2009/387 [Pathname Reparse Points]

[Joerg Schilling]

"Alan.M.Wright" <amw at Sun.COM> wrote:

> That would seem to resolve the PATH_MAX question asked earlier since
> the length interpretation is implementation dependent.
>
> An interpretation change request assumes that we would want this to be
> adopted as a standard, which may be desirable, but this proposal doesn't
> rely on reparse data being used for local path resolution (at this time).
> This proposal is intended to support reparse point aware services.
> Regular applications should be unaware that the text is anything other than
> a dangling symlink.

If you like this case to be established as POSIX standard, you would need to 
write a much more in depth explanation than you did.

There are many questions that need to be answered before you could present it 
to the standard commitee. If you make a proposal that will cause existing 
applications to missbehave, you should be prepared that your proposal will be 
rejected the same way as Microsofts proposal 8 years ago was rejected when they
first tried to introduce special semantics to the ':' in file names and later 
tried to introduce a directory "..." with special meanings in order to 
implement their file streams.

Let me ask the questions that you need to answer:

-	What exactly is the purpose of the new techique?

	-	How will it be used?

	-	Where is it used?

	-	What will happen based on what data?

	-	What is the benefit of having this technique?

	...

-	How _exactly_ will it be implemented?

-	Is is based on technology that is currently part of POSIX?

	-	If the technology is not in POSIX, it needs to be 
		described in detail and first offered to POSIX

	-	If it is in POSIX already, does it need to be enhanced?
	

-	Which existing applications will fail or missbehave?

	-	Note that making existing applications to missbehave
		is a no-go....

	-	Is there a way to prevent miss-behavior?

...

There are many more questions. I just like to give you a cookbook for a better
proposal. Looking at *BSD implementations does not verify that existing 
applications are not affected. Note that e.g. FreeBSD did introduce large file 
support without implementing a default fallback for old applications that don't 
know about large files. This was done because the FreeBSD guys did modify all
related FreeBSD software. POSIX intends to take care of existing applications...

Jörg

-- 
 EMail:joerg at schily.isdn.cs.tu-berlin.de (home) Jörg Schilling D-13353 Berlin
       js at cs.tu-berlin.de                (uni)  
       joerg.schilling at fokus.fraunhofer.de (work) Blog: http://schily.blogspot.com/
 URL:  http://cdrecord.berlios.de/private/ ftp://ftp.berlios.de/pub/schily