Sean McGrath, CTO, Propylon

Sean McGrath's Weblog.

Saturday, April 16, 2005
    Anna Livia Blogger Eile
We numbered more than 10 and less than 20 (I reckon). We met in the IFI meeting room on Eustace St. and we talked of many things. First prize in the "best array of quality geek gadgetry" category was Bernie Goldbach.

A rumour is going around that a rather famous personage with very a high visionary and geek quotient might be attending Semantic Web 2005 in Galway in November of this year. If it happens, I'll be there. Even if it doesn't, I might still be there. Its only a rumour though, so it stops here.

posted by Sean 1:05 PM
[Link]
. . .
Friday, April 15, 2005
    Dublin Python Meetup
Last nights Dublin Python meetup was revectored at short notice to a pub we knew Paul Everitt of Zope fame would be in :-)

It was good to put some faces to some names and to finally meet Paul Everitt F2F. The location was, in my opinion, streets ahead of the Dame Lane location and has the added benefit that the famished can get food with their drinks if required.

Lots of interesting tidbits from Michael Twomey about PyCon 2005. I must make a point of getting to one of those.

posted by Sean 12:40 AM
[Link]
. . .
    Dublin Bloggers Meeting
Hot on the heels of last nights Python Meetup in Dublin (of which more anon) comes a meeting of bloggers. There is a non-zero probability that I will be there.

posted by Sean 12:38 AM
[Link]
. . .
Wednesday, April 13, 2005
    HTTP URIs rather than URNs for identifiers
Dave Orchard questions some of the orthodoxy concerning when to use URNS versus when to use HTTP URIS.

As Dave says, context is the critical thing. Context is king. The notion that a URN is somehow "better" because HTTP URIs tie things to places/versions/whatever is simply not true out there. Just looking around at the way HTTP URIS are really used on the Web shows that it isn't true. Context is king. Identifiers are dereferenced at a time and a place with an abritrary rich resolving context. The act of resolution can - and frequently does - take full advantage of all that contextual information.

Any would-be rigid designator, needs a state of affairs to be dereferenced against. HTTP is a well known resolver of rigid designators given a state of affairs:-)

Most ironic in all the MMTT URN shenanigans is that so many URN resolving schemes end up doing a quick shimmy to convert the URNs into unique HTTP URIs :-)

posted by Sean 12:29 AM
[Link]
. . .
Tuesday, April 12, 2005
    Critical machine selection criteria - Emacs Pinky Position
So, from the comments on my recent posting enquiring about a good laptop for Ubuntu Linux, an IBM T42 looks like a contender.

A critical question remains - is the Ctrl key in the bottom left hand corner of the keyboard?

Why is this critical? Because pinky position is critical to Emacs without which, I am essentially useless at a keyboard.

I cannot make out where the Ctrl key is on the T42 from the images I found on the Web. Can some kind soul illuminate me?

Aside: I have more than once exacerbated sales guys in shops by not caring to ask questions about their beautiful laptops other than to press "Ctrl-x-s" to see how it feels :-)

posted by Sean 9:48 AM
[Link]
. . .
    Ubuntu, Laptop - just work - which?
If you wanted Ubuntu Linux on a notebook/laptop. And if you wanted all the bits like XWindows and WIFI and suspend-mode etc. to "just work". What notebook/laptop would you get?

I've been reading through some of the excellent compatibility reports out there but have yet to find one that says "Yup. Just do this and everything will work clean as a whistle."

Is there such a notebook/laptop?

posted by Sean 4:57 AM
[Link]
. . .
    Use the tests, Luke
Musings on the relationship between Test Driven Development and Documentation this week on ITWorld : Use the tests Luke!.

posted by Sean 12:59 AM
[Link]
. . .

. . .
Weblog Commenting by HaloScan.com