On Fri, 2012-02-03 at 15:20 -0500, Stephen Gallagher wrote:
On Fri, 2012-02-03 at 14:32 -0500, Simo Sorce wrote:
> On Fri, 2012-02-03 at 10:31 -0500, Stephen Gallagher wrote:
> > Sorry for breaking the thread, but my email server seems to have lost
> > track of the original emails (they're not showing up in my folders).
> >
> > Jakub, thanks for catching the ioa_common.c issue. Fixed.
> >
> > I didn't bother differentiating the messages in the API because
they're
> > A) unimportant and B) not sufficiently different to be worth adding new
> > translatable strings for.
> >
> > I did fix the manpages though. Those SHOULD be documented.
> >
> > New patch attached.
> >
> > Also, this is designed to apply atop my patches for the RootDSE/search
> > base fix (which still needs reviewing). See "[PATCH] LDAP: Do not fail
> > if RootDSE check cannot determine search bases"
>
> This is a matter of taste I guess but I find it difficult to read and
> remember th anme you cage to the options.
> Why entry_cache_user_timeout and not just user_cache_timeout (and so on
> for other maps) ?
> It sounds more readable to me.
>
>
> Also I understand why you do these changes:
> - timeout = dp_opt_get_int(state->opts->basic, SDAP_ENTRY_CACHE_TIMEOUT);
> + timeout = dp_opt_get_int(state->opts->basic, SDAP_SEARCH_TIMEOUT);
>
> But I think they belong in a separate patch, as they seem to fix using
> the wrong timeout but are not directly related to the change the patch
> is about ?
>
> Everything else looks fine.
Sure, I agree. When I push them I'll split them into a separate patch.
For posterity (and to make it easier on me to keep track of what needs
pushing), here are the two new patches.
I left the option names alone, for the record. I prefer the shorter
names, and it's a purely cosmetic decision.