4:43 p.m.
At 08:15 AM 9/10/2001 -0700, Karr, David wrote:
Could you clarify exactly what this means? Do you mean that
"find-file"
will understand "c:/dir/file" (or even "c:\dir\file") and do
completion on
it, and still allow the same behavior for Unix-like paths? If so, this will
be very nice.
Correct on all counts.
andy
9:05 a.m.
New subject: [21.4] Enable win32 filenames under cygwin
Andy Piper writes:
At 08:15 AM 9/10/2001 -0700, Karr, David wrote:
>Could you clarify exactly what this means? Do you mean that
>"find-file" will understand "c:/dir/file" (or even
"c:\dir\file")
>and do completion on it, and still allow the same behavior for
>Unix-like paths? If so, this will be very nice.
Correct on all counts.
And it is nice. Now I can ditch the chain of 57 scripts it takes to
patch gnuclient into explorer. Some questions...
Will the DDE integration use gnuserv/gnuclient?
Is it possible to set up DDE without running netinstall?
Are you planning to canonicalise the file or buffer names so that
c:\dir\file and /cygdrive/c/dir/file are treated the same?
- Phil
5:01 p.m.
New subject: [21.4] Enable win32 filenames under cygwin
At 09:05 AM 9/11/01 +0100, Philip Aston wrote:
And it is nice. Now I can ditch the chain of 57 scripts it takes to
patch gnuclient into explorer. Some questions...
Will the DDE integration use gnuserv/gnuclient?
No. It connects to XEmacs directly and enqueues the filename into the event
loop. gnuserv/gnuclient is not necessary.
Is it possible to set up DDE without running netinstall?
Yes. Its just an explorer thing. Various people have posted scripts to do this.
Are you planning to canonicalise the file or buffer names so that
c:\dir\file and /cygdrive/c/dir/file are treated the same?
I've thought about this. Yes we should probably do this but it would be
interesting to have some others thoughts on this. Incidentally the dde
support canonicalizes filenames already, its only filenames on the command
line that aren't.
andy
6:40 p.m.
New subject: [21.4] Enable win32 filenames under cygwin
Andy Piper writes:
At 09:05 AM 9/11/01 +0100, Philip Aston wrote:
>And it is nice. Now I can ditch the chain of 57 scripts it takes to
>patch gnuclient into explorer. Some questions...
>
>Will the DDE integration use gnuserv/gnuclient?
No. It connects to XEmacs directly and enqueues the filename into the event
loop. gnuserv/gnuclient is not necessary.
>Is it possible to set up DDE without running netinstall?
Yes. Its just an explorer thing. Various people have posted scripts to do this.
>Are you planning to canonicalise the file or buffer names so that
>c:\dir\file and /cygdrive/c/dir/file are treated the same?
I've thought about this. Yes we should probably do this but it would be
interesting to have some others thoughts on this. Incidentally the dde
support canonicalizes filenames already, its only filenames on the command
line that aren't.
Common Lisp has just the facility to deal with this. It splits the
filename up into different components and creates a pathname object.
All lisp functions that deal with files use these pathname objects
and, therefore, don't really care about what the printed
representation of the pathname is. For more information on CL
pathnames see
http://www.xanalys.com/software_tools/reference/HyperSpec/Body/chap-19.html
The full Common Lisp HyperSpec is
http://www.xanalys.com/software_tools/reference/HyperSpec/
Maybe XEmacs should have its own pathname object.
Craig
10:55 p.m.
New subject: [21.4] Enable win32 filenames under cygwin
Andy Piper wrote:
>Are you planning to canonicalise the file or buffer names so
that
>c:\dir\file and /cygdrive/c/dir/file are treated the same?
I've thought about this. Yes we should probably do this but it would be
interesting to have some others thoughts on this.
My main thought is that Windows has so many different aliasing
mechanisms that the main issue is which ones you're going to allow for
and which ones you're going to ignore; you'll never get them all.
If you want a list, search the BugTraq (and similar) archives for
vulnerabilities involving the ability to circumvent filename/URL
access controls by using aliases. Off the top of my head, you have:
Case folding ("foo" == "FOO", usually but not always)
8.3 filenames ("C:\Program Files" == "C:\PROGRA~1")
Multiple dots ("..." == "..\..", sometimes)
Device files ("<anything>\LPT1" == "<anything
else>\LPT1")
Fault-tolerant Unicode (trailing invalid characters are discarded)
Host aliases in UNC paths ("\\localhost\foo\bar" ==
"\\127.0.0.1\foo\bar" == "\\0x7F000001\foo\bar" == ...)
I'm fairly sure that this isn't all of them.
Some of these aren't relevant outside of a security context (e.g.
device files), but others may be.
--
Glynn Clements <glynn.clements(a)virgin.net>
12:04 a.m.
New subject: [21.4] Enable win32 filenames under cygwin
At 10:55 PM 9/11/01 +0100, Glynn Clements wrote:
My main thought is that Windows has so many different aliasing
mechanisms that the main issue is which ones you're going to allow for
and which ones you're going to ignore; you'll never get them all.
True, but cygwin has the same problems - its still windows. If we always
canonicalize the filename at the end to something cygwin gives us then its
not our problem anymore. On the other hand, if you type c:/ maybe you want
to still see that rather than some weird mount point.
If you want a list, search the BugTraq (and similar) archives for
vulnerabilities involving the ability to circumvent filename/URL
access controls by using aliases. Off the top of my head, you have:
Case folding ("foo" == "FOO", usually but not always)
Actually I've not enabled case folding.
8.3 filenames ("C:\Program Files" ==
"C:\PROGRA~1")
So if there is a drive we do whatever the native version does here.
Multiple dots ("..." == "..\..", sometimes)
Again, if there is a drive we do whatever the native version does,
otherwise cygwin.
Device files ("<anything>\LPT1" ==
"<anything else>\LPT1")
Fault-tolerant Unicode (trailing invalid characters are discarded)
Host aliases in UNC paths ("\\localhost\foo\bar" ==
"\\127.0.0.1\foo\bar" == "\\0x7F000001\foo\bar" == ...)
I'm fairly sure that this isn't all of them.
RIght, but I'm fairly sure that the current native version doesn't support
these either.
Some of these aren't relevant outside of a security context (e.g.
device files), but others may be.
I suppose I'm not understanding what your point is. Are you saying
canonicalization is doomed or native filenames in general are doomed or
that the native version of XEmacs is doomed?
andy
12:42 a.m.
New subject: [21.4] Enable win32 filenames under cygwin
Andy Piper wrote:
I suppose I'm not understanding what your point is. Are you
saying
canonicalization is doomed or native filenames in general are doomed or
that the native version of XEmacs is doomed?
The closest of those options is "canonicalization is doomed", but
"doomed" is probably too strong a word (unless you really need to get
it right, as is the case in a security context).
--
Glynn Clements <glynn.clements(a)virgin.net>
5:50 p.m.
New subject: [21.4] Enable win32 filenames under cygwin
Andy Piper writes:
At 09:05 AM 9/11/01 +0100, Philip Aston wrote:
>And it is nice. Now I can ditch the chain of 57 scripts it takes to
>patch gnuclient into explorer. Some questions...
>
>Will the DDE integration use gnuserv/gnuclient?
No. It connects to XEmacs directly and enqueues the filename into the event
loop. gnuserv/gnuclient is not necessary.
Cool. Anyone know whether there's anything I can use from the command
line รก la gnuclient?
>Is it possible to set up DDE without running netinstall?
Yes. Its just an explorer thing. Various people have posted scripts
to do this.
Found them. Time to turn my xemacs-nt subscription back on.
Thanks,
- Phil
8472
days inactive
8474
days old
7 comments
4 participants
participants (4)
-
Andy Piper -
Craig Lanning -
Glynn Clements -
Philip Aston