Colin's Journal: A place for thoughts about politics, software, and daily life.
There’s been no posts for a while, and there will be few over the next couple of weeks because of the Christmas break. A Happy Christmas to you all!
Unlike Dave we didn’t get snow here on Christmas day – the temprature was around 10C and it rained, but despite this it was a really good Christmas day. I’ve read the Night Watch (newish book by Terry Pratchett), almost finishing it Christmas day, but instead polishing it off today.
Here’s a fairly good description of what’s wrong with JSPs (via Hack the Planet). JSPs are exceedingly popular at the moment as part of the J2EE environment, mainly because they are so well supported and documented. They are also extremely powerful, and so programmers and architects will initially think that they are the best way of doing a web application. The HTML designers usually get no say in the selection of architecture, even for a web based application, and so all of the issues that JSPs have don’t come out.
What’s an alternative? The ideal replacement needs to be valid HTML so that the resulting file can still be edited/viewed in the standard tools of a web designer. It should also have a syntax that’s easy enough to learn that those doing the HTML design can easily incorporate the required elements into the page. The best example I’ve seen of this is TAL. I’ve actually written a new implementation of TAL, that isn’t dependent on Zope (or rather some of the python extensions that come with Zope), for this very weblog. TAL is nice because there is very little logic that can be included in the template, and the syntax of the template is non-intrusive to the HTML (only additional tags are used).
As you can see if you are reading this post I’ve updated my weblog and main page to a new look. The formatting is basically the same, but the new selection of colours should make reading posts much easier. It also looks much more proffesional (at least to my eye!) than the previous design.
There could be other candidates for the worst form of taxation possible, but the one described in this BBC article seems like a front runner to me. The idea is that it’s bad for university graduates to have an individual debt to pay off (i.e. a loan) and that instead it’s some how preferable for graduates to pay a special tax for the rest of their lives.
Now it could be that the duration of a graduate tax, as proposed, would be limited, but that in turn doesn’t make much sense. It would mean that the best thing to do after graduation would be to take a cheap job until this time limit expires, at which point you can get your high paying job and not pay the extra tax. Unless they tie the duration of the tax to how much you pay, in which case it’s effectively the same as individualised debt.
I used to be a big fan of free university education, after all I benefited from it myself, but it’s become clear that the current system is unsustainable. General taxation can not rise enough (and can not be justified) to cover the costs of providing a world class university education. The idea of some sort of graduate tax doesn’t work, and so the only real option left is loans that are tied to the individual that received the education. To make them palatable they should have only minimal interest, payments only triggered once a certain wage threshold is reached, and a limited life span so that they are written off if you still have them after 20 years.
Once this issue is resolved the next major step for the UK in terms of university education will be whether private universities can develop. From the point of view of generating more choice and giving the existing state universities some competition it seems like a good idea. Once the fees for the top state universities reach the levels that are required to keep top notch facilities and staff, the potential private universities will not be priced out of the market as they are today.
A good article on what happened to the cod in Newfoundland in 1992. It’s a story I hadn’t heard before – about how Canada (with a little outside help) managed to fish the waters so completely clean that the entire industry collapsed. The article is meant as backdrop to the current debate on setting the EUs fish quotas issue, and it does a good job of it.
The best piece of the article for me is a couple of quotes from a marine scientist:
“If you look at the data on the catches-per-unit of the trawler fleet, the highest ever recorded in this fishery were in 1992, when the stocks were on the verge of collapse,” said Professor Rose.
“So if fishermen are still saying they can find concentrations, that’s good news for now, but it should give no reassurance that you couldn’t take those last bits of fish down and push the whole thing right over the edge.”
It just goes to show that the “common sense” response (there’s lots of fish – see!) isn’t always superior to scientific advice.
A couple of interesting weblog posts on EU expansion (via Davos Newbies). I have to agree with the first on Turkey, as I’ve posted before, I see no reason to limit entrance to countries based purely on geography.
I’m not so convinced on the question over whether the EU will now become too large to continue along the federal route. The EU as it stands today is considerably further along the path to a true federal state than most people would have predicted at the beginning; after all todays members are not known for their history of peaceful agreement. With the enlarged EU I think we will see the same thing – countries will find themselves heading more and more in that direction (at least on certain matters), and end up going further than they would have initially signed up to.
On one front however I’m sure that we agree – the idea of financial harmonisation seems like a bad plan. Having some level of federal spending is not a bad thing, and I would not be surprised to see this element increase over time (it’s actually very small as a proportion of GDP), but keeping financial decisions at the state level will in the long run be for the best. This way each state can compete in terms of taxation, minimum wage (something promoted by the political right), and also in terms of promotion of quality of life through public services and regulation (something generally forgotten by the political left). I’m looking forward to the day that British politicians are willing to campaign on a platform of emulating another member states policies (on say health or level of taxation) rather than trying to invent something new as though these issues have not been tackled by anyone else before.
I’ve just read (a little belatedly) the latest Joel on Software. The point that the article makes is that to really know a platform inside out takes a huge amount of time and energy, and so the people that do know a platform inside out are extremely valuable. It’s all summed up at the end as “So for now, my advice is this: don’t start a new project without at least one architect with several years of solid experience in the language, classes, APIs, and platforms you’re building on.”
There is however another side to the coin – the people that know lots of different platforms and technologies reasonably well, rather than in depth, are able to adapt to new situations better than those that have concentrated on one area to exclusion. When architecting out a new product you need to pick a platform (or platforms) and then define the functional areas of responsibility (this bit talks to web clients, this bit talks to the databases, etc) and corresponding API layers. Determining coding standards and support libraries is also important (otherwise you end up with 3 different types of logging none of which can be configured) at this early stage. While it’s extremely helpful to have platform experts working with you at this level there is a danger if they have the final say because the vision they will layout is the one they have always used, not necessarily the one that is best for this particular product.
However the role of platform expert is far less important in integration work. When integrating different systems on different platforms you are almost always going to be working with things you have never seen before. In these situations the best kind of person you can have is one that has seen and understands how multiple different platforms work. This general background in multiple different platforms enables you to easily define in generic terms the terminology used by a particular vendor/system, as well as understanding the limitations as they are described to you. I still see plenty of room in IT for those that wish to deal with multiple different platforms – especially as the integration of existing systems or new systems into an established environment is the mainstay of the IT world.
A short series of slides from the BBC on poverty in the UK. The most interesting to me was slide 3 which shows the percentage of those receiving means tested benefits as a break down by category. The vast majority are composed of pensioners (46%) and the sick and disabled (29%). Lone parents come in at 19% with the unemployed coming in at 3%.
Those that can reasonably be expected to eventually get out of poverty through employment (the unemployed) make up a tiny percentage of those receiving benefits, compared with those that can not reasonably be expected to work out of poverty. A few things come to mind:
On the first point I’m not sure the measure used here is reasonable. Saying that anyone with less than 60% of the median wage is in poverty seems highly artificial. If the cost of living went down, and the median wage went up, the numbers would show an increase in the level of poverty. Admittedly if you do not use a arbitrary measure such as this you then have to make a judgement as to how much money is required to be deemed ‘above the poverty line’. Any such judgement is going to be difficult to make, but it could result in significantly more or less people being regarded as living in poverty and so change the proportions of those affected.
The second point is one of concern – what if those that are receiving means tested benefits are not those in poverty? For example there could be large numbers of people from any of these groups that do not receive means tested benefits and yet are classed as being below the poverty line. There’s no figures given by the BBC that answer that question, but if you find the original report you can see that they do not correspond in the slightest. The report quotes that “Just 4 per cent of households where all adults are working have low incomes, and 19 per cent of households where one of the couple is in full-time work but the other is now working. In contrast, three-quarters of households where all the adults are unemployed have incomes below the low income threshold”.
In fact the more you read through the report the more it becomes clear that the absolute levels and figures quoted are not that important (too many arbitrary assumptions), rather it’s the change in the figures over time that can tell you things.
The International Herald Tribune has an article praising Giscard d’Estaing and his handling of the European convention. I agree with the first half of the article – he has certainly ensured that all aspects of the debate are being heard after a long time when few would dare raise the topic. Their take on the question of admitting Turkey however seems strange to me.
If Turkey does join the EU (and I would like to see it happen) then indeed other countries in the region may well consider applying. Having a growing EU though does not in my mind dictate that the union has to be a loose coupling with only an economic scope. In fact the freedom of movement of people means that the EU will always be more than just an economic union. If other countries outside of (geographical) Europe wish to join, and meet the criteria for doing so, then it seems to me that it’s to the advantage of existing members to let them. If countries currently outside the EU do not want to be part of a tightly coupled union then they will not apply to join. The EU should be defined in terms of what it does; the number of countries that wish to be party to it will result from that definition, not the other way around.
My guess is that a lot of countries will see advantage in being member of a tightly coupled union if the EU is defined in terms of equal rights for all EU citizens, free trade, security, and co-operation. At the end of the day though the definition of the union should be in the hands of the current and about to join members, without worrying about those that may wish to join in some hypothetical future.
In a post last month I talked about the “Ransom model” for software development. This was partly inspired by the effort to purchase the source code to a text editor called pepper, and release it as open source. It appears that this effort has failed (via Hack the Planet) and that those who donnated are to have their money returned.
Email: colin at owlfish.com