I am looking for someone running a UNIX version of IDL 7.1 to confirm that the problem described in this article dealing with Dialog_Pickfile has been fixed.
Specifically, if you specify both a file name and a path when you call Dialog_Pickfile, and then do nothing else except hit the Accept button, the file name that is returned has two paths appended to it and is bogus.
I am interested in what is in the Selection box, and what is returned.
I am, say we say, dubious. :-)
Thanks,
David -- David Fanning, Ph.D. Coyote's Guide to IDL Programming (www.dfanning.com) Sepore ma de ni thui. ("Perhaps thou speakest truth.")
> I am looking for someone running a UNIX version of > IDL 7.1 to confirm that the problem described in > this article dealing with Dialog_Pickfile has > been fixed.
> Specifically, if you specify both a file name and > a path when you call Dialog_Pickfile, and then do > nothing else except hit the Accept button, the file name > that is returned has two paths appended to it and is > bogus.
> I am interested in what is in the Selection box, > and what is returned.
In IDL 7.1 in Linux, I see the path with the file appended to it in the selection box, and the return value when I just accept it is correct.
On Aug 18, 7:17 pm, David Fanning <n...@dfanning.com> wrote:
> I am interested in what is in the Selection box, > and what is returned.
Hi David,
I think that the include-the-path-in-the-Selection-box behavior has been there for a long time - at least since 5.4. I've always treated it as an annoying platform-dependence rather than a bug.
Mike writes: > I think that the include-the-path-in-the-Selection-box behavior has > been there for a long time - at least since 5.4. I've always treated > it as an annoying platform-dependence rather than a bug.
Which is obviously wrong. So maybe it is just a bug in how wildcards are handled. Anyway, it sure looks and feels like a bug to me. :-)
Cheers,
David
P.S. Can you check the test case above in IDL 7.1 and let me know what you see? It will probably help with my report to ITTVIS. You will probably have to substitute your own directories for mine.
-- David Fanning, Ph.D. Fanning Software Consulting, Inc. Coyote's Guide to IDL Programming: http://www.dfanning.com/ Sepore ma de ni thui. ("Perhaps thou speakest truth.")
> Which is obviously wrong. So maybe it is just a bug in > how wildcards are handled. Anyway, it sure looks and feels > like a bug to me. :-)
> Cheers,
> David
> P.S. Can you check the test case above in IDL 7.1 and let me > know what you see? It will probably help with my report to > ITTVIS. You will probably have to substitute your own > directories for mine.
Indeed, it messes up the result when there is a wildcard in the path:
On Aug 19, 1:25 pm, David Fanning <n...@dfanning.com> wrote:
> Well, you are probably younger than I am. :-)
Maybe - plus I'm used to oddities in the unix/linux and unused to the idea that a software company might actually act on bug reports.
Either way, you inspire me to heights of idealism and usefulness in the world. :-)
> P.S. Can you check the test case above in IDL 7.1 and let me > know what you see? It will probably help with my report to > ITTVIS. You will probably have to substitute your own > directories for mine.
Except for the inexplicable expansion to my username, instead of yours, I saw the same result:
Mike writes: > Maybe - plus I'm used to oddities in the unix/linux and unused to the > idea that a software company might actually act on bug reports.
Well, given the amount of effort I had to go to to even convince anyone there might be a *problem*, let alone a bug, I wouldn't put too much hope in a rapid solution. (Shouldn't technical support people be *required* to read the e-mails we send them!)
> Either way, you inspire me to heights of idealism and usefulness in > the world. :-)
We all need more hope. Lord knows there is a deficiency in the world!
Cheers,
David -- David Fanning, Ph.D. Coyote's Guide to IDL Programming (www.dfanning.com) Sepore ma de ni thui. ("Perhaps thou speakest truth.")
On Wed, 19 Aug 2009 11:25:22 -0600, David Fanning wrote: > P.S. Can you check the test case above in IDL 7.1 and let me know what > you see? It will probably help with my report to ITTVIS. You will > probably have to substitute your own directories for mine.
I'm sure coyote noticed this, but file_search has an option to /expand_tilde.
Perhaps that could be used in some manor or fashion to solve the problem "graciously" (like beating ITTVIS over the head :-)
I had encountered with this bug a while ago on Red Hat release 4. The way around to this was to use EXPAND_PATH() function. IDL> theDirectory = EXPAND_PATH('~/IDL/coyote/') IDL> theFile = 'fsc_surface.pro' IDL> selectedFile = Dialog_Pickfile(FILE=theFile, PATH=theDirectory)
The file selection box still doesn't look right; as you said it doesn't look right, but it acts right.
On Aug 19, 12:17 am, David Fanning <n...@dfanning.com> wrote:
> I am looking for someone running a UNIX version of > IDL 7.1 to confirm that the problem described in > this article dealing with Dialog_Pickfile has > been fixed.
> Specifically, if you specify both a file name and > a path when you call Dialog_Pickfile, and then do > nothing else except hit the Accept button, the file name > that is returned has two paths appended to it and is > bogus.
> I am interested in what is in the Selection box, > and what is returned.
> I am, say we say, dubious. :-)
> Thanks,
> David > -- > David Fanning, Ph.D. > Coyote's Guide to IDL Programming (www.dfanning.com) > Sepore ma de ni thui. ("Perhaps thou speakest truth.")
