[BUG] Can't write files with Squeak 3.2.7Beta5 on MacOS X
10.1.2
Andrew P. Black
black at cse.ogi.edu
Thu May 23 08:15:02 UTC 2002
At 17:08 -0700 05/22/2002, John M McIntosh wrote:
>
>having a repeatable test case would be nice.
Well, this is repeatable on my system (OS X 10.2.1) -- but only with
my copy of the Squeak3.2gamma-4857.image. Just drop it on Squeak
3.2.7Beta5; in the Squeak system that starts, go to the change set
browser and try to file out the (empty) change set "New Changes".
I get a walkback when header is sent to nil.
But when I downloaded the zip file again and got a fresh
Squeak3.2gamma-4857.image, the problem does not re-occur. The only
thing that I can think of that may have messed things up is that the
first time that I started that image, I had forgotten to copy the
sources file into the same directory, and got the usual warnings. I
think that I saved before quitting and restarting ...
>
>Look for issues with file or folder name in the path exceeding 32 characters.
>Perhaps the file is opened and never closed, thus maybe an error is
>now tossed on the open again. Do repeated fileouts to the same
>changeset name 'foobar.cs' result in a problem?
Yep, the same problem occurs
>
>Squeak 3.0Alpha21MTCarbon uses the macintosh file open api which has
>different behavior than the Unix file api used by the mach-0 or
>Cocoa versions of the Mac VM, also it's large file name aware.
>
>Note of course the primOpen:writable: should really throw an
>exception here? Versus handing off a nil object that later results a
>does not understand. Somehow catching the error via walkback would
>be nicer when it happens.
I'm not overly fond of exceptions. I think that
open: afilename forWrite: aBoolean onFailure: aBlock
should be the interface that one uses if one wants to react to the
failure to write the file, and open:forWrite: should be defined in
terms of open:forWrite:onFailure: with a notifier as the failure
Block.
>
>PS Any aliased folders in the path?
No
Thanks for you help and forcing me to work though this.
Andrew
More information about the Squeak-dev
mailing list
|