[Gnu-arch-users] Re: Case insensitive file systems [was: tla1.2 on cygwin]

[Dustin Sallings]

On Mar 10, 2004, at 9:48, Parker, Ron wrote:

> Secondarily, is it possible to provide a way out of this morass once it
> exists?  I don't know from Dustin's description if his archive has  
> actually
> been corrupted or not.  If it has I realize there may be nothing that  
> can be
> done, but that takes me back to the "First, do no harm" statement.  I  
> would
> hope since tla tries to only write new files to an archive without ever
> replacing existing files that this might not be the case and  
> correcting it
> might be doable.

	The branch is corrupt starting from the first commit I did on the  
system with a case insensitive filesystem where my category name was  
different in case from the ``winning'' category name.  The different  
case was used in the {arch} tree (one of ViewARCH or ViewArch) whereas  
my category was viewarch.  When trying to check the tree out with a  
case sensitive filesystem (NetBSD's ffs), you get the following error:

* patching for revision:  
address@hidden/viewarch--py22--0.0.7--patch-1
Error calling `vu_lstat' for  
"/tmp/,,get.vap22.1078944931.7935.1/,,next-patch.1078944934.7935.40/ 
new-files-archive/./{arch}/viewarch/viewarch--py22/viewarch--py22- 
-0.0.7/address@hidden/patch-log/patch-1" (No such file or  
directory)

	The immutable nature of patches means that nothing was broken up to  
that patch, though.

> Thirdly, and I really don't know enough to even guess at an answer.   
> Is it
> possible to consolidate mismatched categories, branches, etc. on (or  
> from)
> case-insensitive systems?

	This was what I was getting at.  The categories and branches belong to  
archives, in my mind.  Representing them that way in {arch} prevents  
the possibility of them colliding.  I think it's the best solution for  
cleanliness and consistency, but I don't claim that it's the only way  
to solve this problem.

--
Dustin Sallings