Open source software

July 27th, 2006 by daniel pett

The Scheme runs on a rather tight budget, ergo we have quite a small pot for IT resources. The majority of our money has gone on buying development time from our excellent team at Oxford ArchDigital, who have produced our database and the procurement of two dedicated blade servers.

Our database uses open source technology on a Linux, Apache, MySQL and PHP platform (LAMP for short), and has decreased costs considerably.
The rest of our web resources have been created mainly by myself and also have used free and opensource technology. I’ve used quite a broad range of products, and the below details each one and provides a link to the place where you can get it. We also try to follow webstandards, valid HTML and CSS, tableless layout etc where ever possible, and we make a stab at being as accessible as possible.

  1. PHP
  2. MySQL
  3. Apache
  4. WordPress – used on this blog
  5. Textpattern
  6. Simplepie RSS parser
  7. Zoomify – not really OS – used for the zooming images in the coin guides
  8. Feedreader – really good desktop RSS feed reading
  9. fGIS – for creating maps – unfortunately no longer available
  10. Gallery2
  11. Filebrowser – lightweight thumbnailing programme
  12. Vanilla – forum software
  13. PHPBB2 – forum software – ours is pretty much dead now, and I hate it with a passion.
  14. Firefox
  15. Thunderbird email client
  16. Opera
  17. Microformats – not really software but coding standards
  18. Web mapping server
  19. Filezilla – opensource sftp programme. My favourite at the moment.
  20. Putty for server admin

So from that list you can see our diversity of software, no mention of Microsoft (not to say I don’t use their products!) I usually recommend that the Scheme’s staff use Opera or Firefox for their browsing, which has proved problematic to say the least with council IT departments, who are Microsoft or nothing. It should be recognised that IE sucks by now.

Some of these products, I haven’t used exclusively for the Scheme, but I have employed them elsewhere on Museum based projects. For example, Textpattern has been used as the CMS for the Eric Gill Society’s new website and I also installed Vanilla as their forum software and included a wide range of plugins built by other people.

I’m quite keen on the Textpattern CMS software at the moment, as you can convert it from Blog to dynamic website quite easily and it allows for authored workflows. It’s obviously not in the same league as large scale CMS software, but it suits many people and can produce rather stunning websites. For example: http://www.hicksdesign.co.uk. There’s loads more out there. Obviously, these CMS progs need quite a bit of work to be integrated with custom databases, but anything is possible.

There are obviously associated risks with going down the OS route to software procurement – security/ continuity of product etc; however, our experiences have left us able to create new web resources rapidly and with flexibility.

I’ve customised themes, functionality and styles for many of the products listed above, and some have proved to be very successful. Our forum is now all but closed due to the administration overhead; the gallery is added to by several users outside of the Scheme, as is this blog (not enough though….) So a community ethos is slowly developing. I think when the British Museum website gets developed in the coming year, and our prominence rises, we will attract more users physically and virtually. That remains to be seen.

You could say we’re trying to get towards a web 2.0 (horrible phrase – surely the web doesn’t need versioning it just evolves) type organisation. Tom Goskar {Wessex’s Archaeology IT guru} has recently talked about the Zoomr application, geo tagging photographs, and this is the sort of functionality that we’ll be implementing come September.

As it stands now, our system is able to produce GIS maps from any search where spatial data is attached {the only downside is the clunky client application.}
Our image records are all associated with an object, which is associated with a findspot. The findspot comes with geographical co-ordinates, so by using our Web Mapping Server and a GIS front end, we can produce geo tagged location maps based on your image search. We won’t be using Google Maps as our map provider, as we can’t afford the dev time to make changes if their API evolves – which it does! We’re also planning to allow users with registered accounts to tag images, records etc with their interpretation and thoughts about objects that we have recorded. There is quite alot of emotive or empathy that can be generated from the geolocation/object type/ chronology and this can be tapped into.

So we’re saving public cash, and that has to be good…..apart from the database, our grand total of spending for getting the websites and resources running at finds.org.uk has been £0 for software.

Tags: , , , , , , , , , , , , , , , , , ,
Posted in Techie stuff | Comments (1)

Information Architecture – a subtle rejig

July 24th, 2006 by daniel pett

As we now use the Google Analytics tools, in conjunction with analysis of our weblogs, it now makes it easier for the Scheme to work out if the architecture of this website is actually working. I’ve been monitoring the positioning of page “furniture” for a while now, and this has shown that some things were not actually working well. Therefore a subtle rejig has been needed and should be replicated throughout the site by tomorrow.

For example:

  1. Database icon – too far down the page, people weren’t finding it easily. Now it is moved to the primary navigation bar.
  2. Reviews of the Scheme – a new one was released a few weeks ago, and they couldn’t be found easily. Remedy – new category.
  3. Headley Trust information – really should be under Treasure.
  4. Added tag cloud on the sidebar.

There’s a few more, but let’s see if the rejig has any effects. I’m considering moving Pastexplorers to Textpattern at the moment, as an interim measure, whilst it is considered if it can be shifted to the new British Museum Content Management System (a long way off.)
Web stats are pretty good still, but freshness and content will keep people coming back. The Medieval coins guide will launch properly soon, and the Roman coins guide will be updated with biographies and information on “Reece Periods” now that we have a new coins adviser.

Tags: , ,
Posted in News, Techie stuff | Comments (0)

Quick search database fields/ plugin construction

July 19th, 2006 by daniel pett

Search plugin - brooch entered

Earlier today I wrote about the search plugin for firefox, our quick search system can interrogate these fields.
It isn’t a huge amount of fields, but they can be configured quite simply to be made searchable. If you have any suggestions of fields that you think would be useful to add to these (like I said earlier, cross table searching hasn’t been enabled so you can’t search find spot data.) Ideally parish/district/ county/ region will be added at some point soon.

  1. Find Id – for example PAS-A123AD
  2. Object type – for example coin
  3. Inscription – for example search on JESU
  4. Ascribed Culture – eg. Viking
  5. Broad period – eg. roman

With regards to building the Firefox plugin, it goes a little something like this:

Code, which forms the file PAS.src – located at www.finds.org.uk/documents/pas.src

An icon image which is located at www.finds.org.uk/documents/pas.gif
PAS icon

Both of these files are installed into the searchplugins folder of your mozilla installation. So if you want to create your own plugin for a web database, you may want to break down the construction of this plugin.

A guide to creating the syntax for search plugins can be found on O’Reilly and more resources are available on the Mycroft Moz dev website.

If you want to setup a link on your page to automatically install the files into the plugins folder, then you can make use of some JavaScript and a link. Obviously, you need to wrap this script in the javascript tags….

function errorMsg()
{
alert("Mozilla or Firefox is needed to install this plugin - get a better browser!!!");
}
function addEngine(name,ext,cat)
{
if ((typeof window.sidebar == "object") && (typeof
window.sidebar.addSearchEngine == "function"))
{
//cat="Web";
//cat=prompt('In what category should this engine be installed?','Web')
window.sidebar.addSearchEngine(
"{Address to plugin}/"+name+".src",
"{Address to icon}/"+name+"."+ext,
name,
cat );
}
else
{
errorMsg();
}
}

You then need to add the following as your hyperlink -

javascript:addEngine('{filename for the src file}','{file extension for the icon file}','{Category that you want the search plugin to be in}','0')</code>

Thus for these two files, my javascript link is:

javascript:addEngine('PAS','gif','Web','0')</code>

Simple, no? It’s obviously a bit of a fluff this extension, but it is good for quick searching our dataset without being logged in.

Link to results set

Tags: , , , , , , , , ,
Posted in Database, Techie stuff | Comments (0)

Searching our database

July 19th, 2006 by daniel pett

The Scheme’s database is growing rapidly these days and is now starting to attract attention from a wide range of researchers. I’m now issuing around 4 higher level research accounts per month, and the staff at the British Museum are beginning to make use of these data. In the past, the quality of our data has been criticised by many archaeological researchers; for example Cool, Crummy and Cunliffe (3 Cs!) We’re now more confident that the quality of our work is improving and is sufficient for many researchers to produce valid work. In the last few months, Tom Brindle, our FLO for Northamptonshire has been accepted for the PhD at Kings that was advertised on here and extensively criticised by a few over at Britarch. John Naylor’s work on the VASLE project at York is now starting to wrap itself up, and the preliminary results discussed at the April Conference at the British Museum, were interesting.

However, back to the database itself as I have digressed. There are some significant changes planned for August/ September where are wizards at OAD are going to change some functionality, and inshallah, the search mechanism should change extensively!
We’re planning to completely overhaul our numismatic searches (introducing Reece periods at the behest of our new Roman coins adviser – Sam Moorhead) and tying all our dropdowns (JavaScript) together for a more intuitive experience. Usability has been looked at, and we’re adding more features and functions that should make it easier to navigate and hopefully our userbase will increase. I’d like to introduce some more user generated content, such as tagging images and producing tag clouds, comments regarding the objects recorded and various other bits. Can’t tell you them all as that will steal our thunder.

One thing that I have created in the last few days, is a plugin for Firefox’s search box, so if you use this great piece of software, you can add our search engine to the dropdown list. This searches in the same manner as our quick search feature and so you can retrieve objects by entering the following for example:

  1. Roman coin – note not coins as we don’t do plurality!
  2. NMS
  3. CCCE01 – if you just use the alphanumeric code after a record ID it is far easier to find.

That’s just a few examples of what you can search on; I’ll post an attachment detailing what fields you can actually interrogate using the plugin later. At the moment, geographical locations aren’t included – they will be soon if I get my way!

So, to get this plugin, go to this location: add firefox search, and then restart Firefox. This should now appear in the search drop down list as finds database. I’ll be adding it to the mozdev archive of search plugins shortly. Any feedback?

Tags: , , , , , , , , , , ,
Posted in Database, Techie stuff | Comments (0)

Gold Fragment and the TVC

July 10th, 2006 by venicone

Less than 11 months after I found and handed this tiny piece of excitement over to Caroline my FLO , it has now been before the Treasure Valuation Committee on the 30th June. Prior to that two separate valuations were done by the “experts” who are called into the British Museum by the DCMS to actually see the objects. I think it was a hard call for them to come up with any figure bearing in mind there is nothing else to base a valuation on apart from coins from the same era.

gold runic fragment

Now its down to Robert Rose and the Braintree Museum where it will go on display to find the funding needed to acquire this small fragment of gold. I am tickled by the thought that my small gold find will perhaps be in the same display cabinet as a certain gold Roman phallus!!

The final valuation figure is £1,100 which I am very happy about and I hope the landowner is happy with this too.

The find has been a true find of a lifetime and not one I am likely to repeat! It has also been a rollercoaster of a ride – not least of all because just days after I handed this to Caroline, my dear Dad died after a long illness. I have pondered long and hard on whether to disclaim my share of any reward, but have decided that any money I receive will go towards a headstone for my dear POP.

A Treasure for a Treasure…….

Tags: , , , ,
Posted in News, Treasure process | Comments (2)

The content contained within the Blog's pages do not represent an official position from any of the organisations associated with the Portable Antiquities Scheme. They are solely those of the post's author.