* Cache system
@ 2016-10-09 18:24 sylvain
2016-10-09 18:51 ` john
0 siblings, 1 reply; 3+ messages in thread
From: sylvain @ 2016-10-09 18:24 UTC (permalink / raw)
Good evening,
As far as I understand it the cgit cache system relies only on the
path of a request (the PATH_INFO var to be precise).
The cgit setup I made is not compatible with this cache system because
I use http for public repositories and http*s* for private ones on the
same machine.
Basically with my setup, paths could collide between http and https,
especially the index page which lists the repositories.
I was wondering if you could think about making the cache system rely
on the whole url instead of just the path.
Regards.
--
Sylvain Rabot <sylvain at abstraction.fr>
^ permalink raw reply [flat|nested] 3+ messages in thread
* Cache system
2016-10-09 18:24 Cache system sylvain
@ 2016-10-09 18:51 ` john
2016-10-09 19:47 ` sylvain
0 siblings, 1 reply; 3+ messages in thread
From: john @ 2016-10-09 18:51 UTC (permalink / raw)
On Sun, Oct 09, 2016 at 08:24:20PM +0200, Sylvain Rabot wrote:
> As far as I understand it the cgit cache system relies only on the
> path of a request (the PATH_INFO var to be precise).
>
> The cgit setup I made is not compatible with this cache system because
> I use http for public repositories and http*s* for private ones on the
> same machine.
>
> Basically with my setup, paths could collide between http and https,
> especially the index page which lists the repositories.
>
> I was wondering if you could think about making the cache system rely
> on the whole url instead of just the path.
I'm not sure we can reliably receive this over the CGI interface. Even
if we do, I don't think this change is a benefit for most users.
Can't you run with separate config files for HTTP and HTTPS, using the
"include" directive to keep everything except the cache directory the
same?
^ permalink raw reply [flat|nested] 3+ messages in thread
* Cache system
2016-10-09 18:51 ` john
@ 2016-10-09 19:47 ` sylvain
0 siblings, 0 replies; 3+ messages in thread
From: sylvain @ 2016-10-09 19:47 UTC (permalink / raw)
On Sun, Oct 9, 2016 at 8:51 PM, John Keeping <john at keeping.me.uk> wrote:
> On Sun, Oct 09, 2016 at 08:24:20PM +0200, Sylvain Rabot wrote:
>> As far as I understand it the cgit cache system relies only on the
>> path of a request (the PATH_INFO var to be precise).
>>
>> The cgit setup I made is not compatible with this cache system because
>> I use http for public repositories and http*s* for private ones on the
>> same machine.
>>
>> Basically with my setup, paths could collide between http and https,
>> especially the index page which lists the repositories.
>>
>> I was wondering if you could think about making the cache system rely
>> on the whole url instead of just the path.
>
> I'm not sure we can reliably receive this over the CGI interface. Even
> if we do, I don't think this change is a benefit for most users.
>
> Can't you run with separate config files for HTTP and HTTPS, using the
> "include" directive to keep everything except the cache directory the
> same?
Indeed, I did not think of that. Thank you!
--
Sylvain Rabot <sylvain at abstraction.fr>
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2016-10-09 19:47 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2016-10-09 18:24 Cache system sylvain
2016-10-09 18:51 ` john
2016-10-09 19:47 ` sylvain
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).