Thursday, March 29, 2007

NSF Cyberinfrastructure Vision Document Available
National Science Foundation Cyberinfrastructure Council's Cyberinfrastructure Vision for 21st Century Discovery document has been released. It is revised from the 2006 draft document but appears to have mostly the same content. I haven't read it yet, but it is beautifully designed with some amazing photographs and visualizations.

This is a very significant document, basically a blueprint for cybersinfrastructure for science over the next five years, with implications beyond this time. From the document, it is focusing on:
...a vision for cyberinfrastructure in four overlapping and complementary areas: 1) High Performance Computing, 2) Data, Data Analysis, and Visualization, 3) Cyber Services and Virtual Organizations, and 4) Learning and Workforce Development.

Tuesday, March 27, 2007

AAAI2007's workshop on Semantic e-Science CFP deadline extended
I see that the CFP for this workshop at AAAI2007 (in Vancouver, July 22-26) has been extended until April 12 2007.

Saturday, March 03, 2007

OOXML @ ISO: "When is a standard not a standard?"

Free Software Magazine has an excellent point-by-point response (Edward Macnaghten's blog) to the ECMA's response to the contradictions (ISO standards term roughly for 'objections') of JTC1 members (twenty: a record) to the submisssion of OOXML (I'm out of breath!). For anyone wondering whether the objections were real and substantial (and whether the 6000 page standard was a disaster or not), this is an excellent examination of the points of contention with a good understanding of past (good) standards and the impact of standards on the real world.

A couple of the objections point-out tags such as:
  • footnoteLayoutLikeWW8 (Emulate Word 6.x/95/97 Footnote Placement)
  • useWord2002TableStyleRules (Emulate Word 2002 Table Style Rules)

Macnaghten also highlights some of the spin that is being spun by the submitter OOXML, Microsoft: the claim that the resistance to OOXML is actually a "proxy for product competition in the marketplace" and that the objections are merely political. By claiming the process has been politicized by its market competitors, MS seeks to claim the high ground.

Macnaghten concludes:
"I do not believe Microsoft are really interested in open standards, but just paying lip service to them. The history of the OOXML and ODF "discussions" supports this theory. OOXML is not as open as it's name applies: it ignores existing standards and it only fully caters for Microsoft Office programs making it near impossible for a competing product to use. By Microsoft's own admission it is simply an "XMLization" of existing closed .doc, .xls and .ppt formats. It should not be adopted by ISO, and certainly not "fast tracked".

ODF is genuinely open. There is no reason why Microsoft cannot use it. Adoption of ODF will benefit everyone who is not trying to maintain a near monopoly in the Office Production Suite domain."

Friday, March 02, 2007

Software, light bulbs & standards: Banning incandescent bulbs akin to banning FLOSS

I have been closely following the incandescent light vs. compact fluorescent light (CFL) debate (BBC, /., ABC , The Hindu) which has moved into the legislatures of many provinces (here in Canada: Ontario, New Brunswick), states (in the U.S.: California, ), countries (Australia) and regions (EU).

I find this debate has interesting parallels to the software world, in particular the Free/Libre Open Source Software (FLOSS) world.

The light bulb debate has been focused on (banning) a particular implementation technology (incandescent) when it should be focused on metrics (i.e. the proposed "standards" say something like "any light bulbs except incandescents"). Instead of banning a particular technology you instead want to put in place sensible, reasonable metrics and apply these as your standard. In this case, the standard should be some combination of energy/light output efficiency (light output per energy input, perhaps in lumens / watt) and luminous efficacy (you might also include something banning or limiting or imposing recovery of, nasty things like heavy metals, organochlorides, etc. and level of recyclability, etc. in your standard...). By creating a metrics-based standard, it is open to be used to measure innovation.

Creating a "standard" through the banning of a technology restricts innovation. And GE's announcement of high efficiency incandescent bulbs shows how short-sighted and misdirected implementation-based standards are.

This is very similar to how FLOSS software has been excluded from many governments and other organizations: the "standard" that is applied is one defined by a particular implementation, such as "must comply with Software X from Big Vendor Y", not a measurable metric or agreed-on open standard. With such fuzzy, incomplete or completely opaque "standards" -- often obfuscated by interested parties -- compliance is not possible and the standard itself is an impossible moving target.

There are also parallels between the software standards world and light bulb standards in the area of intellectual property (sorry Russell!): let's say I invented the incanfluorobabar light bulb which was 100 times more efficient than any existing light bulb, was inexpensive, safe, lasted 50k hours, produced perfect sun-like light, and was made from recycled newsprint, CO2, and straw (!), and I had the technology locked-up in patents world-wide. If legislatures around the world legislated standards which said that only incanfluorobabar were allowed to be sold in their countries, then everyone would have to pay me royalties to implement this technology, and competing technologies would have little incentive to innovate as it would be to difficult/expensive to change the legislation. Even if it was not legislated but expressed as a de facto (read CLOSED) standard in procurement, other technologies would suffer. Do you see parallels to the FLOSS universe?

[To be clear, I very much want a standard put in place to increase the energy efficiency of light bulbs and I am in no way an incandescent bulb fanboy...]