Squeak-dev February 2002

squeak-dev@lists.squeakfoundation.org

202 participants
665 discussions

Re: A Tool for managing modules
by Henrik Gedenryd 01 Feb '02

01 Feb '02

Ned Konz wrote: >>> And I tried to import my Connectors change set (after changing all the >>> category names to be 'People-NedKonz-Connectors'). >>> >>> And I imported the change set. When it executes the postscript of the >>> change set, I get a walkback on a doesNotUnderstand:. Modules shouldn't be able to define doits. They contain static program definitions. (Use an initialization method.) >> I have tried to reproduce this behavior, but I couldn't. Where can I get >> your Connectors change ? Actually I have used the Connectors as one of my test cases. It's too bad the SqF server had gotten a bad configuration so we cant use it right now to distribute modules easily. We'll see if things change over the weekend. >>> But this is the strange thing: the Package Browser shows that I've added >>> the new method to the class in question (this is a one-method extension >>> to MessageSend). >>> >>> It's in a module called >>> #(People NedKonz Connectors 'Squeak Language Core Objects delta') >> >> Strange name for a module, but anyway I will try to fix that. This is a DeltaModule, its name is automatically generated from who defines it and what base module it affects. > I didn't name it. The import did. I tried to import to #(People NedKonz > Connectors) and it put my extensions in modules with names like that. Exactly. The importer recognizes class/module extensions automagically. >>> How am I supposed to use the system if I can't extend existing classes? But you just did! >> Which system do you refer (ModuleTool is simply a gui, a front-end to >> ModuleInstaller) ? > > The module system; I'm sure this isn't something you did. I like your GUI. > >> Are you sure to want to use a such name ? > > No, but the module system seems to want to. Torsten.Bergmann(a)phaidros.com wrote: > Has someone be able to track down the error of evaluating the > code twice before getting it into the image. It was not an error as such, the files had the wrong names, see my first posting in this thread. The fact that it works the second time may be a bug ;-) Henrik

1 0

[updates] 4 more; mostly whiplash
by Dan Ingalls 01 Feb '02

01 Feb '02

Folks - I've just posted 4 more updates, but they are mostly repairs for my having missed some important input from the harvesters. - Dan ----------------------- 4727initLeftMargin-ls -- Lex Spoon -- 14 January 2002 Make sure leftMargin gets initialized in CharacterScanner; this bug was brought out recently when Scamper is showing a page with indentation, and the mouse is waved over an indented line." 4728includesIdentity-sw -- Scott Wallace -- 30 January 2002 Fixes up one missed reference to References in the module work." 4729existenceOfPath-cmm -- Chris Muller, Scott Wallace -- 25 January 2002 The crux of this update is Chris's fix to FileDirectory>>assureExistanceOfPath:, which did not previously account for platform-specific differences in case-sensitivity, causing it to take a path of logic that quickly leads to a primitive failure on Win32 machine (attempting to create a directory that already exists). The 'existance' misspelling is corrected to 'existence' and the other method name with the same misspelling is also fixed, as of course are all senders." 4730FixUpdateGlitches-di -- Dan Ingalls -- 30 January 2002 In merging the various submissions from harvesters, I omitted several useful tweaks and reviewed decisions. This update attempts to set them straight, with apologies to Scott Wallace for my oversights. In particular, Restored the solution of 4728includesIdentity 4272InitLeftMargin supplants CharacterBlockScannerFix (simpler, more general) ShrinkWrapObjectsTool withdrawn until we can find a way to avoid jerkiness. Retracted 4705 and 4707. Improved wording of Brace explanation in 4714.

2 1

mac vm carbon 3.2.2b4
by John M McIntosh 01 Feb '02

01 Feb '02

I've posted a new mac VM 3.2.2b4 with the previously mentioned changes. That should show up in the usual places sometime soon. As mentioned the big issue change is FFI support. Usually I post VM's to few beta testers first, however in this case it goes out early to you all because the number of people I know that are using FFI is quite small (one). Hopefully there is more than one of you that needs this feature. Also I'm looking for someone to test the AppleScript support under OS-X, before I tackle getting the AppleScript plugin to work. -- -- =========================================================================== John M. McIntosh <johnmci(a)smalltalkconsulting.com> 1-800-477-2659 Corporate Smalltalk Consulting Ltd. http://www.smalltalkconsulting.com ===========================================================================

2 1

iteration - don't optimize it
by Mark Mullin 01 Feb '02

01 Feb '02

Ahh - that got your attention. No, I'm not really opposed to improving iteration speeds, by all means have at it. Terrific idea. My own vote is cast for a hashing class based on SHA5. I think you could get interesting results from using a perfect hash,even if the thing is 120 bits. On the other hand, maybe I didn't make my original point too well. My thread started over "can one language do everything" and my response was "no, languages are tools and you shouln't pound in a nail with a power saw". >From my own perspective, I live in a world where we drop down to native platform machine code in a few places, and C++ code may often be written to deal with little issues like "Pentiums usually predict you'll take a reverse branch over a forward branch." When performance gets to be that much of an issue (we have clients that require many billions of high level linear algebra calcs per second) you tend to survey your language tools very carefully. For the record, C++ iteration is what we use to deal with problems that really vex the computer but that the user could care less about. Smalltalk has a richer and more powerful set of iterators and collections than STL, but the costs of such a dynamic flexible solution are more than we can bear. Nothing I've said should be construed to mean I'm talking about collections the user interacts with directly. In my first post I observed that C/C++ was prefered over Smalltalk when the computers oddest peripheral wasn't directly involved :-) I really only keep responding to this because I do feel it's counterproductive to go "I don't care what you got, Squeak can do it better". I'm just arguing for a slight modification - If you don't care enough to profile it to death, then Squeak will do perfectly. Recognizing this means we can say, we don't care what your problem is, what Smalltalk can't do well directly it can do by being extended. Legacy systems, databases, whatever. But I rather violently disagree with those that claim you should just use Smalltalk cause it's iterators aren't as fast as C/C++. Bull. A C/C++ iterator/loop construct is a lot more expensive to write, a lot more expensive to maintain, and a lot less accessable. It's also a hell of a lot faster cause it's compiled, and cause if performance is the central thread, it's using static storage. For example, in such a world one has a tendancy to use a vector to aquire data of unknown length, and then immediately reconstruct it as a fixed size array as soon as the length is known. On reflection, if there are those who think I'm arguing for recoding the collection classes or using plugins to do collection -- neat idea, but no. I like Smalltalk collections just fine, I was only using collections/iteration as an example of why languages were tools. What is particularly interesting these days is that Smalltalk and it's close relatives seem to be the only ones that truly recognize that the class hierarchy is in fact a derivative language, and implicitly, a tool in its own right. In the real world of C++ it's hard to keep two class libraries from either stepping on each other or being incapable of efficient interoperation, depending on namespace:. In Smalltalk land, we regularly seem to throw 30 peoples work into one unified whole with a lot less sweat. I wrote a book in 89 under duress about how to use C++, and my primary focus was on 'if you do these things this way, you too can have the ability to create and share class libraries like we do every day in Smalltalk'. Funny, things haven't worked out that way. Yes we have more libraries, but nowhere near what I have expected. Now that's what I think is most unique of all about the Smalltalk tool. Why is it so damn good at collecting, merging, and sharing the work of many participants and everything else just sucks soooo bad. Mark

3 2

Re: [SqueakDot] Latest news
by goran.hultgren＠bluefish.se 01 Feb '02

01 Feb '02

1. Added description to polls. 2. Added non anonymous polls (only for account holders, can only vote once and only when logged in). But you can't see who has voted, and the model doesn't even track what you voted. 3. Further refined the HtmlBuilder approach which actually turned out to be pretty cool, just look at the source - you can build pages in amazingly little code. I have converted #addpoll:, #newaccount: #accounts: #settings: to this new approach. I hope that perhaps SqF could come up with more features for this so that it could be used for "real" polls/votes. Well, enough for today. There has been very little activity during the day but perhaps you are all asleep... :-) regards, Göran

1 0

Re: Tail Call
by Anthony Hannan 01 Feb '02

01 Feb '02

Scott A Crosby <crosby(a)qwes.math.cmu.edu> wrote: > BTW, what about tailcall out of a block? > > foo: > ^[^self bar] > > Is that a different bytecode? Tail call out of a block would be harder because the caller's receiver (the block) (which we are replacing) holds the return context and callee's (#bar) localReturns would have to be changed to a remoteReturns. So a new method would have to be made on the fly and the receiver would have to be wrapped in a new block closure with the return context. Is tail recursion required for continuations? I really don't understand continuations, can you give me some pointers? Cheers, Anthony

4 3

Re: [Challenge] large files smart compare (was: Re: Squeak for I/O and Memory Intensive tasks )
by danielv＠netvision.net.il 01 Feb '02

01 Feb '02

So merge whole records, not a problem. It's even not that important if you only use one field to sort, because you'll make the problem much smaller by eliminating many trivial matches. At which point, you can use the existing, expensive, but precise algorithm. See Richards very interesting description of this solution, though it depend on being comfortable with the unix toolbox. I think for this particular problem, if you need to do this task often, changing the algorithm will give you great results. As a general rule, and as a better answer to your challange, exploring what work was done on using BDB might be more generally useful. Have fun. Daniel Yoel Jacobsen <yoel(a)emet.co.il> wrote: > > --Boundary_(ID_vUEaGTJTEo1YvPamN1n8gQ) > Content-type: text/plain; charset=us-ascii; format=flowed > Content-transfer-encoding: 7BIT > > This is not correct since I need to compare entry to entry not pair to > pair. Sort will only damage the LDIF files. > > Yoel > > danielv(a)netvision.net.il wrote: > > >>0. Any good idea about how to make it practical for 450K entries (18M > >>lines)? What should I use for persistence? > >> > > > >Assuming that the entries have to be string equal to be equal and thus > >"not differences" and thus boring: > >1. Use some generic sorting utility like unix 'sort' to sort both > >inputs. They're pretty good at doing this for big files. > >2. Do something akin to a phase in merge sort - read both files in a > >synched manner. Any lines that match from both files, ignore. Any lines > >without matchers, keep. If you have many matchers, don't keep in memory, > >but write them to a file. > > > >This should be fast, and more useful than the python code. > > > >> Thanks > >> Yoel > >> > > > >Daniel > > > > > > > > > --Boundary_(ID_vUEaGTJTEo1YvPamN1n8gQ) > Content-type: text/html; charset=us-ascii > Content-transfer-encoding: 7BIT > > <html> > <head> > </head> > <body> > This is not correct since I need to compare entry to entry not pair to pair. > Sort will only damage the LDIF files. > >     Yoel > > <a class="moz-txt-link-abbreviated" href="mailto:danielv@netvision.net.il">danielv(a)netvision.net.il</a> wrote: > <blockquote type="cite" cite="mid:0GQQ00FGD4CDGX@mxout1.netvision.net.il"> > <blockquote type="cite"> > <pre wrap="">0. Any good idea about how to make it practical for 450K entries (18M lines)? What should I use for persistence? </pre> > </blockquote> > <pre wrap=""> Assuming that the entries have to be string equal to be equal and thus "not differences" and thus boring: 1. Use some generic sorting utility like unix 'sort' to sort both inputs. They're pretty good at doing this for big files. 2. Do something akin to a phase in merge sort - read both files in a synched manner. Any lines that match from both files, ignore. Any lines without matchers, keep. If you have many matchers, don't keep in memory, but write them to a file. This should be fast, and more useful than the python code. </pre> > <blockquote type="cite"> > <pre wrap=""> Thanks Yoel </pre> > </blockquote> > <pre wrap=""> Daniel </pre> > </blockquote> > > </body> > </html> > > --Boundary_(ID_vUEaGTJTEo1YvPamN1n8gQ)--

1 0

Replacing mouse pointer
by olmeca＠netscape.net 01 Feb '02

01 Feb '02

Hi there, I remember having read somewhere that the mouse pointer is actually a morph (in morphic projects), so I was wondering if I could replace it by another morph. Can anyone point out how to do this? Rudi Angela -- __________________________________________________________________ Your favorite stores, helpful shopping tools and great gift ideas. Experience the convenience of buying online with Shop@Netscape! http://shopnow.netscape.com/ Get your own FREE, personal Netscape Mail account today at http://webmail.netscape.com/

2 1

[BUG] VirtualRootModule<<cachedClassNames not updated
by Craig Latta 01 Feb '02

01 Feb '02

Hi Henrik-- It seems that VirtualRootModule<<cachedClassNames doesn't get updated (it doesn't get nilled when classes are added or removed), so, e.g., new classes aren't found with "find class..." in the system browser. -C -- Craig Latta composer and computer scientist craig.latta(a)netjam.org www.netjam.org crl(a)watson.ibm.com Smalltalkers do: [:it | All with: Class, (And love: it)]

2 1

Re: Persistence & DTSTTCPW: ZODB clone?
by goran.hultgren＠bluefish.se 01 Feb '02

01 Feb '02

cg(a)home.cdegroot.com (Cees de Groot) wrote: > <goran.hultgren(a)bluefish.se> said: [SNIP] > Note that CATS wants you to subclass transactional objects (IIRC); that's Are you sure? No, I think you needed to "enable" your classes using a special compiler. But I may be remembering Stephen's demo wrong. Persistence should of course (most seem to agree on that nowadays) be orthogonal to type/class. > maybe something you don't want. However, I don't know another way of doing > what CATS does with routing instance variable accesses to the transaction > (except for some transparent proxy magic, which would work but maybe doing > this through DNU is bound to be slow) > > >I know he has been thinking about all the other parts too and a small > >bird has whispered in my ear that there might be stuff coming... > > > Could you wisper a bit louder, bird? I'd hate to do double work ;-) Stephen will have to whisper himself I think. ;-) Drop him an email. > >As another sidenote, I have been playing with ImageSegments a bit and > ><plug> if you sign up for an account at SqueakDot </plug> you can > >actually see it in action right there - you can > >checkpoint/revert/analyze your account using ImageSegments and you will > >see some performance numbers right away. > > > The only thing I keep asking myself is whether ImageSegments are good for > fine-grained storage (say a persistent objects and the reference graph to > its non-persistent contained objects, as opposed to a big project). They seem > to be a bit heavy-weighted, don't they? Yep, they are a bit on the heavy side. But they may be useful nevertheless! :-) They are darn fast coming in, a bit slower going out of course given how it's done, you pay a price for the overhead of a markphase of course. But it's neat. Note: I use export segments at SqueakDot and that costs a bit more of course. It's because I want to be able to start fresh from an empty image. > (OBTW: David Gorisek is contemplating an OmniBase port to Squeak; however, > what's keeping him until now is that he cannot seem to do file locking from > Squeak - anything coming there? Is there a lowest common denominator between > the Squeak platforms for file locking?) All the file stuff needs a big overhaul I guess. There are people pondering that. regards, Göran

1 0

Jump to page: