rfc: remove "revision specs"

[Robert Collins]

On Tue, 2005-08-30 at 13:05 +0300, Nir Soffer wrote:
> On 30 Aug, 2005, at 9:47, Lalo Martins wrote:
> 
> > Now I'm thinking, if you don't actually expect people to *type* this,
> > but rather, you support it for the sake of tools like pqm and for
> > uniform python api - then I'd go with the semicolon, precisely 
> > *because*
> > it's bad shell - therefore it's less likely to be a legitimate 
> > filename.
> >  (It's even illegal on windows, isn't it?)
> 
> Not all operating systems have illegal characters, it will not be wise 
> to design a feature around this limitation in a multi platform tool.

; is not illegal for the file system, its just illegal for most shells.

It doesn't matter whether the delimiter is a legal file character or
not: std66 provides % escaping for all control and delimiter characters.

the path /home/foo/bar,rev=1040 has a file url of:
file:///home/foo/bar%2crev%3d1040
If we also provided a rev parameter to it it would become:
file:///home/foo/bar%2crev%3d1040,rev=1054

Note that this is file system independent and unambiguous. But: its not
something that users can construct easily in the event that their file
path contains must-escape characters -  which is why its not an ideal
end user API, though its great for programmatic queries where the client
can generate the url predictably.

Rob

-- 
GPG key available at: <http://www.robertcollins.net/keys.txt>.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part
Url : https://lists.ubuntu.com/archives/bazaar/attachments/20050831/714e92aa/attachment.pgp