I hosed up with vi

Walter nospam at please.not.me
Tue Jan 6 19:39:24 PST 2004

Matthew Dillon wrote:

    You should be able to boot with the CD.  If the system
    is not booting the CD when you reboot get into the BIOS
    and change the boot try ordering so it tries the CD
    before the HD.
    Once you are able to boot from the CD again you should
    be able to fix up the hard drive issues.  There are a lot
    of things that can result in a disk-full problem.  When
    you say you only have 80MB of RAM do you mean that you
    only have 80MB of RAM or an 80MB hard disk?  80MB of ram
    should not be a problem, but an 80MB hard disk would be a
    We should probably include a sample disklabel on the CDRom.

    Disklabels will accept sizes in megabytes, and '*' for
    auto-calculation.  So, for example, you could do this:
8 partitions:
#        size   offset    fstype   [fsize bsize bps/cpg]
  a:   256m         0    4.2BSD 
  b:  1024m   *      swap
  c: (keep this number intact) 0 unused
  d: 256m *    4.2BSD
  e:  256m  *    4.2BSD
  f:  8192m  *    4.2BSD
  g:  *  *    4.2BSD

    disklabel might get confused with so many wildcards.  If
    so, leave off the 'g' partition, save it, then run disklabel
    again to set up the 'g' partition.
    In this example I am assuming at least a 10G hard drive.

Hi Matt,

   Sorry if I did not explain properly.  I had booted from CD
and was starting the procedure (commands given in the README
file) to initialize, prepare and copy DfBSD to the HD.  I hosed
up with vi at the "disklabel -e" part.  I do have 80 MB of
RAM and a 3.2 GB HD.  The README file seems to expect 4GB for
the /usr slice, but I am hoping to get by with less because
I'm just playing with it right now; I have FBSD on a 2.something
GB HD on another computer.  I was able to reboot from the CD
after cycling power.
   I probably typed in some long numbers in vi without getting
into insert mode first.  Maybe it did a repeat of subsequent
keys strokes.
   I'll try using the asterisks next time (but not tonight).
Maybe someone will be nice and write a small program to help
the user fill in that data in some future release (as in FBSD)?
   But don't you think that ctrl-c should eventually kill a
program regardless of errors?  Likewise ctrl-alt-del should
eventually reboot and not endlessly spit out an error message?
My 2c.


More information about the Bugs mailing list