The Cathays Cemetary Project

This blog has been set up to allow discussion on a potential transcription project for the Cathays Cemetary in Cardiff, Wales.

The proposal is to photograph and transcribe the contents of the cemetary, a no mean task considering the size of it, and publish those results for researchers to benefit in locating their ancestors.

I'll start by posting my interest in the project and thoughts on how we should proceed, how the data can be accessed etc.


For a couple of years now I've been photographing the graveyards in Monmouthshire and Glamorgan and publishing them on the OGRE website (which can be found at http://www.cefnpennar.com). For the usual churchyard sized cemetary this format is adequate and easily maintainable and there are usually no more than a hundred or so graves to be catalogued.

For larger municipal graveyards, this format becomes unwieldy as you cannot list 10,000 graves on a single HTML page as it would be impossible to maintain and would take ages to load when viewed.

Thus, the first question is how to store and access the data? The two obvious options are an on-line database and query pages, or a set of static, linked HTML pages in alphabetical order, for example.

The next obvious issue would be organising and managing the photography of the headstones. From experience I know that I can do an average graveyard in a couple of hours. By the time you've waded through brambles and all sorts of obstructions (probably not an issue in the Cathays cemetary), got the camera to focus and fiddled with the zoom etc you are looking at perhaps a couple of shots a minute maximum.

Now a number of those shots will be duds due to lighting and focus issues, therefore we need to be able to re-shoot images on demand for those that do not come out properly. Obviously the graveyard needs to be sectioned and organised. Some of this may be possible to review given the level of detail in the Google Map satellite imagery, but I suspect a fairly detailed site visit would be required to examine how to use natural boundaries to give small, workable areas.

The volunteers who can help photograph the graves will obviously need access to the kind of equipment which can give a minimum image resolution, and the ability and technical knowledge to be able to crop and rotate those images and submit them to a central repository (such dreary things as file naming conventions to prevent overwriting images, and backups come to mind here).

Now comes the hard part, transcribing the images. This is the most tricky part, and requires many cups of black coffee and late nights... Welsh transcriptions should really be handled by Welsh speakers as I know when I transcribe welsh data it slows me to a snails pace checking the spelling and translating the dates etc to English.

All transcriptions need to be checked for accuracy, as it is very easy to transpose the data when you've typed in thousands of surnames.

And this brings us back to the data format. A couple of methods strike me as the way forward, firstly, to use an on-line form to submit and review data, or a simple spreadsheet.

The online form will be the harder way to do it, and would require an on-line database to provide access. This would require an amount of effort to set up and check before it could be used.

A spreadsheet is easier to handle and mail, and we can use it to generate HTML or a set of data for a database, and can be up and running fairly swiftly.

Finally, I would prefer to see this offered to the end user as a free resource that can be accessed without charge. Something to consider for this is that a free website is usually rubbish due to the small bandwidth provided and the lack of service. If this is to be published on the web it needs to have a small downtime (all computers crash sometimes), it needs to be reliably back-up by the service provider in case they have a disaster, and it needs a good bandwidth allowance to allow people to get at the site to see the data.

So, having said all that, let's get a discussion going and see where it takes us.

Cheers
Richard, aka The OGRE.