[PATCH] walkrepo: recurse only if ** in [paths] or [collections]

Matt Mackall mpm at selenic.com
Tue Dec 9 14:04:57 CST 2008


On Tue, 2008-12-09 at 11:31 -0800, Brendan Cully wrote:
> On Tuesday, 09 December 2008 at 13:28, Matt Mackall wrote:
> > On Tue, 2008-12-09 at 09:44 +0100, Ben wrote:
> > > [Adding CC to the list as it seems to have been forgotten]
> > > 
> > > On Mon, Dec 8, 2008 at 23:42, Dirkjan Ochtman <dirkjan at ochtman.nl> wrote:
> > > > Thanks for working on this! I have some comments below.
> > > >
> > > > On 08/12/2008 22:21, Benoit Allard wrote:
> > > >>
> > > >> So the behavior is as follow:
> > > >> collections: recursive discovery
> > > >
> > > > I think this should be non-recursive, as before.
> > > 
> > > We're having a political issue there as the release note for 1.1 is
> > > mentioning recursive collections. As for now, path = '*' or '**' are
> > > behaving recursively, the former could be considered as s bug and
> > > corrected into flat discovery. But I'm afraid we'll have to keep the
> > > CompatibilityRules on collections.
> > 
> > CompatibilityRules doesn't apply to things that themselves break
> > CompatibilityRules. hgweb is now taking 5 to 30 seconds just to start
> > for some users when it used to be fast. In other words, hgweb is broken.
> > 
> > We need to return to 1.0 behavior *as soon as possible*. We can add **
> > for recursion after.
> 
> Should we recur by default unless there's a dirstate?

Is it possible it will take >1s in some setups? Yes. So no.

Remember, this is overhead on every single web requests in most setups.

-- 
Mathematics is the supreme nostalgia of our time.



More information about the Mercurial-devel mailing list