Docu update

[Oliver Osburg]

Hi Carl,

thanks for interest! Best would be to put these things in the wiki. If
you do not have an account (I did not check) mail to:

faiwikimgr at informatik.uni-koeln.de

regards,
oliver

* Carl J. Van Arsdall <cvanarsdall at mvista.com> [070213 20:24]:
> Oliver Osburg wrote:
> >Hi,
> >
> >I plan to do the messy work to update the documentation. My first step
> >was to work through the wish- and buglist on
> >
> >http://faiwiki.informatik.uni-koeln.de/index.php/Etch-doc-todo
> >
> >Basic plan is first to fix all of the bugs in it (wrong paths in
> >descriptions, etc, etc) and then exend the guide _a little_ according to
> >the wishlist.
> >
> >Basic comments, suggestions?
> >  
> 
> I don't know if I missed this in the manual or if its just not there, 
> but I thought it would have been useful to have a "Scripting for FAI" 
> type of guide.  This guide would define various internal variables like 
> $target, the appropriate way to exit a bash script for FAI (like a 
> description of the method used in the examples) things like that.  
> Perhaps even a simple and a complex script demonstrating the 
> capabilities and the documentation to support things that are unique (or 
> necessary) to FAI.  That might go beyond "a little", I'd be willing to 
> help/edit too, but I don't know enough of what's going on to write that 
> document myself.
> 
> There's also a howto on the wiki about configuring/compiling extra 
> kernels, I think that would be a useful addition to the manual "how to 
> make a kernel" that covers necessary or recommended options for install 
> and client kernels, and adding them to your local apt repository using a 
> command like reprepro.  (again, all stuff someone could figure out, but 
> that would have helped me when i was most clueless, especially if all of 
> that information had been in a single place instead of scattered).  
> That's another one I can help with (or update the wiki page, and put a 
> link to it in the manual, i don't know how you guys like to approach docs).
> 
> A third thing on my wishlist: supporting heterogeneous computing farms 
> (having one fai server doing both i386/amd64 and the special steps to 
> take in order to set that up).  Maybe an entry in an FAQ?
> 
> And the last thing, a high-level "scripts and defaults put into fai".  
> This might be one i missed in the manual too (i feel like i read it a 
> million times, but something always eludes me and it has been a while, I 
> don't remember all of what is in there).  What i'm talking about here is 
> how by default nodes will get the class: FAIBASE.  Following that, the 
> scripts in /srv/fai/config/scripts/FAIBASE are executed and may alter 
> settings that a user cares about.    Again, I know you could just go and 
> read the scripts, but it would be cool to know how the system would work 
> right out of the box and what scripts you may want to edit before really 
> digging into each script.
> 
> Ah, sorry for so much, I guess I really like documentation.  I'm also 
> happy to help where I can.
> 
> -carl
> 
> -- 
> 
> Carl J. Van Arsdall
> cvanarsdall at mvista.com
> Build and Release
> MontaVista Software
> 
--
 *  Oliver Osburg    -     Software and System Architect  *
 *  email: osburg at osburg.org     phone: +49 179 50 76 784 *
 *  note: All above is my opinion unless said otherwise   *
 * One Unix to rule them all, One Resolver to find them,  *
 * One IP to bring them all and in the zone bind them.    *