Interesting article on the Longhorn/Vista development process- how broken and un-modular it was; how modular they’ve tried to make it. From Scoble.
Similarly interesting article on the economics of Nintendo innovation. Worth reading for those interested in innovation in niches that $COMPETITOR (MS, Google, Office, whoever) might suck up later.
Webcast went off OK this morning, after much scraping and fear. As a piece of technology that works, MS Media Encoder impresses me; as something that I had to make work, I was not so impressed. We spent a lot of time pulling our hair out to get this to happen.
Had a brief meeting today with Jimbo Wales. Totally skewed meeting summary: social software is Hard. :) Perhaps we’ll see some fruit of that later.
After that, he led a talk/discussion with the berkman fellows on ten things that will be free. Obviously that list starts with the Encyclopedia. #2 on the list is the dictionary. Notes that mediawiki is adding functionality to make it easier to do text which has specific structure, like a dictionary entry (and basically unlike an encyclopedia.) #3 is curriculum- say, wikibooks. Notes that wikibooks is doing a good job at creating encyclopedia-sized textbook modules, but not necessarily at book-length modules, and that other projects may be more promising at that. #4 is music and scores and arrangements- not necessarily of new stuff, but of classical, public-domain works. #5 is maps, though he thinks that this will be more difficult (all of them more difficult, really) if maps.google or foo.google comes along with beer-free open APIs. Tangentially, he seems to get lots of things. #6 is public art- paintings, sculptures, etc., that are not under copyright but owned by museums, etc. National Portrait Gallery in London has emailed C&Ds to wikipedia about 400 year old public domain images of Shakespeare. Their response is that it is fair use; if not fair use, it is public domain; if not public domain, he shames them, often by quoting from their own mission statement. #7 file formats; considers these substantially worse than unfree software. He has asked Steve Jobs personally to make ipod support ogg, which is madly cool. This went into a big discussion, with people giving Jimbo lots of good justifications for why this is bad. #8 is product identifiers- i.e., make ISBNs, etc., a public standard and database. If product identifiers are an open standard, it frees up interchange of data about goods, both in the review sense and in the sales and buying sense. #9 free, transparent search engines. In this case, he thinks this will be free; not that it should be free. A long discussion ensues; perhaps the most interesting point to me was JZ’s note that if semantic web succeeds in providing common structured data, this particular opening becomes much easier. #10 is free communities- basically, web communities should demand that data they give to third-party hosters be available and licensed in such a way that they can switch hosters/providers whenever they want.
He notes that they have been blocked in China; their media strategy is not to go to the Western press (which they already have plenty of) but to meekly go to the Chinese with the ‘we’re just an encyclopedia- surely it must be a mistake that we are blocked?’ approach.
