Mail Archives: cygwin-developers/1998/09/18/14:09:04
On Fri, Sep 18, 1998 at 02:21:47PM -0500, Parker, Ron wrote:
>> From: Christopher Faylor [mailto:cgf AT cygnus DOT com]
>> Sent: Friday, September 18, 1998 1:34 PM
>> To: Parker, Ron; Corinna Vinschen; DJ Delorie;
>> cygwin32-developers AT cygnus DOT com
>> Subject: Re: New winsup snapshot 9/14/98
>
>...snip...
>
>> Does anyone know if there is any "slop" in the directory entries for
>> VFAT that would allow us to shove some extra information in?
>
>VFAT in the driver. I assume you mean FAT|FAT32 directory entries.
No, I actually meant VFAT. VFAT adds "special" entries to the directory
to allow long filenames.
>By slop do you mean unused bits and bytes?
Yes.
>>>If a Windows 9x machine accesses a share that contains an index file,
>>>it may use and even modify the index file. Obviously if we allow index
>>>files to be modified remotely, they can not be cached locally. Either
>>>way, they should never be cached remotely because the serving machine
>>>can still the index file them.
>>
>>So, if a Windows 9x machine accesses an NTFS partition over the net, it
>>will never see the "UNIX permissions" that are associated with the
>>directory? Hmm.
>
>...snip...
>
>If someone has a better solution I'm open to it. One possibility is to
>use an index file on all platforms. Which may be the best solution, but
>I don't personally like it.
I don't either. There are too many places in cygwin where accomodations
are made for Windows 95 at the expense of Windows NT. I'd hate to add
one more.
Another thing that just occurred to me is to create a "phantom directory
structure" so that if you're running Windows 95 you could have something
like:
/usr/bin
and
/__cygwin__/usr/bin
/__cygwin__/usr/bin would be hidden and would mirror /usr/bin except
that the files associated with filenames would contain only with
permission and device information.
--
cgf AT cygnus DOT com
http://www.cygnus.com/
- Raw text -