Richard Adams wrote:
> Hi Kenneth Stephen. et all.
>
> Here are a few comments about the newbie faq, i hope i am not being to
> picky, but i belive that such a document should give the correct
> information.
>
Absolutely. I was unsure about some of the answers provided, and so I left them
alone. It is good to get feedback so that I can improve the FAQ.
> You do not even need to reboot, if you do the System will automaticly send a
> message to all connecties/users that "The System is going down". Even using
> 'init 1' will still send other users the same message.
>
> No dont reboot, just use 'init 1' which has the same effect but does not
> reboot the machine.
> ]
>
'init 1' is certainly a better and more succinct solution than the one
mentioned in the FAQ.
> [
> Not exactly true, mounts can be done by users, its up to the sysadmin to
> edit /etc/fstab to allow users to use mount on certain devices.
>
> Futher more it would be a good idea to re word the whole paragraph, as you
> say at the beginning, "but when I do an 'ls' in the /cdrom directory," this
> will prompt some unknowing fellow to go into the /cdrom dir, and issue the
> mount command, which will lead to failure, you cant mount a device when you
> are in the directory it's to be mounted in, it will result in something like;
>
> mount: /dev/cdrom already mounted or /cdrom busy
> ]
>
Yup. Rewording is definitely indicated.
> 1.6
> If you are still at the initial login prompt, <Shift + PgUp> and
> <Shift+PgDn>
> should allow you to scroll through the bootup messages. If you have already
> logged in, you can try using the 'dmesg' command.
>
> [
> There are a lot of Distributions which clear the screen just before the
> login prompt, thus renering the <Shift+PgUp> idea unusable.
> ]
>
Thanks. I didnt know that.
> 1.7
> do a 'ps aux' to and look through the output to see which line corresponds
> to the process you are trying to kill. The second column on the relevant
> line is the process id. Now do a 'kill -9 <process id>'.
>
> [
> do 'ps ax | grep command' Where command is the name typed which caused the
> problem. (Saves having to sift thro' all those prosecces).
> Doing a kill -9 PID could possably leave child processes running and cause
> more havoc, use kill -9 ONLY as a last ditch effort.
> ]
>
This point was already mentioned by in a previous post. I will be fixing it.
> 3.7
> [
> Contradictory, in 1.4 you state it can only be done by the superuser.
> ]
>
You got me!
> 4.4
> [
> Your explanation that there are 6 ttys is not correct, you can define more
> or indeed less ttys in /etc/inittab;
> 7:2345:respawn:/sbin/mingetty tty7
> 8:2345:respawn:/sbin/mingetty tty8
>
> OR just comment out what you dont need/want.
> ]
>
Technically you are correct. But I felt that this is one detail that did not
need to be explained in the answer to this question. I was going by my RedHat
and Debian installations which both seem to have 6 tty's defined by default.
> 5.4
> 5.4 Sendmail sometimes stalls for a long time. Why?
> --- To Do ----- Please send me the answer. - It should have something
> -- to do with no fqdn in /etc/hosts -- Will the person who told me -- about
> also give me an answer please?
>
> [
> Sendmail stalls because it cant map hostnames to ipaddresses via a DNS
> server. The need for a defined dnsserver in /etc/resolv.conf and a valid
> route for dns lookups. Same goes for the SMB server.
> ]
THANK YOU!!
>
>
> 5.7 What does "kernel symbols not found!..." message at boot up mean?
> That you have no System.map
> You get that file when you compile the kernel, and you must put it as
> System.map or System.map-kernel version in either '/' or '/boot'.
>
> [
> No you are wrong here, must be a typo, as you explane about System.map in
> section 8.17.
>
> The System.map is for debugging kernel funtions and programs, it contains the
> addresses of kernel funtions and is used in conjuntion with GDB, the GNU
> debugger and also decijfering a Kernel Oops when running it tho' a small
> script called ksymoops.
>
> What you possably mean is /boot/map which is normaly defined in
> /etc/lilo.conf That contains information for booting.
>
> Or do you mean the Error messages about Unresolved Symbols caused by stale
> modules being found.??
>
> ]
I had the feeling that this answer was incorrect. I need to do some serious
reading of Lilo documentation.
> There is no mention of "modules" anywhere or am i missing that one.
> I cannot express the importance of having some sort of text about modules, as
> when a system is installed the accompanying kernel is modular, which has just
> about all drivers and filesystems anyone could need, just think about how
> many mails are posted to linux-newbie about compiling a new kernel, when
> realy 95% of the time its absolutly unnessasacry.
>
I shall try to remedy this by the time the next edition is posted.
Kenneth
--
There is no such thing as luck. 'Luck' is nothing but an absence of bad luck.
