problems with today's rawhide

[Jim Cornette]

Alexandre Oliva wrote:

>In preparation to test1, I've just given today's rawhide a try.  Here
>are some problems I've run into:
>
>rescuecd boot screen still says Fedora Core 1
>
>vg* tools in rescuecd are now useless, should ideally be mapped to lvm
>(2) cmds
>
>pump fails to obtain dhcp configuration for rescue or install.
>offering network info by hand enables it to mount the install image
>over NFS.  network card is rtl8139.  dhcpd on the server says:
>5 bad udp checksums in 5 packets
>fallback_discard: Resource temporarily unavailable
>
>booting with `linux ks', it attempts to determine network info twice:
>before asking any questions, and after asking for install location
>this probably only happens when the initial DHCP query fails
>
>lots of `numerical sysctl 1 {23,49} is obsolete' in vt4
>
>while installing in text mode (and also in rescue mode), when it looks
>for existing installations, it prints `/dev/hdc: open failed:
>Read-only file system' and garbles the screen a little bit.  In disk
>druid, it prints it several times.  /dev/hdc is the CD drive.  This
>messages seems to come from lvm, that prints this error whenever it
>scans for physical volumes.  The presence of LVM1 physical volumes
>might be a trigger.
>
>Genius EasyMouse+ PS/2 mis-detected as Whell Mouse in anaconda
>start-up
>
>system-config-users won't create new users.  clicking on ok produces:
>Traceback (most recent call last):
>  File "/usr/share/system-config-users/userWindow.py", line 227, in on_userWin_ok_button_clicked
>    if self.homeDirCheck.get_active() == gtk.TRUE:
>TypeError: argument 2 must be list, not str
>
>Account created with useradd (during kickstart %post) can't log in.
>gui and ssh logins exit immediately, and so does VT login without
>security context.  with security context, it asks for role and type,
>and nothing seems to work.  selinux doesn't seem to work, at all.  cat
>/selinux/context fails with invalid argument, and getconlist fails to
>write to it similarly.  I couldn't find other ways to create accounts
>that would get a functional login.
>
>vi won't run on a full install, it can't find libperl.so.  It looks in
>5.8.2 internal perl lib dir, but only 5.8.{0,1,3} are available.
>Karsten is aware of the problem already.
>
>  
>
I was only able to get as far as the Fedora background before it dumped 
on an issue related to pango.

When I tried an http install with yesterdays boot disc, I ended up 
getting a signal 11 and than an abrubt exit.

When I tried an ftp install with today's boot.iso, I ended up with the 
crash. (Photos submitted on bug report)

https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=113358

I didn't have any issues with getting a dhcp address from my router. I 
was using the realtech ethernet card.

Jim