Re: [fix] devel/gitea, bash as rundep + socket path creation

2023-01-09 Thread Pavel Korovin
On 01/09, Stefan Hagen wrote: > The problematic part is, that "[repository]" is still commented. This > means SCRIPT_TYPE is in "[service]" (previous open section) and has no effect. > > So what we actually need is not bash, but something like the sed in > do-install below. Stefan, thank you! I

Re: [fix] devel/gitea, bash as rundep + socket path creation

2023-01-09 Thread Stefan Hagen
Pavel Korovin wrote (2023-01-09 12:53 WET): > On 01/09, Stefan Hagen wrote: > > /var/gitea# fgrep -r 'env bash' . > > ./data/gitea-repositories/sdk/dotfiles.git/hooks/pre-receive.d/gitea:#!/usr/bin/env > > bash > > ./data/gitea-repositories/sdk/dotfiles.git/hooks/pre-receive:#!/usr/bin/env > > ba

Re: [fix] devel/gitea, bash as rundep + socket path creation

2023-01-09 Thread Pavel Korovin
On 01/09, Stefan Hagen wrote: > /var/gitea# fgrep -r 'env bash' . > ./data/gitea-repositories/sdk/dotfiles.git/hooks/pre-receive.d/gitea:#!/usr/bin/env > bash > ./data/gitea-repositories/sdk/dotfiles.git/hooks/pre-receive:#!/usr/bin/env > bash > ./data/gitea-repositories/sdk/dotfiles.git/hooks/up

Re: [fix] devel/gitea, bash as rundep + socket path creation

2023-01-08 Thread Stefan Hagen
Pavel Korovin wrote (2023-01-08 23:00 CET): > On 01/08, Stefan Hagen wrote: > > I took me a minute to figure out that I need SUBST_VARS to make VARBASE > > work in PLIST, but now it works as intended. > > > > I committed the bash fix already. So this is only for the socket dir. > > > > OK? > >

Re: [fix] devel/gitea, bash as rundep + socket path creation

2023-01-08 Thread Pavel Korovin
On 01/08, Stefan Hagen wrote: > I took me a minute to figure out that I need SUBST_VARS to make VARBASE > work in PLIST, but now it works as intended. > > I committed the bash fix already. So this is only for the socket dir. > > OK? Hello Stefan, I don't see a good reason for creating socket di

Re: [fix] devel/gitea, bash as rundep + socket path creation

2023-01-08 Thread Stefan Hagen
Stuart Henderson wrote (2023-01-07 17:24 CET): > On 2023/01/07 16:53, Stefan Hagen wrote: > > But actually, I don't see a reason for this directory at all. > > The rc script starts gitea with uid _gitea, so it doesn't have > access to write to /var/www/run with default perms. > I took me a minut

Re: [fix] devel/gitea, bash as rundep + socket path creation

2023-01-07 Thread Matthieu Herrb
A propos gitea, anyone intested to switch to forgejo (gitea fork because of unfriendly takeover of gitea) ? https://forgejo.org/ https://forgejo.org/faq/ -- Matthieu Herrb

Re: [fix] devel/gitea, bash as rundep + socket path creation

2023-01-07 Thread Stuart Henderson
On 2023/01/07 16:53, Stefan Hagen wrote: > But actually, I don't see a reason for this directory at all. The rc script starts gitea with uid _gitea, so it doesn't have access to write to /var/www/run with default perms.

Re: [fix] devel/gitea, bash as rundep + socket path creation

2023-01-07 Thread Stefan Hagen
Klemens Nanni wrote (2023-01-07 16:04 CET): > 07.01.2023 14:48, Stefan Hagen пишет: > > Hi, > > > > I've set up a Gitea Server and noticed two inaccuracies in the port. > > > > 1. gitea sets up a post-receive hook which is used internally and > >requires a bash. Without it, it errors out on

Re: [fix] devel/gitea, bash as rundep + socket path creation

2023-01-07 Thread Klemens Nanni
07.01.2023 14:48, Stefan Hagen пишет: > Hi, > > I've set up a Gitea Server and noticed two inaccuracies in the port. > > 1. gitea sets up a post-receive hook which is used internally and >requires a bash. Without it, it errors out on the first repository >push. OK kn > > 2. After set

[fix] devel/gitea, bash as rundep + socket path creation

2023-01-07 Thread Stefan Hagen
Hi, I've set up a Gitea Server and noticed two inaccuracies in the port. 1. gitea sets up a post-receive hook which is used internally and requires a bash. Without it, it errors out on the first repository push. 2. After setting up the socket path as described in the README, gitea do