In a previous post, I related a tale of woe of getting RMA support on a Seagate hard disk drive which I had purchased from Scan Computers International Limited.
I implied in reply to comments (from mrben and others) that it wasn’t a simple catalogue error on Seagate’s part. Here’s why. On 13th October Seagate informed me that the drive had been marked “missing in transit”, and that it was stolen before arriving at their resellers. The drive shouldn’t have made it to retail. The police investigation was ongoing, but on 2nd November Seagate confirmed I could contact Scan about the issue. Unsurprisingly Seagate won’t support RMA requests on their hardware that has been nicked. The only way for me to get a refund or a replacement drive would be from Scan themselves.
So, on 7th November I wrote to the Managing Director of Scan Computers International Limited, enclosing all the related e-mail documentation on the order. On the 10th I received a standard automated RMA e-mail, giving me seven days to return the items. No personal comment or explanation about the alleged dodgy status of the disk. I posted the drive back on the 14th and received e-mail confirmation on the 15th that it had been tested and failed their diagnostics. Scan’s RMA page showed that I should be receiving a refund, but none was forthcoming.
On the 21st November I e-mailed Scan to enquire about the refund. I haven’t received a reply, but on 24th November I received a full refund for the drive to my bank account. I haven’t yet received any communication from a real person about this whole mess. I will be boycotting Scan from now on.
Our builders managed to cut the telephone cable at home, yesterday. This is a bad thing. Hopefully it will be repaired today, but I have my concerns as to how the ADSL will be affected. E-mail is queuing up on the backup MX, but if you need to get hold of me more urgently, please leave a comment.
A new service from mysociety.org is a web-based petition system, hosted by the Prime Minister’s office. Some of the petitions submitted are clearly from loonies and the socially maladjusted, but some are of genuine interest and importance. One of the top five most popular petitions is:
We the undersigned petition the Prime Minister to create a new exception to copyright law that gives individuals the right to create a private copy of copyrighted materials for their own personal use, including back-ups, archiving and shifting format.
Almost 1,500 signatures have been added to the above petition so far. If you’re a British citizen and support the motion, then please take a minute (and that’s all it does take) to sign up here. Another top five petition is:
We the undersigned petition the Prime Minister to scrap the proposed introduction of ID cards.
This petition can be signed up to here.
Recently this domain has started to be used by e-mail spammers. (No, I do not run an open mail relay, they are simply spoofing random addresses with my domain as the extension.) As a result I’m getting delivery failure messages bouncing back to me from mail servers that are configured to return such things. So, I looked into implementing SPF on my domains. It won’t completely eliminate bounce messages, because not all mail servers check SPF records, but at least I will have done what I can.
My outgoing mail goes through my ISP’s mail servers, as many mail servers block mail sent directly from home Internet connections. It’s easy enough to use an “include” statement in the SPF record to allow another domain’s SPF recorded mail servers to relay mail for your domain. However, my ISP doesn’t publish SPF records for their mail servers. So the only other option is to include the IP address of their mail servers in my SPF record. The problem with this, is that Force9 have at least 9 outgoing mail servers, each of which has to be added individually and kept up-to-date. One mail server changing IP address could result in intermittent delivery failures which could be hard to detect.
I’ve tried raising the issue on the Force9 forums, but there hasn’t been any interest. So, Force9 / Plus.Net, please implement SPF on your mail servers!
I have proof!
There have been suggestions from LUG Radio and Nik Butler that the lug.org.uk project needs a President. The further suggestion is that Alan Pope is the man for the job. Whilst Alan has all manner of personal qualities that would suit a leadership role, the wider question is whether such a role is needed. lug.org.uk has survived, even thrived, since 1998 without a formal, democratically elected leader. Sure, Mark Lewis was essentially in charge in the early days because he was paying for the service to be set up and doing a lot of work on it. But for a few years the team has basically lead itself.
That’s not to say that lug.org.uk shouldn’t have a “leader” internally. Having someone who could make management decisions if opinion is split between the admin team could be a good thing. However, at the moment, with only a handful of people actively involved in administering the system and developing ideas there’s no urgent need. It may also be useful to have a named press contact to redirect incoming enquiries. Certainly at points in the last few years some internal leadership might have beneficial in motivating the sysadmins and speaking to potential hardware donors.
But what isn’t needed is some sort of President to speak forth on behalf of LUGs. Most LUGs are essentially anarchies. Fill a room with geeks and you’ll get as many different opinions on Free Software topics as there are bad haircuts. Presenting a consolidated view of one LUG is hard enough (I’ve been there), but doing so for all UK LUGs would be an impossible task. Many LUGs would object to being spoken for in this way. It could lead to LUGs who disagree with “the lug.org.uk viewpoint” leaving the service and going it alone. Rather than bringing the community together under one figurehead, it could split it.
There are practical problems in electing any sort of “President” too. There is no formal definition of a LUG, no formal definition of what LUGs should be able to vote on lug.org.uk matters (what about LUGs who don’t use the service?), no formal definition of membership of most LUGs, no formal definition of who represents a LUG to other LUGs (sometimes this is elected, sometimes a Benevolent Dictator), no formal definition of how different size LUGs should be represented within the lug.org.uk community (does a large LUG deserve more votes?) and most importantly no formal definition of an electoral procedure for any such post. Because it doesn’t exist.
Perhaps most importantly, there hasn’t been a clamour for such a President. One or two voices have championed the idea, but there doesn’t seem to be any demand from the LUGs themselves.
I use Kino to capture DV from Alan’s DV camera when I’ve used it to record the talks given at our LUG. It worked pretty well under Dapper. However, when trying to capture DV using Ubuntu Edgy, Kino returned an error saying
WARNING: raw1394 kernel module not loaded or failure to read/write /dev/raw1394!
Actually the module was already loaded and the permissions on the device node were just fine. It had read-write access to the “video” group, and I was in that group. The diagnostic program testlibraw reported it could talk to the devices OK, so I began to suspect that Kino was actually at fault. After doing some poking around using strace it turns out that after accessing /dev/raw1394, there’s a different device node Kino is trying to access:
open("/dev/ieee1394/dv/host0/PAL/in", O_RDWR|O_LARGEFILE) = -1 ENOENT (No such file or directory)
The device node above didn’t exist on my system. Some googling turned up these French Ubuntu forums and related mkdir and mknod commands to rectify the problem:
mkdir -p /dev/ieee1394/dv/host0/NTSC
mknod -m 666 /dev/ieee1394/dv/host0/NTSC/in c 171 32
mknod -m 666 /dev/ieee1394/dv/host0/NTSC/out c 171 33
mkdir -p /dev/ieee1394/dv/host0/PAL
mknod -m 666 /dev/ieee1394/dv/host0/PAL/in c 171 34
mknod -m 666 /dev/ieee1394/dv/host0/PAL/out c 171 35
I only created the PAL related nodes and Kino then worked fine. As Ubuntu uses udev, these device nodes will not survive a reboot, so frequent users of such equipment might want to add these lines (or similar ones with slightly more secure permissions) to /etc/rc.local. The actual cause of the problem isn’t so clear, whether it’s a udev issue or upstart or what, I don’t know. Using a Dapper kernel didn’t resolve the issue though.
Contrary to some people’s experience, I’ve not had any serious problems upgrading to Ubuntu Edgy. I upgraded my desktop a couple of days after Edgy’s release and it’s great. I had to run “aptitude dist-upgrade” a few times, but I just accepted the suggestions each time and after three or four iterations, I had a lovely Edgy desktop.
The only problem I had was with VMWare Workstation, which I run on my desktop. The first thing is to make sure you have the latest release of VMWare Workstation. That’s 5.5.0 build 29772. The kernel modules supplied with VMWare will not build against Edgy’s kernel with any older versions. I was already running the latest version of VMWare Workstation, so it was just a matter of running the usual vmware-config.pl script that follows a kernel upgrade. There were errors related to creating the application menu icons, but I still had them from before. For new installs this might be a problem.
The problem I had was that when I started VMWare Workstation from the command line having done this.
$ vmware
/usr/local/lib/vmware/bin/vmware: /usr/local/lib/vmware/lib/libpng12.so.0/libpng12.so.0: no version information available (required by /usr/lib/libcairo.so.2)
And then it just sat there with the CPU at 100%. For ever. Now, I’d had the libpng warning when running VMWare Workstation under Dapper, so I didn’t think this would be the cause of the problem. I did some investigation using strace myself, and then came across this thread at the Ubuntu Forums, which basically suggests removing libdbus-1-2, as Edgy has libdbus-1-3 and having both versions installed makes VMWare Workstation cry.
Using aptitude to remove libdbus-1-2 also removed libnautilus-burn3, which was the second suggested solution. Again, Edgy has the newer libnautilus-burn4, so this should be OK. So this problem could catch out quite a few people who upgraded to Edgy from Dapper, but shouldn’t be an issue for a fresh install of Edgy.
My good friend Alan Pope is interviewed on the latest episode of LUG Radio talking about UK LUGs. Alan did a very good job of representing the current state of LUGs in the UK. Download it and have a listen! I had been due to go on the show to talk about the recent goings-on at HantsLUG, but got passed over at the last minute in favour of Al.
I can think of no-one nicer to be abruptly shit-canned for. 
Update: Alan has posted a blog entry that follows up, expands upon and reflects the interview. It’s well worth a read, and pretty much reflects my views on lug.org.uk, especially the hard work done by Andy Smith and the other admins. I have volunteered to help out lug.org.uk not by doing any admin work (the guys are much better at this than me and really have it all covered) but with things like the lug.org.uk BOF I organised at LUG Radio Live this year.
