WebMink

Leaving the mountains [PhotoTravel]

Blogging has been sparse, but the travel has been intense. Today I am in San Jose about to travel to the OMA meeting, having arrived here last night from Denver via Dallas (how wonderful to have to use American Airlines). One of the pleasures of travelling is the ability to meet in person people you've only previously met online - yesterday was the chance to share coffee and chat-space with Eric Norlin, although I rather fear I did all the talking. He did describe his current project though - PingID, a Visa-style digital identity federation network (as in keiretsu) which seems to me to be an essential pre-requisite usable digital ID.

posted at 4:59 PM (UK) | Comment? (0 so far) | links to this post | |

Thursday, November 7

Fair use needs leeway [SocioPolitico]

David Weinberger has a perceptive and important insight into the issue of fair use in connection with digital rights management. He points out that much that might be termed 'fair use' is made possible by the exercise of leeway:
If we build software that enables us to "negotiate" usage rules with content providers, the rules can be as favorable as we'd like but their enforcement will necessarily be strict, literal and unforgiving. Binary, not human.
A comment on David's blog expands on this:
The better the society, the less law there will be. ... The worse the society, the more law there will be.
No amount of precision can actually protect fair use, because by its nature it is created by imprecision, leeway and generosity. If we have to resort to rules or law to protect fair use, the battle is already lost.

posted at 10:13 PM (UK) | Comment? (0 so far) | links to this post | |

Tuesday, November 5

Register back on track [WebTech]

Congrats to Andrew on his J2EE vs .NET update. I'm not entirely sure about the conclusion of his comments though:
What Öberg does establish is that Microsoft wrote its .NET Pet Store for speed, while Sun wrote its Pet Store to show programming techniques. We really do need a neutral party to establish some transparent benchmarks for this kind of exercise.
Benchmarks are a sad game. Even in an apparently uniform space such as processor performance they are hard to trust. For example, the contrast between x86 and SPARC architectures means that higher clock speeds don't necessarily mean better application performance. One might conceive of a benchmark for J2EE implementations (after all, there are plenty of them and they are all by definition architecturally similar).

But in a space where there is a radical architectural and implementation gulf, I struggle to see how one can even use the word 'benchmark'. What we really need here is a recognition of the fact that customers choose between J2EE and .NET for reasons other than performance. They are after vendor choice, or application selection ease, or some other benchmark-proof attributes - it's like choosing between a Mac and a PC. So to even attempt to reduce the discussion to the levels that the TMC Petstore Fiasco attempts is an insult to real customers.

posted at 10:29 PM (UK) | Comment? (0 so far) | links to this post | |

Monday, November 4

Tsunami warning [WebTech]

Having been excused with a slapped hand and thus effectively told that there's no problem with ethically suspect behaviour under the Bush administration ("Don't you know there's a war on?"), I'm expecting to see a wave of dubious practice like the TMC Java Petstore fiasco (so subtle that even The (normally sharp) Register got suckered - time they fixed that, surely?) breaking on the industry. Heck, it started just before the verdict was announced, with the court of all things setting the precedent (personally I didn't see it until just before 4pm PST). Those of us who have been vocal critics of Microsoft need to batten down the hatches, we can expect targetted attacks like the ones Rickard Öberg is already facing and worse.

posted at 8:31 AM (UK) | Comment? (0 so far) | links to this post | |

Sunday, November 3

Are cartels always wrong? [SocioPolitico]

Visiting the ancient city of Winchester yesterday, I visited a favourite book shop - P & G Wells. Once frequented by the likes of Jane Austen, and still offer book binding as a service, this store is located near the rear entrance to Winchester Cathedral and next to Winchester College. Inside is a splendid selection of childrens' books, an eclectic mix of fiction and many one-off selections to catch the eye and inspire reading.

It is a rarity in Britain these days - a local bookstore. They used to be everywhere - there was a fine book shop called Wiseman's nearby in Portswood once - but the end of the book publishers' cartel (the Net Book Agreement) meant that book discounters could move in and, as forecast, the degree of choice in Britain has dipped to near zero. Southampton has two branches of Waterstones, which is not a bad bookseller, but the selection they carry is impersonal and uniform with best marketing practice and the closure of Gilbert's Books means they are the only choice left. By contrast, Wells' has an invigorating selection of books I would not otherwise have encountered (I checked - Waterstone's doesn't carry them, even Amazon says 'yet to be published' on one!) as well as good, solid mainstream stuff. I'm guessing their location and history have preserved them from the same fate as Wiseman's. But Southampton is an impoverished place for the lack of choice.

Which leads me to ask: was the cartel that protected them really wrong? It wasn't like the RIAA - it was more a league of decency that allowed local booksellers to make enough profit from their business, and without it we have the joint power of Amazon and Waterstone's dominating the market. I like Amazon but I like spontaneous book-buying too. I was always sceptical of my father's defence of the net book agreement, but more and more I see its demise as damaging to society rather than as a victory for capitalism.

posted at 8:32 PM (UK) | Comment? (0 so far) | links to this post | |