[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [OctDev] Test web-site of new octave-forge web-site including all pa
From: |
John W. Eaton |
Subject: |
Re: [OctDev] Test web-site of new octave-forge web-site including all packages |
Date: |
Tue, 10 Oct 2006 22:05:32 -0400 |
On 10-Oct-2006, Bill Denney wrote:
| John W. Eaton wrote:
| > On 10-Oct-2006, John Swensen wrote:
| >
| > | If you are not averse to this idea, I would be willing to work on it
| > | (probably using wget or curl, I think most machines have one of these
two).
| > |
| > | Suggested commands:
| > | pkg netlist
| > | pkg netinstall
| > | pkg netupgrade
| >
| > Octave now has urlwrite, which is based on curl, so you could do the
| > download step directly in Octave.
| >
| > See also this thread:
| >
| > https://www.cae.wisc.edu/pipermail/bug-octave/2006-October/000877.html
| >
| A related functionality that I had been thinking about recently is that
| this modification to fileparts could be helpful. I tried to look up
| what the "version" output from fileparts was supposed to be so that it
| could be compatible, but mathworks' own documentation only returns
| empty. This one will see if the end of the name looks like a version
| (i.e. anything that's a digit or a decimal point) and return that as the
| version.
I suspect the original reason for attempting to handle the version
part of a file name was for VMS-style file versions:
foo.bar;1
foo.bar;2
etc. As I recall, these were handled by the filesystem and were not
part of the filename itself.
I hesitate to add anything to fileparts that attempts to parse the
filename and decide what might be a version number. Doing that could
break code or at least produce surprising results for people who are
not expecting it.
jwe