Squeak-dev May 2000

squeak-dev@lists.squeakfoundation.org

137 participants
680 discussions

Re: Time to make a Freshmeat kind of Squeak site ?
by Lex Spoon 02 May '00

02 May '00

Karl Ramberg <karl.ramberg(a)chello.se> wrote: > What I sugeste is a Freshmeat ( freshmeat.org ) kind of site on a swiki > and > some simple rules to how to categorize/describe what the stuff does and > then try to get as many as possible to put theyr struff there. I agree. We actually have two contenders already for the site: 1. The Smalltalk archive at UIUC. 2. The Swiki At this point, we are still lacking a critical mass of projects posted. There aren't enough projects posted, that someone is likely to go searching through either of the above sites for interesting Squeak code. As a smaller improvement, it sure would be nice if the Swiki pages were better organized. Do we really need separate Goodies, Packages, and Tools pages? Furthermore, the current listings are pretty hard to read, because they just dump the information in a long list. Anyway, once such a thing gets started, it should grow like mad. -Lex

1 0

[BUG] Error opening MagnifierMorph in 2.8alpha
by Doug Way 02 May '00

02 May '00

With the latest updates (through 2042), if you try to open a MagnifierMorph through the "new morph..." menu, you'll get a debugger window. (Same problem with FishEyeMorph, too.) The problem seems to be in MagnifierMorph>>sourcePoint... checking "versions" shows that it was modified 3/24/2000 by sma. (By the way, I use MagnifierMorph quite a bit... it's great when you need to examine how things are drawn.) - Doug Way EAI/Transom Technologies, Ann Arbor, MI http://www.transom.com dway(a)mat.net, @eai.com

2 1

Re: [updates] 10 more for 2.8alpha
by Bruce O'Neel 02 May '00

02 May '00

Hi, <ftp://st.cs.uiuc.edu//Smalltalk/Squeak/2.7/files/Squeak2.8a-beo.zip> has these changes cheers bruce Dan Ingalls <Dan.Ingalls(a)disney.com> wrote: > Folks - > > Herewith 10 short updates. > > - Dan > ------------------------- > 2043PreserveSelectionFix-di -- Dan Ingalls -- 27 April 2000 > Prevents loss of editor state when moving text with active keyboard focus. While this may sound unlikely, it is what happens when you type into a field that gets realigned as a result of the edit. > > 2044DropShadowTweaks-di -- Dan Ingalls -- 27 April 2000 > Several improvements in the control over morphs with drop shadows > (thanks to Scott Wallace for helping to track down the loose ends): > The halo properly shows the name of the target object. > Resize control goes properly to the shadowed morph. > The morph menu now includes three items for control over the shadow. > > 2045PNGTweaks-dm -- Duane Maxwell/EntryPoint -- 27 April 2000 > This is a changeset to Squeak 2.8a #2040. > It deals with a couple of bugs in the PNGReadWriter class. > 1) fixes bug with unusual-depth odd-sized non-interlaced indexed images > 2) fixes bug with extremely small interlaced images (even Netscape has a > bug here) > With this changeset, all of the non-corrupt images at: > http://www.cdrom.com/pub/png/pngsuite.html > should be correctly rendered to the degree we support PNG's esoterica. > Note that this page contains three intentionally corrupt images which may > cause walkbalks. > > 2046AutoResizeScreen -- Dan Ingalls -- 27 April 2000 > Causes both Morphic and MVC projects to resize their Display when the OS window has been resized. > > 2047NewSources1-hmm -- Hans-Martin Mosner <hmm(a)heeg.de> -- 26 April 2000 > First part of a modification that concentrates all mapping from source pointers to actual source files in one place. This is needed for more sophisticated source code management schemes, such as that needed for Collage." > > 2048NewSources2-hmm -- Hans-Martin Mosner <hmm(a)heeg.de> -- 26 April 2000 > Second part of the modification of the SourceFiles mechanism. In this change set, all methods that think they know about the structure of sourcePointers are replaced by versions that delegate to SourceFiles." > > 2049DigitSymbols-bf -- Bert Freudenberg -- 27 April 2000 > Fixes printing of one digit symbols (e.g., #'3')." > > 2050FixResidentFonts-di -- Dan Ingalls -- 28 April 2000 > Fixes artifacts in bold and italic fonts due to lack of leading on 1-wide characters introduced in update #1975. > > 2051TurtlePen-ar -- Andreas Raab -- 27 April 2000 > Fixes a problem with a nil turtle pen when a PasteUp was saved with its pen down. The PasteUp did rely on turtleTrailsForm isNil = turtlePen isNil which was broken by the #prepareToBeSaved. The fix simply removes the initial assumption and restores the turtle pen if needed." > > 2052FileListMenuItems -- Dan Ingalls -- 28 April 2000 > It appears that update 2011PNGstuff sustained a bit of damage in some line-wrapper, causing the text for the 'read image into ImageImports' menu item to appear on two lines, and thus fouling up the rest of any menu with that option. Andreas reported the problem, and this should fix it.

1 0

Re: Comanche/Swiki
by Bijan Parsia 02 May '00

02 May '00

--On Tuesday, May 02, 2000, 12:52 PM -0400 "John Duncan" <jddst19+(a)pitt.edu> wrote: > Hi guys, > > I'm vainly searching through the image for that little "aha!" that lets me > understand Comanche and Swiki 9. Bolot, others, perhaps you could give me a > few pointers on where to start exploring these seemingly powerful tools? > I've installed them, and I've played with the default Swiki, but I need to > get well beyond that stage to coming up with novel applications using > Comanche and the swiki stuff already written. > > Any help would be greatly appreciated. Well, I have a little simple Comanche app which I wrote for the purpose of testing a friends theories about using HTTP/XML without a RPC layer. http://www.unc.edu/~bparsia/DWChat.8.cs I plan to enhance/comment it for the sake of the Squeak book's networking chapter. A simple modification would be to make it use HTML rather than XML. If you want to start with a Swiki, I suggest peeking at the actions folder for starters. And the templates. Did you have a particular project in mind? (Note it took me about two days to figure out comanche with no docs for the DWChat. This included figuring out what we wanted to do :)) Cheers, Bijan.

1 0

Comanche/Swiki
by John Duncan 02 May '00

02 May '00

Hi guys, I'm vainly searching through the image for that little "aha!" that lets me understand Comanche and Swiki 9. Bolot, others, perhaps you could give me a few pointers on where to start exploring these seemingly powerful tools? I've installed them, and I've played with the default Swiki, but I need to get well beyond that stage to coming up with novel applications using Comanche and the swiki stuff already written. Any help would be greatly appreciated. -John

1 0

Re: Celeste design (was Re: [ENH] Celeste attachments/displaying a thingie)
by Bijan Parsia 02 May '00

02 May '00

--On Tuesday, May 02, 2000, 9:01 AM -0500 "Lex Spoon" <lex(a)cc.gatech.edu> wrote: > >> IMHO, MIMEPart should be renamed MIMEDocument. It is a Composite, possibly >> containing other MIMEDocuments. MailMessage should be reserved for, well, >> mail messages. Maybe it could go. In the correspondents framework, there's the following hierarchy: NetMessage (1) EmailMessage (2) NewsMessage (3) > Sounds reasonable. The only odd situation is data+headers vs. data w/o > headers. Currently we have MIMEDocument when there are no headers, and > MailMessage which is a MIMEDocument plus some headers. Note, by they Well, one sor tof MIMEDocument + headers, right? > way, that a multipart message doesn't make sense unless there is at > least one header to specify the separator string. So, a multipart > MIMEDocument right now is a really useless critter, because you can't > separate the parts. But having minimal headers merely to specify separator string doesn't make one a mail message, does it? > So MIMEDocument should have at least one header. But once you do that, > why not just add a full complement of headers? If that one header is > legitimate, then it came from a real header and it can't hurt to record > those other headers along with it. Unless you're making something that might have a variety of different headers depending on how it gets served. Then you might very well *want* to have a minimal set of headers, and convert on demand. Cheers, Bijan Parsia.

1 0

Re: Celeste design (was Re: [ENH] Celeste attachments/displaying a thingie)
by Bert Freudenberg 02 May '00

02 May '00

On Tue, 2 May 2000, Lex Spoon wrote: > So, a multipart MIMEDocument right now is a really useless critter, > because you can't separate the parts. Yeah, I saw the code that just grabs the first line remotely resembling a MIME boundary ;-) > > All in all, this is great! Unfortunately, I don't see an easy way to make > > Celeste cooperate with the normal Unix mail handling tools ... > > But then, does it matter that much? Most Unix stations have POP > available as well. I prefer leaving the mail on the server. I guess what I really want is IMAP, but if I say this, someone will point at me and sing that darned Nike slogan ... -Bert-

1 0

[Bug] DrawShadowMorph
by Karl Ramberg 02 May '00

02 May '00

When the shadow gets a translucent color it does not get removed when you move the morph. Seems to be something with the offset? Karl

1 0

RE: VRML in Squeak? was: XP course at UIUC using Squeak
by Bruce O'Neel 02 May '00

02 May '00

Hi, The GeoVRML website is at http://www.ai.sri.com/geovrml cheers bruce "Adam Hill" <adamhill(a)pobox.com> wrote: > VRML in a nutshell: > > VRML - Been around for a while, declarative, file based 3D scene > technology. Has bindings to Java, JavaScript and COM. Is an ISO > standard. Mainly used in a web browser plugin style. There are loaders > for Java3D. > > Lots of hype in the beginning, but has simmered down over the years. > People now actually build useful things with it. Currently being > upgraded and componentized, syntax is being changed for XML'ization. > > GeoVRML is a recent addon by some people at SRI. NIST is invloved in > conformance suite. > > Check out http://www.web3d.org. > > adam... > > > > -----Original Message----- > > From: John-Reed Maffeo [mailto:rlpa80@email.sps.mot.com] > > Sent: Monday, May 01, 2000 10:38 AM > > To: squeak(a)cs.uiuc.edu > > Subject: VRML in Squeak? was: XP course at UIUC using Squeak > > > > > > A quick peek at Altavista found little GeoVRML information. Please > > elaborate on why you think that GeoVRML would be a useful addition > > to Squeak. > > > > Would someone please compare and contrast VRML and Alice? > > > > Thanks, > > > > John-Reed Maffeo > > Mesa, AZ > >

1 0

RE: XP course at UIUC using Squeak
by Baveco, Dr. J.M. 02 May '00

02 May '00

I would like to second this. It would really be nice to be able to create AliceWorlds from carthographic data. As an alternative, GeoVRML adds maybe a dozen new objects to VRML; if GeoVRML were supported in Squeak (as far as I know VRML support is at least partly present already) we would have a (maybe more generic) solution to creating 3D object worlds from carthographic data. cheers, Hans > -----Oorspronkelijk bericht----- > Van: John-Reed Maffeo [SMTP:rlpa80@email.sps.mot.com] > Verzonden: vrijdag 28 april 2000 20:16 > Aan: squeak(a)cs.uiuc.edu > Onderwerp: Re: XP course at UIUC using Squeak > > Roger, > > This project is very interesting to me. I don't think I will be able > to attend - I would need much more lead time to get away from work for > so long:( But I would like to provide a vision of what I want to > be able to do one day and let you decide if there is any synergy > possible as the projects move forward. > > Preamble > > 1. The USGS has made a huge amount of cartographic source data > available on the web for free. > > 2. The data is stored compressed in a well defined standard format. > http://mcmcweb.er.usgs.gov/sdts/ > > 3. The data can be used to create planemetric and topographic maps. > > 4. Planemetric data is presented as a topologicaly structured graph and > the topographic data is presented as an array of elevations. > These data can be convolved to produce a very high quality > 3D topographic map. > > > > Vision > > Point a browser to a known dataset and materialize a 2 or 3D map of > an area as described by the data. > > Point a browser to a pair of known 2 and 3D datasets > covering the same geographic area and materialize an Alice World > of that area (or a subset of it) and enable the user to > explore it. > > Discussion > > There is no place on the web that I have found that provides this > kind of service. There are places where you can get digitized map > images but they are high bandwidth raster graphics and offer no > user interaction other than zooming and panning. > > The SDTS data standard is a recent evolution in geospatial data > distribution and forms the basis of future work on the ITS > (Intelligent Transportation System). > > Squeak has the required infrastructure in Baloon Graphics and > Alice. > > > Best Regards, > > John-Reed Maffeo > Mesa, AZ > > > > > > > > > Roger Whitney wrote: > > > > Ralph Johnson and I are planning to offer a course on extreme > programming > > (XP) May 15 - June 9 at UIUC in Champaign-Urbana Illinois. Students will > > learn XP by working as an XP team to build a multiuser multimedia system > in > > Squeak. > > > > There is a course description at > > http://wiki.cs.uiuc.edu/cs497rej/XP+workshop. See > > http://www.xProgramming.com/ for more information about extreme > programming. > > > > While the course is designed for UIUC students, others are welcome to > > attend the course. Those with experience in Squeak or XP would not have > to > > attend the entire time to benefit from the course and contribute to the > > project. If you are interested in attending all or part of the course > > contact me at whitney(a)cs.sdsu.edu for more details. > > > > If you do not attend the course you can help us out by acting like a > > customer. Take a look at the project description at > > http://wiki.cs.uiuc.edu/cs497rej/ObjectWeb and suggest > features/extensions > > to the project. So here is your chance to suggest to a team of UIUC > > students to implement some Squeak feature you always wanted. Of course, > as > > is the XP way, the on-site customer will rank all features (stories) in > > order of importance. > > > > Roger Whitney Mathematical & Computer Sciences Department > > whitney(a)cs.sdsu.edu San Diego State University > > http://www.eli.sdsu.edu/ San Diego, CA 92182-7720 > > (217) 328-0824 > > (619) 594-3535 > > (619) 594-6746 (fax)

3 2

Jump to page: