[[!meta copyright="Copyright © 2013 Free Software Foundation, Inc."]]
[[!meta license="""[[!toggle id="license" text="GFDL 1.2+"]][[!toggleable
id="license" text="Permission is granted to copy, distribute and/or modify this
document under the terms of the GNU Free Documentation License, Version 1.2 or
any later version published by the Free Software Foundation; with no Invariant
Sections, no Front-Cover Texts, and no Back-Cover Texts. A copy of the license
is included in the section entitled [[GNU Free Documentation
License|/fdl]]."]]"""]]
/!\ Incomplete, but hopefully helpful for future reference.
# Packages
* apache2-mpm-prefork (used to be apache2-mpm-worker but changed because of some threading issues with CGI)
* ikiwiki libcgi-session-perl libtext-csv-perl libcgi-formbuilder-perl
libauthen-passphrase-perl libnet-openid-consumer-perl
liblwpx-paranoidagent-perl libterm-readline-gnu-perl libgravatar-url-perl
librpc-xml-perl libtext-wikiformat-perl libhighlight-perl perlmagick
graphviz texinfo
* libemail-send-perl (for my *sendmail vs. ikiwiki* patch)
* libsearch-xapian-perl xapian-omega (for ikiwiki's search plugin)
* libyaml-perl libyaml-syck-perl (for ikiwiki's YAML field plugins)
* gitweb highlight
sudo ln -s ~hurd-web/hurd-web.git /var/lib/git/
* git-daemon-sysvinit
Enable as per `/usr/share/doc/git-daemon-sysvinit/README.Debian`. Also set
`GIT_DAEMON_OPTIONS=--export-all`, and `GIT_DAEMON_DIRECTORY='/var/lib/git'`.
# `~hurd-web/`
$ mkdir hurd-web.git && GIT_DIR=hurd-web.git git init
# `~tschwinge/`
/!\ TODO.
$ mkdir tmp/backup && chmod 0733 tmp/backup
# `/var/www/robots.txt`
/!\ TODO.
This file used to contain:
User-agent: *
Disallow: /gitweb/
Disallow: /cgi-bin/
... which I've now changed to:
User-agent: *
Disallow: /
The goal is that robots rather index the official pages,
, instead of the staging area on
.
# Restore Backup
/!\ TODO.
## `/etc/apache2/mods-enabled/`
`rewrite.load`, `userdir.conf`, `userdir.load`
# IRC, freenode, #hurd, 2013-02-09
We need an httpd (Apache used to work), and ikiwiki and some
such stuff.
This has its own git repository.
This was on a separate virtual machine.
Then there was the Git repository on flubber used for people to
push to.
Ho -- let me actually try to remember the setup. Has been some
years...
what machine currently hosts the wiki ?
Anyway, there is no requirement for the web server to be on a
separate machine; your decision.
braunr: http://www.gnu.org/software/hurd/public_hurd_boxen.html
snubber
That was the web server.
isn't it gnu.org ?
And flubber had the repository for developers to push to.
No, gnu.org is updated manually (by me).
ah
It'S a snapshot of the wiki so to say.
ok so, is this wiki really meant to be modifiable from a browser ?
http://www.gnu.org/software/hurd/contributing/web_pages.html#index5h2
Yes.
i see
I should still be able to access the data from Barry's zenhost
(including all the VMs it hosted), so I should be able to replicate that
quite easily.
do you think it could be hosted on darnassus, or would you like a
separate vm ?
The repository for people to push to and pull from (used to be
on flubber) would be on darnassus now.
About the web server, hmm.
It's basically a security concern.
And it might get hammered by bots from time to time.
it won't need much resources i suppose
No. The web server (snubber) was running with 242 MiB of RAM,
and had uptimes of several weeks typically.
tschwinge: otherwise, could it use the web server running on the
host ?
The host being darnassus?
no
shattrath, the linux system
Ah.
Sure.
There is no requirement this to be a Hurd system -- was just
nice to show to people.
i think it is too
what's the problem with darnassus ?
yçou mentioned security
The web server being a public-facing component which might be
broken into.
how ?
it's so much easier to just ask for an account .. :)
Web server bugs, CGI script bugs, etc.
Sure.
I just wanted to make you aware of it. :-)
oh don't worry
ok so darnassus it is
was it running apache? maybe, if other (lighter?) web servers
are tested to work on hurd, they could be used
pinotree: Yes, and yes.
doesn't ikiwiki need php ?
Only requirement (I think) is abaility to run CGI scripts.
braunr: No. It's written in perl.
ok
i still think i'll use apache
it's really not that heavy
lighter servers matter when the number of concurrent clients get
very high
Then I'll figure out how exactly the setup was between flubber
and snubber.
ok
it's good to finally get that going :)
braunr: Of course ;-) -- I had some parts of the process
documented:
http://www.gnu.org/software/hurd/public_hurd_boxen/installation/snubber.html
If both Git repositories are to be on the same machine
(darnassus) we might not actually need inetutils-inetd and netcat.
Still trying to figure out what I had done there... ;-)
OK, I again understand the setup. Last been touched in the
2008/2009 timeframe. ;-)
:)
braunr: Please use the following ikiwiki packages: dpkg -i
~tschwinge/tmp/ikiwiki_3.20110608_all.deb
what makes this package special ?
Some patch that I added to get rendering of our news pages
correct.
ok
I have not updated it ever since (and the patch was not yet in
a suitable form for upstream).
Nothing major.
tschwinge: why is the ikiwiki package status hi ?
braunr: I set it to hold.
ah ok
so you finished your pat i suppose
i'll install apache
part*
I'll add a hurd-web user.
So... I actually have to locate a backup of the files from
flubber related to the wiki,
* tschwinge goes searching his backup devices.
i added userdirs on darnassus' apache
braunr: I just noticed when I wanted to add it myself. ;-)
braunr: Do you know about CGI scripts?
yes
braunr: snubber had these in /var/www/cgi-bin/; darnassus now
in /usr/lib/cgi-bin/.
ikiwiki needs to install one CGI script.
ok
Does this go into /usr/lib/cgi-bin/ then? Or into ~hurd-web/
and a symlink somewhere?
ikiwiki should have installed it where it's appropriate
normally in /usr/lib/cgi-bin/
It's a CGI script that is generated per ikiwiki instance, so
specific to hurd-web.
where does it install it by default ?
$PWD ;-)
ah
it seems a bit silly to me to generate cgi scripts :/
i don't care much actually, we won't have virtual servers
so anywhere is fine
What does the +SymLinksIfOwnerMatch Apache option mean?
apache will normally not follow symlink
unless the owner of the symlink is the same as the target's
(with this option)
That's enabled for CGI scripts. So would it work to have a
symlink /usr/lib/cgi-bin/hurd-web.cgi -> ~hurd-web/hurd-web.cgi?
the traditional way to access cgi scripts is to explicitely refer
to them as http://server/cgi-bin/script
using *.cgi may allow too open access to cgis
(although normally, the userdir conf should disable them)
hm not sure it does
so put it in /usr/lib/cgi-bin/
So the hurd-web ikiwiki instance just needs to be configured
accordingly with the URL where the CGI script will be found, and then it
will render the pages accordingly.
OK.
and just named hurd-web
## IRC, freenode, #hurd, 2013-02-10
http://darnassus.sceen.net/~hurd-web/
Have at it!
braunr: ^
:)
great
And push to/pull from darnassus:~hurd-web/hurd-web.git for Git
access.
Will update the web pages tomorrow, and all that.
braunr: And also install gitweb on darnassus, so one can view
diffs of the wiki pages, etc. OK?
tschwinge: there are still links towards bddebian
history for example
just fyi, we can look at this tomorrow
braunr: Yes, that'S what I need gitweb for.
braunr: gitweb installed, hurd-web URLs fixed
(s%bddebian%darnassus), also some more ikiwiki-related Perl pacakges
installed (openID login, for example).