ArchCamp Blog/Wiki

Proceedings

Session 1 @ Cafe 1001:
We kicked off the ArchCamp with a discussion of some of the things we've been up to:

MySociety.org has a whole bunch geocoding tools that could be pretty useful including getting lat/long, postcode etc.

The group spend some time dwelling on the need for Parish and County on OASIS record sheets given that you already have to give a National Grid Reference. Surely it should be possible to have a webservice that could provide it automatically. Of course that would mean free access to political boundary maps first.

Session 2 @ The Big Chill
Andrew gave a demo of Ruby on Rails which looks like a godsend for those sick of/baffled by creating webby interfaces to their DBs. It's a framework for autogenerating Web frontends for DBs and has functions to automatically create an abstration layer from the DB structure, data views and all the client interface you'll need. Although Ruby seems to be a little cryptic as a language, it doesn't seem to be too bad for tweaking, which is fine as 90% of the code is autogenerated. Andrew had a nice example of KML feed he'd knocked together for a Votive Deposit database. In his opinion, you can get an interface for your database together in minutes which looks like the last nail in the coffin for Access.

MetaCarta
there's a new service where you can show 2 OpenLayers layers together, enabling you to upload an ungeoreferenced raster, georeference it against another layer and then download the world file for it.

3D skeletons
Leif showed what OA have been doing with recording skeletons - using a total station to shooting in points of 'stick-men' using a 3D approach - by shooting the different elements of the skeles in order they can show the direction of each 'element' of the skeles. This can then be displayed using volumetric polygons (in the first case pyramids) to enable you to view it in a 3D viewer (such as ArcScene) - each of these can be attributed up (i.e. type of bone/condition/etc.) and you can then use varying symbologies to display these more clearly as well as how they relate spatially to one another and other entities such as small finds or terrain topography (land surface, grave cuts etc.)

He has written a bit of Java that takes the total station data and then it pops out a nice 3D DXF that can be exchanged etc.

We then talked a bit about the use of ortho-rectified/georeferenced photos for recording skeletons and contexts, etc. Leif suggested that the simple use of taking the points and displaying the simple data in 3D is actually extremely useful and do we really need to digitise every bone in 3D? GH pointed out that is completely depends on your research question.

Regex NGR Scraper
Leif then went on to demonstrate his regex scraper that OA have used to go through old legacy documents for mass retrieval of NGR point data in projects which haven't been spatially plotted. It uses a Shareware document parser to go through their old site reports, etc. to pull out any instances of a particular regular expression - and then give you a list of the documents in which that occurs. He said he would supply the code!

This brought on a discussion about whether or not we actually need to tag any data at all - or just use the modern search functionality (i.e. Google) to simply parse through every bit of data and retrieve results. Is tagging dead?!

It's also worth mentioning that the NMR can also provide NGRs for any project that's ever been archived with them in gazetteer format so if anyone wants to use webmapping to show the sites they've worked on but hasn't recorded their locations electronically then have a word with Paul Adams there.

FASTI Online
LP Archaeology's demonstrated their new open layers based web interface to fasti online. It uses ARK as a backend for the data but bundles text into single table to enable indexed searching with a (cool) relevancy score. The assembled participants were quick to discover every brothel in Italy. Sorry, every ancient brothel excavated since 2000. The site uses data from data from the International Association of Classical Archaeologists and covers Italy, Morocco, Romania, Bulgaria, Macedonia and Malta which means that it is also multi-lingual and supports multiple character sets and aliases take care of translation. It also has a custom timeline function. Minnesota MapServer and MySQL and a customised Ka-Map application are used to cache tiles so as to speed up map generation.

Spatio-Temporal Periods and sites etc.
We had a chat (after getting beer and ordering burgers) about the problem of 'periods' being both a temporal AND spatial entity. i.e. the Roman period in Scotland not really existing... we then decided that there are many ways to do it - but that it gets very very complicated - as you then also have to say Dr. So and So thinks that the Roman period in England starts in 43AD and finishes in 500AD but Prof. Blah says it starts in 55BC, etc. Its a headache but LP are having to approach it for the fasti-online project so watch that space!

Funding opportunities
Leif outlined what OA are doing in terms of funding for Open Source projects

The ISCHIA project (Implementing e-Science for Cultural Heritage In Archaeology) is a collaboration btwn Oxford/Wessex/S'oton Uni Electronic and Computer Sciences/Reading Uni Classics/Ure Museum of Greek Archaeology to work on an Open Source data-sharing infrastructure. The ultimate aim is to make it as easy as possible for any organization storing and/or curating cultural heritage data to share it electronically. They've applied for funding from the AHRC-APSRC-JISC for a three year project which would pay for researchers and couple of full-time developers but the project is commited to keeping the community in the driving seat.

He also pointed out that the Methods Network might well be a great source of advice and possible funding for small projects which disseminate both information and useful resources throughout the archaeological sector. To that end we agreed to get in touch with MN to find out more about they can help.

What is Antiquist?
Currently a mailing list - but what should it be?

Is it going to become a Britarch-esque talking-shop or is Antiquist going to become more of a archaeological software development shop where people say "I want to do this thing (i.e. a quick uploader for OASIS forms)". Perhaps we could define certain things we want, maybe apply for funding for them and them distribute work between the memebers of Antiquist to help develop them.

What projects do we want to do??
After alot of discussion we decided that perhaps we could revive the old FISH/EH project, the HEEP project. HEEP effectively works like a WMS server but for heritage data. Therefore you have a HEEP server that reads in heritage data (for instance from a Fasti dataset) it then offers that data to a client in various formats (for instance MIDAS XML).

The HEEP server receives a GetCapabilities request and returns the formats that it can poutput. It then recieves a GetFeature (or whatever) request for a specific geographic area or 'period' or SQL query and then goes and returns the data in the requrested format.

It was proposed that we could apply for funding for a week long seminar/workshop/hack-fest from the Method Networks to pay for a proper session where at the end we produce a HEEP server that reads in data from at least two sources (Fasti and PAS?) and then outputs both datasets into one single format (MIDAS XML?). Dissemination and documentation are a big thing and must be done as we go along. Also something must be produced by the end of the week - a concrete working example!

Leif is going to talk to Method Networks to ascertain what would be a suitable way to apply for funding.

Conferences
Need to push forward organisation for ArchCamp2 at the CAAUK - and also was suggested that we (as Antiquist) should present a paper on the Antiquist movement. We should also have a think about a pub-meet in Berlin for Antiquist.