Squeak-dev April 1999

squeak-dev@lists.squeakfoundation.org

130 participants
319 discussions

RE: Weak support for directory manipulation in Squeak (?)
by Peter Crowther 09 Apr '99

09 Apr '99

> I think it's silly that deleting a directory > fails if it's not empty - it should recursively delete its contents. I know of no OS that does this. If the application wishes to recursively delete a large area of filestore, then it's up to the application to do so. Partly, that's because it's easier for the OS to implement that way round! Deletion can get particularly interesting on Windows, where the Explorer hides file extensions and you can therefore have a directory with the same (apparent) name as a file. If you ask the system to delete the file, using its visible name, you would instead delete the directory. And, yes, I *do* know someone who's done this. I'll get a load of responses to this saying 'but that's just a GUI issue that arises due to the appalling UI design in Windows 95'. Yes, but there are plenty of other counter-intuitive examples in other systems. I'd rather have the safety of a *primitive* delete-directory function only removing an empty directory, and requiring non-primitive code to walk the directory structure to do a recursive delete. - Peter

1 0

Re: Squeak -> Objective-C translation
by Marcel Weiher 09 Apr '99

09 Apr '99

Hi Les, yes, producer was one of my inspirations. Thanks to your pointer I was able to take another look at it, but it was just like I remembered it: old and crufty. A squeaky-clean solution is just so much nicer! Thanks again, Marcel > From: Les Tyrrell <tyrrell(a)canis.uiuc.edu> > > You may want to look at Producer, which I believe was a Smalltalk to > Objective-C translator done way back when, probably for ObjectWorks. > It used to be available on the UIUC archive, although I am having > trouble finding it now.

2 1

Re: Weak support for directory manipulation in Squeak (?)
by Tim Rowledge 09 Apr '99

09 Apr '99

On Thu 08 Apr, Stephen Travis Pope wrote: > Well, if there's general agreement about this, then we could simply take the > code that makes the call to PBHDeleteSync() out of the dir_Delete() function > I sent out earlier today and put it in the function sqFileDeleteNameSize in > sqFilePrims.c (making that file less portable -- does anyone know if rmdir() > works on the Mac?) Put it in sqXXXDirectory.c and then the portability is no worse than before - there's already platform spread. tim -- Strange OpCodes: DMK: Destroy Memory-protection Key Tim Rowledge: rowledge(a)interval.com (w) +1 (650) 842-6110 (w) tim(a)sumeru.stanford.edu (h) <http://sumeru.stanford.edu/tim>

1 0

Re: Weak support for directory manipulation in Squeak (?)
by johnm＠wdi.disney.com 09 Apr '99

09 Apr '99

I'd like to keep sqFilePrims.c portable. I think the best solution is to add a directory delete primitive, as Stephen first suggested. I'm not eager to add a directoryExists primitive, since there is already a way to get that information (as Stephen's filein demonstrates). Basically, the code Stephen sent out looks fine and I see no reason not to fold it into the baseline. -- John At 1:02 PM -0700 4/8/99, Stephen Travis Pope wrote: >Duane Maxwell wrote: >> >> I think they should be treated identically except for creation. The >> differentiation of behaviors for directories and files should be handled >> within the primitives, with the "natural" thing being done, even if it >> means throwing an error. > >Well, if there's general agreement about this, then we could simply take the >code that makes the call to PBHDeleteSync() out of the dir_Delete() function >I sent out earlier today and put it in the function sqFileDeleteNameSize in >sqFilePrims.c (making that file less portable -- does anyone know if rmdir() >works on the Mac?)

1 0

Re: Weak support for directory manipulation in Squeak (?)
by Duane Maxwell 08 Apr '99

08 Apr '99

Stephen Travis Pope asks: >does anyone know if rmdir() works on the Mac? Yup - it ultimately turns into a PBHDeleteSync() call. You can find the source in the "{CodeWarrior}Metrowerks Standard Library:MSL C:MSL Mac:Source" directory in "unistd.mac.c". =================================================== Duane Maxwell dmaxwell (at) launchpados.com CTO http://www.launchpados.com Launchpad, Inc. (619) 578-8500 x226 Information contained herein is my personal opinion and not necessarily that of Launchpad, Inc. ===================================================

1 0

RE: Weak support for directory manipulation in Squeak (?)
by Norton, Chris 08 Apr '99

08 Apr '99

Duane Maxwell wrote: > I think they should be treated identically except for creation. The > differentiation of behaviors for directories and files should be handled > within the primitives, with the "natural" thing being done, even if it > means throwing an error. I think it's silly that deleting a directory > fails if it's not empty - it should recursively delete its contents. > [Norton, Chris] Yikes! The user should probably be warned first! When I was a Unix newbie, I accidentally executed the following command, thinking that it would delete all of the core files from all subdirectories, and boy was I mistaken! It nuked my whole machine and it took a couple of hours to restore everything back. rm -rf * "core" (Note: the actual syntax may have changed... I haven't touched Unix in over 7 years) ---==> Chris

2 1

RE: Weak support for directory manipulation in Squeak (?)
by agree＠carltonfields.com 08 Apr '99

08 Apr '99

You can also use DirectoryEntry>>isDirectory. For example: FileDirectory default entryAt: 'ReadMe.txt' isDirectory You can drill to the DirectoryEntry object about a bazillion different ways. -----Original Message----- From: MIME :Andreas.Raab@disney.com Sent: Wednesday, April 07, 1999 3:08 PM To: squeak(a)cs.uiuc.edu Subject: RE: Weak support for directory manipulation in Squeak (?) Stephen, > In porting the Minnestore OODB to Squeak, I'm running into problems with > directory manipulation. There are methods in class FileDirectory for > creating > directories, but not for deleting them (FileDirectory deleteFileNamed: > won't > work) or asking whether a directory exists (FileDirectory fileExists: > answers > false), and there's also no method for asking whether a given name is a > directory or a normal file (though this can be written using existing > methods). > Have a look at FileDirectory>>fileAndDirectoryNames vs. FileDirectory>>fileNames. From both it should be easy to derive a #directoryNames or #isDirectory: method. Andreas -- PLEASE NOTE NEW EMAIL ADRESS! +===== Andreas Raab ========= (andreasr(a)wdi.disney.com) ==+ | Walt Disney Imagineering Phone: +1 818 544 5016 I I Glendale, CA Fax: +1 818 544 4544 I +======< http://isgwww.cs.uni-magdeburg.de/~raab >========+ << File: ENVELOPE.TXT >>

3 4

2 fixes for PositionableStream
by Stephen Pope 08 Apr '99

08 Apr '99

These two methods appear to both have been broken (and thus to have kept ReferenceStreams from working) in 2.3 with the recent updates. 'From Squeak 2.3 of January 14, 1999 on 07 April 1999 at 3:55:02 am'! !PositionableStream methodsFor: 'nonhomogeneous accessing' stamp: 'stp 04/07/1999 03:54'! nextNumber: n "Answer the next n bytes as a positive Integer or LargePositiveInteger." | s | n <= 2 ifTrue: [s := 0. n timesRepeat: [s := s * 256 + self next]. ^s]. s := Integer new: n neg: false. 1 to: n do: [:i | s digitAt: n + 1 - i put: self next]. "reverse order of significance" ^s normalize! ! !PositionableStream methodsFor: 'nonhomogeneous accessing' stamp: 'stp 04/07/1999 03:54'! nextString "Answer a new String read from the receiver. The first byte is the length of the string, unless it is greater than 192, in which case the first two bytes encode the length. " | aString length | length := self next asciiValue. "first byte." length >= 192 ifTrue: [length := length - 192 * 256 + self next]. aString := String new: length. 1 to: length do: [:i | aString at: i put: self next asCharacter]. ^aString! ! -- stp Stephen Travis Pope stp(a)create.ucsb.edu http://www.create.ucsb.edu/~stp/

1 0

Directory management -- first pass
by Stephen Pope 08 Apr '99

08 Apr '99

Here are some system extensions in response to my query about directory manipulation. At the moment, I can't compile a new VM to test the primitive, do any Mac (or other platform) VM makers have the time to look this over. 'From Squeak 2.3 of January 14, 1999 on 08 April 1999 at 5:55:08 am'! !FileDirectory methodsFor: 'testing' stamp: 'stp 04/08/1999 05:39'! fileExists: filenameOrPath "Answer true if a normal file of the given name exists. The given name may be either a full path name or a local file within this directory." "FileDirectory default fileExists: 'Nomad:Squeak:2.3' " "FileDirectory default fileExists: 'Nomad:Squeak:2.x' " "FileDirectory default fileExists: 'Nomad:Squeak:2.3:ReadMe.txt' " | dir fname | FileDirectory splitName: filenameOrPath to: [:filePath :tail | fname := tail. filePath isEmpty ifTrue: [ "file in this directory" dir := FileDirectory default] ifFalse: [ "file has its own path" dir := FileDirectory on: filePath]]. ^dir fileNames includes: fname! ! !FileDirectory methodsFor: 'testing' stamp: 'stp 04/08/1999 05:38'! directoryExists: filenameOrPath "Answer true if a directory of the given name exists. The given name may be either a full path name or a local file within this directory." "FileDirectory default directoryExists: 'Nomad:Squeak:2.3' " "FileDirectory default directoryExists: 'Nomad:Squeak:2.x' " "FileDirectory default directoryExists: 'Nomad:Squeak:2.3:ReadMe.txt' " | dir fname | FileDirectory splitName: filenameOrPath to: [:filePath :tail | fname := tail. filePath isEmpty ifTrue: [ "file in this directory" dir := FileDirectory default] ifFalse: [ "file has its own path" dir := FileDirectory on: filePath]]. ^dir directoryNames includes: fname! ! !FileDirectory methodsFor: 'testing' stamp: 'stp 04/08/1999 05:36'! fileOrDirectoryExists: filenameOrPath "Answer true if a file or directory of the given name exists. The given name may be either a full path name or a local file within this directory." "FileDirectory default fileOrDirectoryExists: 'Nomad:Squeak:2.3' " "FileDirectory default fileOrDirectoryExists: 'Nomad:Squeak:2.x' " "FileDirectory default fileOrDirectoryExists: 'Nomad:Squeak:2.3:ReadMe.txt' " | dir fname | FileDirectory splitName: filenameOrPath to: [:filePath :tail | fname := tail. filePath isEmpty ifTrue: [ "file in this directory" dir := FileDirectory default] ifFalse: [ "file has its own path" dir := FileDirectory on: filePath]]. ^dir includesKey: fname! ! ------ New Prim for directory delete -------- C Source -- put in sqMacDirectory.c int dir_Delete(char *pathString, int pathStringLength) { /* Delete the existing directory with the given path. */ Str255 name; HParamBlockRec pb; int okay, newRefNum, newVolNum; int i; for (i = 0; i < pathStringLength; i++) { name[i] = pathString[i]; } name[i] = 0; okay = lookupPath(pathString, pathStringLength, &newRefNum, &newVolNum); c2pstr((char *) name); pb.fileParam.ioNamePtr = name; pb.fileParam.ioVRefNum = newVolNum; pb.fileParam.ioDirID = newRefNum; return PBHDeleteSync(&pb) == noErr; } --- ST source 'From Squeak 2.3 of January 14, 1999 on 08 April 1999 at 6:00:06 am'! !Interpreter methodsFor: 'file primitives' stamp: 'stp 04/08/1999 05:59'! primitiveDirectoryDelete "Delet the directory whose name's on the top of the stack." | dirName dirNameIndex dirNameSize | dirName _ self stackTop. self success: (self isBytes: dirName). successFlag ifTrue: [dirNameIndex _ dirName + BaseHeaderSize. dirNameSize _ self lengthOf: dirName]. successFlag ifTrue: [self success: (self cCode: 'dir_Delete((char *) dirNameIndex, dirNameSize)')]. successFlag ifTrue: [self pop: 1]. "pop dirName; leave rcvr on stack"! ! 'From Squeak 2.3 of January 14, 1999 on 08 April 1999 at 6:01:26 am'! !Interpreter class methodsFor: 'initialization' stamp: 'stp 04/08/1999 06:01'! initializePrimitiveTable "This table generates a C switch statement for primitive dispatching." .... insert the following in the right place (163 primitiveDirectoryDelete) (164 168 primitiveFail) --------------------------------- stp Stephen Travis Pope stp(a)create.ucsb.edu http://www.create.ucsb.edu/~stp/

1 0

Re: [VM] NCM performance report
by shaping＠bigfoot.com 08 Apr '99

08 Apr '99

----- Original Message ----- From: Peter Crowther <peter.crowther(a)it-iq.com> To: <squeak(a)cs.uiuc.edu> Sent: Tuesday, April 06, 1999 5:08 AM Subject: RE: [VM] NCM performance report > > From: Dan Ingalls [mailto:DanI@wdi.disney.com] > > Sent: 25 March 1999 06:15 > [...] > > We have been considering > > th NCM changes since Tim first got them working. Simplicity > > and malleability are certainly positive features. > > Performance, if it's real, is as well. > > > > The main reason I have balked at this change so far is the > > cost in space for small implementations > [...] > > A competing change we have been weighing for a long time is > > to do away with compact headers. This would add about the > > same amount of space (70k) to the mini image, but would yield > > greater simplifications and speedups in the virtual machine > > (IMO). > > [I know, I'm late as usual. I've been off-site for a couple of weeks and, > while I could get at my mailbox, I couldn't mail out] > > For interest, I tried bodging the SystemTracer to write out an image with > 3-word headers for every object in the distribution 2.3 image. It took a > 3.5Mb image to a shade over 5Mb if I made the correct changes --- a large > penalty on a PDA. The flip side is that some of us [hi Adrian] trying to > run multi-user systems on Squeak are willing to pay that space for a large > speedup if we could get one. I am, too. I've had my head in the object allocation and > GC stuff, so may have a biased view, but it looks like some considerable > speed could be gained by doing away with the twisty turny maze of tests > about object formats. Has anyone done any experiments in this area? If > not, it looks like I'll have to do this anyway at some point, as adding an > optional fourth security word to the header causes all kinds of merry hell > if you try to keep the existing variable-length headers. Simplify. Cheers.

4 5

← Newer
1
...
23
24
25
26
27
28
29
...
32
Older →

Jump to page: