Ok, lots has been going on again. Finally did a multi day moto trip, had the most successful Phnom Penh Mapping Meetup yet, started using the server at work (at last!) and got the ball rolling on the relational database.

WORK

So, first up I have begun to introduce the staff to the process of using the server for mapping and spreadsheet work. This was met with some resistance since they were thinking of only using it as an archive for final data. This is the result of miscommunication, since I only recommended it with every day, job data use in mind, as well as for storing final/reference data. If we had communicated properly and I'd understood they only intended to use it as an archive I would have only recommended we get a large hard drive. Now that we have a server, though, I am keen at least to show them the benefits of using a server for job data as well. That is, for storing intermediate processing data so that one can easily share current progress on a job. A good example is a shared spreadsheet where more than one user can make updates and the server will regularly update it with the latest data from all users. It's also much easier to share data when telling a manager where one is up to. One only has to (install clipboardpath.exe (or) and then) hold down control, right click on a file and you'll have the UNC path to the file (starting with \\ rather than the possibly unique drive letter you've mapped it to) for sharing in an email. This beats zipping it all up and emailing it as an attachment!

There are valid reasons for people's reluctance, which I thought of myself as well:

1) power failure means the server only stays online for about 30 minutes until the UPS expires
2) storage space is only 1 terabyte at the moment (1000 gigabytes), and each staff member has up to about 20 gigabytes of data at the moment
3) there is no plan for backing up the system at the moment

The reasons in support of using the server for everyday processing are:
1) As recommended by another volunteer in 2009, it is good to have a central location for all data, making it easier to search, back up and update
2) A central location for GIS reference data is recommended because it is easier to update it rather than having to update it on individual computers
3) Team work is easier because one does not have to copy data between computers, people just show the location of the files on the server to others that they want help from. This was demonstrated today when I was accessing copies of the relocation site datasets on the server,
which I had set as shared spreadsheets. I then emailed the file paths to these documents to attendees of the meeting and they can see them now too.
4) It is easier for managers to keep track of activity if it all happens in one location
5) We are developing a database of all point location and associated demographic survey data produced by the NGO, this will need to go on the server because multiple teams will be updating, managing and exploiting various tables of this database.

Speaking of databases, I spent a morning introducing staff to my proposed entity relationship diagram for the database (see picture). This comprises a few tables of attribute data hanging off a central table of locations of interest to the NGO. This means we have a lot of one to many relationships where there are multiple things, essentially, that happen at each location. For example, at one location there might be a community profile survey, then there might be a record of a threat of eviction, then there might be a record of an eviction, then there might be a record of the relocation site that they went to. Possibly not all went to the same relocation site, so we might link it to multiple relocation sites. I have not gotten any further making the database as we haven't finished the datasets. We're 90% there on the evictions dataset (yay!) and 100% there for the latest relocation sites dataset. We can make the table schemas since we know all of the fields, though.

PHNOM PENH MAPPING MEETUP 4

I keep saying these are a great success whenever they happen but this one was really the best so far. We had 30+ attendees! I was particularly chuffed with the fact that half of them were Khmer, though. We had two great presentations. The first was a tutorial on how to use Quantum GIS, by Math Pellerin from LICADHO. The best part about it was that he went through the process of adding provinces and roads of Cambodia (just vector data), demonstrating naming features using Khmer Unicode along the way, and finally showing how to compose a simple map with this data. This way the audience could compare it with what ArcGIS, the program most would have been more familiar with. Given that it's so similar to ArcGIS but has more robust support for Khmer Unicode and is also open source, I hope the presentation encouraged many to use it as an alternative.

We than had a presentation by Agus Rachmat from the US Naval Research Institute about mapping disease outbreaks (dengue, malaria and flu) across Cambodia and associated environmental factors. This was a fascinating presentation (excuse my bias!!!) because we were seeing how their team went about mapping things on the ground and the processes of systematically surveying villages where outbreaks had occurred. He also showed us some national maps about malaria. I recorded a video of the presentations, although it is very large so have not had time to upload it yet. I also met Tim Coulas, Project Field Manager and Land Registration Expert at Cambodia Land Administration Support Project who wants to present at the next meetup on 25th April! We may also have our first Khmer presenter (yay!) at the next meetup as Agus said that his counterpart needs to give us part two of the presentation where they will explain some of their spatial analysis in more detail.

MAKING TROUBLE IN PHNOM PENH

I was invited on a five day moto ride with three others last week. We slept in a mate's new guesthouse (basically a treehouse) by the Mekong in Kampong Cham the first night. Then went via some very bumpy dirt roads to Siem Reap via Skon after crossing the river in the morning. We spent two nights in Siem Reap, the first day mostly involved sleeping and running around trying to fix the motos. The second day we did a short trip around the West Baray, a big womanmade lake west of Angkor Wat. It's at least 10km to ride around. We had a lot of fun burning along the twisty dirt road on the northern side. One of our motos ran out of fuel at the south western corner but we were close to the road and a fuel station so it was ok. Not the last fuel scare though! We then headed up to Preah Vihear province the next day, making it to Sra Aem, a little town I had to add to OpenStreetMap 30km southeast of Prasat Preah Vihear, the World Heritage listed wat at the centre of a border dispute between the Thairs and Cambodians. We saw a lot of military activity, including a big dance! Yes, the military, military police and normal police got together at the temple whilst we were there and danced with a few celebrities from Bayon TV. Possibly the best timing ever! Whilst videoing it we were asked to join them and I got to dance with one of the most beautiful women I've ever seen. I am glad I didn't know how bad I looked at the time.

Oh, I also went in my first graded mountain bike race on Sunday! It was on Koh Dach (Silk Island) in the Mekong. I was meant to be in B but got there too late so went in C, this was a good thing as I was b*****ed after (a pathetic) two laps of the 6.5km course! It was a lot of fun doing mountain biking - done plenty of road and track races but going on a bumpy, jumpy course makes it exciting! I'd still say track racing is best (nothing like climbing the wall and going kamikaze) but it was still good to try out this form of the sport at last. Now I've just got to try cyclocross and I'll have done them all (yes, I've done a few time trials). Oh, well, I guess I should also try downhill racing...

Another trip is planned for the next two days, Phnom Penh to the Cardomoms and back. Should be good.