Re: nginx woes

2015-03-31 Thread frantisek holop
Stuart Henderson, 30 Mar 2015 23:10: A no_chroot flavour does seem reasonable, do you fancy sending a diff? i think it would be logical to add patch-man_nginx_8 to the nginx_chroot* external patch... then i can remove PATCH_LIST here is a wip patch, but i am getting stuck on the upgrade path

Re: nginx woes

2015-03-31 Thread frantisek holop
Theo de Raadt, 30 Mar 2015 18:12: The most important question of the conversation was: Why did nginx not accept a diff to add chroot to mainline? All of you, feel free to ask their developers. Yeah, I know this will come off as political. But this is a significant and intentional oversight

Re: nginx woes

2015-03-30 Thread Kevin Chadwick
On Sun, 29 Mar 2015 22:49:00 +0200 frantisek holop wrote: and i have no beef with the chroot patch. it is in fact a nice addition, but it is not part of nginx itself, Why not? I think that is the real issue but perhaps upstream has refused on some odd basis? I don't use nginx but it seems

Re: nginx woes

2015-03-30 Thread Stuart Henderson
On 2015/03/29 22:49, frantisek holop wrote: and i have no beef with the chroot patch. it is in fact a nice addition, It's a little buggy (see -p /) and needs the manpage changes. Perhaps when those are fixed maybe we could see if upstream might be interested in taking the diff. but it is not

Re: nginx woes

2015-03-30 Thread frantisek holop
Stuart Henderson, 30 Mar 2015 09:58: when nginx was part of base, it was expected that it would receive all kinds of local patches. but just as openbsd apache is not apache anymore, this nginx is also not nginx anymore. It's quite a different situation, the old OpenBSD httpd was

Re: nginx woes

2015-03-30 Thread Stuart Henderson
On 2015/03/30 22:33, frantisek holop wrote: Stuart Henderson, 30 Mar 2015 09:58: when nginx was part of base, it was expected that it would receive all kinds of local patches. but just as openbsd apache is not apache anymore, this nginx is also not nginx anymore. It's quite a

Re: nginx woes

2015-03-30 Thread Theo de Raadt
A no_chroot flavour does seem reasonable, do you fancy sending a diff? I am staying out of ports matters. Changing to not chrooting by default is more problematic because existing users will loose chroot (and may not notice this, especially if updating from an OpenBSD version with nginx in

Re: nginx woes

2015-03-29 Thread Bryan Steele
On Sat, Mar 28, 2015 at 01:02:25PM +0100, frantisek holop wrote: so i have just been bitten by the implicitly included openbsd chroot patch for nginx. the issue is, that i _remembered_ there is a parameter to turn it off, but upon doing man nginx, i could not find it. then i proceeded to

Re: nginx woes

2015-03-29 Thread frantisek holop
Bryan Steele, 29 Mar 2015 11:18: sometimes i just want a vanilla nginx, just like upstream, so i am not going to be bitten by this when i deploy stuff to linux servers... actually, i think it should be the other way around, nginx-chroot being a flavor, and not vice versa -f

Re: nginx woes

2015-03-28 Thread Jason Tubnor
On 28/03/2015 11:04 pm, frantisek holop min...@obiit.org wrote: so i have just been bitten by the implicitly included openbsd chroot patch for nginx. the issue is, that i _remembered_ there is a parameter to turn it off, but upon doing man nginx, i could not find it. then i proceeded to

nginx woes

2015-03-28 Thread frantisek holop
so i have just been bitten by the implicitly included openbsd chroot patch for nginx. the issue is, that i _remembered_ there is a parameter to turn it off, but upon doing man nginx, i could not find it. then i proceeded to read the pkg-readme as well, and not a word about it. yes, it is in