Installed today's rawhide boot.iso

[SlowJet]

Online install from boot.iso into F11 Host with VBox-3.0.4.

Initialized disk img - ok
Partitioning - use enire disk (default) - prated.py messages when creating sda1.

rebooted - hung on searching for scsi?

rebooted using alternate xdrv vesa
Back to partitioning
Selected Replace exiting system (choice #2)
Anaconda proceeded but nothing showed about sda1, only sda2. However it was ok.

Install completed and first boot ok
Did some conf while cron runs - elapsed time 5 hrs 40 mins - need about 20 more before rebooting into vesa bigger screen, configure mail, WORKGROUP, change netwrok, do some rpm updates.

Very stable compared to <before today's> rawhide.
Did not do a mkinird -f and it worked ok.
Basic /ext/X11/xorg.conf was there, thank you.Typed in my stuff with nano.

Sj

[Demz]

guess there's no point in seeing what the boot time was. SJ when you downloaded snap1 LiveCD did you get KMS? i didnt

also SJ. do this test with Epiphany if you can, check to see if it can get to these forums some reason the snap build i couldnt get epiphany to go to these forums, but iit got to any other site

[SlowJet]

Well, second boot did not come up, gdm and / or xorg.
Tried to mkinitrd but cpu was to busy to get anywhere.
shutdown for now.

SJ

[Demz]

in other words wait for Snap2 , i thought they have done away with mkinitrd

[SlowJet]

I think I'll wait about a week or more.
The "replace existing system" did not replace sda1, but started at sda2, the LVM is on sda3.
So default install choice #1 and #2 is buggy.

Then I see the bad fonts report on devl-list. Yikes.
I bet this is the cause of many VBox crashes and hangs.

mkinitrd is not gone, yet.
And dracut is not the default. (So why is img so big and screwed up? No one has explained that, it just happened.)

SJ

[scottro]

One major (at least for me) fix with the latest boot.iso is that once again, it is responding to askmethod. For a couple of weeks there, it wasn't. and giving no choice of installation save for hard drive.

[SlowJet]

OK, I had some coffee and tried again.:)

1. The mkintrd works ok.
2. The kernel prams are all jammed together with no spaces.
So, no quitfontkeybdlang, but this probably wouldn't matter except I messed with it trying to remove font (witch cross the end of the line) and that may have been the hang up.
3. After kernel line spaced and fixed, xorg.conf had one more typo.
reboot and gdm, xorg, Logon and Gnome started.

Only error so far was setroubleshoot bombed on python errors.
Looks good otherwise.

Demz:
Epiphany started ok, but I don't know what to do with it? :)
What do you mean, "get to these forms?" irc? What's the trick?

SJ

[SlowJet]

Updated to new kernel-2.6.31-0.204.rc9 and discovered

1. I never had package pref noarch from kernel packages. Now big img works, no need to mkintrd.
2. quietLANGSYSFONT run together and so SYSFONT LANG get repeated on kernel line, but not quiet.

So that needs to be cleaned up and with pref package it boots.(but takes longer.)

at-spi crashed on reboot? It is listed twice in the start up programs?
SJ

[scottro]

Well, it *is* alpha. I spoke too soon about the boot.iso, anyway, turns out that when I choose URL to install, it just goes into reboot.

This one really seems to be a rougher alpha than the previous versions, but that could just be my memory being hazy.

[Demz]

remind me not to install an Alpha an they wonder why not many people install it, i would of thought they would of made Dracut the default in the Alpha to get it thoroughly tested before the beta. just wonder if it'll make its way into F12

[scottro]

Heh, you coward. :D :D

Seriously, I remember issues with the F11 betas, and I know some of this is due to changes in anaconda (I almost wrote Amazon--I must be tired). However, I've run into the following issues here and there.

Won't boot, saying it can't find root partition--if I change it from a UID label to /dev/sdwhateverWhatever, it would start to boot, but then die on being unable to open console (I think).

That one was quickly fixed. The boot and/or net.iso not allowing networks is, as per James L., due to anaconda problems, and I'm watching the bug.

On VM's, with almost each iteration, I've had to do nomodeset. Etc. Nothing that I think is a showstopper for them (save for that not finding root, and that one seems fixed), but just more than usual.

On the other hand, I seem to be having better luck than SlowJet. 90 percent of the time, I get a workable installation.

There's a few *nice* (IMHO) things I want. to try--for example, when customizing the installation, one can choose a fluxbox environment. I want to try that and see if things like NM and the like work, but first, they have to fix it a bit. (I was going to try that today with the boot.iso, but as mentioned above, didn't get past getting a network address.)