver doesn't support as much as my OSS driver does; it doesn't
> have recording, and it doesn't have volume control.  But it might be
> enough to interest some people.
>
> Along the way, I've rearranged the sources a little so that you can more
> easily add new audio drivers.  So if somebody wants to shoot for ESD or
> ARTS or whatever else, go for it!
>
>
> -Lex
>
>
> --__--__--
>
> Message: 24
> Date: Sun, 30 Sep 2001 16:46:46 -0400
> Subject: Re: Storing and Retrieving Point
> To: tim at jump.net
> Cc: squeak-dev at lists.squeakfoundation.org
> From: raymondasselin at sympatico.ca
> Reply-To: squeak-dev at lists.squeakfoundation.org
>
> Tim Olson <tim at jump.net> wrote:
> > >Hi Bob can you explain a little bit the construct {'text'. pt} what is
> > >the semantic of these accolades '{ }' I did'nt find them in the
> > >Smalltalk books I get. What they means and for which uses were they
> > >made? Same for the meaning of the '.' between 'text' and pt.
> >
> > See:
> >
> >      http://minnow.cc.gatech.edu/squeak/90
> >
> >
> >
> >      -- tim
>
> Sadly I'm still used to check my books instead of Swiki...
> Thanks for the point, and reference.
>
> Raymond
>
>
> --__--__--
>
> Message: 25
> Date: Sun, 30 Sep 2001 17:10:19 -0400
> To: squeak-dev at lists.squeakfoundation.org
> From: "David N. Smith (IBM)" <dnsmith at watson.ibm.com>
> Subject: Re: Why we should remove {} from Squeak
> Cc: <danielv at netvision.net.il>,
> "squeak-dev at lists.squeakfoundation.org"
<squeak-dev at lists.squeakfoundation.org>,
> ducasse stephane <ducasse at iam.unibe.ch>
> Reply-To: squeak-dev at lists.squeakfoundation.org
>
> Stef:
>
> But......
> I use it all the time. It's a neat idea. Here is code I found by searching
all source for #{.
>
> EToyIncommingMessages class>>allTypes
>
> ^MessageTypes ifNil: [
> MessageTypes _ {
> self typeKeyboardChat.
> self typeMorph.
> self typeFridge.
> self typeStatusRequest.
> self typeStatusReply.
> self typeSeeDesktop.
> self typeAudioChat.
> self typeAudioChatContinuous.
> self typeMultiChat. } ]
>
> HeadMorph>>addSpikyHair
> | hair |
> hair _ PolygonMorph
> vertices: {83 at 3. 81 at 30. 91 at 27. 111 at 23. 97 at 32. 112 at 37. 99 at 45. 114 at 52.
95 at 53. 55 at 43. 10 at 50. 1 at 40. 14 at 40. 8 at 26. 24 at 37. 15 at 11. 29 at 29. 30 at 16. 36 at 30.
41 at 6. 49 at 31. 54 at 8. 61 at 32. 64 at 1. 70 at 27}
>
>
> Next is a fragment of a demo. The names Babble, Donut and Marble are
humorous, but descriptive.
>
> d := BabbleDonutMorph new largeDonut; name: '- Commons Area -'.
> d addAllMarbles: {
> (BabbleMarbleMorph new mediumMarble name: 'Wendy').
> (BabbleMarbleMorph new mediumMarble name: 'Dave and Archie at home').
> (BabbleMarbleMorph new mediumMarble name: 'Tom in Minneapolis').
> (BabbleMarbleMorph new mediumMarble name: 'Mark').
> (BabbleMarbleMorph new mediumMarble name: 'Big Screen').
> (BabbleMarbleMorph new mediumMarble name: 'John in the Lab') } shuffled.
>
>
> These are just a few examples I picked to show 'interesting' uses of {
and }. I found 794 methods in my image that contained at least one ${. A few
were commented out C code. Most were from the array constructor. It's used
widely and the result of using it makes the code much more readable than
'Array with:'.
>
> The construct is useful and widely used. If it has definitional and
theoretical problems, then certainly look at how to fix those; but let's not
discard a useful construct.
>
>
> At 8:49 +0200 9/30/01, ducasse stephane wrote:
> >Hi daniel
> >
> >I send it into the mailing list too becuase this is important
> >
> >From a language design perspective
> >
> >#() is necessary because it is compiled statically and cannot not be
> >simulated by other construct. Storing #() in Stack frame is not really
goo
> >but this is an implementation aspect.
>
> I don't think this is true at all.  #() holds symbols, strings, characters
numbers, and in some Smalltalk systems it holds true and false. There is
nothing in a #() that you can't write as a literal elsewhere, so:
>
>    #( abc 'abc' 4  3.4  $r )
>
> can be written as:
>
>    (Array with: #abc with: 'abc'), (Array with: 4 with: 3.4 with: $r)
>
> #() is a great convenience, but it is not absolutely necessary.
>
> >#{} is not necessary because this is shortcut macro-expansion for Array
new
> >add: ....
> >
> >So why having only this one. There is two ways:
> >    or you have a way to describe macro-expansion like in Scheme where
you
> >    can define any macro in a clear way (you know the semantic of the
stuff
> >
> >    or we remove it because why not having
> >        {{}} for Set new add:
>
> This already has a definition: an empty array in an array, like:  #(#()).
>
> >        {#[]} for OrderedCollection new....Because I use more
> >                OrderedCollection than array
>
> #[] isn't used in Squeak, but some Smalltalk systems (and maybe the
standard) use it as a way of writing ByteArray literals:  #[ 0 255 76 98 4
67 2 ]. We should carefully decide if we want to differ from others in
syntax. (Actually, 'we' already have; for some odd reason '#[' is a symbol!
I would have expected to have to write #'[' to get a symbol with a special
character.)
>
> I use more OrderedCollection's too, but it was interesting to note that
after looking at a couple of hundred uses of {} I don't recall seeing even
one message send to convert the array to something else. I think that OCs
are widely used when code is building an indexed collection, but rare when
the collection is built ahead of time by a literal array or array
constructor.
>
> >... SNIP ...
> >
>
> Dave
> --
> _______________________________
> David N. Smith
> IBM T J Watson Research Center
> Hawthorne, NY
> dnsmith at watson.ibm.com
>
>
> --__--__--
>
> Message: 26
> Date: Mon, 01 Oct 2001 00:04:35 +0200
> From: Torge Husfeldt <jean-jacques.gelee at gmx.de>
> To: squeak-dev at lists.squeakfoundation.org
> Subject: Re: Getting accents into Squeak
> Reply-To: squeak-dev at lists.squeakfoundation.org
>
> Dies ist eine mehrteilige Nachricht im MIME-Format.
> --------------80A37A49A5880464F2294B39
> Content-Type: text/plain; charset=iso-8859-1
> Content-Transfer-Encoding: quoted-printable
>
> Hi John,
>
> I don't think you're right with your observation that locale's aren't an
> option in your case. Most non US/UK-keyboards are somewhat
> internationalized in a way that you can access Characters from foreign
> countries with little effort. The simplest way to do this is the
> "DeadCharacters" support. Now this can be rather annoying because you
> have to type $^ $  to make the caret appear because the caret is a dead
> Character and the logic cannot decide wheater it will be combined to a
> %eacute; or not before the next key is pressed. The other solution I saw
> was the Sun-keyboards wich use a dedicated combine-key. Once this key is
> pressed the next two characters will be combined to one within the
> possibilities of the codepage used. This on the other hand will take
> three keystrokes instead of two for every character that is not on the
> keyboard, and that can be very annoying too.
>
> No matter, I implemented a very simple but effective "DeadCharacter"
> suport for squeak based on 2.8alpha-1789 - I hope it wont break to many
> things when filed in. Feel free to modify the code or the bindings as
> you pleases. =
>
>
> John Hinsley wrote:
> > =
>
> > I've just noticed this (the localisation stuff may have some kind of
> > bearing on it, but......).
> > =
>
> > There I was trying to put a name with "=FC" (that's u with a dieresis, =
> I
> > think) in my Rolodex. Of course, you can copy and paste into a workspac=
> e
> > and it'll appear correctly, but you can't copy and paste into a Text or=
>
> > ShowEmptyText morph (maybe I should take a look at that -- I think it
> > would be useful). But is there any way (and here I'm working in Linux -=
> -
> > I guess the old number pad method works in Windows) of typing it in
> > directly, given a UK keyboard? (Here, locales or foreign keyboards are
> > not really a viable option: I'd need some bizzare combination of
> > Italian, German and Spanish.
> > =
>
> > Cheers
> > =
>
> > John
> > =
>
> > --
> > Can't cope anymore? Desperate for help?
> > Join the 12 step program for those who yearn to give up Microsoft:
> > http://home.earthlink.net/~penguinrox/index.html
>
> P.S.: I think the french keyboard is the best starting point. they
> already have most of the characters there and the diaresis is a separate
> key so you can combine all the umlauts and tremata.
> --------------80A37A49A5880464F2294B39
> Content-Type: text/plain; charset=iso-8859-1;
>  name="DeadCharacterSupport-th.2.cs"
> Content-Transfer-Encoding: quoted-printable
> Content-Disposition: inline;
>  filename="DeadCharacterSupport-th.2.cs"
>
> 'From Squeak2.7 of 5 January 2000 [latest update: #1789] on 30 September =
> 2001 at 11:57:46 pm'!
> "Change Set: DeadCharacterSupport-th
> Date: 30 September 2001
> Author: Your Name
>
> Adds dead character support in ParagraphEditors and their heirs.
>
> ToDo: =
>
> 1 - make it a preference.
> 2 - think of ways to produce the remainig characters as well
> 3 - maybe switch to compose-key logic
> 4 - general cleanup of code
>
> Dead characters are a relatively straightforward way to type in Character=
> s that are not directly available on the keyboard. This approach is being=
>  used by several Keyboard-drivers. In this simplistic approach there is p=
> lenty of room for improvement. =
>
> Here is a list of characters that i have not been able to map to the give=
> n dead/character/character combinations. If you find a way to map them, g=
> o on.
> '$=A7 $=AE $=AF $=BE $=BF $=CE $=CF '
> A more thorough approach would surely make use of a combine-key (cf Sun-k=
> eyboards)"!
>
> Magnitude subclass: #Character
> instanceVariableNames: 'value '
> classVariableNames: 'CharacterTable DeadCharacters '
> poolDictionaries: ''
> category: 'Collections-Text'!
> ScrollController subclass: #ParagraphEditor
> instanceVariableNames: 'paragraph startBlock stopBlock beginTypeInBlock =
> emphasisHere initialText selectionShowing otherInterval lastDeadCharacter=
>  '
> classVariableNames: 'ChangeText CmdActions CurrentSelection FindText Key=
> board ShiftCmdActions TextEditorYellowButtonMenu TextEditorYellowButtonMe=
> ssages UndoInterval UndoMessage UndoParagraph UndoSelection Undone '
> poolDictionaries: 'TextConstants '
> category: 'Kernel-ST80 Remnants'!
>
> !Character methodsFor: 'converting' stamp: 'th 9/30/2001 22:20'!
> combineWith: aCharacter
> "Dead Character support"
> self isDeadCharacter ifFalse:[^ String with: self with: aCharacter].
> ^(self class deadCharacters at: self) at: aCharacter ifAbsent: [String w=
> ith: self with: aCharacter]! !
>
> !Character methodsFor: 'converting' stamp: 'th 9/30/2001 22:22'!
> isDeadCharacter
> ^ self class deadCharacters includesKey: self! !
>
>
> !Character class methodsFor: 'constants' stamp: 'th 9/30/2001 23:36'!
> deadCharacters
> "Dead characters are a relatively straightforward way to type in Charact=
> ers that are not directly available on the keyboard. This approach is bei=
> ng used by several Keyboard-drivers."
> "In this simplistic approach there is plenty room for improvement. For a=
>  beginning you might want to implement this as a class variable so it won=
> t get re-built from scratch on each invocation."
> "second: here is a list of characters that i have not been able to map t=
> o the given dead/character/character combinations. If you find a way to m=
> ap them, go on."
> "$=A7 $=AE $=AF $=BE $=BF $=CE $=CF "
> "A more thorough approach would surely make use of a combine-key (cf Sun=
> -keyboards)"
> ^DeadCharacters =
>
> ifNotNil:[DeadCharacters] =
>
> ifNil: [DeadCharacters _ Dictionary newFrom:
>    { $~ -> (Dictionary newFrom:{$n -> '-'. $A ->'=CC'. $O -> '=CD'. $N ->=
>  '?'. $a -> '<'.  $o -> '>' }).
> $^ -> (Dictionary newFrom:{$a -> '?'. $e -> '?'. $i -> '?'. $o -> '?'. $=
> u -> '?' }).
> $=A1 -> (Dictionary newFrom:{$A -> '?'. $a -> '?' }).
> $=AB -> (Dictionary newFrom:{$a -> '?'. $C -> ','. $E -> 'f'. $c -> '?'.=
>  $e -> '?'. $i -> '?'. $o -> '-'. $u -> '?' }).
> $` -> (Dictionary newFrom:{$A -> '=CB'. $a -> '^'. $e -> '?'. $i -> '?'.=
>   $o -> '~'. $u -> '?' }).
> $=AC -> (Dictionary newFrom:{$a -> 'S'. $A -> '?'. $O -> '?'. $U -> '?'.=
>  $e -> '?'. $i -> '*'. $o -> 's'. $u -> 'Y'. $y -> '=D8' })
>    }]! !
>
>
> !ParagraphEditor methodsFor: 'typing/selecting keys' stamp: 'th 9/30/2001=
>  23:40'!
> normalCharacter: t1 =
>
> | nextChar |
> nextChar _ sensor keyboard.
> lastDeadCharacter isNil
> ifTrue:[nextChar isDeadCharacter
> ifTrue:[lastDeadCharacter _ nextChar]
> ifFalse:[t1 nextPut: nextChar]]
> ifFalse:[
> t1 nextPutAll: (lastDeadCharacter combineWith: nextChar).
> lastDeadCharacter _ nil].
> =
>
> "t1 nextPut: sensor keyboard."
> ^ false! !
>
>
>
> --------------80A37A49A5880464F2294B39--
>
>
>
> --__--__--
>
> Message: 27
> Date: Sun, 30 Sep 2001 19:03:58 -0400
> From: Kevin Fisher <kgf at golden.net>
> To: squeak-dev at lists.squeakfoundation.org
> Subject: Re: Computerchannel.de: Squeak 3.0 tested
> Reply-To: squeak-dev at lists.squeakfoundation.org
>
> On Sun, Sep 30, 2001 at 09:00:36PM +0200, Karl Ramberg wrote:
> >
>
> [snip]
>
> > The way I learned was by finding a simple morphic demo or program,
> > change a few parameters, open the morph and see what had happened...
> > Do this over and over again until you know what to change to get
> > what you want. Then expand by cutting and pasting in code from other
> > places, testing and debugging a few thousen times you will suddenly
> > just know how to do it the next time.
>
> Ditto for me, too.  I can't quite remember where I started...I think
> I began dissecting some of the simpler stuff like Tetris, and then
> Celeste.  I learned a lot in a very short time, which really amazed me.
> And this includes the classes that don't even have class comments!
>
> It always amazes me how Squeak-- Smalltalk-- is so 'self educating'.
Heck,
> the only Smalltalk book I have is the old Purple book, which I've only
read
> about half of...and I only recently acquired both the "white" book and the
> "dark blue" book (which are all excellent reading, I must say).
>
>
> >
> > I allways browse with optinal buttons on and use the senders of and
> > implementors of all the time to see how stuff is used elsewere in
> > the image.
> >
> > Karl
> >
> >
>
>
> --__--__--
>
> Message: 28
> From: danielv at netvision.net.il
> Date: Mon, 1 Oct 2001 01:00:43 +0300
> Subject: Re: [modules] Ginsu in 3.1
> To: squeak-dev at lists.squeakfoundation.org,
> Joseph Pelrine <jpelrine at acm.org>, Sarkela <sarkela at home.com>
> Reply-To: squeak-dev at lists.squeakfoundation.org
>
>
> --==CelesteAttachment89133==
>
> Hans-Martin Mosner <hm.mosner at cityweb.de> wrote:
> > Perhaps we need some kind of module and repository concept here :-)
> Have you looked at SCAN? it doesn't suck too bad... ;-)
>
> > BTW, I'm now working on producing a complete first modularization of the
> > 3.1 image, but this will take some more time... Running the module
> > builder again and again is somewhat time consuming.
> So if it's three times faster, you'll finish real soon? :-)
>
> I promise it's faster. I'm pretty sure it's correct, too...
> But if you have the tests loaded, I'd run them first. ;-)
>
> > Cheers,
> > Hans-Martin
>
> Daniel Vainsencher
> --==CelesteAttachment89133==
> Content-type: application/octet-stream
> Content-transfer-encoding: base64
> Content-disposition: attachment;filename="GinsuSpeedup.1.cs.gz"
>
> H4sIAAAAAAAAAJ1TsW7bMBCdNegfzungZHEkx6kRBh0KFyk8uA1gZAqcghbPEROKVETKgYF8
> fI+UZCkN2qGTyDvdu/feHcc3lSlg/VIjf76YpFyVOQezgzmseJXlME2SFO4Vd2gd1KWgA4NP
> s4vZfANGQwo/M2e2WDU/cgfplF1+ZukV8GI8iuPRCl1uxDfcSS2dpJIiBOyNqRiMPazUj2Ow
> jhclBcQe0uQ8PQ9wScJmV4SizZ3emlqLH7xAu9QM+MIohZkHjKM36JPwFkeDG/sy/HUSRw8W
> 1Q4q3GGFOkPxXZktVxa4UmvupN0dGNwz5CT9HazUmaoFWgY+t4lHQOJInRG1wveayMvq8E9R
> 0zmJ2ksrtwoXilsbenghZUW8XlolcBLI5ly5yUkcDXOkK+SI9m0I14TmsKFjIY4iqZ9INIOl
> QO2kO6zRgcbXkHGGRAKzFOqkUjzydy7EV6UalfCB4jUcTF35zhvv5ZARAXS1p4HasJjyWTOF
> 0LjtGT6aqjdn11TeQbfWLlRtHS3W/3jbD66beGaK0miywg68OR2a8ypdTqutpTrrTWJZS5/Z
> zJTotyuKulint0n1PQcutVpuefbMH//YkxxV2ev7u6LLhBQVZo9LvfadVi0IPdw7XXBNwIJB
> 3R39GukwksDIU26uv5qV6XlO2qdCmWM1iONLtf7/ZmSMos2S+MPH5xgabI54wgweERWFk1fg
> OQ+ogve49+k3msYvdYwEAAA=
>
> --==CelesteAttachment89133==--
>
>
> --__--__--
>
> Message: 29
> Date: Sun, 30 Sep 2001 19:10:10 -0400
> From: Kevin Fisher <kgf at golden.net>
> To: squeak-dev <squeak-dev at lists.squeakfoundation.org>
> Subject: [OT] aRts (KDE) and Squeak...one year later
> Reply-To: squeak-dev at lists.squeakfoundation.org
>
> Hi folks:
>
> I think about a year ago I was experimenting with 'artsdsp' and Squeak.
> Arsdsp magically forwards all of Squeak's sound calls to a sound daemon.
> A year ago, it worked rather poorly with the nasty side-effect of driving
> CPU usage up to 100%.
>
> I've been playing with it with the latest artsd shipped with the latest
KDE
> and everything seems a lot better.  Even speech synth (which didn't work
> a  year ago) is working over aRts.  I haven't seen the 100% CPU problem
> happen either.
>
> So, for any aRts (or KDE, as I use aRts with Window Maker) users out
there,
> you can just do 'artsdsp squeak SomeImage.image' and Squeak will magically
> work with aRts...I haven't had any problems so far.
>
> I wonder how difficult a VM plugin for aRts would be?
>
>
> --__--__--
>
> Message: 30
> From: Daniel Joyce <daniel.a.joyce at att.net>
> To: squeak-dev at lists.squeakfoundation.org
> Subject: Re: Help with sound
> Date: Mon, 1 Oct 2001 17:30:38 -0500
> Reply-To: squeak-dev at lists.squeakfoundation.org
>
> On Sunday 30 September 2001 10:57 am, you wrote:
> > On Saturday 29 September 2001 09:15 pm, Daniel Joyce wrote:
> > > The Midi works great.
> >
> > Under Unix? How? The MIDI code is just stubbed out. You must be on a
> > different OS.
> >
>
> I mean the dsp synthesized MIDI, sound playing works fine.
>
> It seems to be a issue with SB cards and recording in general, none of my
> Sound apps can record from ALSA .5 and SB Live!. It seems every other
release
> breaks that feature.
>
> I'm looking on the ALSA list now.
>
> Daniel
>
>
> --__--__--
>
> Message: 31
> Date: Sun, 30 Sep 2001 19:13:33 -0400
> From: Doug Way <dway at riskmetrics.com>
> To: squeak-dev at lists.squeakfoundation.org, BEOneel at MindSpring.com
> Subject: Re: Which Squeak is preferred for Mac OS X ?
> Reply-To: squeak-dev at lists.squeakfoundation.org
>
>
> This reminds me that the OS X information on the squeak.org download
> page is a little out of date... it doesn't mention anything about the
> Carbon version.  Perhaps the text could be updated to point to the
> 3.0/mac directory of the ftp site to find the current Carbon/Cocoa VM's
> rather than hardcoding a link to an old VM.
>
> While I was thinking about this, I updated the DownloadForMacintosh
> swiki page with some info on OS X downloads.
>
> - Doug Way
>   dway at riskmetrics.com
>
>
> Doug Way wrote:
> >
> > For most everyday use, I'd say the Carbon version is the one to use.
Pretty much everything seems to work as far as I've tried, and it's better
than the Classic version (when using OS X) since you don't have to start up
the Classic environment, etc.
> >
> > The Cocoa version is promising, but last I checked it still has a
serious problem with events and/or screen updates, which makes the UI feel
quite slow.  Plus there are still a few missing features e.g. full screen
mode.  Eventually these things will be addressed, I'm guessing.
> >
> > (I personally switched back to using Squeak on OS 9, solely because
there wasn't 3-button mouse support available on OS X.  Although I believe
Marcel was adding support for this to the Cocoa version.)
> >
> > - Doug Way
> >   dway at riskmetrics.com
> >
> > Andy Stoffel wrote:
> > >
> > > [I couldn't find this information anywhere (mailing list or minnow)]
> > >
> > > If I understand it correctly there seem to be several different
> > > versions:
> > >
> > > - Carbon
> > > - Cocoa
> > > - Classic
> > >
> > > Is there a preferred version (I've only played with
> > > the Cocoa version so far) ? And a list of
> > > which one has what (if any ) feature running under Mac OS X ?
> > > [eg. I've noticed full screen doesn't seem to work in the
> > >  Cocoa version - is anything else not working in it?]
> > >
> > > Thanks
> > >
> > > -Andy-
>
>
> --__--__--
>
> Message: 32
> From: danielv at netvision.net.il
> Date: Mon, 1 Oct 2001 01:55:41 +0300
> Subject: Re: Skip List in #4308
> To: squeak-dev at lists.squeakfoundation.org
> Reply-To: squeak-dev at lists.squeakfoundation.org
>
> Well, I really cannot say who said what, or who I'm answering to, but -
> In Celeste, this could make recovery after image crashes faster. With
> SortedCollections re-adding the n logged actions done before the crash
> takes n^2 time, with SkipLists, presumably, n*log(n).
>
> Or we could tweak Celeste to not update this list during recovery, and
> resort it once later.
> Still nicer to just have nice behaving data structures...
>
> Anyone know what this thing eats in memory usage?
>
> Daniel
>
> Karl Ramberg <karl.ramberg at chello.se> wrote:
> > On Tue, 25 Sep 2001, Andreas Raab wrote:
> >
> > > Folks,
> > >
> > > I noticed them coming into Squeak a while ago and always since I
wanted to
> > > ask this question: What the heck are they good for?! ;-) I don't see
them
> > Basically, Skiplists (and/or balanced tree implementations of
> > SortedCollection) should be used anywhere where you have a lot of
> > inserts/removals from a sorted collection, and profiling shows that that
> > is a problem.
> >
> > Scott
>
>
> --__--__--
>
> Message: 33
> Date: Sun, 30 Sep 2001 16:23:03 -0700
> From: Tim Rowledge <tim at sumeru.stanford.edu>
> To: squeak-dev at lists.squeakfoundation.org
> Subject: Re: Skip Lists?!
> Reply-To: squeak-dev at lists.squeakfoundation.org
>
> Scott A Crosby <crosby at qwes.math.cmu.edu> is widely believed to have
written:
>
>
> > Nope. Most cachelines are 16 bytes.
> I doubt it. ARM uses a variety of cache line sizes ranging from 0 to 256
> bytes. I think PPC uses 32 bytes. Sparc uses a variety depending on step
> level (and just to add fun, the line range spec in the the flush
> instruction varies; some levels use a closed interval, some a half-open
> interval. Imagine the amusement derived from debugging a code generator
> when SUN forgets to tell you this ).
>
> tim
>
> --
> Tim Rowledge, tim at sumeru.stanford.edu, http://sumeru.stanford.edu/tim
> Useful random insult:- All booster, no payload.
>
>
>
> --__--__--
>
> Message: 34
> From: "PhiHo Hoang" <phiho.hoang at home.com>
> To: <squeak-dev at lists.squeakfoundation.org>
> Subject: Re: [ENH][Win32]RE: Font Support
> Date: Sun, 30 Sep 2001 21:09:56 -0400
> Reply-To: squeak-dev at lists.squeakfoundation.org
>
> Hi Andreas,
>
>     Thanks for the SoundPlugin. It works (almost ) like a charm for
> Squeak-D3D and not at all for SqM (I hope it's just another property
getter
> incident ;-)
>
>     I chose to load 'Times New Roman' and somehow I got extra little
> vertical rectangles everywhere this font is used. (Update #4347 -
> Squeak-D3D - Win2K-Pro)
>
>     Cheers,
>
>     PhiHo.
>
> ----- Original Message -----
> From: "Andreas Raab" <Andreas.Raab at gmx.de>
> To: <squeak-dev at lists.squeakfoundation.org>
> Sent: Saturday, September 29, 2001 5:55 PM
> Subject: [ENH][Win32]RE: Font Support
>
>
> > Dave,
> >
> > Thanks for the complaint. Perhaps to get started, attached is something
> that
> > might be of tremendous help for at least the Win32 people. The changes
> (and
> > the plugin) allow you to import any font you have on your system into
> Squeak
> > (and have it there as StrikeFont). All you have to do is to evaluate:
> >
> > HostFont textStyleFromUser.
> >
> > choose the font you'd like to import and get it. But be warned: The
> imported
> > fonts blow up your image... they tend to get large since a variety of
font
> > sizes is imported. The usual disclaimers apply (blabla, mumble mumble,
may
> > not work, may not look right, etc. etc. etc.)
> >
> > Cheers,
> >   - Andreas
> >
> >
> > > -----Original Message-----
> > > From: squeak-dev-admin at lists.squeakfoundation.org
> > > [mailto:squeak-dev-admin at lists.squeakfoundation.org]On Behalf Of David
> > > N. Smith (IBM)
> > > Sent: Saturday, September 29, 2001 1:02 PM
> > > To: squeak-dev at lists.squeakfoundation.org
> > > Subject: Font Support
> > >
> > >
> > > All:
> > >
> > > The single most troubling problem in Squeak is the lack of
> > > real font support. I've just finished a large, packaged
> > > program and had to make do with NewYork, and images of large
> > > characters in a font cut from Illustrator and pasted into
> > > Graphic Converter, then twiddled, and then saved as GIFs, in
> > > three large sizes, for about 30 characters each, then read
> > > into the program and then laid out by code I wrote. (There
> > > are other solutions; I disliked this one the least.)
> > >
> > > With real font support I could address the font directly.
> > >
> > > I know one issue is portability and that strike fonts move
> > > readily between platforms.
> > >
> > > Maybe there should be some way to:
> > >
> > > * Query the current platform for fonts with certain
> > > characteristics. (Answer all serif, answer all monospace,
> > > answer all bold Courier, etc.  X-Windows has a mechanism for
> > > doing such searches that might be useful to look at.)
> > >
> > > * Specify that a certain font is wanted; it then acts like a
> > > built-in font.
> > >
> > > * Squeak then asks the platform to draw the characters.
> > >
> > > There are issues wrt moving an image to a new platform, but I
> > > think they can be handled so long as 'Squeak itself' uses
> > > built-in, portable fonts.
> > >
> > > Maybe I'm all wet here, with a quickly proposed solution that
> > > might not work. Regardless, there needs to be SOME solution
> > > to the Squeak font problem.
> > >
> > > The current font support sucks. The TT converter doesn't work
> > > on a Mac, it does work on Windoze but only at 32 point and
> > > above. It doesn't fix the problem.
> > >
> > > One cannot claim to have the worlds greatest multimedia
> > > content builder, then admit to having only three fonts with
> > > fixed sizes.
> > >
> > > Dave
> > >
> > > Chris, thanks for reminding me that I've been going to complain too.
> > >
> > >
> >
>
>
>
> --__--__--
>
> Message: 35
> Date: Mon, 1 Oct 2001 13:55:15 +1200 (NZST)
> From: "Richard A. O'Keefe" <ok at atlas.otago.ac.nz>
> To: squeak-dev at lists.squeakfoundation.org
> Subject: Re: Localization in code
> Reply-To: squeak-dev at lists.squeakfoundation.org
>
> Jecel Assumpcao Jr <jecel at merlintec.com> wrote:
> Any reference for ILEX?
>
> http://cirrus.dai.ed.ac.uk:8000/ilex
>
>
>
> --__--__--
>
> Message: 36
> From: "PhiHo Hoang" <phiho.hoang at home.com>
> To: <squeak-dev at lists.squeakfoundation.org>
> Subject: Re: [ENH][Win32]RE: Font Support
> Date: Sun, 30 Sep 2001 21:09:56 -0400
> Reply-To: squeak-dev at lists.squeakfoundation.org
>
> Hi Andreas,
>
>     Thanks for the SoundPlugin. It works (almost ) like a charm for
> Squeak-D3D and not at all for SqM (I hope it's just another property
getter
> incident ;-)
>
>     I chose to load 'Times New Roman' and somehow I got extra little
> vertical rectangles everywhere this font is used. (Update #4347 -
> Squeak-D3D - Win2K-Pro)
>
>     Cheers,
>
>     PhiHo.
>
> ----- Original Message -----
> From: "Andreas Raab" <Andreas.Raab at gmx.de>
> To: <squeak-dev at lists.squeakfoundation.org>
> Sent: Saturday, September 29, 2001 5:55 PM
> Subject: [ENH][Win32]RE: Font Support
>
>
> > Dave,
> >
> > Thanks for the complaint. Perhaps to get started, attached is something
> that
> > might be of tremendous help for at least the Win32 people. The changes
> (and
> > the plugin) allow you to import any font you have on your system into
> Squeak
> > (and have it there as StrikeFont). All you have to do is to evaluate:
> >
> > HostFont textStyleFromUser.
> >
> > choose the font you'd like to import and get it. But be warned: The
> imported
> > fonts blow up your image... they tend to get large since a variety of
font
> > sizes is imported. The usual disclaimers apply (blabla, mumble mumble,
may
> > not work, may not look right, etc. etc. etc.)
> >
> > Cheers,
> >   - Andreas
> >
> >
> > > -----Original Message-----
> > > From: squeak-dev-admin at lists.squeakfoundation.org
> > > [mailto:squeak-dev-admin at lists.squeakfoundation.org]On Behalf Of David
> > > N. Smith (IBM)
> > > Sent: Saturday, September 29, 2001 1:02 PM
> > > To: squeak-dev at lists.squeakfoundation.org
> > > Subject: Font Support
> > >
> > >
> > > All:
> > >
> > > The single most troubling problem in Squeak is the lack of
> > > real font support. I've just finished a large, packaged
> > > program and had to make do with NewYork, and images of large
> > > characters in a font cut from Illustrator and pasted into
> > > Graphic Converter, then twiddled, and then saved as GIFs, in
> > > three large sizes, for about 30 characters each, then read
> > > into the program and then laid out by code I wrote. (There
> > > are other solutions; I disliked this one the least.)
> > >
> > > With real font support I could address the font directly.
> > >
> > > I know one issue is portability and that strike fonts move
> > > readily between platforms.
> > >
> > > Maybe there should be some way to:
> > >
> > > * Query the current platform for fonts with certain
> > > characteristics. (Answer all serif, answer all monospace,
> > > answer all bold Courier, etc.  X-Windows has a mechanism for
> > > doing such searches that might be useful to look at.)
> > >
> > > * Specify that a certain font is wanted; it then acts like a
> > > built-in font.
> > >
> > > * Squeak then asks the platform to draw the characters.
> > >
> > > There are issues wrt moving an image to a new platform, but I
> > > think they can be handled so long as 'Squeak itself' uses
> > > built-in, portable fonts.
> > >
> > > Maybe I'm all wet here, with a quickly proposed solution that
> > > might not work. Regardless, there needs to be SOME solution
> > > to the Squeak font problem.
> > >
> > > The current font support sucks. The TT converter doesn't work
> > > on a Mac, it does work on Windoze but only at 32 point and
> > > above. It doesn't fix the problem.
> > >
> > > One cannot claim to have the worlds greatest multimedia
> > > content builder, then admit to having only three fonts with
> > > fixed sizes.
> > >
> > > Dave
> > >
> > > Chris, thanks for reminding me that I've been going to complain too.
> > >
> > >
> >
>
>
>
> --__--__--
>
> Message: 37
> Date: Mon, 01 Oct 2001 02:59:03 +0100
> From: John Hinsley <jhinsley at telinco.co.uk>
> To: squeak-dev at lists.squeakfoundation.org
> Subject: Re: Getting accents into Squeak
> Reply-To: squeak-dev at lists.squeakfoundation.org
>
> Torge Husfeldt wrote:
> >
> > Hi John,
> >
> //snipped//
> >
> > No matter, I implemented a very simple but effective "DeadCharacter"
> > suport for squeak based on 2.8alpha-1789 - I hope it wont break to many
> > things when filed in. Feel free to modify the code or the bindings as
> > you pleases.
>
> Thanks Torge:
>
> I'm not absolutely sure what I'm supposed to type: the file in went
> without a hitch -- at least, it doesn't seem to have broken anything
> (but maybe something broke it? ;-) ). -- An admission, I never get on
> well with keyboard modifiers: the only ones I seem to be able to use are
> in gvim and a rather nice word processor on the Atari. (Neither of which
> are a great deal of use here!) Treat me as very stupid in this respect!
>
> I open up a Workspace, and run a "print it" on Characters deadCharacters
> and get:
>
>  a Dictionary($^->a Dictionary($a->'?' $e->'?' $i->'?' $o->'?' $u->'?' )
> $`->a Dictionary($A->'À' $a->'^' $e->'?' $i->'?' $o->'~' $u->'?' ) $~->a
> Dictionary($A->'Ã' $N->'?' $O->'Õ' $a->'<' $n->'-' $o->'>' ) $°->a
> Dictionary($A->'?' $a->'?' ) $´->a Dictionary($C->',' $E->'f' $a->'?'
> $c->'?' $e->'?' $i->'?' $o->'-' $u->'?' ) $¨->a Dictionary($A->'?'
> $O->'?' $U->'?' $a->'S' $e->'?' $i->'*' $o->'s' $u->'Y' $y->'ÿ' ) )
>
> I do a series of print its:
>
> $ isDeadCharacter false
>  Nothing more expected ->~ isDeadCharacter
> # isDeadCharacter #isDeadCharacter
>  Nothing more expected ->@ isDeadCharacter
>  Unmatched string quote ->' isDeadCharacter
> | isDeadCharacter Vertical bar expected ->
>  Nothing more expected ->? isDeadCharacter
> $^ isDeadCharacter true
> ^ isDeadCharacter nil
>
> (this in the belief that US keyboards have the $ in a different place to
> UK ones -- but for the life of me, I can't remember where!)
>
> Hmmm:
>
> Nothing more expected ->£ isDeadCharacter
>
> (Ah well, worth a try!)
>
> I simply don't get it!
>
> (This has always seemed to me to be one of the biggest holes in Linux:
> we have menus in languages from Hungarian through to simplified Chinese,
> but no easy way of getting even the simplest foreign character! Aaargh!)
>
> Cheers
>
> John
>
>
>
>
> --
> Can't cope anymore? Desperate for help?
> Join the 12 step program for those who yearn to give up Microsoft:
> http://home.earthlink.net/~penguinrox/index.html
>
>
> --__--__--
>
> Message: 38
> From: Daniel Joyce <daniel.a.joyce at att.net>
> To: <squeak-dev at lists.squeakfoundation.org>
> Subject: Caveats
> Date: Mon, 1 Oct 2001 19:15:03 -0500
> Reply-To: squeak-dev at lists.squeakfoundation.org
>
> I will say the biggest caveats I still have is the 'dirtiness' of morphic.
>
> Hard coded extents, or improper clipping cause all sorts of
text/buttons/etc
> to appear outside the bounds of their containing morphs.
>
> It's ugly in some places.
>
> Is there anyone working on cleaning this up?
>
> Also, has the Morphic rewrite started? Morphic still handles some
> translations/rotations inconsistently.
>
> If there is any work being done on rewriting morphic transforms, I'd be
> intrested in helping.
>
> Daniel
>
>
> --__--__--
>
> Message: 39
> From: Daniel Joyce <daniel.a.joyce at att.net>
> To: squeak-dev <squeak-dev at lists.squeakfoundation.org>
> Subject: Augh...
> Date: Mon, 1 Oct 2001 19:33:19 -0500
> Reply-To: squeak-dev at lists.squeakfoundation.org
>
> Okay, how I play a sound under Squeak?
>
> What class is used by the system to play sounds when morphs are
> added/removed?
>
> I've tried looking in Morph, but the class is a mess. ( Is anyone working
on
> a rewrite? ).
>
> I find it sad that for a system that supports so much easy OOP, that
things
> are not terribly clean, or refactored.
>
> So,
>
> A) I'd like to know the best place to look on where sounds are played, and
> how they are stored in Squeak ( what's the buffer used? ).
>
> B) I'd like to help to rewrite Morphic ( It needs a lot of help, it seems
> the base Class Morph is BALLOONING with features every release ).
>
> Squeak seems to suffer from too much enthusiasm, and too little design
> sometimes... ;)
>
> Daniel
>
>
> --__--__--
>
> Message: 40
> Date: Mon, 1 Oct 2001 14:38:37 +1200 (NZST)
> From: "Richard A. O'Keefe" <ok at atlas.otago.ac.nz>
> To: squeak-dev at lists.squeakfoundation.org
> Subject: Re: Localization in code
> Reply-To: squeak-dev at lists.squeakfoundation.org
>
> Karl Ramberg <karl.ramberg at chello.se> wrote:
> I find icons and pictures a interesting way to deal with
> this. Numbers are language independent. So is music notation.
>
> Numbers may be language-independent, but digits are not.
> In Unicode 3.1.1 there are no fewer than 27 versions of "DIGIT ZERO";
> No fewer than 5 of them are "fonted" variants for mathematics (I guess
> the Unicode maintainers have decided that "the thing about rules is,
> when you break them, break them hard").  But there are 17 genuine
> culturally different versions of "DIGIT ZERO", seventeen.
> Then of course there are decimal points and commas:
>     US ',' '.'
>     IEE  ' ' '<middot>'
>     EUR '.'     ','
> and quite a few other variations.  The underlying structure is
> nearly language-independent, but the details of the writing system aren't.
>
> As for music notation, it may be language-independent, but it is not
> culture-independent.  Byzantine music notation, old Western music
notation,
> classical Chinese music notation, and modern "Western" music notation are
> all quite different.  (Do I need to mention that modern music notation
> includes many words, which are sometimes translated, and sometimes not?)
> Come to think of it, when you see chord names, sometimes you see
> "Fmin" and sometimes you see "Fmoll".  Language difference.
>
> As for icons, they are just a hieroglyphic code.  By now we have enough
> icons to make learning Egyptian (which at its peak had about 700 different
> signs) look easy, and some of the commonest ones don't look the same under
> different operating systems.  Heck, they don't even look the same under
> different *programs*.  I'd be lost without tooltips.
> Think about this:  I have two keyboards on this desk.  On one the "power"
> key is left-hollow-triangle.  On the other, it's "bicycle padlock" (or a
> "C" lying on its back with an "I" between its jaws).
>
> Pictures are interesting, but inherently ambiguous.
>
>
>
>
> --__--__--
>
> _______________________________________________
> Squeak-dev mailing list
> Squeak-dev at lists.squeakfoundation.org
> http://lists.squeakfoundation.org/listinfo/squeak-dev
>
> End of Squeak-dev Digest





More information about the Squeak-dev mailing list