Squeak-dev June 2000

squeak-dev@lists.squeakfoundation.org

134 participants
686 discussions

RE: DSA changes open for comment
by Raab, Andreas 01 Jun '00

01 Jun '00

[Bob wrote] > As best I can tell the two Mac VMs that Dan placed on uiuc > recently do not have the *new* LargeIntegers code in them. I > cannot be sure about Andreas' Windows VM, but his > announcement of it sounded like it did not contain the > LargeIntegers plugin either. The latest Windows VM (2.8alpha) does contain LargeIntegers. It's been built based on the changes that were current at this time. [Stephan wrote] > I think most people update their VMs, if there is a official > (not alpha) version number change. > Currently we are staying at VM2.7. > > Is this true or may I be wrong here? That depends. People who are not on the aggressive update road through the alpha updates should certainly stay at 2.7. Those who want to have the very latest stuff should make sure they got the latest VMs - this is important since more and more code now relies on the fast primitive failure for named primitive (most importantly from LargeIntegers ;-) - Andreas

1 0

Re: DSA changes open for comment
by Bob Arning 01 Jun '00

01 Jun '00

On Wed, 31 May 2000 22:59:33 +0200 Stephan Rudlof <sr(a)evolgo.de> wrote: >I think most people update their VMs, if there is a official (not alpha) >version number change. >Currently we are staying at VM2.7. > >Is this true or may I be wrong here? It's hard to say - who knows how many lurkers out are there and what they are doing. Personally, I tend to lag behind in getting new VM's simply because it's another step, but in this case, Dan warned us that we could see significant degradation in performance if we took all the new image updates and were missing some optional plugins. On Friday he wrote: ---- In the same 2.8 folder on UIUC where I put the interim 2.8beta VM for the Mac, there is now a "real" one, by the name of SqueakVM2.8beta.sea.bin This VM... IS built from all the latest plugin sources IS "fat" (ie it will run on a 68K) DOES NOT have the new LargeInteger primitives (but it rocks with the plugin!) DOES have fast recovery from missing primitives (so it still runs fine without the plugin) DOES support 'Smalltalk listBuiltinModules' and related functions Thanks to all the VM construction workers, and to John for compiling this VM. ---- So I got that one. The story on other platforms may vary. Cheers, Bob

2 1

Re: DSA changes open for comment
by Bob Arning 01 Jun '00

01 Jun '00

On Wed, 31 May 2000 14:56:21 EDT JArchibald(a)aol.com wrote: ><< Note: if you have a DSA plugin for your platform, but do NOT have the >LargeIntegers plugin, then you probably do not want to use this. >> > >Bob, > >I don't think this is quite correct, Bob. If you have the latest VM, at least >on Windows, you have the LargeIntegers code even without a LargeIntegers >plug-in. I haven't checked the latest VM that Dan posted, but I think that >it's probably true for that one, alse. Jerry, What I was trying to say was that if you do NOT have the LargeIntegers plugin compiled C code (either as a separate shared library/DLL/.so OR bound into the VM), then the second of these changesets may dramatically slow down DSA execution. It was not my impresssion that any of the regularly available VMs currently have the LargeIntegers plugin internally, although that is certainly a possibility with the new plugin architecture. Cheers, Bob

2 1

Re: Proposal3: Make $_ a valid identifier character
by Richard A. O'Keefe 01 Jun '00

01 Jun '00

This is historically wrong. (See eg. "The early history of Smalltalk".) Smalltalk did not adopt ascii until ST-80. Before this they used their own set, including "eyes" and smiley faces in ST-72. What Smalltalk*72* used or was able to use is one thing. The fact remains that the Altos did have an arrow key, did not have a smiley key, and did not have an underscore key. Hence, they could obviously use whatever they wanted. Not *and* share files with Mesa or Lisp people. Not *and* share files with Bravo. A look at some ST-76 code shows that the intercappedStyle was used already then, Only relevant to the discussion *if* underscore was available. I think it is a definite possibilty that underscores came to be used because there were no lowercase characters avaiable, only uppercase. (Lowercase? Luxury.) While *some* equipment didn't have lower case, Flexowriters did. Perhaps more interestingly, the American military codeset FIELDATA had lower case before 1963. WeDontWriteEnglishThisWaySoWhyShouldWeWriteProgramsThisWay? Note in particular that the underscore was placed in character sets to support typewriter-style underlining (ASCII-63 and ASCII-67 are both paper-type/typewriter codes rather than punched card/line printer codes). The hypothesis that it has some connection with upper case is falsified by APL, which uses an up triangle to separate words (LIKE/\THIS), even though it has an underscore (the up triangle in APL is a single character, the same width as the letters).

1 0

Current VM locations. Was: DSA changes open for comment
by JArchibald＠aol.com 01 Jun '00

01 Jun '00

>From previous mail: => 5/31/00 4:41:54 PM EDT, JArchibald(a)aol.com => [corrected[ The latest Squeak2.7 Virtual Machines have been posted to ftp://st.cs.uiuc.edu/pub/Smalltalk/Squeak/2.7/files/SqueakExe.zip for Windows (I have verified this one) and ftp://st.cs.uiuc.edu/pub/Smalltalk/Squeak/2.8/Squeak_VM_2.8-NOT.sea.bin for the Macintosh (though I haven't personally tested this one -- the VM~2.8). I don't believe the Linux versions have been completed and made available yet. Jerry. ____________________________ Jerry L. Archibald systemObjectivesIncorporated ____________________________

2 1

[FIX][DSA] DSA-Fix-sr
by Stephan Rudlof 01 Jun '00

01 Jun '00

"Change Set: DSA-Fix-sr Date: 1 June 2000 Author: Stephan Rudlof Fixes a bug in DigitalSignatureAlgorithm>>initRandomFromUser, which could lead to malformed randoms from user. " -- Stephan Rudlof (sr(a)evolgo.de) "Genius doesn't work on an assembly line basis. You can't simply say, 'Today I will be brilliant.'" -- Kirk, "The Ultimate Computer", stardate 4731.3

1 0

Jump to page: