Sunday 31 August 2008

Byzantine links with post roman britain ...

The simple view of the post roman history of Britain is that the army left sometime before 410, and in 410 the cities and communities of Britain were told to fend for themselves. This they failed to do and collapsed under the weight of hordes of land hungry anglo saxon migrants. What of course this view does not show is the fact that there must have been an ongoing conflict for at least two centuries as the the anglo saxon communities pressed westward and the romano british retreated, yet were capable of mustering the effort to build fortifications such as Wansdyke. We akso know, both from literary sources, such as Gildas and Nennius that there were kingdoms in the west of britian, perhaps based originally on old roman local government divisions which themselves were based, loosley, on pre-Roman tribal boundaries. And that these stateles contained towns, certainly with eveidence that there was a roman style town functioning at Wroxeter till sometime after 500. There are arguments as to how romanized Roman britain was and to what extent romanization was only skin deep - he construction of towns in Roman Britain mainly because 'had to have them' and how majority of popultion in west and north continued to live in tribal villages, areas that were less romanized than the south and east. Interestingly, there's a similar example from Morocco. Most of Tingatania was abandoned by Rome in the face of the Vandal advance in the 400's but Volubilis remained occupied until being abandoned after an earthquake and then re-occupied with a smaller walled settlement on the edge of the town next to some fresh water springs, the town aqueduct being one of the casualties of the earthquake. These people were not Romans, even though some of them were buried with gravestones with latin inscriptions, and whose deaths were still dated from the founding of the province. Nor were they Arabs, their arrival had to wait until the coming of Islam. Most likely they were berbers, whos great grandparents may have had a patina of romanization but whose descendants were not, but who treated Latin as the 'official' language for business. And if anyone should doubt that Rome had an influence on the Berber's simply look to theBerber calendar, the names of the months and the celebration of Yennayer 1 as New year's day on 14 January, neatly paralleling the Orthodox Julian calendar. So one can say that it is quite probable that there were functioning post Roman statelets in the west of Britain. Like Bereber Volubilis, they were probably Roman in name only, even if their elites gave themselves titles such as 'protector' which drived from late Roman official titles, and the towns were only large native vilages perhaps with a few Roman style buildings built of wood, not stone. Now there statelets cannot have existed in a vacuum. Historians tend to concentrate on the saxon ascendancy and the conflict with the Romano British, yet we know that churchment travelled from the still british west to mainland europe, and given that these churchmen sailed on boats, that there must have been some sort of trade. And not just with Gaul. Byzantine coin finds are more common in the west than the east of england suggesting greater trade links.
(Question - how does this compare to the distribution for frankish coins?), but certainly suggesting that there was direct contact between the Byzanitine empire and the British successor state in Cornwall, based around Tintagel. Of course Byzantium does not mean Istanbul. The sixth century Byzantine empire had successfully reconquered North Africa and the grain ships sailed from Egypt and Carthage to feed the population of Constantinople. And paralleling the coin distribution, North African pottery is more common in west of england than east - pointing to trade route via north africa for supply of items such as wine and olives. And again the finds are focused around Cornwall and Tintagel. But why would anyone bother to sail to Tintagel from Carthage to trade with a gang of smelly celts who spoke bad latin and claimed to be Roman. Certainly not out of altruism. But the smelly celts had one thing that was in short supply elsewhere - tin - needed for making bronze. And in much the same way that minoan and phonecian traders before them found it worthwhile to risk the long sea journey to trade for tin so must it have been for the byzanitines, trading luxuries for tin ingots. And there are modern parallels to this scenario. During the second world war the danish colony in Greenland was cut off from Denmark, but managed to keep going and pay for the necessary imports by having something to trade, in the Greenland case cryolite that they could sell to the US and then use to pay for imports. Now this is all circumstantial. But someone with links to the Byzantine Empire was trading with Tintagel, where people did also make grave markers with inscriptions in bad latin. And the journey must have been worth their while - the more interesting question is what other forms of contacts were there and did they include any degree of cultural exchange.

Faberfinds

Books get published, get read and then go out of print. You then end up trawling second hand bookshops and these days the internet to track down a copy at a reasonable price. And of course publishers are faced with the costs of warehousing the inventory, something that's increasingly expensive, so the old, the obscure, and the plain boring end up being dumped onthe second hand and remainder market really early, or if you're unlucky, pulped.

Some university and academic publishers have gone to a print on demand model, where the text is prepared for printing and copies are only printed as one offs as required, which in these days of cheap high volume laser printing is a really compelling way to go - no warehousing or inventory management costs.

Now comes news of Faber Finds - a mainstream UK publisher giving print on demand to its back list - basically you get a bound printed copy of a book from the back list on request. Of course this costs money, but it does provide an interesting change in the way of providing access to out of print texts, and incidentally to scanning and digitally archiving these books.

Tuesday 26 August 2008

Viral spread of webmail...

Back in April I blogged about how, for students at least, email had come to mean webmail.  

Well we've seen an interesting phenomenon. Our current public webmail system is Sun Java Communications Suite 5 using the UWC client, but we also ran up the new version 6 (Convergence) on a test system a few weeks ago to see how it went and to play with it internally. We didn't bother protecting it, or restricting access, as we were wanting to do some user testing on the interface.

Well we've certainly got that. Somehow, even before we've started any formal testing of the system the url has leaked out and spread through the student community with sixty or seventy people logged into it at any time during the day. I think we might have got user acceptance ...

[if you're interested in Convergence, Sun have a demo system. You will need a username and password. I don't think they're very secret, either your local Sun account manager can get you them or else look on the Sun Communications suite website : at a pinch mail me]

Monday 25 August 2008

Reconstructing Minoan wall paintings ...


Interesting article about a group at Princeton who have developed a computer system for pattern matching to reconstruct Minoan wall paintings on Thera. Interesting. Very simmlar in concept to the system developed in Germany to digitally glue together the shredded Stasi files - I suppose the questions are - can one (a) uses such a system to combine papyrus or manuscript fragments held in different collections to do virtual reconstructions of the documents, and (b) could one then pipe the reconstruction into a handwriting recognition system to recover the text for further analysis?

Language diversity in the Caucasus

On the back of the unfortunate conflict in Georgia, an interesting piece in the IHT on the degree of linguistic diversity in the Caucasus. Despite having studied Russian years ago, including slogging through and later enjoying Tolstoy's short stories set in the Caucasus, I'd never quite clicked that there were that many ethnicities, languages, cultures in the area. Nor had my interest in Byzantine history helped much despite the close links between Armenia and Georgia, to name but two and the Byzantine empire.

One thing that did resonate with me was the comment about the Ossetian lexicon being burned. nearly twenty years ago now, shortly after the Iraqi invasion of Kuwait, a small Yemeni man turned up at the Computer Centre at York where I was working, doing data recovery and document format conversion, with a bag of floppy disks. He had been working in Kuwait and the disks contained all that was left of his research notes - most had been stored securely on a server at Kuwait University, like they should where they were backed up properly, etc, except that the Iraqi's decided to use the server and its disk stack for target practice.

I did get most of his data back, and he thanked me with a present of wonderful fresh coffee beans from his father's farm - a wonderful thank you and something that makes it all worthwhile.

Tuesday 19 August 2008

reading old documents

reading old documents can be difficult - incomplete text, blobby type and all the rest. There have been fairly successful automated attempts, abot which I've previously blogged elsewhere:

Searching Manuscripts Electronically posted Mon, 13 Feb 2006 09:29:51 -0800

 Digitisation of historical records is fine, but all you end up with digitisation projects for historical documents is a series of high resolution images which may be easier to work with and increase access but doesn't do anything for search.  Printed documents are more or less OK for search. Scanned and OCR'd versions of printed books, even very old printed books, such as those from rennaisance Italy are fine, even if you do need to sometimes 'teach' the OCR software how to deal with a non standard font.  Manuascripts have however, up to now been a no-no. 

The only way to make an electronic text was to type it in by hand and mark it up using an encoding schema such as those developed by the Text Encoding Initiative Consortium.  

Laborious.  

Now comes news of a really clever idea. Allan Smeaton's reserach group has been looking at shape recognition software - basically software to recognise objects such as cars and planes in photographs as cars or planes - trickier than it seems as you need to be able to find an object and recognise it from any angle, something humans can do easily, but machines find hard.  For example from the window of my office I can see a car park containing a mixture of sedans, hatchbacks, SUVs, all by different manufacturers and all different colours, but I can recgnise them as cars.  The clever thing about Allan Smeaton's software is that it can look at an image and twist it to match a category, so it can tell that a Peugeot hatchback, a big Ford Sedan, and a Subaru Forester are all cars.  

On a whim Smeaton fed digitised images of George Washington's letters into his software and it recognised an 'A' as an 'A', a 'B' as a 'B' and so on - all of which was pretty impressive, because while George Washington was taught to write in an age where legibility was prized, being as handwriting was the only real means of communication other than face to face discourse, like of all of us his handwriting got a bit more sloppy (and variable) as he got older and busier.  

Smeaton has also tried this on digitised medieaval manuscripts. These were actually easier to handle as the monks were going for legibility, and hence repeatibility.  

Smeaton has now obtained funding from Google, among others, to develop this as a search tool for digitised manuscripts - essentially a sort of plastic OCR that copes with variation.  Copperplate and other highly repeatable handwiting - and I would guess not just in Latin script - appears to be in reach, but I would guess that dealing with highly variable scribbly script, such as in diaries, especially now from C20, C21, are not. This would be because in the last hundred years or so handwritten documents were usally for personal consumption only, with most other documents being typescript or latterly computer printed, and hence subject to greater variation (aka scribbly).  This may mean that the TEI-C folks are still in business, either doing difficult cases by hand or by correcting errors in shape recognised texts. 

[Years ago, I came across another object recognition project which was to write a naked people detector. Apart from the use of the algorithm as an engine for censorware for the prurient, it's a genuinely hard problem given that people are all shapes and sizes and are photographed from all sorts of different angles, and come in two sexes, both of which have nipples, and meaning you can't simply cheat by guessing it's a torso and then if its got nipples (like one or two round redbrown circles two thirds of the way up it's naked and not for public consumption]

Now there's an interesting alternative method - use the human eyeball by extracting text from old difficult to process document and then use the extracted text in captchas. which is an interesting idea.
Now if you put theses two approaches together would it work for reading ancient manuscripts or cuneiform tablets, and could you make the system self learn?

Monday 18 August 2008

finance spam and angst

we all know about spam, and the endless tedious adverts for viagra and the like, which must hint at a deep level of sexual insecurity among some american males, otherwise we wouldn't see any such thing. (Spam, while very cheap to send, does cost something to send, and if it doesn't help sell snake oil or genital enhancement the logic of the market is they'd stop sending it if it didn't bring a return.)

So to finance spam. Before the credit crunch it was all about dodgy loans and remortgage deals to redraw on your equity to spend it on pink furry car seats or whatever - 'live richer, live above your means' - now it's all about debt consolidation and credit card payoffs. Still dodgier than a three dollar note, but it's obviously speaking to the angst of middle america.

There must be at least a sociology masters in this somewhere ... 

Tuesday 12 August 2008

darwin and blondeness ...


one of the substrands of this blog is the evolution of blondness, and the social and cultural implications thereof.

Interesting article in today's SMH suggesting that Darwin was also interested in this topic, and wether blondeness had a role in sexual selection...

Sunday 10 August 2008

tracing knowledge transfer

News of an interesting project tracing and mapping how knowledge ws transferred around the Mediterranean in early classical times.

Interesting in its own right for telling us about early trade routes and island hopping - eg the spread of the greek culture an colonies to southern Italy and even as far as Marseilles, and by contact with the celtic peoples living there, up the Rhone valley.

As I say intrinsically interesting as it would allow us to trace ancient knowledge networks and by using GIS techniques tie them into the landscape etc.

I was always struck, while staying at the Crawfordjohn campsite in Scotland, the Roman road, the old road, the drove road, the nineteenth century trainline and the modern freeway all followed essentially the same line as that was what the geopgraphy of the landscape required.

Anyway back to the project. There's an interesting rider to this. They also plan to use the same mapping techniques to the spread of information technology and concepts, and modelling how ideas ripple through - a Kevin Bacon experiment for knowledge transfer and mapping global cyber-social networks.

Friday 8 August 2008

Gmail outages

Gmail was out yesterday - off line for roughly fourteen and a half hours. 

This was irritating, but since it was my personal mail service no big deal. The question that interests me is what would happen if one had outsourced email to Google and had such an outage?

Now I'm not picking on Gmail here - I have windows live and yahoo accounts, but I don't use them frequently enough to have any real feel for how reliable they are. Gmail is very reliable, this is the first major outage I've experienced in four years.

On the other hand our corporate mail service has had two six hour outages in the last three years - both caused by SAN problems, and neither of them were comfortable experiences. Is one fourteen hour outage better than two six hour outages?


Thursday 7 August 2008

Gmail's out


strange how you get dependent on things. This morning, gmail's been out for around an hour so far, leaving me feeling oddly bereft.

Because my email goes via a forwarding service I could of course repoint it at another account if I was desparate, but I can survive. 

Socialogically, I always think it's really interesting how dependent we are on what is a free service with no implied contract. We only use it because it's perceived to be reliable and because it's perceived to be reliable we're out of sorts when it doesn't ...

If it was a paid for service we'd be no better off except there would be someone to should at and provide notional compensation. Like all outages the best solution is patience - my problem is that I'm not a patient person ;-)

Wednesday 6 August 2008

Google StreetView ...

Well, Google Streetview is here and if you look up our address you can see a really tatty picture of our house, probably taken sometime in November or December last year. Problem is, it was not long after we moved in, and as the tenants didn't look after the garden, it was all run down looking. Question is, what is the impact on property prices if we were to sell and people look up StreetView to get a 'real' impression of the house as opposed to the glossy real estate agent pictures ...

Monday 4 August 2008

Virtualising interactive shell accounts ...

Historically, universities have provided people with interactive shell accounts, sometimes called login accounts, to some form of time sharing system. Originally this used to be to all the computing resources on campus, and was typically a Vax/VMS or Unix system of some sort. Computing meant command line and meant login accounts.

Of course with the rise of personal computing and cheap hardware, and widely available GUI or WIMP environments that all went away. After all, why use a nasty command line system when you could have one with pictures and clicky things. The world has moved on, and while there is still a need for shell accounts they are mostly in niche areas and not in general purpose computing.

However, there is a perception that universities should have a general purpose login server providing shell accounts. Or sometimes two, one for staff and one for students, the student one being a bit more locked down.

Usage of these services is on average very low, but they are quite often hosted on individual boxes, often because that's how it's always been done but sometimes also because the operating system provided was not readily virtualisable. And as the boxes were small and relatively cheap, there wasn't a great imperative to do anything about them.

Now in these straitened times, rack space is a precious commodity as is the power consumed by the server and the cooling. Vastly disporportionate to the number of active users. And while no one would build a virtual infrastructure just to virtualise the login servers, if you have one already the overhead of virtualising these two little used servers is minimal.

But that leaves the question of what to do about computer science. Computer Science departments typically provide a login server for undergraduates to learn to do geeky things on, but these days more and more of their work is done on workstations and the amount of command line geekery required is less and less as industry in the main wants people who know about pointy and clicky things and how to program them. And of course there's always the fear that the trainee geeks would bring down something important or overload the system in some way.

Well if your virtualisation layer is any good they should be able to crash their virtual server without breaking anything else. But strangely no one seems to have tried this. Googling doesn't bring up anything useful ...