Julien Valroff

It seems that Josselin Mouette managed to fix bugzilla #327249!

I am now able to have a shared public folder, using Posix ACLs. However, there is a major limitation which I haven’t paid attention to before: neither cp nor mv do respect default ACL set on the parent directory.

I have opened a bug in the BTS against coreutils, as I am pretty sure this is not the correct behaviour.

I received yesterday my new Dell Inspiron 1300 which is already running Debian Etch (I could not accept Windows EULA). I have chosen hestia for its name.

Setting it up was really problem-free, although I havent tried to install GNOME and the like. I haven’t tested anything except the wireless NIC with WPA2 (works like a charm thanks to bcm43xx-fwcutter and wpasupplicant, just needed to add [code]wpa-ap-scan 2[/code] to my /etc/network/interfaces file to deal with my hidden SSID.

I will thus be able to use this new laptop as an internal server, ie. not accessible from the outside world, It will host an LDAP directory for my contacts, and other things to be thought of (not yet decided). I wanted a wifi-enabed laptop to be able to set it up anywhere in my flat, the battery will be enough in case of power outage.

However, I was quite surprised by the old-looking box of the laptop. It looks a bit like my 6 or 7 years old Compaq Armada e500! The 14.1″ screen is also very small compared to the surface of the case. I wouldn’t recommend this laptop for every-day use, except if you can’t afford to pay an extra 50€ or 100€ for an Acer or Toshiba.

Note that I won’t write any article on setting up GNU/Linux on this laptop, as complete reports can be easily found.

Although I have been using successfully suspend2 patch and hibernate script for some time now on my Toshiba laptop, I noticed last week-end that resuming from suspend-to-ram did not re-initalize my video card correctly, leading to a blank screen. I couldn’t find any solution nor explanation to this anf finally gave a try to µsuspend (known as the uswsusp package in Debian)… and this works perfectly, and is far more reliable as suspend2/hibernate.

As a bonus, the time needed to suspend (both to ram and to disk) is considerabily shorter, resuming from s2ram is almost instantaneous!

I had to recompile my kernel to include initramfs support, as well as CONFIG_SOFTWARE_SUSPEND. I also unapplied suspend2 patch from my kernel tree.
As I didn’t previously use initramfs, I had to alter /etc/kernel-img.conf to add:
[code]ramdisk = /usr/sbin/mkinitramfs-kpkg[/code] and still have to recall to use the --initrd option when calling make-kpkg!

I will update my article on the TS M40X during the week-end.

The next step is to use the same set-up on my desktop machine, which currently does not support suspend-to-ram.