Hello,
I noticed that after using a bootable CD with WC, version 10.2 and also 11.0, at the next boot of the PC from the HDD (using a bootable floppy with the PLOP-bootmanager), the PC starts with the message that the C and D partition needs to be checked (so a CheckDisk).
This behavior worries me. If WC is not touching the HDD, why is the PC then starting with this check, something must have been changed on the HDD.
Is the a possible explanation?
PS
To stop WC I have to push the PC-out-button (a power-down).
Then I push this PC-button again to start the PC again, however then with a bootable floppy with the PLOP-bootmanager (I cannot open de CD-tray when WC is active).
From the PLOP-bootmanager, which shows all bootable options, I then boot through the HDD. Then the CD-tray opens again.
Can you not include in your WC to stop/exit WC, and then offering to open the CD-tray and remove the bootable-CD. So then I do not have to power-down the PC to stop WC.
Best Regards
Jan
This behaviour would worry me too Jan. Can anyone confirm this?
The only thing I can think that maybe doing this is a udev script I
dropped in more than a month ago that automounts disks it finds on
insertion.
This was to realise the "upload photos at Internet cafe" use case.
Tbh I have had nightmares about this, since it opens an attack vector.
Hence I've fell short of blogging / announcing this loudly, which is
really bad of me. I wanted to make sure at least it mount files
without execute permissions, to somehow circumvent people have uxterm
on their disk and somehow (I'm not sure how, but I'm thinking
"security in depth"), executing it.
> Can you not include in your WC to stop/exit WC, and then offering to open
> the CD-tray and remove the bootable-CD. So then I do not have to power-down
> the PC to stop WC.
I think this can be fixed by removing "quickreboot" option from the
boot command line.
I quickly built a version without udev scripts and quickreboot, please
give it a try:
http://custom.webconverger.com/overbeek.iso
Let me know if you can reproduce the disk issue, then I will know it's
not udev at least. Thank you!
Kind regards,
Ok I've permanently removed the offending udev automounting code:
http://anonscm.debian.org/gitweb/?p=debian-live/config-webc.git;a=commitdiff;h=ea18dff8af2b6800c580b70ce4219aacae9dae73
I don't know if there is a way of writing that udev script, so it only
mount on insertion of a disk.
> However I now had another odd experience.
> Later after another boot into WC, I had the same in-responsiveness (the
> Google-icon was gone). However now after clicking X the normal working
> internet reappeared.
I need some time to try reproduce this.
> Now the power-down/WC-CD removing/rebooting isuue.
> When I push the power-down-button on my PC (WC does not have a
> power-down/reboot window?) WC indeed opens the CD-tray.
> However after the enter-button my PC powers down. So I have to push the
> power-button again to power up for booting into my HDD.
> At closing the WC-session after removing the CD I prefer the get a choice to
> really power-down or reboot into my HDD.
> It is maybe a minor issue but it annoyances me.
Tbh I think it's more annoying for most to have a dialog to eject the
CD, so I am not keen to make ejecting the CD default behaviour. Or
what do others think?
Kind regards,
Ok I've permanently removed the offending udev automounting code:http://anonscm.debian.org/gitweb/?p=debian-live/config-webc.git;a=commitdiff;h=ea18dff8af2b6800c580b70ce4219aacae9dae73
I don't know if there is a way of writing that udev script, so it only
mount on insertion of a disk.
Tbh I think it's more annoying for most to have a dialog to eject the
CD, so I am not keen to make ejecting the CD default behaviour. Or
what do others think?
I've run a 11.0 instance all day on my X201 Thinkpad connected by wifi
and it was stable, however I think judging by the `dmesg` log it
re-associated with the AP once or twice. I suspect your hardware is
flaky. There are unfortunately a lot of poor quality wireless
equipment in the market.
I highly recommend TP-LINK equipment given a choice.
You can debug by going into debug mode and tailing the logs. Perhaps
you can see where the problem is.
Kind regards,
Exit of the WC-session (on my PC) by a power-down, is not a real deal-breaker.
ChkDsk probably does some raw dd style checking whilst mount opens the
filesystem for use and probably writes something to the metadata.
I think I need to establish a poll whether people want/need to upload
from Webconverger. In order to warrant looking into how to only mount
inserted disks, as opposed to what udev discovers. :)
Kind regards,
Good suggestion. Kind of worried http://webconverger.org/boot is
getting a bit crazy.
I was thinking of having more of a menu on the boot screen, so people
could choose language et al. But with all the options, it turns into
some UI nightmare. :)
http://www.spinics.net/lists/hotplug/msg05324.html
Some things in Linux suck a lot.
Kind regards,
> Just followed up with udev upsteam and I'm evidently doing it all
> wrong trying to mount via udev. _sigh_
Perhaps pydsm might help (just stumbled upon it, haven't actually tried
it yet)?
Gr.
Matthijs
Python, sourceforge and CVS aren't good signs. :) Will look.
I think I should study http://www.chromium.org/chromium-os sources and
maybe even BSDs to see how they tackle these sorts problems.
Anyway, I plan to release 11.2 very soon without mounting support.
Kind regards,
The release is delayed for Iceweasel (Firefox) 10 to hit and give it
some testing with
https://github.com/Webconverger/iceweasel-webconverger
http://www.mozilla.org/en-US/firefox/10.0/releasenotes/
Was pleased to see 1155 new 11.0 users last month.
http://ping.webconverger.org/2012-01.txt
I think 11.2 will be much better, tempted to call it 12. :)