[Trac-bugs] [PC-BSD Trac] #230: PCBSD-8
PC-BSD
trac at pcbsd.org
Mon Feb 1 22:33:29 PST 2010
#230: PCBSD-8
-----------------------+----------------------------------------------------
Reporter: GeeKen | Owner: kris
Type: defect | Status: new
Priority: major | Milestone: 8.0
Component: Installer | Version: 8.0-BETA
Keywords: |
-----------------------+----------------------------------------------------
Comment(by GeeKen):
Replying to [comment:1 kris]:
== ==
I've now fixed the e-mail log script, it'll prompt you to copy it offsite
now, and let you view the log at the same time. However e-mail won't work
since we don't have smtp settings for each user :P
'''''OK - great. Problem is I don't install it every day ! I have some
text from part of the error if this helps, see below. So will TRAC BASE
receive the error reports automatically and directly ???'''''
>
== ==
> Can you re-test the x86 install, and this time do a "more" of the
logfile when it fails? If I can see the actual error, it should be easy to
fix.
''''The error text that I thought was major was as follows :''''''''''
"
ZFS NOTICE : Prefetch is disabled by Default on i386 - - - to enable add :
"vfs.zfs.prefetch-disable=0" to /boot/loader.conf
ZFS WARNING : Recommend minimum kmemsize is 512 MByte : expect unstable
behaviour. Consider tuning vm.kmem_size and vm.kmem_size_max in
/boot/loader.conf
ZFS f/s Version 13
ZFS storage pool version 13.
etc etc etc
"
== ==
The i386 installer has several problems :
1) The above, which is terminal - - -
2) Another massive error is if I try the "upgrade option" (Beta to RC1)
where it unmounts everything first. During this process it finds something
unmountable because it can't find the item. This causes it to stop. (this
error sequence was repeatedly reported in the email log that you have now
fixed from above - hopefully you have this record). If it can't find
something that it wants to remove, why stop when it must be already
removed !
3) Install always needs to be run in Mode 6 and/or 2, so that the install
doesn't end up with a
non-responsive monitor and sometimes disabling ACPI allows the install to
go forward beyond these setup erros to the major halting errors. Here I
would suggest that all installs BEGIN IN VESA so that no video halt
occurs. And if the ACPI error occurs then it should retry with ACPI
disabled.
4) At all times the user can be consulted to avoid a freeze- for eg., a
screen could come up showing the undesirable situation, offering optional
solutions :
Message = "I have halted - suggest that you disble ACPI - shall I proceed
with no ACPI ?"
LINUX MINT-8 installer seems to make all of these kind of decisions by
itself - successfully
too.
== ==
BSD-8 RC1 works great on my AMD 64 bit HP. The bugs that I keep seeing
are:
1) The Mouse wheel can cause screen jumping when the wheel is scrolled up
too far. Its
nervous on the Mouse Wheel. FAR BETTER THAN THE BETA VERSION (which was
quite psychotic - - -).
2) When using Firefox 3.5.7 (maybe this is fixed on version 3.6 ?), any
slows in the Internet
feed causes the whole machine to lock up, while it waits. My machine is a
dual-core, surely the other core can carry on ? Following on from this
though process, there needs to be a review of how BSD-8 handles multi-core
processors. Maybe there is more speed to be had - - -
Handling the Firefox "lock-up" problem is essential. As it stands I just
have to wait for a minute for the Internet feed to rebegin or for the
block of material that's pending to download.
== ==
Anyways, hope this data is useful.
BSD-8 is the Best. The Linux-MINT-8 I am testing is also great but it
fails upon upgrade as the upgrades are uncontrolled. This is the main
reason why PCBSD will win in the end - - -
Hey - what happened to TURTLE ?? - that was a cool program for teaching
kids and doing small APPS. CAN WE HAVE IT BACK ??
Thanks for your amazing input - keep up the great work !
== ==
Regards
Ken Gibbs
Engineer
* * *
--
Ticket URL: <https://trac.pcbsd.org/ticket/230#comment:2>
PC-BSD <http://trac.pcbsd.org>
PC-BSD Project Management
More information about the Trac-bugs
mailing list