Squeak-dev March 2000

squeak-dev@lists.squeakfoundation.org

131 participants
580 discussions

RE: [ENH] Case independent numbers
by Lex Spoon 01 Mar '00

01 Mar '00

Bert Freudenberg <bert(a)isgnw.CS.Uni-Magdeburg.De> wrote: > On Mon, 28 Feb 2000, Jarvis, Robert P. (Contingent) wrote: > > > > >I wrote: > > > >who of you really would expect that 16r1e5 equals 1048576 while 16r1E5 is > > > >485? If not too many, I propose to make hex numbers case independent, and > > > >exponent notation, too. > > > > > > Dan wrote: > > > I have a problem with this, Bert. > > > The way it is, there is a simple syntactic rule for numbers. > > > Your changes imply a rule that is not syntactic, but value-dependent: > > > > > > 16r1E5 = 485, but > > > 14r1E5 = 537824 By the way, the standard seems to suggest that uppercase letters are to be used, and that the base should only go up to 36. (section 3.5.6) Lowercase 'e' is used to designate the exponent of a floating point number (as are 'd' and 'q'). It is illegal to specify an exponent for what is otherwise an integer--eg, 10e10 is illegal, while 10.0e10 is legal. Anyway, if anyone is planning to go through cleaning this stuff up, maybe they should make it fit the standard. Lex

2 1

RE: [VIRUS WARNING] Re: Check this
by agree＠carltonfields.com 01 Mar '00

01 Mar '00

Here is the culprit: http://www.symantec.com/region/uk/avcenter/venc/vbs_freelink.html

3 2

Re: Not infectable...
by Lawson English 01 Mar '00

01 Mar '00

On Tue, Feb 29, 2000 5:39 PM, Charles-A. Rovira <mailto:crovira@wt.net> wrote: >Its times like these that I'm glad I use a Mac and DON'T use any other >language that Smalltalk. :-) > > Ditto. Seems to me that this is a matter for law enforcement agencies, actually. Is it easier to track users of mailing lists than individual e-mails or can Mr. "Garcia" get away without punishment? ------------------------------------------------------------------------- "If everyone lived with a sense of wonder, their lives would be filled with joy." -Last words of Doug Henning, 5/3/47 - 2/7/00 -------------------------------------------------------------------------

4 3

Comments on comments on comments (no elipsis) (Was Re: [CCC] . . .)
by agree＠carltonfields.com 01 Mar '00

01 Mar '00

May I suggest reserving [CCC] and Re: [CCC] title lines to actual proposals for comments, and actual proposals for amendments thereto, perhaps after a discussion? Obviously comments on documentation invite substantial discussion -- and that is good, but in time someone will need to edit these documents, and it would be nice if he/she could rely on the headings alone to assure that he/she had the final word on the consensus of the community, without having to browse through extensive discussion. > -----Original Message----- > From: MIME :crovira@wt.net > Sent: Wednesday, March 01, 2000 11:00 AM > To: squeak(a)cs.uiuc.edu > Subject: Re: [CCC] a few comments on the comments on my comments... > > > Hello Chris, > > the var name classes it could assume are a pain to get right post > implementation. For some things its very difficult to get > right (and that's > only structural description,) the really hard part is "why?" > > I think we may very well cause the sources file and change > sets to become huge. > So what? :-) > > -Charles-A. > > "Norton, Chris" wrote: > > > Hi! > > > > Dwight Hughes [SMTP:dwighth@ipa.net] wrote: > > ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ > > "Believe me, I would like to see things go far beyond what > they are now (and > > I have a number of ideas of what and how), but all of that > is just smoke and > > noise until we create the core documentation to begin with." > > ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ > > > > Here here! You are doing a fine, contentious job Dwight! > More power to you > > and many thanks from me! > > > > Although Mr. Rovira has some good points, a lot of > formalism can be added > > post-factum. In particular, I think an > auto-comment-generator (see Dan & > > Vassili's ideas in that direction) should be added to Squeak > > _in_addition_to_ the traditional, hand-crafted comments. > Then we'll get our > > cake and eat it too. > > > > Thanks for your helpful additions to Squeak! > > > > ---==> Chris > > > > PS> I particularly liked your 2nd definition of Object! Nice job. > > > >

1 0

Re: [CCC] a few comments on the comments on my comments...
by Charles-A. Rovira 01 Mar '00

01 Mar '00

Hello Chris, the var name classes it could assume are a pain to get right post implementation. For some things its very difficult to get right (and that's only structural description,) the really hard part is "why?" I think we may very well cause the sources file and change sets to become huge. So what? :-) -Charles-A. "Norton, Chris" wrote: > Hi! > > Dwight Hughes [SMTP:dwighth@ipa.net] wrote: > ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ > "Believe me, I would like to see things go far beyond what they are now (and > I have a number of ideas of what and how), but all of that is just smoke and > noise until we create the core documentation to begin with." > ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ > > Here here! You are doing a fine, contentious job Dwight! More power to you > and many thanks from me! > > Although Mr. Rovira has some good points, a lot of formalism can be added > post-factum. In particular, I think an auto-comment-generator (see Dan & > Vassili's ideas in that direction) should be added to Squeak > _in_addition_to_ the traditional, hand-crafted comments. Then we'll get our > cake and eat it too. > > Thanks for your helpful additions to Squeak! > > ---==> Chris > > PS> I particularly liked your 2nd definition of Object! Nice job.

1 0

[ENH] Case independent numbers
by Bert Freudenberg 01 Mar '00

01 Mar '00

Hi, who of you really would expect that 16r1e5 equals 1048576 while 16r1E5 is 485? If not too many, I propose to make hex numbers case independent, and exponent notation, too. Changeset is attached. -Bert-

2 2

[VIRUS WARNING] Re: Check this
by Bruce ONeel 01 Mar '00

01 Mar '00

It's with messages like this that I am so happy that I read my mail from Squeak with Celeste. Celeste, of course, doesn't do executable content in email. :-) cheers bruce Duane Maxwell <dmaxwell(a)entrypoint.com> wrote: > This appears to be a Visual Basic virus. DON'T EXECUTE IT! > > >Have fun with these links. > >Bye.

1 0

[CCC] a few comments on the comments on my comments...
by Dwight Hughes 01 Mar '00

01 Mar '00

In general, I would prefer that class comments be reasonably self contained -- so I don't have a problem with the minor redundancy of the True/False comments -- especially since the comments are so short. Also, True and False have the unusual property (in the class hierarchy anyway) of truly being mirror images of each other -- so the repetition is particularly apparent. If the common section was more extensive, then I would want to move it to their superclass and add links to it instead. One thing I'm trying to be mindful of as I compose a comment is to try to give enough detail to prevent some of the various "newbie" problems and misunderstandings that I've seen come up -- and to try to document some of the "gotchas" in general. I am actually not too crazy about having a class comment explicitly mention subclasses of that class (like I did in Boolean) -- but for the core classes like these, it is very important to document their interrelations and structure (which change only rarely), so I think certain exceptions can be made. I would definitely agree that the variables defined for a class should be documented by their purpose in the class -- just documenting the class of object they usually contain is of little value. Without the added meaning, we might as well leave them out of the comment. Now, concerning thoughts about what all we could/should/might potentially turn the comments and their general support and examination structure into -- I would like to ask that we *please* try to stay focused on creating the fundamental class comments for a reasonably comprehensive set of classes for now. Believe me, I would like to see things go far beyond what they are now (and I have a number of ideas of what and how), but all of that is just smoke and noise until we create the core documentation to begin with. -- Dwight

1 0

[CCC] False class comment - v2
by Dwight Hughes 01 Mar '00

01 Mar '00

----------------------------------------------------------------- False defines the behavior of its single instance, false -- logical negation. Notice how the truth-value checks become direct message sends, without the need for explicit testing. Be aware however that most of these methods are not sent as real messages in normal use. Most are inline coded by the compiler as test and jump bytecodes - avoiding the overhead of the full message sends. So simply redefining these methods here will have no effect. ----------------------------------------------------------------- This particular version is my comment as edited by Stefan Matthias Aust and is already in update #1891. -- Dwight

1 0

[CCC] Object class comment - v2
by Dwight Hughes 01 Mar '00

01 Mar '00

---------------------------------------------------------------- Object is the root class for almost all other classes in the class hierarchy (except for its superclass ProtoObject -- and any other classes that directly subclass nil). Object inherits the minimum necessary object behavior defined in ProtoObject and adds the behavior/protocols common to all normal objects, such as access, copying, comparison, error handling, message sending, and reflection. Object has no instance variables, nor should any be added. Several classes of objects that inherit from Object either have special implementations (SmallInteger and UndefinedObject for example), or their structure and layout are known to the VM. Because Object is the root of the normal inheritance tree, general utility methods are often defined here to give all objects special behaviors needed by certain subsystems or applications, or to respond to certain general test messages. Class Variables: DependentsFields an IdentityDictionary Maps each object to its dependents. Allows all objects to take part in the dependency mechanism, even if they do not have an explicit dependents field. See Object>>dependents and its senders. ---------------------------------------------------------------- Thanks to all who have contributed their thoughts on this - I hope I captured the essence of most of your contributions. -- Dwight

1 0

Jump to page: