Is this thing on?

It doesn't really matter much these days which browser you are using. On the whole I find Opera less annoying to use than Firefox, and Opera additionally has some nice perks like better keyboard access and Opera Link.

Once in a while there are things that really makes a difference. On top of my list is that a browser should never lose my data. On the whole Opera is pretty good in this regard, but it has a huge gaping hole in its armour. If you type in text and the window is closed, you can restore the window but the text is lost. This is almost adding insult to injury, as you can see the empty space where the text you worked so hard at used to be. This excruciatingly horrible bug has a name, 155102, and has been known for a while now, but for a number of reasons it has taken time to fix.

The workaround is simple: Never put anything valuable directly into a textarea. I have used Notes, I have used mail compose windows, I have used external programs. They all have their disadvantages. Copy to note (Ctrl+Shift+C) is particularly convenient and very robust, but you end up with your notes full of half-written texts, and with Opera Link those texts are synchronised to every Opera client you have, which in my case is quite a few. Opera Mail and external programs often mangle whitespace. The biggest problem with all workarounds is that you don't always know you're going to write a long text, and the web page text entry boxes are alluringly convenient.

I am fairly active in the forums, and while I could do the workarounds above, or make sure to preview often, which usually does the trick, in practice I am too lazy to do either. So when I lost and had to retype a longer post yesterday, and lost another one today, I decided I couldn't wait any longer, so I downloaded Firefox 3 and Opera 10 alpha. Firefox is reported to restore user entered text, though when I tested it it didn't work. Maybe I have to find some particular extension. The bug isn't fixed in 10 alpha either, so I am back to Opera 9, looking for a better way out. Suggestions would be appreciated.

The Masters of My Opera could add an AJAXY Javascript to automatically save the posts-in-writing, but those scripts are fairly intrusive and liable to fail with different browser configurations, as well as slowing things down. This really ought to be done by the web client, only that none seem to do that now.

New Scientist recently published a special issue, The Folly of Growth, ostensibly an expose of flawed economic theory.

As you can deduce from this blog and writings elsewhere, New Scientist is my favourite magazine, and I have read practically every issue the last twenty years. While generally of quality there have been some duds among the articles and issues. This issue is among the worst.

New Scientist has changed through generations of editors, but retaining many of the strengths and weaknesses. It has never done economy well, and most of the time it has been weak in information technology, doing better these days.

Whatever your viewpoints this issue was bad economics, anyone reading it would not learn anything about economy from it, and whatever they learned would be more likely to be wrong than right. Through a Global warming thread in the forums I was pointed to a rebuttal from The Register (of all places), as the errors were basic and numerous that saves me some time.

What dismayed me more with this issue was bad science. I would love to see a scientific outlook on the "dismal science", where you can find much folly indeed. New Scientist here wasted a good opportunity. I know how hard it is to make a special issue work well, you depend greatly on the contributors and the editors to turn the disparate articles into a coherent whole, and reject the articles that can't be improved. I don't know who declined to participate, but of the contributions several articles should never have been published in New Scientist.

Several contributors were recycled from the old-style ecological movement. This in itself was bad judgement, they have repeatedly been shown to have poor domain expertise. Traditionally ecologist have had poor grasp of economy and vice versa, which can seem odd given that both are working with complex dynamic systems that in many ways are similar.

I would pick out Tim Jackson's article Why politicians dare not limit economic growth as particularly badly written, he should stick to writing letters to the editor, or at least do some research. I pick on him for an article with no substance and horrible style.

Originally posted by Tim Jackson:

The message from all this is clear: any alternative to growth remains unthinkable, even 40 years after the American ecologists Paul Ehrlich and John Holdren made some blindingly obvious points about the arithmetic of relentless consumption.

The Ehrlich equation, I = PAT, says simply that the impact (I) of human activity on the planet is the product of three factors: the size of the population (P), its level of affluence (A) expressed as income per person, and a technology factor (T), which is a measure of the impact on the planet associated with each dollar we spend.

Take climate change, for example. The global population is just under 7 billion and the average level of affluence is around $8000 per person. The T factor is just over 0.5 tonnes of carbon dioxide per thousand dollars of GDP - in other words, every $1000 worth of goods and services produced using today's technology releases 0.5 tonnes of CO2 into the atmosphere. So today's global CO2 emissions work out at 7 billion × 8 × 0.5 = 28 billion tonnes per year.

I in particular take exception to this language, I don't expect to see arguments by the "blindingly obvious" in New Scientist, this is a scientific journal, not a religious one. If this had been a post in our own Debates & Discussions he would be promptly shot down.

This is not to say that there was no thought-provoking content in this issue, but what was there to be found was badly edited and hidden by the chaff. The editor of New Scientist, Jeremy Webb, may right that NS should be an ideas magazine and not do proper science the Custodian of Truth. The open questioning approach of NS is why I am a fan of the magazine after all, other magazines tell you what science "has discovered", no questions asked. The Folly of Growth had preciously little opennness however.

As an example of an almost-good article asking an important question I would highlight Does growth really help the poor? Unfortunately the article is only half-done, but it was a good start. Fundamental questions like this, testing whether the economic system actually achieves what it proclaims to do, would be within the domain of New Scientist to ask. Unfortunately they forgot to be a science magazine this time.

Some nights ago, at a pub in Žižkov that used to be called U smrku (The Pine), I was chatting with a friend while not listening seriously to the songs being played. I asked her what would be the ten most decadent songs in history, without really coming up with suggestions of my own.

I tentatively put up "Crazy frog" (Axel F) in position #8, but I am really not good with lists, particularly not with music lists. Maybe you can come up with some suggestions?

I am interested in the concept of employee ownership, and early on while working for Opera I set a goal of a 1/1000 ownership. I never quite reached that goal, the rest of the world has more money than I do and I have not been willing to borrow money to invest in shares, but still I am more economically exposed to the fortunes of Opera Software than anything else I own. So when Opera posted excellent results, the market reaction made me more paper money in a day than I can expect to make in real money in this year. ;cheers: I'd say.

This 17% rise in a single day and a similar fall earlier when a disappointing result was posted is fair enough, this can be seen as a reaction to real performance by Opera, but most share price movements have not been a result of changes in Opera's profitabilities or potential. My experience of Opera Software is as a serious longterm company, my experience of Oslo Stock Exchange (OSE), where Opera is listed, is the very opposite. OSE has had a terrible reputation in the past, and there is no indication that it is any better in the present. In my view OSE has been, is, and will be gamed, and there is not much to lead you to believe that the Norwegian government will apply the same rules of transparency to it as it does to itself. OSE is no more trustworthy than most emerging market bourses.

As an Opera employee I reflected that it was probably possible to make (or lose) more money on short-trading the Opera stock than what it would be as an employee. Due to insider trader consideration short term trading was inconvenient to say the least, but it was pretty obvious that the market had little understanding of the consequences of an event, reacting strongly to an inconsequential event and ignoring an important one. It was also a useful lesson in why not to invest in the company you are working for, not so much due to spreading risk as due to the delay in buyng/selling stock.

I could go into trading, at least on a hobby basis. Two things hold me back, high transaction costs and a distrust of OSE. That said, longer term fundamentals always matter, some of the time.

This post is a reaction to xErath's comment to an earlier rambling.

Originally posted by xErath:

There's enough info online about the deficiencies of ooxml. I don't need to quote them here.

Just to have it over with, we agree. I may be a Microsoft groupie at heart, but I don't have time for OOXML. People who want to discuss OOXML should feel free to do so, elsewhere. More interesting is this part:

HTML5 is a big patch over HTML4, made to be backwards compatible with the complete mess that is the web.

ODF on the other hand is a clean approach to define an abstract model of document formats. Then ODF has many features that html5 will never have nor will make any sense. I doubt html5 will ever include native spreadsheets support, slide transitions, page footers and headers... you name it. ODF will not support local databases nor globalStorage.

First, I have come to realise that the phrase "it is a complete mess" is a shorthand for "it works very well, and I don't think it should". In fact the Web is a remarkably sane place. We have millions of monkeys making Web code every day, and it works. The Web is more like an ecosystem than an engine, you don't stop and "repair" it, it is survival of the fittest. If you look at Web code written today and ten years ago the old code was worse and still achieved less. The Web has evolved and it keeps evolving.

Read more...

That should more properly be "having left Opera", as I am no longer an Opera employee, but that doesn't work as well as a title. There is nothing dramatic about the decision, and I think Opera is an excellent employer, but I had been at Opera for seven and a half years and that was long enough.

Much of my work at Opera has been with standards, and standards don't matter. Having standards does, but as long as they are reasonably sane it doesn't matter what they are. The latest standards debacle in Norway wasn't related to browsers but with office products, word processors and the like, with the competing standards called ODF and OOXML.

ODF is not a good standard. You can read through the entire spec and will find nothing clever there. Anything ODF can do HTML5 can do better. Add cursor position to HTML5 and it could have been called ODF 2.0. What it has going for it is the absence of bad. Microsoft makes good standards, much of the time. OOXML is not one of those times, what it is lacking is the absence of bad. Could it be fixed? Probably, but to me it isn't worth it.

As for Web standards I think it should be an optional for Opera. Opera should encourage the presence of standards, and follow them unless they are bad, but it shouldn't necessarily form them. Opera should do what makes its users happy.

Opera Mini makes me happy. It lets me do things I couldn't do before. This entry was intended to be typed in on Opera Mini while I was on the move, but in the end it was typed in on a PC. It wasn't written in Opera Mini because Opera Mini isn't data loss safe, without copy&paste or save I can easily lose what I write, and data loss does not make me happy.

Being unemployed makes me happy as well, for now anyway. It's been a long while since the last time, as the last few times I changed jobs I went directly from one to the next. It is almost the same elated feeling as being homeless. I haven't actually been homeless in the sense that I own one flat and rent another, but I have adapted to a mobile lifestyle and from time to time I've not known where I will spend the next night and that is a strong feeling of freedom (until nightfall) — I can go whereever I want. I have used to claim that the bag in my one hand is my office, containing my laptop and other work stuff, and the bag in my other hand is my home, containing clothes and other private stuff. For now I can move with one bag less.

There have been many language threads and digressions; I should know, I have participated in most of them. Maybe it's time to start talking about talking: What is language, where did it come from and for what reason? How do languages compete, cooperate, coopt each other? Where are they going? Is one language better than another? What about dialects, sociolects, idiolects, jargon?

Follow the discussion here

And so it begins... On the circuitous route from the outskirts of Europe to the centre, going from Oslo through Linköping and Wrocław, Opera has finally reached Prague.

Read more...

This is my week off for the upcoming party, and it is really getting there. As am I, of course.

The CSS 2.1 specification was recently refused entry into the (almost) safe harbour of Candidate Recommendation (CR) status again (for what is it? the fourth time?). This means that the spec is still in a you-can't-believe-a-word-of-this Working Draft state, just like it was seven years ago. This can be summed up in the word frustrating.

To be fair the goalposts have not only been shifted from when Tantek Çelik originally suggested CSS 2.1 one Sunday afternoon, but completely replaced. Originally the idea was just to remove the parts of CSS 2 that nobody wanted to implement. This could have been done in months. It would essentially be a profile of CSS 2, "CSS 2—The useful bits".

Instead CSS 2 was fully rewritten in tedious detail, the goal was now the interoperable CSS 2. This in itself was a good thing. It is much better that the spec is unambiguous than having incompatible implementations of the same, vague, spec. The ones having already implemented CSS 2 would need to adjust their products, but for new CSS 2.1 user agents getting it right the first time is much cheaper, not to speak of the millions of web designers that don't have to make specific hacks for every browser in existence.

But there is a diminishing return. The optimal time/clarity moment was probably when it was first submitted for CR, three years or so ago. This is not to say that what was done after that was unnecessary, but the extra clarity we have gotten for increasingly obscure issues have simply not been worth the additional years. They would have to be resolved at one point or another, but are errata or CSS3 material. The fatigue in the working group and impatience in the designer community is also starting to show, those that haven't given up on it years ago.

It is also slightly unfair. SMIL 2.1 had its first (and last!) working draft in February 2005, and a finished recommendation in December 2005. Were they that more efficient? No, but evidently their goal was to push out a spec as soon as possible rather than solving underlying problems, and crucially nobody outside the working group cared.


With this as a background it is none too surprising that a discussion on wither working group, even in some cases whether working group, has started. One solution I find particularly unconvincing is the argument that the CSS working group should follow the lead of the new HTML working group which currently has 405 self-invited experts, and the number is growing. That is the classic "the project is late, let's throw in many more people" approach. This behind the scenes report may be the best summing up of the situation.

The other symptom is an earlier call for CSS 2.2, reminiscent of my suggestion in 2002 or thereabout of a CSS 3.1. The big problem was as evident then as now. CSS 2.1 may have taken way more time and effort than was originally intended, and now we suffer from it. But sooner or more likely later it will be done. CSS3 on the other hand is a monster many times bigger than CSS 2(.1), the chances of it being fully done, let alone implemented, any time soon is nil. So what about the designers that have waited patiently on rounded corners for a decade and now have the outlook of having to wait another decade so that at least their children will be able to specify it in their style sheet?