About Me

My photo
Web person at the Imperial War Museum, just completed PhD about digital sustainability in museums (the original motivation for this blog was as my research diary). Posting occasionally, and usually museum tech stuff but prone to stray. I welcome comments if you want to take anything further. These are my opinions and should not be attributed to my employer or anyone else (unless they thought of them too). Twitter: @jottevanger

Wednesday, February 25, 2009

Amazon Exposes 1 Terrabyte of Public Data

Hmm, never did anything with this when I saw it and amn't going to. It's a cool development, though: lots of new datasets through Amazon from (mainly) US sources, from DNA sequences to census and traffic data. It's more evidence that Amazon is shaping up as Google's real competition in the coming years, not Microsoft or Yahoo! (put that one in for those asleep at the back).

What can we offer from museums?

Anyway, here's the link to the RWW story: Amazon Exposes 1 Terrabyte of Public Data - ReadWriteWeb

Tuesday, February 24, 2009

Luke's Elegy and Ascent

Well I'm not the best placed to review Saturday evening's performance of Luke's new work, Elegy and Ascent, what with being a musical ignoramus and a very biased brother. I enjoyed it immensely, though, and his excellent talk beforehand (and our chats), and his reflections on how he reached this point in his artistic development, added a lot to the listening experience. It was fascinating to hear how he had, in the first half of the composition, taken part of a piano piece and redrawn it for orchestra. The original piece was composed using the chance (the drawing of cards) and Luke used a set of rules and his aesthetic sense to thread his way through this in a way that, as he pointed out, could only be his - another composer would have found another route through the chaos and made a different sort of beauty of it. To orchestrate this then, I suppose, added an extra layer of "human intervention" to the original randomness, as of course do the processes of conducting and performing the music. The second half of the Elegy and Ascent was very different but complementary: ordered but complex, and built around many interlocking, subtle and erudite ideas, as is usually the way with Luke's music. I won't say more for fear of sounding like I'm flattering my little brother, and we really can't have that. It would be even worse and kind of irrelevant to say we're proud of him, but...

I do hope that out there on't interweb somewhere somebody competent has done a proper job of reviewing the performance. One thing I have found, though, Karl Henning's blog where he previewed the concert and included Luke's notes. Also, if' you like to read scores, you can find lots of Luke's work on Scribd here (although not as yet Elegy and Ascent, I think). Hopefully in due course there will be a recording to listen to, too.

Wednesday, February 18, 2009

Digital preservation for da masses..? Archive Team, meet the DPC

Jeremy Keith (Adactio) has been fretting over the loss of his bookmarks with the dead-pooling of Magnolia, not to mention Pownce last year. Fortunately for Jeremy, he's a true alpha geek (not to mention trail-blazing, platform-pounding, left-clearing, good guy), so his response is to wonder how he can scrape, spider, and grab by hook or by crook all the stuff of his that he leaves scattered across the Web. Turns out, perhaps unsurprisingly, that he's not the only one with both the coding prowess and the strong sense of proprietorship over their stuff to try to do something like this. The spanking new Archive Team is up for it, guerilla-stylee, and hatching plans to help yer average Joe: as they put it, "We're gonna rescue your shit". Which is nice.

I like this. There's an active and sophisticated digital preservation community that has developed over many years. It has its roots, as far as I know, largely in the research and higher education communities, libraries, and the IT industry, all with their own priorities and preoccupations (museums are late-comers and minor players at best right now). And of course, lay people must have always played a part in some of this stuff, and some initiatives that could come under the DP banner - like the Internet Archive, I expect - rely heavily on informal communities to stash and curate stuff. Over the years ad hoc groups have sprung up to try to resuscitate corpsed forums, to salvage stuff from dying web-rings, to plan exit strategies for flaky virtual worlds and so on. But I don't think that there's really been a concerted approach to help individuals to keep a-hold of their stuff, spread as it often is across an array of sites from your many Google services, to Flickr and Delicious, to Twitter and your blog, Facebook, Slideshare and as many more as you care to mention. This is a huge variety, and what we might want to preserve in each will vary, as will the context required to make sense of it, so each service may need its own assessment for each individual (convergent evolution of the "significant properties" concept seems inevitable). It's a huge challenge but the Archive Team seem up for it.

I hope that these guys (and Jeremy K) and the "formal" DP community (DPC/DCC etc.) hook up and share knowledge. My guess is the latter have lots to offer in terms of technical expertise, connections, hardware and software, good theorisation, whilst the former group have wicked coding skills, motivation, inside-out knowledge of social software and its users, and a focus on the needs of individuals. They could offer each other a lot.

Monday, February 16, 2009

Something for the weekend

An(other) off-topic post intended for those small number of you who either (a) hang about in Leicester (hello!), or (b) are long-standing fans of the work of Luke Ottevanger (wow, how the hell did you find out about him?)
There's a concert this coming Saturday at the Fraser Noble Hall on London Road, Leicester, conducted by Michael Sackin and featuring Luke's Elegy and Ascent sandwiched between a Beethoven piano concerto and symphony. Luke is, let's say, not the greatest at publicising his work, and it's very rare that we in his family (or anyone else for that matter) get to hear what he has composed, so this is really exciting for us, and we have Michael to thank for asking Luke to write this piece, which I think he developed out of two previous works. Luke will be talking before the concert, hopefully explaining how the direction of his work has changed since its last public outing probably 10 years ago. Should be good, so come and hear the première!
Read more here.

Wednesday, February 11, 2009

There must be a name for this

That's it, I've had enough! Browsing the latest flurry of flim-flam from TechCrunch I found myself getting more and more cheesy at the current crop of crap names. OK so we're all used to mini-trends in the names that no-hoper start-ups go with - like the batch of second-ratrs (sic) that followed Flickr's success and ended their name with "[$consonant]r". Dopplr, Grazr, Etceterr. Funnily enough it's the fact that Twitter chose to name itself using the final vowel in the midst of that lame craze that makes me hate it less.

Nowadays it's all about the Ell: crap names ending with "...l" or "...le". OK so some names have been round for ever and make sense in their own way - to whit, Apple, Google - and others are more recent but might make sense still - Clickable, for example (no, "for example" is not a name, yet), even Pipl. I sort of forgive Amazon's Kindle (it's hardware anyway) but despite the fact it sort of relates to it's function I start to lose patience with Huddle, and having passed through Moodle, Oodle and Wonderfl we reach Trackle. Truly execrable. And slightly off the racing line, there's Twhirl.

Finally on the subject of appalling names from startups which will 90% certainly go to the wall, how about TwtQpon. Say it to yourself: TwtQpon.


Hey ho, perhaps I should lighten up and rename this blog Doofkl.

The new NPG website: lots of stuff to see

Well I've got little to say about this, not having actually checked it out yet, but it's exciting to hear that London's National Portrait Gallery has a brand new website which now gives access to what sounds like most or all of their collection. Here's the 24HM story. Doubtless there'll be interesting things to learn technically and organisationally about the thing (for one, it was outsourced to CogApp - I'm interested now to know about the skills they maintain in-house). Basically though I'm just excited to be able to go and use it myself. Cool!

Friday, February 06, 2009

Museum of London API enhancements

I've been extending the MOLA publications database lately so that it now includes archaeological sites to which the publications relate. For pretty much all of these we have decent geographical point data (as we should, having dug most of them) and of course this is great for playing around with the data. So I've got a KML output going for that web API, which returns placemarks for sites corresponding to publications returned by your query. In the placemark are links to all the publications related to the site. You can also go the other way round: either query by the publication's ID to get its details and (in KML mode) related sites; or query by site, which returns simple XML for each related publication or, if you choose KML, returns all points related to books linked to that site (potentially lots of points, not just the one in your query). What it all means is that you can see your search results in Google Earth or on a map if you like by pasting in the URL for the KML e.g. http://www.museumoflondon.org.uk/museumoflondon/food/rest.aspx?source=pubs&mode=kml&period=roman
So far I've not put this into the public interface because I think to do so requires some consideration, but this will come. I'm pleased in part because I can start to use this API in my own behind-the-scenes integration. For example I've also just done a load of work on the site summaries that we publish for all the work that MOLA does. The KML for these is cleaned up and the old ASP/XSLT thing I did to search across these (and other reports) by borough has been refreshed. It now lets you search by site code, and now that I can get at publications via site code it's a pretty small step to get it also to return related publications, which should come very soon (perhaps before going-home time). Because they're in very different data sources (XML files and SQL Server) it wasn't so straight forward before, now it is. The publications API is mainly for internal consumption like this (it also runs the user-facing publications page itself, at a lower level). I'd be really pleased, though, if others find a use for this, and any thoughts about the rights and wrongs of how I've done it would be gratefully received.

Thursday, February 05, 2009

Ideum's multi-touch table

Nice to see an alternative multi-touch table to the MS Surface one coming out of the ever-interesing Ideam: Multi-touch, Multi-user Table Prototype. This is a company working with museums very much at the centre of their concern. I don't know if that has any impact on how an interface like this might be designed, but in any case it looks like there's going to be choice in the market. Fingers crossed I might get to see it at MW2009, though the chances of wheedling the necessary cash from the accountants' talons look slim.

Wednesday, February 04, 2009

Being for the benefit of Mr Shite

...like me. This is just a post for the poor .Net coder out there in some remote corner of the world who, like me, will find themselves wondering why the hell their XSL transformations using disable-output-escaping are ending up with their output being escaped. For me this meant invalid KML, where hte CDATA sections I needed weren't being output correctly. I knew the transformation worked just by putting the XML and XSL together, but my cheapo web service did the transformation via C# and was screwing it up.

So if you're that coder of modest capacity like myself, here's where I got my clues:

  • This thread told me that the XmlTextWriter which is part of my Transform class would be ignoring "disable-output-escaping", and that I should write to a stream instead.
  • Never having actually done much with streams, I found these utility methods perfect for cut-n-paste MemoryStream-to-string conversion

Job done.

If you happen to be wondering how you can output a CDATA section in your XSLT, I wanted to put that in here too but unsurprisingly Blogger doesn't like that code. I'll see if I can find a way around that. Google "disable-output-escaping CDATA", that should get you somewhere!