Profile
Blog
Photos
Videos
Another fortnight has rolled by I'm definitely in the thick of it. Well, more accurately I'm reminded of struggling in the towering whitewash and rips at Goolwa Beach, South Australia as an easily submerged 10 year old. This entry covers the ongoing effort to organise the various datasets at the NGO so that we don't duplicate things, and spend minimal time retrieving information that we have stored. I also cover a well earned holiday in Sihanoukville.
We have a frustrating issue with our eviction, eviction notice, community survey and relocation site datasets. They all make reference to each other, but often do not do so by the same name, or spelling if the names are actually the same. I think this is my first encounter with 'siloing' in the wild. We have had a team do a survey of all urban poor communities in Phnom Penh, published 2009. This also covered all the communities under threat of eviction at the time and most of the relocation sites. Then we have had a separate team remaining abreast of the eviction threats/notices issued throughout the city and, to their credit, much of this dataset is linked to the urban poor community survey via the relevant site code. Where new sites have been discovered, a new code has not been made, however, only a district abbreviation followed by coordinates. Then we have the evictions dataset. This competes with the relocation site dataset for being the most orphan like. It has been managed by another team (we're up to three now) and, despite being about sites in the urban poor community survey, often has different names for locations with the same coordinates. There is no cross linking between datasets using site codes, such as where an eviction has occurred at an urban poor site. The eviction site dataset is made from three principle sources, an in house advocacy team member's knowledge of the issue and the knowledge of the organisation's two founders. They do not always agree. They do not always provide the same depth of information about each site. I've tried to funnel all of these sources to finish off the dataset and it has taken about six rounds of consultation, with our second trip to the field tomorrow. Then there is the relocation sites dataset, the second the organisation has produced in four years. The first was produced by a different team than the last (siloing within a silo!) Needless to say, there is no cross linking between the relocation site dataset and the 35 relocation sites covered in the urban poor survey from 2009. The eviction sites mentioned in the relocation site dataset, or locations where families came from when the relocation site was established, frequently do not match the spellings in the actual eviction site dataset, when they are found in it at all.
It is fitting that that paragraph was far too long. The approach to fix this all is two pronged. Number one is to make a big (well, 9 tables actually) relational database (RDB). The entity relationship diagram has been designed and we will be discussing it further this week. The table schemas we don't have to worry so much about, as the different datasets for all of this information can be used for this. This brings me to the second prong. Finishing, as much as we can, the various tables that will go in the RDB. We have been working hard at this second, lesser prong, and it's important now having equal focus on the first since this is what will reduce the chances of going insane tracking down those cross references.
Not much else to describe work wise. It's the hard yards at the moment. Chugging through. Some ways I feel like it's my calling, I've always liked setting things straight. Hopefully I can end up having properly lived the dream. It would not be good to leave with this kind of a job half finished. Although that brings me to the other main development this episode. My capacity as a capacity builder has increased. I have begun to get better at handing over a process, and dataset, to relevant staff. I will be less stressed due to this, as well as recognising the agency of staff members. As many critics of the #Kony2012 video stated, a frequent fault of aid workers from the West is that they deny the capability of aid recipients to effect change themselves. Whilst I cited literature about capacity building to get this position, where I had to demonstrate my awareness of the need for an attitude of respect toward the agency of my counterparts, I have taken this whole affair as a wake up call. I realised that it's time to make a feature of my work shifting responsibility to staff members for processes and standards that we establish and projects that I start. Inevitably I have been doing a fair bit of the ground work in a number of areas as this is the first, not the second half, of my placement. It's not ok to coast along though, it's time to get them in the driver's seat. Today I put this into action by handing over the final integration of the urban poor community survey with the eviction notice dataset to our IT officer. He will be preparing it for upload to UrbanVoice. It was nice to show him what we needed to do, then sit and relax whilst he went about getting it done. Passenger seat on at least one dataset: tick.
Another fantastic bit of progress is that we now have a mirror of the UrbanVoice website setup. This will soon be totally Khmer so that we have a site easy to read for Cambodian citizens! http://khmer.urbanvoicecambodia.net/ - watch this space.
Oh, I almost forgot! Last week I did a 2 hour workshop for journalists from the Phnom Penh Post and the Myanmar Times. A friend of mine from the PPP invited me along to show them some basic online mapping procedures and data sources to help them make better news reports. I showed them how to make a basic location map in Google Maps, as well as how to map with a spreadsheet by collecting coordinates from Google Maps (try right clicking on the map and then click 'what's here?'). A few of them managed to keep up ;-)
Making Trouble in Phnom Penh
I went in the inaugural Sihanoukville Bike Race and Ride! It was a most mediocre performance, I only did 40km of a 60km race! Fail. But it was a hilly circuit, on a mountain bike without lockable forks and I hadn't trained one inch. Also had a wonderful ride on my moto from Phnom Penh to Kampot via Kep. After a night in Kampot and meeting there about Phnom Penh eviction sites in the morning with a former organisation head, off I rode to Sihanoukville. What a place. The beaches. The cocktails. The jetskis. The hills! Phnom Penh is so flat. There are hills there. All the tuk tuks are pulled by 150cc motos instead of the piddly little Daelims everywhere here. Had a great time, but the ride back was long and stressful. I hallucinated there was a bridge and somehow thought we must have therefore been near the Tonle Sap when really we'd just come to the big roundabout west of the airport!
I also went to my second Khmer wedding. This continued Cambodia's reputation for interesting wedding locations (my first was at a gas station if you recall) as it was on an abandoned railway line. The one going west of Boeung Kak up to the abandoned oil storage tanks near the Tonle Sap. It was the wedding of a friend of one of my counterparts at the NGO, and sobering to realise how much poverty there is all around us in Phnom Penh. The expat delusion of brunches at Namu and a lazy afternoon on the couch at Blue Pumpkin is safe if you don't go 5km in any direction from either locale.
The fourth Phnom Penh Mapping Meetup is on Thursday week - 5:30pm onward, 29/3/12 at Equinox Bar, upstairs. The theme is economic land concessions, Math from LICADHO will present about making maps in Quantum GIS. He's the one responsible for the map that caused all the ELC controversy this past month.
Khmer invitation:
ការប្រជុំពីការគូសផែនទីនៅភ្នំពេញលើកទី ៤
ទីកន្លែងៈ Equinox ផ្លូវ ២៧៨, ជាន់លើ
គេហទំព័រៈ http://bit.ly/wNKnuX
កាលបរិច្ឆេទៈថ្ងៃព្រហស្បតិ៍ ទី២៩ ខែមិនា
ម៉ោងៈ ចាប់ពីម៉ោង ៥ កន្លៈទៅ
ប្រធានបទៈ ផែនទីដីសប្បទានសេដ្ឋកិច្ច
អ្នកធ្វើបទបង្ហាញៈ Mathieu Pellerin
Quantum GIS
ការបង្កើតផែនទីដោយប្រើ Open Source Software
ការប្រជុំពីការគូសផែនទីនៅភ្នំពេញជាការប្រជុំប្រចាំខែសំរាប់អ្នកជំនាញការនិងអ្នកដែលមានចំណាប់អារម្មណ៍ទាក់ទងនឹង
ផែនទី។
- comments