bluebones.net

I’ve got zillions of 500 errors in my log files (well, a handful each day). All are from Internet Explorer 5 on Windows NT (specifically the User-Agent is “Mozilla/4.0+(compatible;+MSIE+5.0;+Windows+NT)”. I’m pretty certain that the problem is ampersand character entities in anchor tag href attributes. It seems that IE5 is barfing on & being used in querystrings rather than just an ampersand. But of course the W3C’s HTML validator won’t have anything but the character entity without telling me off. Is there an answer?

Fucking fantastic. Mind you they could have played a bunch of Wham! covers and I think I would have enjoyed it.

Set list (NOT in order):

• Caribou
• Nimrod’s Son
• Holiday Song
• Bone Machine
• Broken Face
• Something Against You
• Gigantic
• River Euphrates
• Cactus
• Vamos
• Debaser
• Tame
• Wave Of Mutilation (LP)
• I Bleed
• Here Comes Your Man
• Dead
• Monkey Gone To Heaven
• No. 13 Baby
• Hey
• Gouge Away
• Velouria
• Subbacultcha
• Winterlong
• In Heaven/Wave (UK Surf)
• Into The White

That’s 8/18 from my perfect Pixies setlist. Plenty enough to keep me grinning all night.

Not since Elvis has one for the money been so good.

As part of their “Copyright versus Community” event I saw Richard Stallman and Cory Doctorow talking (separately) about various copyright issues.

Stallman lived up to his reputation as an uncompromising proponent of free software (he invented/popularised most of its basic ideas, after all). He gave a simple but effective critique of “Open Source” as opposed to “Free Software”: in a nutshell, that Open Source advocates claim open source is the best way to develop software but that they do not have any ethical objection to closed source. On the other hand, Free Software advocates see closed source software as inherently wrong. So if you develop some closed source software and it is better than the open source software then the Open Source advocates have no objections.

He managed to fit all the cranky behaviours he has become famous for into a two hour stretch: talking about how GNU/Linux should always be called “GNU plus Linux” and not just Linux; correcting each and every use of the phrase “open source” where the speaker meant “free software”; that “piracy” is not the right word for “sharing” or “helping your neighbour”. He even managed to get in a row with Cory Doctorow before his speech about whether it was useful to umbrella patent and copyright issues (he thinks not, Cory thinks so). This is a man that will not bend, even a little, from what he sees as right. You could say that’s unreasonable, but have you given us Emacs, gcc and all the rest?

Cory Doctorow was full of energy despite having jumped off a plane from Barcelona an hour before. He has a great turn of phrase and peppered his speech with neologisms, refusing to talk down to his audience. You could tell he was a writer from the delight he took in really using language. He gave a very interesting speech largely about the history of the “copyfights”. A copyfight being defined here as a situation in which a new democratising technology comes up against the power of vested interests (the printing press versus the Church, player pianos versus US government, VCRs and later P2P filesharing versus the entertainment industry).

The speech was full of vivid anecdote and with real detail behing the arguments. He is clearly a great force for the EFF and I’m glad that he is now their “man in London”. He also showed a real sense of humour. He reminded me of the page in Paul Arden’s It’s Not About How Good You Are, It’s About How Good You Want To Be: “Energy. It’s 75% of it. If you haven’t got it, be nice.”

Ravensbourne College is some kind of “design school” and I think very technically slanted. One guy was drawing a desert nightscape in Photoshop on his laptop while listening to one of the talks. Their domain is rave.ac.uk which shows some kind of “trendy” thinking. Quite how they managed to get these guys down there I don’t know but it was a worthwhile afternoon.

A colleague of mine last Summer decided he would learn C# rather than Java because, “that’s where all the jobs are.” I disagreed and showed him the jobserve listings for that day that seemed to prove him wrong. “Well, that’s where all the jobs WILL be.” he replied. Thus was born jobfight.

I think enough time has gone by (10 and a half months) to have a look and see if we can discern any trends. The graph belows shows the statistics I have gathered on UK programming jobs for the keywords “C#” and “Java”:

It seems that Java is growing faster in terms of total number of jobs and at nearly the same rate as a percentage of the existing jobs.

Paul Graham thinks Java “[smells] supicious”. One of his reasons is:

4. No one loves it. C, Perl, Python, Smalltalk, and Lisp programmers love their languages. I’ve never heard anyone say that they loved Java.

But I think some people do love java. In fact if you put “I love Java” into google you get approximately 2,530 results. “I hate Java” did get 768 results though. This got me thinking about how some programming lannguages are loved, like Perl in 1999 (before the backlash about it being “write-only”). PHP, then Python and Ruby emerged in a similar way from the hacker community and also seem to be loved. I decided to extend my unscientific test.

Results of putting “I love x” and “I hate x” into Google for various programming languages.

language	love	hate	ratio
ruby	1,550	76	20.39
python	820	82	10.0
c#	287	30	9.57
smalltalk	131	24	5.45
php	4,070	772	5.27
assembler	72	14	5.14
ml	30	7	4.29
cobol	43	13	3.31
java	2,530	768	3.29
perl	1,990	670	2.97
haskell	20	7	2.86
c	1,640	582	2.82
scheme	65	27	2.41
lisp	181	78	2.32
sql	164	98	1.67
fortran	17	18	0.94
prolog	21	24	0.88
c++	311	477	0.65
vbscript	17	30	0.56
javascript	179	432	0.41
vb	522	1,270	0.41

Some languages had to be excluded from the test because their results were just noise: Icon, Joy, Eiffel (apparently there’s a band of that name). Sisal, T-SQL and Intercal (at least) don’t really have enough information even for this rough test (a very small number of “I love”s and no “I hate”s making them appear to be the most popular languages).

Ruby’s outstanding result should not be taken too seriously as it was definitely more polluted by noise than a language like Smalltalk (no one says, “I love Smalltalk” unless they mean the programming language). Nevertheless a quick skim through the results seemed to show that most of them related to the programming language.

Three of the top five languages by love/hate are the “community” languages Ruby (1), Python (2) and PHP (5). Perl, however is way down at 10 (below Java). Of the other languages in the top five, Smalltalk was developed by a small group at Xerox Parc and the rapidly expanding C# is, of course, a Microsoft creation (unless you count the 10 years of prototyping done by Sun).

The languages that inspire much more hate than love are the perennial whipping boys VB and VBScript as well as JavaScript (hatred due at least in part to the terrible environment in which it usually operates – incompatible browsers) and (hilariously) C++ in which an awful lot of desktop apps are written.

Java comes ninth. That’s above Perl in tenth, C in twelfth and Lisp in fourteenth – three of Paul’s “loved” languages.

PHP has the most people willing to proclaim their love (4,070) but also a significant number of detractors (772). In fact, by volume Java lovers are the second largest group (2,530). Java has obviously picked up a lot of adherents since April 2001 (when the article was written).

If you liked this you will probably like the follow up, Does Anyone Love Java?

You may also find my jobs by programming language page interesting.

So you cancel a print job using Document, Cancel. But it just won’t disappear and nothing will come out of the printer (all queued behind the “deleting” job). Simple answer: restart the “Print Spooler” service. Either in Administrative Tools (Control Panel or Start Menu) or using net stop "Print Spooler" then net start "Print Spooler" at a command prompt. Easy, but it was driving me mad.

I’ve put together an article on the latest program I’ve been working on and the statistics on newlines, semicolons and other productivity indicators that I’ve kept while doing the work.

I was having trouble finding the correct syntax for the cref attribute of the see tag (and the exception, seealso and other tags) in .NET doc comments. I tried Namespace.TypeName.MethodName and Namespace.TypeName.MethodName(ArgType) to no avail getting the error “XML comment on ‘foo’ has cref attribute ‘Namespace.TypeName.MethodName’ that could not be found.” Eventually I found this in Google’s cache (not on site anymore) that made things clearer:

Use the cref attribute to link to a type or member or the langword attribute to specify a language keyword. The body of the tag is ignored. Cref attributes have the form: a one-letter prefix (N, T, C, M, P, F, E), a colon, and a value. Here are some example crefs:

• “N:” for Namespaces, example N:System
• “T:” for Types (classes, structs, interfaces, enumerations), example: T:System.Byte
• “C:” for Constructors, example: C:Gtk.Button()
• “M:” for Methods, example: M:System.String.Substring(System.Int32,System.Int32) (the argument list is optional)
• “P:” for Properties, example: P:System.AppDomain.CurrentDomain
• “F:” for Fields, example: F:Gtk.TreeIter.Zero
• “E:” for Events, example: E:Gtk.Button.Clicked

Common langword usages: <see langword=”null”/>, <see langword=”true”/>.

So I just needed to put “M:” on the front of my cref et voila no more compiler warning.

The original page, should it come back online, was at http://www.nullenvoid.com/mono/wiki/index.php/ECMAStyleDocumentation.

Everybody knows that Tetris is the greatest computer game of all time. The version I have played so much lately is Tetrablocks which is a very nice version. Sadly the speed of the game is tied to the speed of the machine you play it on which means scores are not really comparable across machines but if you can get more than 350 lines then you’re probably better than me.

The reason I mention Tetris is because there is a documentary about the game showing on BBC4 tonight (2100 in the UK) and replaying through the week.

One of my favourite Tetris things is this 12.8MB MPEG from the 2001 Japanese Tetris Championship (will take at least 8 mins to download from this site, as long as 40 mins over dial up). How fast is that guy?

And finally there’s a great site about Tetris AI.

Looking on eBay I can see that my tickets for the Pixies could be sold for up to 100 pounds (nearly 200 US dollars). Some of the buyers have decent feedback ratings so I don’t think it’s a case of people creating dummy accounts to mess up the touts.

That’s ridiculous. I bought four tickets for 115 pounds. I could sell them for nearly 400 pounds and I’d have done about 15 minutes work (including entering on eBay and posting the tickets) maybe 30 minutes if I include a walk to the post office to send the tickets recorded delivery. If I had 10 credit cards I could conceivably have bought 40 tickets. Then my profit would be 4000 pounds for not much more work, certainly less than an hour. If I did that for Reading Festival, Glastonbury Festival and any other events where tickets are guaranteed to sell out then I could give up my job and just work for about an hour or two every few weeks!

That leads me to two questions really. 1) What is going to be done about this? Neither organisers nor fans want it like this. And 2) Why don’t ticket sellers make their tickets more expensive if people are willing to pay that much? It’s surely not altruism – companies like Mean Fiddler or Ticketmaster don’t know what that means.