27th Jan 2006: The IT Crowd

Channel 4 have a new series: The IT Crowd. I was quite interested to hear about this, and to see how a mainstream sitcom could be made about IT.
Ironically, the "internet preview" requires MS Internet Explorer and Windows Media Player.
So far, one episode, one techie reference (if you listen carefully, there was a mention of loading device drivers). Other than that, it seems to be the same disappointing stereotypes about tech support being inferior sex-crazed imbeciles and women having no idea about IT. The geeks are portayed as dim-witted idiots, being played the fool by the (also idiotic) management.

One geek is an extrovert imbecile, the other is a dull imbecile. I'm not a professional comedy writer, but I thought that some contrast was a key part of comedy. This isn't a riveting contrast.

"Jokes" (marked with quotation marks for clarity) include:

"Alright - a plan. Let me put on my slightly larger glasses." (it then turns from a plan to a scheme, so he puts on even larger glasses)

Before meeting their new (female) boss, the geeks giggle at the phrase "we do not want to go into there half-cocked"

"This original ZX81" ... I have never heard of a fake ZX81.

To be fair, the manager is also shown as incompetent (doesn't realise that her phone is not plugged in, then later that her PC is not plugged in, but on both occasions is pretending to use them). Despite all this, the techies take a long time to spot her incompetence from these clues. A comedy where the entire cast are idiots is surely a bad formula.
The early comment about device drivers (and the RTFM t-shirt) are the only evidence of any research.
The manager eventually admits that she knows nothing about computers (another sterotype, of course - she is a woman; IT is the male domain) When a female user does turn up (to beat the techies with a stilletto), the two women then talk about shoes. (yet another stereotype).
This is Bastard Operator From Hell done very badly. In fact, is BOfH without the ... well, without anything.

My hopes were not terribly high, but they were higher than this.

On a side note, Google's advert system seems to have a hard time with this page - it's offering me adverts for timing belts (well yeah, I can see that), but also for Italian leather belts, money belts, as well as belts for indirect-drive record players (remember those?)

24th Jan 2006: Polo Fuel Fixed?

It turns out than an epoxy resin can be easily added to fuel tanks - I ordered mine from lb-restoration.co.uk today. I'd give a direct link, but I can't find one - I found the website (via a forum found with Google) and phoned them; the person I spoke to talked knowledgeably about the problem and recommended a solution (for £6.50, at that.) It's cheap and easy (if the gap is small and accessible.) It also saves spending about £200.00 on a new fuel tank. It works like the strong glues you can get - you combine the two compounds together, which quickly set into a combination as strong as steel (so not much like glue, after all!) I knew C21 wouldn't let me down!

This issue had caused me to doubt the sanity of the 'buy a cheap old car and who cares if it has problems' approach, but now I've solved this issue also, I'm becoming a stronger fan of this policy. If I'd ended up needing to spend £200 on a £500 car, I'd be in trouble; spending a tenner (incl. P&P) to fix the problem is no problem. It's more for convenience than functionality - I'm still getting over 30mpg even with the leak, but when you step out of the front door in the morning, the first thing you smell is petrol!

22nd Jan 2006: Polo

I'm torn. I've got two grilles for the Polo (the one I got it with was cracked, so I got a new one). I painted the old one red, to see if it would look better (if uncracked). I'm still not sure. Here are the photos (taken with my Nokia phone, which seems to turn things rather bluish) - click on the images for VGA scale:

The red one is cracked, I know that much - it's the old one which has been sacrificed for the sake of cosmetic experimentation. The colour from the Nokia is very poor - it doesn't really show that the red grille has a badly-painted gold VW badge; if I went with a red grill, it would use the silver badge, of course.

The red grille reminds me of a Pug 205, but the black one looks more dated to me. I am not convinced about the merits of either grille. I can not see any options beyond red or black; it has to be one or the other. Opinions appreciated :)

I must also confess to a previous inaccuracy; the squeaking noise was from the Aux belt, not (fortunately) the drive belt. The squeak returned; it seems that the garage had not tightened the belt fully. The Aux belt transfers the engine power to the Alternator, and was getting loose, and as a result, the rubber belt was squeaking when the engine power suddenly increased (such as when changing downgear). VW provide an intuitive method for tightening the belt: move the Alternator by traversing a cog along its (unmarked) line, after loosening its tension screw. Haynes recommend 8Nm for new belts, and 4Nm for old belts, but since the line is unmarked, that advice seems entirely useless. Conveniently, these are 21mm and 13mm respectively. Luckily, my Dad recently gave me a universal ratchet set which allows you to access most bolts from most positions - without that, I'd have had to have resorted to a couple of pairs of pliers.

On a more serious note, the stench of petrol (and the reason that PFS pumps often overspill) turns out to be due to the hole I found in the pipe between the fuel filler cap and the fuel tank. Oh dear. There's a limit to how long I can block that with kitchen tissue! As the fuel tank has no provision for draining, welding a patch is not going to be a solution, but presumably in the 21st Century there is a way of patching metal which doesn't involve high temperatures - some kind of polyfiller for metal?

Anyway, please let me know which radiator grill (black or red) you think looks best...

22nd Jan 2006: Snooker

John Higgins has done a wonderful job of taking the Masters. It was close all the way, with pot averages constantly above the 90% mark, and neither Higgins nor Ronnie O'Sullivan ever getting a 2-frame advantage. Higgins had a very stressful final frame, with a red hovering over the corner pocket before finally dropping in. An excellent canon lead to a first-class cleanup in very difficult circumstances.

22nd Jan 2006: WMF - the GRC Takedown

The Register have (not so) kindly done what I was not so cruel to do - point out (as if it was necessary) that Gibson is a nothing more than a self-publicising idiot (I kind-of hinted at that with the label "shy genious" in my earlier entry about WMF). Greene is also rather unkind about Microsoft, in a way which cannot be called inaccurate, but is rather unsporting. I can see a journalistic need to provide balanced reporting, but pointing out Microsoft's previous errors does not really achieve that. Simply pointing out what a fool Gibson is (and I am surprised that Greene didn't mention the many flaws in Gibson's much-publicised Shields Up! utility) was quite adequate for this discussion.

21st Jan 2006: Google

Apparently Google have refused to provide what AOL, MS and Yahoo! provided to the US Feds - from what I hear on Radio 4, they asked for (1) All search terms for one sample week, and (2) One million random websites. I can not see a privacy issue with any of those requests, but I can not see how that could possibly help any work against child pornography, which is apparently the reason for these requests. According to The Register, the PATRIOT act allows the US Govt to demand such data. Be that as it may, I fail to see how it could possibly help the US Govt do anything - combat child pornography, or anything else - even if MS, AOL and Yahoo! have already complied. As the Register article already mentioned notes, AOL use Google for their search, so they have probably got much of what they want anyway (though I suspect that many AOL users actually use Google).

17th Jan 2006: WMF

Microsoft blog the WMF problem with some rather entertaining excuses:
Excuse #1: It was designed for co-operative multitasking that is, the feeble attempt used in Windows 3, which could only run one process at a time, and could only switch from one to another with the permission of the currently-running process. (If an app didn't give permission, then no other applications could do any work.)
This method meant that an application passing control to the printing subsystem could not monitor progress, so the application had to call the printing subsystem with a message saying "go off and print this; any problems, just call me on this number and I'll wake up to deal with it". The app could then provide a chunk of code which would deal with the error (alert the user, etc). Ignoring that the whole concept of co-operative multitasking is pretty poor, there is nothing fundamentally wrong with this - it's the only workaround really possible. The printing subsystem would volunteer to stop running and pass control back to the calling application if an error had to be dealt with by the application.
WMF seems to have been designed slightly differently, though - instead of the application providing the callback function, the callback is in the WMF file itself. Microsoft just about manage to admit that this was a design flaw: "these metafile records were all completely trusted by the OS" - even though they could have come from anywhere!

Excuse #2: Internet Explorer isn't vulnerable as it ignores META_ESCAPE records, which are necessary to run this code. That's skirting around the truth - Internet Explorer implements IFRAMEs, and passes them off to for Windows to do whatever it likes with them - by default, passing them to the Windows Picture and Fax Viewer, which does not ignore META_ESCAPE. So it's stretching the truth to say that Internet Explorer is not vulnerable.

Excuse #3: SetAbortProc was written in assembly language when it was ported to Windows 3.0. That is probably the most feeble of all the excuses, especially as it implies that it has since been rewritten (presumably for Windows NT) without anyone even realising that there is no longer any need to work around the weaknesses of Windows 3. Worse still, the "Trustworthy Computing" initiative announced back in Jan 2002 was supposed to have trawled through the existing code with the explicit goal of looking for security flaws.

Windows 95, 98, Me are apparently not Critical because of the Microsoft definition of Critical - if the user has to do anything at all (eg, press "OK") to trigger the vulnerability, then by Microsoft's standards, that's not a Critical flaw (!).

Shy genious Steve Gibson (author of the ShieldsUp! utility) ploughed into the discussion feet-first by ignoring the published exploit code and writing his own, in such a way as to trigger a non-security-related implementation flaw in the WMF code (if you define your SetAbortProc function at the end of the WMF, it doesn't get triggered until another WMF gets processed). He deduced from this, that the vulnerability only works if you have an incorrect "size" field in the WMF file. If he had looked at the other published working exploit code, he would immediately have seen that he was wrong. Instead, he decided that this must be a deliberate "back-door" written by Microsoft to allow them to run arbitrary code on any PC which visited (say) www.microsoft.com. That's a pretty big claim to make based on some very poor evidence. To be fair to Microsoft, they treat Gibson quite gently in their rebuttal.

14th Jan 2006: Microsoft and Standards

MS Exchange / Office has certain features, like emailing, replying to, and accepting appointments, etc. Probably more which I have not yet discovered. These don't naturally map to (say) Thunderbird (email only), or calendar managers (I don't know, I don't use them, but I believe that Evolution offers such capabilities), but they could be documented, and configured to use (say) certain URLs (eg appointment://18jan2006/9am/title=Meeting&details=foo) so as to be interoperable with a variety of client systems. That would work across MS Outlook and other suites, regardless of OS. I would recommend a UTC timesource in real life, of course, this is just an example.

This can't really map onto the RFC system, which traditionally requires 2 real-life, published-source implementations, and MS wouldn't realistically join into the RFC system anyway - they never have previously done so, to my knowledge. MS seem to see RFCs only as a Read/Embrace/Extend resource.

A system whereby these features were documented are what I, personally, as an enforced Windows user, would find useful for interoperability in a Microsoft-centred environment. That would make it possible for me to use alternative OSes in a MS environment without losing functionality.

This functionality along (not dissimilar to, nor (AFAIK) incompatible with, the GNOME VFS layer) would allow me to manage my email and appointments within a MS environment with a non-MS client OS. SAMBA and other teams have already done much of the other work required; The OpenOffice.org team have done wonders, also.

14th Jan 2006: Booze Bird

An interesting series on BBC 3 - along the lines of SuperSize Me where a guy ate only McDonalds for a month, this 40-something journalist went binge-drinking on over 10 pints per night, matching some early-20s women drink-for-drink. This is fascinating stuff, though I'm not sure exactly what it's supposed to prove, other than that 40-something women cannot keep up with 20-something women. Not that I can condone such heavy drinking, the 20-somethings are clearly doing okay for now, and not showing the symptoms suffered by the journalist. The younger women do not appearr to have the young children to take care of (though the single-parent journalist did make it clear that provision was in place to look after her children).

I know what this is supposed to show - that excessive alcohol abuse is bad for the body - but all it actally shows is that it is even worse for a 40-something body than for a 20-something body.

Surely the BBC would be better off showing the damage done by such alcohol levels without damaging one of their own journalists' health?

Bravo for bringing attention to the legislation as it exists, and how it is ignored, but really, the health side of it is totally unscientific - is the current hype about binge drinkers really about 40-something women?

At the same time, she is doing her journalistic job, and interviewing the lawmakers, and the drinks companies, though she is hardly an interrogator, simply asking questions and taking their answers at face value.

4th Jan 2006: MS Outlook

Here's a great example of poor programming. A claim from the Microsoft Outlook email client that 'all folders are up to date', whilst still admitting that it is 'trying to connect'. One of these statements is true. The other, by definition, is not.

2nd Jan 2006: The Sky at Night

I don't know about you, but I do enjoy watching The Sky At Night from time to time, just as I love to gaze up at the stars from time to time. Apparently there is a probe due to set sail to probe Pluto some time around 2015.
There was some footage of white-suited scientists preparing the gold-covered probe for the voyage.

Tonight's programme ended with an invitation to sign up for a newsletter, which struck me as a nice idea - I'm not an amateur astronomer, but I do find it an interesting area, so the occasional newsletter would be interesting. Being the BBC, I thought, I can trust them with my email address, so I was ready to note down the website or email address required to signup for the newsletter. (By the way, the website is http://www.bbc.co.uk/science/space/spaceguide/skyatnight/). So, how do you sign up for the newsletter about such high-tech projects?

Just send a Stamped, self-addressed envelope to:
The Sky At Night, BBC Birmingham, The Mailbox, Birmingham, B1 1RF
High-tech, or what?!