Strange Cat Photo

The thing that gets the highest hit count on this website is not any of the blog entries, none of the photo galleries nor even the URLs that those damn accursed referrer spammers try to go to, but this photo:

Silly cat photo - source unknown

Unfortunately I can’t find any attribution for it – the best I’ve got is that it was first mentioned in this story by Alec who pointed to this URL that went 404. I managed to track down a copy via some method that escapes me and added that as a comment to Alec’s article, which he then posted about here.

The EXIF information isn’t much use, it just says:

$ exif Cat.jpg
EXIF tags in 'Cat.jpg' ('Motorola' byte order):
--------------------+-----------------------------------------------------------
Tag                 |Value
--------------------+-----------------------------------------------------------
Orientation         |top - left
x-Resolution        |72/1
y-Resolution        |72/1
Resolution Unit     |Inch
Software            |Adobe Photoshop 7.0
Date and Time       |2004:05:19 12:49:44
Compression         |JPEG compression
x-Resolution        |72/1
y-Resolution        |72/1
Resolution Unit     |Inch
Color Space         |Uncalibrated
PixelXDimension     |434
PixelYDimension     |377
--------------------+-----------------------------------------------------------
EXIF data contains a thumbnail (3764 bytes).

My guess is that as Photoshop is involved it’s a fake.. But I’d love to find out who did it so they can get proper credit!

SCO Copies Their Own Legal Documents From Groklaw and Tuxrocks – Without Attribution

Here’s an interesting turn up for the books. SCO, who started the lawsuit and who’s lawyers created their side of the documents, apparently have been been ripping off the scans of legal documents from the cases done by Tuxrocks and Groklaw and putting them up on their new “IP” website, without any credit to the original source.

Ironic, isn’t it ?

Frank Sorenson is quoted on Groklaw as writing:


For example, compare SCO’s #156 with the copy from my site. Note (using pdfinfo) that they were both scanned with “AXIS 700 Scan Server” (the scanner/copier at my office), they were both scanned at the same identical time (“Fri May 28 16:51:21 2004”), and that they both have the sme size (85387 bytes). Also note that they both have the same md5sum (3c1064ecedb19073e2194cd88507c349).

Another example is the Declaration of Ira Kistenberg (compare with the copy I obtained, scanned, and posted). In the upper right-hand corner of the page, you can see my handwriting where I noted which particular document this one was (236-G) so that I could remember when I got home from making copies at the courthouse.

New Tasmanian Tiger Photos – Probably Fakes :-(

ABC News is reporting that the new Tasmanian Tiger photos (as yet unpublished anywhere) which created such a stir recently are now believed to be fakes by the curator of the Tasmanian Museum & Art Gallery.

Mr Pemberton says he thinks the photos are clearly doctored.

“The thylacine in the image was very similar to another photograph that we quickly identified,” he said.

“But secondly there were two photographs and there was a time gap between the two of… five, 10, 20 seconds… and the posture of the animal hadn’t changed, which I found very strange.”

RSS Feed Backend Updated to RSS2.0

Hoorah! I’ve just replaced the standard Postnuke RSS 0.91 RSS backend.php with the xBackends 2.6 from LEXeBus to give me an RSS 2.0 feed instead of the older 0.91 version.

The two minor hacks I did was to mimic the behaviour of the old backend.php so I could just drop it in as a replacement by setting the following variables at the top of the script:

$show_content=1;        // Default to showing the content of articles
$feedtype="rss2";       // Default to giving an RSS2 feed

With that done I get a successful validation via FeedValidator, viz:

[Valid RSS]

Gentoo Split Ebuilds for KDE 3.4 and Onwards

Here’s a good thing, those nice Gentoo folks are going to create ebuilds for the various components of KDE, migrating away from the monolithic, all-or-nothing ebuilds for kdebase and kdepim to packages for things like kdebase-kioslaves and kopete. The monolithic ebuilds will be maintained for 3.4, but for KDE 4.0 and later there will only be the split ones.

For the details see the Gentoo KDE Split Ebuilds HOWTO.

Making Postnuke Understand Time Properly

I finally got fed up enough with the dumb way that PostNuke does its time stamping to try and fix it. Basically it uses local time everywhere, then gets you to tell it which timezone its in via its preferences and finally expects users to register and set their timezone so it can re-correct the time for them!

Anyway, this isn’t obvious until you go hunting around for why it doesn’t seem to work if you’re not in the same timezone (or in my case, continent & hemisphere) for where your website is. Googling around for Postnuke “time zone offset” gives some helpful references, especially with respect to a Postnuke Forums posting about fixing it. But before you go off and read that, note that (a) it won’t work on current versions and (b) there’s a simplified variant of the hack. Still kudos to them for working this out!

You’ll need to read on for the guts of the article, as I don’t want to scare the non-techies out there by putting it on the front page directly.. 🙂

Continue reading

Wales do the Grand Slam!

It takes a while for the news to get down here, where they take a rugby ball and make up silly rules about what to do with it, but Wales have done the grand slam! The last time they did that (1978) I wasn’t even in high school!

For those who have not been initiated into the mysteries of the Five Six Nations Rugby Union championship it means they beat all 5 other teams in the championship. I was over the moon when they beat England but never imagined they’d do the Grand Slam…

Kevin Morgan dives over to score a try for Wales, from the RBS Six Nations Website
Wales Celebrate Victory, from the Welsh Rugby Union website

Fork-bombing Linux – a Lesson in Poor Defaults

My good friend Alec Muffett has blogged an article from SecurityFocus about the vulnerability of default Linux system installs to, what he neatly call, "The triumphant return of: main(){while(1)fork();}".

It’s sad to see that many Linux distros (Debian being the notable exception) still ship with bad defaults that don’t prevent a non-privileged user fork-bombing a box. Certainly something that needs to be addressed as it’s all part of the “defence in depth” that any system needs.