[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Questions (Debian Install)

On Fri, May 23 1997 8:57 PDT David Cary writes:
> I started all over Yet Again, and I think I discovered a bug in my "latex"
> distribution that crashed the default setup (but I have documented a way to
> work around it).
Once (a long time ago with Debian 1.1) I had the problem with 
/usr/sbin/mfmake{cmmf,mf}base and with /usr/sbin/texmake*fmt that
they redirected _both_ stdin and stdout of mf resp. tex from/to /dev/null.
The problem was, that that LaTeX (I think) complained about a too old
format file (more than a year old; my latex is now stone old:
Document Class: article 1996/05/26 v1.3r Standard LaTeX document class)
and wanted the user to confirm the installation -- net result: the postinst
doesn't work.
Solution: run the mfmake* and/or texmake*fmt scripts manually or remove
the </dev/null/>/dev/null part.
(Ask again, if you don't know what I mean).

<lot of snipping for brevity>
> They look OK;
> install.html mentions
> [[
> Extended vs. Expanded Memory
> If your system provides both extended and expanded memory, set it so
> that there is as much extended and as little expanded memory as
> possible. Linux requires extended memory and can not use expanded
> memory.
> ]]
> which doesn't make sense to me -- I thought that these options were set up
> via "config.sys" in DOS, *not* the BIOS. (Since Linux doesn't have
> "config.sys", I hope it Does the Right Thing).

Not exactly. Some BIOS allow you to remap some part of the RAM to some
or other places (typically on 4MB machines you can map the 384KB into the
UMA(sp? it's a long time ago, since I knew this really, fortunately :) 
. This is done on the motherboard.
(And then in the beginnings of the PCs there were EMS expansion cards,

> Whoopsies. I get a "APM BIOS ... Unable to handle kernel NULL pointer
> dereference at virtual address c00004e0" message (with lots more
> gibberish).
> ]]]
APM is unfortunately non standardized. If you have your kernel up and running
you may eventually find patches...

> When I gave Linux a 100 MB partition, and installed the default selections
> in the Debian distribution; when I tried that earlier I got
> ......
>   Setting up latex (2e-7) ...
>   Building new latex format(s) using install-fmt-base(8)
>   Rebuilding 'latex' format ...
> seems to take a *long* time.
Seems to be the bug above.
> It hung up for over 30 minutes (!), so I hit control-C which killed it,
> returning me to the 'dselect' menu.
> I go to select, and try to turn off latex ...
> 'dselect' unexpectedly dies, giving a "No space left on device" error and
> leaving me at the # prompt.
Probably /tmp full. (failed postinst output).

> I set the "Bootable" flag on the Linux partition for no good reason.
That's the right thing to do, BTW. Then you can activate the other ``OS''
if you want to do something dangerous which could trash your boot partition.

> "Next: Configure the Base System"
> I take the default "U.S.Keyboard".
> I choose the "US", "Pacific" timezone.
> I am tempted to choose GMT ... but I also want to run other OSes, so I say
>   Is your system clock set to GMT ... ? n
Which means that automatic the daylight saving time switch won't work.

> "Setting up smail".
> I dunno -- I was planning on using my modem to connect to my ISP via PPP,
> and that doesn't seem to be any of the options. I guess I should choose
> "(5) No configuration.... your mail system will be broken and should not be
> used. ... You must... later ... run /usr/sbin/smailconfig ...".
Configure as leaf node.

> I logoff with "exit".
> What is the proper way to shut down ? I've been logging out all active
> shells, then hitting ctrl-alt-delete.

That's perfect, init takes care of powering down safely
(wait until you see a ``System is halted'' or ``System is rebooting''
 message before turning the power off)

> The total at the bottom is "76716" (KBytes, I assume).

> There's one little bit of weirdness; one line says
> "du: ./proc/129/fd/4: No such file or directory".
/proc is a pseudo filesystem. It doesn't contain real files, `just'
the current state of the processes. Ignore the message.

> Hm; I thought there was some way I could ask Linux where on my hard drive
> the minicom program was; something like
>   ls -R minic*
> but that doesn't work ....
> What is the *nix way to "Find file with name:___" ? I know how to do this
> with the Mac OS and several different Microsoft OSes, it's kinda annoying
> me that I can't do this with Linux.
There are several ways to do this:
1. which minicom
2. locate minicom	(if you have a locate database built, e.g find installed)
3. find / -name minicom -print

Hope my answers are of some help,
David Frey                      |Linux --- the choice of a GNU generation!
51F35923114FC8647D05FF173C61EFDE|GE C++ UL+++ P- W-- !w--- PGP++ t@++ R D--- e++

TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to
debian-devel-request@lists.debian.org .
Trouble?  e-mail to templin@bucknell.edu .

Reply to: