Freesteel Blog » 2007 » December

Monday, December 31st, 2007 at 5:36 pm - - Whipping 2 Comments »

The United Nations publishes everything in six languages (English, French, Spanish, Arabic, Russian, and Chinese). Currently, my webpage only downloads the documents in English, and parses the speeches into plain text from the PDF files, as described in the How section of the FAQ.

It would be a beautiful idea if the whole process could be done in French. I don’t know much French, but I could work with someone who did and re-use most of the parser, and perhaps get it running in a couple of weeks of sustained work. I have already solved all the technical problems for English, and I don’t think there can be anything worse with the French language. (I would not say so for Chinese.)

After the French parser is working, someone in France, where they might take the United Nations a bit more seriously than here in England, could build a really good front end webpage for it, and make it completely bilingual so I can throw away my version.

An extra advantage is that many of the UN interventions are in French speaking nations in Africa and the Caribbean. When people who really know what is going on get to see what is said about them, things can change.

It might be possible to get funding for this project on the back of the work done on Statistical machine translation. This requires a “bilingual text corpora” where you have a large body of translated text of two languages that are both aligned. It has been reported that this has been done using the Canadian Hansard which is produced in French and English. But more could be better. It should be possible to get some university or company to very modestly fund the development of the French UN General Assembly and Security Council speeches parser which is aligned with the English text, to be used as input…

<pauses to surf the web a bit>

Bugger it. It looks like some people are well ahead of us

We collected a corpus of parallel text in 11 languages from the proceedings of the European Parliament, which are published on the web1. This corpus has found widespread use in the NLP community. Here, we focus on its acquisition and its application as training data for statistical machine translation (SMT).We trained SMT systems for 110 language pairs, which reveal interesting clues into the challenges ahead.

Also, there’s this claim:

Google used the United Nations Documents to train their machine, and all in all fed 200 billion words. This is brute force AI, if you want – it works on statistical learning theory only and has not much real “understanding” of anything but patterns.

I don’t believe it for the speeches, because an website would be an obvious byproduct. There are word files for some of the documents, but they are incomplete and sometimes not very high quality.

Sunday, December 30th, 2007 at 10:29 am - - Vero 1 Comment »

I am not an accountant, but how hard can it be? It’s only numbers that are supposed to add up, which account for money — the be all and end all of all business these days.

I happened to find myself on a very long train journey from Stuttgart to Berlin last week, and happened to have downloaded all of Vero Software’s annual report and accounts from 2000 to 2006, and read them. You know all those useless tables of numbers you normally flick through because you’re eyes have glossed over? Well, I started pulling some of the numbers out from them and putting them into another table which graphs the changes over time.

In the following table, £30k means £30,000. When two numbers are given, for example the £1007k or £955k for the 2001 development costs, the first is the number in the 2001 accounts, and the second is from the 2002 where they were restated for comparison, and the number wasn’t the same.

These discrepancies might be due to incompleteness of the main accounts, changes in accounting procedures, restatements backdating the values of company acquisitions for comparison, and so forth. If so, I can’t tell, because they don’t show the workings, and you would expect to see inconsistencies in the value of “number of developers”, which you don’t.

Some of the workings remain for the Kleban column, which refers to the amount of money paid to Kleban and Samor Attorneys in New York City for form filling, at which one of the directors was employed, so is declared as a conflict of interest. The 2000 accounts say they were paid £54k, the 2001 accounts say they got an additional £67k making a total of £121k, and the 2002 accounts report “The aggregate of fees amounted to approximately £74,500 (2001: £121,000).”

I did have the column for Acquisitions, but due to “amortization of goodwill”, the numbers are totally useless and all over the place. The rest of the columns are described below.

Year turnover development costs developers Other income Directors Babbs Kleban
1999 £4545k £614k 16 £142k £355k £81k £0
2000 £5216k or £5642k £1077k 22 £1k £418k £103k or £102k £54k
2001 £6128k or £6456k £1007k or £955k 21 £132k £525k £116k £67k (Ivy £14k)
2002 £6984k or £7542k £1282k 25 £458k £531k £138k £74.5k (Ivy £21k)
2003 £8823k £1350k or £1450k 30 £282k £495k £143k £75k (Ivy £43k)
2004 £9698k £1339k 30 £228k £479k £146k £36k
2005 £9936k or £10196k £1446k or £1268k 31 £95k £520k or £531k £161k £26k
2006 £11027k £1471k 37 £75k £563k £166k £1k

The columns are:

Directors – the total pay of the board of directors, which is between 5% and 8% of the turnover per year. Officially they are paid this much to stop them going off to get jobs elsewhere in the industry, at great loss of talent to the company. Since they have all been there for years, it doesn’t look like there has ever been a threat followed by a bluff-calling. The highest paid director, Babbs, has his own column.

Kleban has already been described. Ivy refers to money given to another director’s financial company for financial bureaucracy. More about this some other time.

Development cost and developers is the investment and number of staff in that part of the company. In business, there’s usually a conflict between the customer’s interests and the shareholder’s interests. Put simply, if you buy software from a company, the only thing you get are what the developers have done for you. That and part of the sales support, I suppose — the ancillary materials. But, for example, if the company blows half its turnover on a beautiful headquarters building, or Jaguar cars for the directors, you don’t get those. The development costs tell you what the cost price of the product is.

For me, the most interesting column is other income, because here the numbers really don’t add up at all. There is a paragraph in the accounts that says:

Government grants are credited to the profit and loss account as other income to match the period in which the costs have been incurred provided that the grant has become receivable and there is reasonable assurance that it will be received.

Now, I happen to have tracked down a statement about Eureka grant E!2551- VISI-XX which reports 6.1million euros split 70%-30% between Italy and The Netherlands between 2001 and 2003, listing Vero Software as the main participant. The company was originally founded in Italy before it moved to the UK. If you do the sums, it rounds to about £3million over those two and a half years.

So where has all that money gone? The other income column certainly does not add up to this sum. I am conducting further investigations.

Thursday, December 27th, 2007 at 7:21 am - - UN, Weekends, Whipping 2 Comments »

Well, I’m here in Berlin. I’ve put back on line for the duration, in case I meet up with anyone. I don’t know if I will, but you never know.

Monday, December 24th, 2007 at 11:27 pm - - Machining, Whipping

Found in the business newspaper clipping section of Dublin public library circa 1992:

Britain’s machine tool industry has been decimated by the loss of business resulting from embargoes of Iraq.

In 1986 the UK machine tool industry had no exports to Iraq. This was before the Department of Trade and Industry sponsored British companies to take part in Iraq’s first international trade fair following the end of the Iran-Iraq war.

In the space of two years UK machine tool manufacturers were enjoying sales of £31.4m, promoting Iraq to number three after the US and West Germany in the table of Britain’s machine tool export markets.

Most of Britain’s machine tool makers had exports to Iraq at the time. Many of the exports were being used for the manufacture of munitions.

Seven years later, with the iraqi market destroyed and with the refusal of the government to countenance a new market in Iran, the UK survivors of the high technology end of the industry can be counted on one hand.

Matrix Churchill, the company with the biggest Iraqi orders, is in receivership; Wickman Bennett, of Coventry has also ceased production. It agreed a fine with Customs and Excise over exports to Iraq.

Mr David Phillips, an industry consultant, estimated that two thirds of the UK machine tool industry, representing annual turnover of about £600m, had been wiped out in the Iraqi debacle.

I guess those were the days when the machine tool industry was a political force to be reckoned with. (Before my time, before I even knew what a machine tool was.) Does it explain why there is an unlikely concentration of machine tool software companies in this country? These things get established, and persist for decades.

Anyway, the story from the article corroborates with some Parliamentary questions. For example, from 3 December 1990:

Mr. Dave Nellist (Coventry, South-East) : Leaving aside how British troops in the Gulf must now feel towards the hon. Member for Plymouth, Sutton (Mr. Clark), and the fact that the occasion of this statement is not only a good occasion for his resignation but another good reason for not drifting senselessly towards war in the Gulf, how can this Minister now face the 89 workers and their families from Matrix Churchill in Coventry who this weekend were given notice of their redundancy, which will take effect four weeks today? The Minister seems to have done nothing in the five weeks since I led a deputation of workers to him in his office. Is it not a fact that his Department encouraged general exports to Iraq, including machine tools and lathes from Matrix Churchill and Wickman Bennett? Is it not a fact that in November 1988 the former Secretary of State for Trade and Industry announced £400 million in credit to enable those exports to be made and that 10 days before the present Secretary of State for Health said in the House that the Iraqis were using chemical weapons to bomb the Kurds? Is it not a fact that the Department allowed the Iraqi secret service to take over Matrix Churchill? Is it not also a fact that, after sanctions were imposed in August, the Government washed their hands of the responsibility for workers in Coventry and elsewhere? As it is all the Government’s responsibility, should not the Minister announce today that the Government will take over the firm, take it out of the hands of the Iraqi secret service, and guarantee the jobs and livelihoods of the workers in Coventry and elsewhere?

The Minister for Trade (Mr. Tim Sainsbury): Job losses and redundancies are always a matter for regret and I am sorry to hear what the hon. Member for Coventry, South-East (Mr. Nellist) said about some of the workers in Matrix Churchill. The hon. Gentleman read out a list of “facts”, most of which I did not recognise as facts.

And also, from a written question 24 November 1992:

Mr. Robert Ainsworth: To ask the President of the Board of Trade if he will make a statement about his Department’s handling of Wickman’s export licence to Iran.

Mr. Needham: No. This Department does not make statements in public about individual export licence applications.

Bob Ainsworth is now a Minister of State in the MoD where he has deplored the alleged supply of weapons from Iran over the border in Iraq where they are used “kill our people”.

Presumably, as long there were no British people getting hurt, and the manufactured munitions were killing no more than Iranians or Iraqis (depending on which side guaranteed our jobs), it was perfectly all right.

Monday, December 24th, 2007 at 9:50 pm - - Weekends

Finally got out of Tuebingen for a day and a half. It’s been three solid weeks programming constant scallop routines, with Becka building her plastic models, and both of us getting on each others’ nerves. Becka used every excuse to avoid going, but it had to be done. We’ve traveled thousands of miles to go to the snow in previous years, and here we were with car, Alps a couple of hours away, and no more excuses.

After a bit of looking at maps, rejecting Switzerland because we didn’t want to bother with a different currency, I chose Oberstdorf as the most likely place for not having to cross any borders, go very far, but get to something mountainous. One of the local cavers recommended Tannheim, just across the border into Austria, and not much further.

One thing that’s very handy are these panels of webcams so you can see the conditions.

Anyways, we got there yesterday, narrowly beat the lunch hour at one cross country ski hire place, and headed straight up the valley to Vilsalpsee where we saw some people skating out on the lake from the side and wandered out ourselves. There were all these amazing crystalline cauliflowers of ice on the surface. Later that day we got bitterly cold, and went for a long walk in the dark before finding some place to eat. The ceiling in the restaurant was decorated with what I recognized to be dried sea fans. I’ve never seen that before.

The skiing was not very extensive, and we covered most of it during the next day. Some dog sleds went past.

Monday, December 24th, 2007 at 9:01 pm - - Kayak Dive, Weekends

Over the weekend of 25 November, Becka and I made it to Dublin to stay with a friend who has recently taken up sea kayaking. I have finally downloaded the pictures a month late.

Here is Becka showing how to capsize and not stay under for long enough to get rescued. It worked the second time. I was astounded at the slickness of the procedure for emptying the boat (assisted by water-tight bulkheads fore and aft of the cockpit whose existence I didn’t know about), and re-entry (you lift your weight against the paddle straddling the empty boat and slip your legs in).

The next day we did a kayak dive in Dalkey Sound, it feeling being very exposed to go any further out, and having found a deep bit close inshore to the cliffs (about 18metres). The current was strengthening, Becka got tangled up in fishing line, but got free before I managed to untangle my knife from its holder and then lose it for good trying to put it back (I think this needs some attention). The visibility was similarly atrocious (though there was a lot of life), and the best bit was getting back out and on land.

We even went and sat in a cafe for an hour with a big cup of hot chocolate, which is unusual. Turned out we were both coming down with the winter vomiting virus, which accounted for the slackness.

Kayak diving is always a bit on the edge, and I will be ordering a marine radio for after Christmas. Still, there’s much worse nutters out there, as I heard on the Sea Kayak podcast on the way home, where some guy paddled solo to St Kilda, Shetland, and all the other Scottish islands in one go, and his rescue plan was an EPIRB. Since it takes a few hours to recover the body out at sea (living or dead) when its position is accurately known, and he wasn’t wearing a drysuit (not practical for such long durations), he had a diving wetsuit which he was going to somehow climb into while bobbing out at sea and not so gradually freezing to death. My immediate response was: “It’s been a while since he’s tried putting on a wetsuit.” Even on land it takes a lot of effort.

The other notable thing about sea kayak camping over long durations is that the food is absolutely terrible. They’ll proudly boast how they ate pasta and soup for weeks on end. Maybe being out at sea is such an overwhelming experience you don’t notice it at all.

Saturday, December 15th, 2007 at 10:34 pm - - Whipping 1 Comment »

I’ve been feeling a bit low today, not well enough to focus on my desperately behind machining work, especially after a walk down into town (Tuebingen) with a bad cold. So I’ve cheered myself up (it’s funny what one does to wind down) by editing the Freedom of Information Bill (2000) Publicwhip policy using my newly devised formulation for describing everything in terms of the “Majority”, which has a meaning, rather than in terms of “Aye” and “No”, which are generally arbitrary and misleading (it depends, for example, on whether the motion was: “That the original words shall stand”, or: “That the amendment be made”).

The Parliamentary debates about the FOI Act in 2000 bear little relation to the issues which I have encountered with the Act, which are to do with cover-ups for anything that is “commercially confidential”, which seems to apply to privatizations that are funded 100% by taxpayer money (for example, contracts for buying software and computers for schools). During a lot of the debate MPs were worried about the exemptions for information that would prejudice the conduct of public affairs, which would allow the Government to say “if we told you that, then it would make it much harder to get free and frank information from experts in the future.”

Anyways, here is the panel of comparisons between the MPs and the pro-FOI policy. The Tories come out a lot better than Labour, because they kept voting for tiny issues which would have strengthened the Act, and Labour always (with those few shining exceptions which you can now see) rejected them. The Tories never had time for such legislation when they were in power. Had there been some votes on the less controversial issues, such as whether there should be Freedom of Information at all, they’d all be brighter, because it was bipartisan, popular, and long overdue.

That’s enough waffle about that. What got me started on this road yesterday was the discovery of the Communications Allowance which was first voted for in Parliament on 1 November 2006, and finally passed, following the report by the committee, on 28 March 2007, just days before it came into force on 1 April.

The new rule caps their pre-paid envelope allowance to £7,000 per year, but gives them a new Communications Allowance of £10,000 per year. This is in addition to their Travel Allowance, Staffing Allowance, Staff Travel Allowance, Centrally Purchased Stationery, Centrally Provided Computer Equipment, London Supplement, Additional Costs Allowance, and Incidental Expenses Provision.

The new Communications Allowance alone can reach up to £6.56million in real money per year.

I have helpfully included most of the guidelines into the motion text regarding that division. Go read it. While looking around, I managed to find a report by the Select Committee on Standards and Privileges (15 November 2007) titled Conduct of Mr Elfyn Llwyd, Mr Adam Price and Mr Hywel Williams. Oh dear. It seems they were caught placing advertisements…

“in newspapers circulating in their constituencies in the week preceding the recent Welsh Assembly general election. Those placed by Mr Llwyd and Mr Williams appeared in the Daily Post, a daily paid-for paper circulating in North Wales, on 26 and 25 April respectively. Those placed by Mr Price appeared in the issues of the same week of three weekly paid-for newspapers circulating in various parts of his own constituency but also outside it—the South Wales Guardian, the Llanelli Star and the Carmarthen Journal. In each case, the advertisements consisted of two full-page spreads, with English language and Welsh language versions of the same advertisement on facing pages. They are reproduced here. All the advertisements were paid for out of the House of Commons Communications Allowance.

Also, their use of their party logo was not “proportionate and discreet”. Consequently, they were told to pay their claim for expenses back.

During the debate over the allowance, Jack Straw said:

Websites that are funded from the communications allowance should be used only for parliamentary purposes. To ensure that there is much better control, we have decided to make the rules clear and to say that it will not be acceptable for Members to allow publicly funded web pages to be contained in another domain or website, and vice versa. Members will have a parliamentary website, and they may have separate websites if they wish to fund them from other sources. Parliamentary websites will close down when Parliament is dissolved after a general election is called.

A website must comply with the same rules on the content of material as any other publication, and it must contain on the home page a statement that it is funded from parliamentary allowances. Any links to other sites must make it clear electronically that the reader is leaving the parliamentary-funded website. However, websites are becoming increasingly sophisticated and it is important that the MEC (Member’s Estimate Committee) should have asked the DFA (Department of Finance and Administration) to monitor website content so that the MEC can be assured that the allowances are being used for the stated purposes.

I’m not sure how this relates to the issue that with websites the more different ones there are, the less they get seen — as opposed to to hard copy documents, where the more there are, the more you can see them piling up.

But let me cut this short. £6.5million is a lot of money to be giving to MPs when they have no clear idea what to spend it on. How about an MP putting some of it towards improving the content on Publicwhip?

The software is fine and ticking over with work I do now and then, and the cost of the server, which gets used for a lot of other work, is covered with about £1000 a year. What we really need is help with the Motion Text Editing. In other words, all that work I just did for the Freedom of Information Bill 2000 turning gibberish like this into something more legible like this, which takes a great deal of time, research, skill and knowledge.

Why can’t some MPs get together and pay someone to do this work part-time in order to turn these Parliamentary motions into something that makes sense? I don’t want to see the money. They merely have to spend it on a staff assistant or two who can call me down to London to help train them how to research and edit HTML.

Then this data can feed through to all sorts of other places, like their Parliamentary webpages and so forth. It all helps to promote the understanding of Parliament, which is about the votes in the division lobbies that ultimately sets the political course of this country, but which the public is generally pretty ignorant about.

Some MPs definitely want to keep it this way (for example, the whips). But some others may not (for example, those who vote in accordance with overwhelming public opinion, and not always the way the whips tell them to).

Come on, folks. There are only 200-300 Parliamentary votes per year, rarely more than ten per week (usually much less). To edit a motion takes between 5 and 30 minutes, depending on the complexity. That’s less than three hours of work per week. Extra billable and accountable (we can exhibit the diffs) cash in hand for your staff, and you won’t get into trouble. How about it? Please contact for more information.

Sunday, December 9th, 2007 at 9:22 am - - Machining 6 Comments »

A couple of my colleagues found out about VoluMill last week, but didn’t pass on the news, assuming that I’d know already, seeing as I try to keep on top of these things. So I found out about it during Euromold, and sent them an email. Haven’t heard back yet, mind you. I guess I’m too unimportant. I’ve included a mention of them in the FAQ. No mention of Adaptive Clearing appears on their web-page. Either we’re totally irrelevant, or they believe people won’t find out about it if they don’t link to it. Ever heard of google? At some stage, pretending things don’t exist when they obviously do stops looking clever.

I don’t have time right now to post onto their new User Forum myself, but it remains to be seen if a discussion about Adaptive Clearing there would get cleansed faster than erectile dysfunction spam. I mean, what else can people discuss in a User Forum? Things really don’t take off in this Web 2.0 world without a critical mass. I would advise replacing it by a blog with the comments turned on, as we have here, as the only way to get a bit of constant life.

This year at Euromold people are beginning to talk for a change. We’re getting beyond the childish stage of:

“Oh, you’re a competitor. No, we haven’t heard of you. You’re far too rubbish and insignificant for us to have wasted any of our precious time thinking about you. So, no, of course we haven’t got any questions we have been wondering about, which you might know the answer to. Now go away before you find out any answers to your questions which might give you a competitive advantage. Really, we do not care at all if the entire CAM software industry completely stagnates as a result of no one talking.”

I have noticed that the VoluMill story has an interesting relation to the software patent campaign that was on-going a couple of years ago in the EU, where I testified as a programmer about how it worked to the utter detriment of everything to do with software development, citing as my example the Surfware TrueMill patent application.

I remember commenting that patents worked totally against the interests of the programmers even within the company that took out the patent because it would prevent them from applying — or threatening to apply — their relevant skills under other employment conditions. Under no circumstances would a programmer ever cooperate with a software patent application if they knew what it was like to sell out their own future.

The name “Glenn Coleman” shows up on the Surfware TrueMill 2005 patent application, now appears as the “Chief Product Officer” of “Celerative Technologies” which has developed VoluMill. Obviously, the website claims it’s a brand-new strategy, but it looks similar enough at a glance for a lawyer to cause a great deal of aggravation, should he be paid to do so.

Doubtess, the Celerative folks have taken this into account, having spent a long enough in the business to know that nobody in the CAM industry is in the habit of doing anything with software patents… yet. We’re all good guys, in that respect. Unfortunately, that doesn’t take account of how these things work. There’s nothing to stop some totally evil patent troll sensing an opportunity and dropping by the Surfware offices in Westlake Village with cash, and a promise to help get their own back. The troll buys the patent and then goes back to his air-conditioned offices where nothing of any good is produced, and starts suing the ass off Celerative until they settle for some sizeable chunk of their investment budget in a process that is practically indistinguishable from racketeering.

Meanwhile, I’ve got a lot of stuff other to do (having now entirely broken the scallop strategy in the machining kernel in a process known as “break to fix”). At some point in the future I’ll compile a detailed review of the VoluMill strategy, as well as a few notes about the areas of the business plan that are going to be hard.

Saturday, December 8th, 2007 at 12:07 pm - - Machining, Weekends

The incomprehensible journey began last week from Liverpool, via Kings ‘ Lynn to drop in on grandfather and Lancastria survivor, then stayed the night at a friend’s house in Orpington outside of London. This was where we picked up Martin in the morning.

A couple hours later as we drove through Dover there was a horrible rattling sound coming from the car engine compartment which caused me to pull over immediately and suffer lots of unnecessary friction over likelihood of missing our booked ferry times — which was of no consequence anyway because the ticket turned out to be valid for any time that day. There was oil dripping onto the ground.

By the time the breakdown man found us, we worked out that it was probably the power steering pump which had failed. Not enough the stop the car from driving once you took the “fan” belt off (it had to be cut off since all the de-tensioning bolts were stuck). We could keep going for as long as the battery held out and the ferry company didn’t notice any problem. We had the option of staying in England to get the car fixed the next day, or attempting the same in Dunkirk, France where we didn’t speak the language and didn’t know the score.

Owing to the score in England being so poor, we took our chances and got help from a random friendly French guy in the ferry car park who phoned ahead to a garage and also guided us through town using his GPS. The night was spent in the empty cheap hotel next to the full cheap hotel we had heard of (brands work when you travel away from home) which charged us 35 euros for a three person room with free WiFi and a convenient hypermarket across the car park. This was more like it.

I’m not sure why they felt it necessary to install floodlights to shine directly into our window, though.

The repair bill set us back a lot of money, most of it for the genuine Citroen part (not some refurbished unit which would have been more than adequate) — a stupid power steering pump which I don’t even care about, and could have been solved if I could have bought a shorter grooved fan-belt that avoided it, or cleaned out the entire useless mechanism of the broken pump so that all that was left was a wheel.

We got away at midday, and reached Wiesbaden where one of Martin’s brothers lives in a house built on an allotment in time for an excellent supper. Then Becka and I headed down to Tuebingen in the dark through the rain on the German autobahn with no cat’s eyes or even apparently any of glass beads included in the paint so you could see the white lines from more than five metres away.

It was a hard drive.

Becka had a room booked in a hostel connected to the Max Planck institute. Since we were going to arrive so late, they left the key in a combination safe by the front door which was opened by pulling out and twisting the plastic handle 180 degrees — but not pulling so hard that you snapped it off. I should carry a pair of pliers in the car for the next time that happens. Oh well. We got in after 20 minutes of forcing the stump.

I caught the train back to the north at Frankfurt where the Euromold trade show is every year, and failed to rendezvous with Martin at the train station, so he waited there for a several hours. I don’t have a mobile phone, unlike everybody else in the world, and have noticed that pay-phones are getting more and more broken, rather than clean and pristine because nobody is using them anymore.

I gave up, paid my way into Euromold for a half-day ticket (last year there was a side-exit you could slip through), found some Cimco people, got them to phone Martin who was outside, borrowed one of their exhibitor’s tickets, went back and handed it through the railings, and watched him get busted for trying to get in on a ticket that had been used once that day but not checked out.

He got through at another desk, and we were able to collect our exhibitor’s passes later that day from our friends who were already in there at Hall 8.0. Everything this year was the same as last year, and the year before. The lay out of the companies stays the same. These trade shows seem more and more ridiculous the longer I go to them and recognize the sheer futility, having got past the naive belief that somewhere somehow it makes sense to someone but I don’t know enough to see it.

These trade shows do perform a vital purpose, different to what people usually say they are which is little understood. You don’t need to believe it for it to work. Analyzing it is like analyzing the purpose of the manned space programme. There isn’t one. The reasons for the establishment are often forgotten and out-dated, and retrospective justifications on the basis of unlikely and irrelevant spin-offs don’t really get to the heart of why it persists now. There’s more of a social purpose to them, like a public park in a city whose justification can’t be made on pure economic grounds. If we got to the purpose directly, the whole thing could be done a heck of a lot cheaper and more effectively. But that’s another rant for another day.

Wednesday, December 5th, 2007 at 9:55 am - - Machining 2 Comments »

For anyone who really wants to look at that book I 5% wrote back in 2002 (mentioned in my last post) between the time of leaving NC Graphics, and Cimco discovering that hunting down the actual programmers what did the work could be a lot more useful than talking to the MD who bought and sold him like a used car, it’s here in glorious LaTeX generated PostScript. (Use Ghostscript to read it.)

This was done before I discovered Python. The examples are all in Java, and so was my system for generating the diagrams in TeX, which were also created using lots of Java. I’ve never had a good time with graphics programs and always seem to make a lot more success simply hacking code that outputs basic pen width, w; move x,y; draw x,y systems.

Most of this is badly written, so I’ll make my apologies now. It was the diagrams and the topics as I was developing them that interested me. I gained a lot of respect for textbooks that contain useful drawings. Maybe if I re-implemented my drawing system in Python I’d be able to use it here. At the moment, I’ve been doing drawings in Microsoft Paint (I told you I didn’t get along with drawing programs), because I have still not found out how to do a plain straight line in Gimp. Maybe I’ll get back to SVG soon.

The bit I found interesting was the different ways of representing subsets of the real line for the purposes of computational geometry. There’s a lot of work on 2D and 3D structures, with quad-cells, oct-trees, BRep models, and so forth. But the 1-dimensional case is important to me because it is simple enough that there is a good chance we could categorize every single computational model for it. I found at least half a dozen, and they were usually an analogue of a familiar geometric structure used in higher dimensions.

The question was, was this all of them? Could I find such a model buried in a mess of code and say what it was?

One thing we don’t have in computational geometry is a clear idea of concepts. A course in the field is usually a list of examples, not how they fit into a bigger scheme. When you’re doing work in the private sector, you simply make these models up — whatever first comes to your mind — without much of a thought for the alternatives. You don’t know the alternatives. And if you do, you can only guess according to prejudices how hard of easy they are to program and maintain. And these ill-informed choices early on in the lifetime of the software constrain it utterly.

The taxonomy of geometric models determines how software develops. For each model, there can be several implementations of varying optimality. Conventionally, software development moves between these implementations, leaving the underlying model intact. Sometimes a small tweak can change one model into another, if they share enough characteristics. This gets development into a whole new zone for moving forward.