HOT/Mali Activation Discussion with humanitarians 2013-02-25

Here is an agenda and chat log for the second meeting between the HOT team and the humanitarian workers in Mali.

Pre meeting agenda

Time and topic

Humanitarian OpenStreetMap Team

Mali Activation Discussion with Clusters

2013-02-25 21:30 GMT

General note: Any information we recieve from OCHA or the clusters needs to come with information on what we can do with it. For information to be suitable for import into our database, we need to know what license the information is under to make sure we are compying with copyright requirements. The same is true for pictures/documents if we intend to use them as a source of map data.

Progress Report

We are continuing to focus on building a 'base map' showing town names and outlines, as well as map the main connecting roads between them. The import of names and tracing of town outlines is largely complete in the areas north and east of Mopti. The next priority will be to turn our focus to the Niger Inner Delta region to map the many small settlements that were not associated with a name in the GNS or OCHA name databases, as well as to determine the areas that flood yearly, properly map the waterways, and map the most fertile farmland in this area.

Our crowdsourcing tool, developed by PovAddict, has continued to be a very useful way of finding small settlements that cannot be seen until you zoom way in. The tool has already been used to classify about 50,000 images (many have been seen multiple times) for whether they contain buildings/villages. We have also developed a workflow for taking the results from the tool, filtering out points that represent objects we have already mapped, and then loading into our tasking manager to be processed by mappers into actual map data, rather than just the click location we get from the tool itself. Each small black tile in the task manager is a place where a user clicked indicating an object of interest.

The crowdourcing tool itself: http://stuff.povaddict.com.ar/mali-crowdsource/

The Task manager job of some of the results from the tool: http://tasks.hotosm.org/job/198

Finally, we have put together several 'extracts' of our database for use in your humanitarian efforts which are automatically updated daily. Shapefile data for traditional GIS software is available via the HOT Export Tool (this may be having problems now due to the recent move to 64 bit node ID numbers). A second GIS shapefile exporter is also available for extracting small areas like a city or small region is also available, but we had some issues with the geographic projections in our early testing of this tool, results may vary. Lastly, we have a user who is daily producing Garmin map images suitable for loading onto a Garmin handheld GPS unit.

The HOT Export Server: http://hot-export.geofabrik.de/

The OSM2GIS tool: http://www.osm974.re/osm2gis/

Daily Garmin Files: http://tnm.verkstad.net/~seb/osm/garmin/Mali/ (Note: you do not need the Mali.osm.pbf file for your garmin, this is just the OSM source data used to make the garmin images)

Health Site Geolocation

We have imported the file CSntMali_02072012 containing health facilities in Mali into our database for refinement of their positions.

Although there are many health facilities listed in this file they were very poorly geolocated and their positions need to be corrected (Many are not even in the town that their name seems to suggest that they would be in).

Many of these health facilities have names that start with an acronym, CSCOM, we don't know what this is. If this is some sort of 'regional' health group, rather than an single actual facility, it might explain why these ones are so poorly positioned, as they would just be the center of the region. If this is the case we should either re-tag these in our database or remove them entirely so as not to indicate there is a health facility in the middle of nowhere.

We would like to know what we can do to improve the location of these facilities in our database. If the health facilites have a particular design or layout we might be able to spot them in the satellite imagery, pictures and text descriptions of what to look for would be very helpful (if this is even possible). Also, having good, exact, locations on a few health centers will help us know what to look for so we can locate others. Finally, we have a very comprehensive list of town names in our database which have been located very precicely using satellite imagery, if you have a list of what facilities are in what towns

Niger Inner Delta Imagery

The Inner Niger Delta is well covered by high resolution satellite imagery, but this imagery only provides one 'snapshot' in time. Since the region experiences annual flooding and vegetation/agricultural growth, having imagery at different points throughout the year is critical to properly mapping the area. With this in mind we have dug through the LandSat 7 imagery archive and have begun to identify lower resolution imagery of the area that is available at different times during the year. We are currently in the process of downloading and converting this imagery into a format suitable for use by our mappers. The plan is then to use the high resolution imagery to trace a given ground feature with high precision, and then 'zoom out' and view that area in the different LandSat images throughout the yearly flood season, to determine the proper classification and boundaries of that object.

For example, some rivers flood large areas, while others stay within their banks. Also, some, seemingly empty areas are fertile farmland after the flood. Lastly, low lying areas flood and the water is 'trapped' there to soak into the ground, these areas are prime candidates for well field suppliying drinking and irrigation water during the dry season. Hopefully these images will allow us to develop a better understanding of this region and produce detailed, actionable long term development related data for this region.

What we need from you to improve our work

Information regarding the health facilities geolocating as described above, along with pictures and/or descriptions of how to identify health facilities in satellite imagery.

A list of towns to mark for priority mapping in the upcoming week (OSM permalinks are prefereable for this if you can make them to avoid confusion). Also, for the given towns we would like to know what you would like traced: do you need all buildings, or just streets, will you be using walking papers in the town, or just passing through, etc.

A list of 'general priorites' for the region: what areas are most in need of information, what kinds of information do you most need, keeping in mind our limitations of mostly only having imagery to work from.

Chat log

[16:47:23] <scream> Hey guys....nice to see you all
[16:50:19] <scream> @All: here are the latest updates to keep in mind, which are quite important:
[16:50:38] <scream> 1. WASH Cluster gave us a good set of data which apparently matches your mapping, which is super
[16:50:46] <scream> the data can be integrated in the OSM Database
[16:50:50] <AndrewBuck> Just keep in mind when the flood happens and use the 'search by month' option on the first page to pick out months. Ideally we would get one flood cycle from the same year, but they don't have to be the same year. The ones I picked already are diff years.
[16:50:54] <scream> first point: SUPER
[16:50:54] <AndrewBuck> pnorman: ^^^
[16:51:14] <pierzen> education cluster around?
[16:51:22] <scream> 2. We will go next week to the geographical institute of Mali
[16:51:33] <scream> and talk about imagery available and data they can share
[16:51:51] <pierzen> good. some spots where we have nothing.
[16:51:51] <scream> as well we will go to the civil protection in Mali to have historical data for floods in mali
[16:52:01] <scream> and 3 point
[16:52:46] <pierzen> education should be back soon. connection problem.
[16:52:57] <scream> my colleague says he will start divulgating through his local network and university
[16:52:59] <scream> about our work
[16:53:14] <scream> so we can start trying to get some people on the ground able and available to volunteer
[16:53:26] <pierzen> this is very important for us.
[16:53:43] <scream> I am currently working on a city map for Bamako, Mopti and Tombouctou and will include the Open Street Map logo
[16:53:49] <scream> and would like to thank you all
[16:53:50] <AndrewBuck> yes definitely. We can do a lot remotely but there is no subtitute for boots on the ground.
[16:54:00] <scream> for the great work we are all together
[16:54:03] <scream> pierzen can confirm
[16:54:09] <scream> that a lot is going around
[16:54:17] <scream> and a lot are curious in understanding
[16:54:31] <scream> and a lot are talking about the super work you have been doing in the last month
[16:54:56] <scream> and I would like to introduce you
[16:55:18] <pierzen> And has coordination is concerned, we start being well equiped with the new discussion group.(private).
[16:55:20] <scream> to education
[16:55:34] <scream> which is here with us
[16:55:46] <scream> to discuss with you guys and be "amazed" about your work
[16:56:03] <scream> we will try to keep regular meetings on #irc with you guys with our cluster colleagues
[16:56:10] <scream> so you can have a direct interaction with them
[16:56:15] <scream> hope this is fine with you all
[16:56:20] <pierzen> Education seems to be gone again.
[16:57:29] <AndrewBuck> yes, the weekly meetings seem to be working well.
[16:58:30] <pnorman> So I can get to work, what months are priorities, and what lat/long is the top priority?
[16:58:45] <AndrewBuck> scream: Have any of your teams gotten a chance to try out the walking papers? or taking geotagged photos or audio notes?
[16:59:02] education joined #hot
[16:59:15] <pierzen> hi education, we see you.
[16:59:23] <education> education cluster here
[16:59:28] <AndrewBuck> pnorman: I would try to get one from March, July, and october.
[16:59:34] <education> ok cool
[16:59:36] <pierzen> Good.
[16:59:53] <pnorman> AndrewBuck: pick one
[16:59:55] <pierzen> ------------- HOT DISCUSSION WITH HUMANITARIANS -------------
[16:59:55] <pierzen> Welcome to humanitarians from the cluster and OCHA
[17:00:10] <pierzen> Let's start,
[17:00:29] <pierzen> We have received education files. Anybody to speak on this.
[17:01:13] <pierzen> AndrewBuck: could we move now to the discussion with education,.
[17:01:38] <AndrewBuck> bremy was doing some tracing from the education file we recieved, the results have not been brought into the database yet though.
[17:01:52] <pierzen> education: the education file was very inprecise.
[17:01:58] <AndrewBuck> I think there was some concern about the license on the file used, but I am not sure.
[17:02:21] <education> I am working on getting the license info from USAID
[17:02:22] <pierzen> AndrewBuck: oups, is this health or education we had problems?
[17:02:28] <education> wating to get a response
[17:02:32] <AndrewBuck> yes, accurate location information seems to be a problem for most of the data files we have recieved.
[17:02:50] <AndrewBuck> education: ok, cool. We will hold off on importing until we get the final go ahead then.
[17:03:08] <education> the data from USAID was not good?
[17:03:10] <AndrewBuck> He has traced the buildings around every education point in bamako though.
[17:03:22] <pierzen> bamako yes.
[17:03:31] <pierzen> other file I am not sure.
[17:03:34] <education> mopti, no?
[17:03:37] <AndrewBuck> education: that one he thought was better than the others i think, but none of the files have really good location information.
[17:04:04] <education> bamako was done by REACH
[17:04:07] <AndrewBuck> education: I don't know the exact details as bremy did that work and is unfortunately not here at the moment.
[17:04:16] <education> ok
[17:04:36] <pierzen> but we can write later and give more detail.
[17:04:55] <education> yea, let me know what I can do, if anything on my end
[17:05:33] <pierzen> education: other then having schools on the map, what do you need for your cluster?
[17:06:00] <education> it depends on the region
[17:06:02] <AndrewBuck> education: ultimately what will have to happen is someone will have to go around and actually check that we got the school in the right place.
[17:06:18] <education> right
[17:06:27] <scream> @Andrew if we manage to do Mopti
[17:06:27] <education> for Bamako it should be good
[17:06:32] <scream> I can personally
[17:06:37] <scream> do the field verification
[17:06:40] <AndrewBuck> We can trace all the buildings around the school location given by you for some 'buffer' distance and then you can take a printed map around and write the true location on it.
[17:06:50] <scream> and start using the fieldpapers in Mopti next wekk
[17:07:03] <AndrewBuck> scream: that would be cool.
[17:07:07] <scream> so that I can then show the benefit to the different cluster partners
[17:07:18] <scream> what do you think?
[17:07:21] <AndrewBuck> We might be able to do buildings for all of mopti as it is not that big and already has a lot done.
[17:07:31] <scream> perfect!
[17:07:36] <AndrewBuck> That would make a good 'case study' I think.
[17:07:49] <pierzen> If you need detail for a particular town, you ask.
[17:07:55] <scream> as well, education, what do you have as priority zones/cities?
[17:07:58] <pierzen> Do you need extract from the database ? Could extracts like the exemple I sent this afternoon could be usefull.
[17:08:14] <scream> mopti is fine as i go there next week
[17:08:44] <scream> those extracts are perfect.....i used the HOT export for Bamako
[17:09:05] <education> zones/cities that are priorities are schools in areas of high IDP density in the south
[17:09:27] <education> towns of Bamako/Mopti/segou
[17:09:38] <education> and schools in flood regions as well
[17:09:48] <pierzen> Daily extracts for all Mali are available , shape format and Garmi GPS. This can be copied on a gps.
[17:10:13] <education> all schools in the north are of interest
[17:10:34] <AndrewBuck> education: our school data is pretty sparse. We mostly can only work from satellite imagery and schools don't have signs written on top saying that is what they are.
[17:10:54] <pierzen> Is there a particular architecture, we can say, this is a school? In Congo this was easy.
[17:10:55] <education> yea, that approach will not be easy I imagine
[17:11:01] <AndrewBuck> education: if you can help us by getting us descriptions and/or pictures of what to look for, we may be able to find many more.
[17:11:15] <education> I can look into it
[17:11:22] <education> see what pictures are available
[17:11:25] <AndrewBuck> oh, by the way, here is the meeting agenda notes. Forgot this at the beginning. http://wiki.openstreetmap.org/wiki/HOT/Mali_Activation_Discussion_with_humanitarians_2013-02-25
[17:12:01] <education> otherwise the education cluster is looking to do a few assessments
[17:12:11] <education> at the school level
[17:12:17] <AndrewBuck> pictures are very helpful for us to know what to look for. If they have a unique layout we can find them, but otherwise all we can do is trace generic buildings and rely on your teams to mark the exact location of schools.
[17:12:33] <education> we could try and get GPS coordinates then
[17:13:08] <education> but that would likely be a limted amount of schools
[17:13:43] <AndrewBuck> education: anything precise we can put into the database (if we get permission of course).
[17:13:50] <pierzen> education: One eaysy way to document buildings, is to combine a gps trace with pictures. This we can explain more in detail later. This way we can locate a picture on a map.
[17:14:31] <pierzen> you take a picture with the name of the school. And the gps tells where it is.
[17:14:47] <AndrewBuck> education: with the water wells, we got a few imported from the millenium villages project in ghana, then we saw what they looked like in imagery, and started marking them in mali using this knowledge. We can do the same for schools if they have a unique look, so even a few that are known for sure would help.
[17:15:07] <scream> @education: i have cameras with gps units embedded we can discuss that in the next days
[17:15:09] <education> evans is going to tombouctou this week (a co-worker of mine)
[17:15:20] <AndrewBuck> We know what they look like in bamako and the large cities, but the rural towns don't seem to have similar looking buildings to those in bamako.
[17:15:24] <education> we could take some pictures of schools there
[17:15:32] <pierzen> scream: yes the best is with embedded gps.
[17:15:36] <scream> @education: perfect
[17:15:39] <education> ok
[17:15:47] <scream> let's talk about that tomorrow education
[17:15:51] <education> lets link up with evans before he leaves wednesday
[17:15:56] <education> sounds good
[17:15:58] <scream> and we get pictures in tombouctou
[17:16:10] <AndrewBuck> scream: can you tell education how to get georeferenced pictures like we discussed last week?
[17:16:17] <AndrewBuck> And have you tried that yourself?
[17:16:38] <pierzen> AndrewBuck: gps integrated in the camera.
[17:17:20] <scream> @andrew as pierzen says we have cameras with embedded gps
[17:17:32] <PovAddict> excellent
[17:17:38] <PovAddict> did you take pictures with that already?
[17:17:50] <AndrewBuck> ahh, ok, if you have the cameras with that then that is much easier.  :)
[17:18:15] pierregiraud_
[~pierregir@ver73-2-88-160-44-40.fbx.proxad.net] a changé de pseudo en pierregiraud
[17:18:20] <AndrewBuck> We don't have fancy toys like that so we just do it with a handheld gps.
[17:18:24] <AndrewBuck> :)
[17:18:30] <education> we can also possibly send these cameras out to the UNICEF regional offices in segou and Mopti
[17:18:40] <pierzen> An important point, is when you start a day, you open the camera outside and wait 5 minutes before you move. The unit search for gps and stabilize.
[17:18:42] <education> and they can get some shots of schools in rural area
[17:18:59] <education> UNICEF also has a team in Kayes
[17:19:22] <AndrewBuck> education: The more pictures the better, even just pictures of the town are useful for understanding what we see in imagery.
[17:19:37] <education> right
[17:19:50] <pierzen> We have concentrated in the north first. If you look at some areas not mapped in the south, tell us.
[17:20:28] <AndrewBuck> ideally we would like good coverage over a whole small town that is somewhat representative of towns in the area. A half dozen general shots of the area, pics of the water wells, where they keep their cattle, the school, any barns, public buildings, etc.
[17:20:49] <pierzen> We also have to look if we have imagery for all southern regions (south of Mopti).
[17:20:52] <scream> @all: would you like to show a couple of mapping examples to education
[17:20:55] <AndrewBuck> The towns all look very similar, so if we get a good "survey town" we can extrapolate from there.
[17:20:55] <scream> so that you can impress him
[17:20:58] <scream> what do you think?
[17:21:41] <AndrewBuck> Here is a good example of some mapping we have done in detail... http://www.openstreetmap.org/?lat=15.10531&lon=-2.6385&zoom=16&layers=M
[17:22:04] <AndrewBuck> We heard you had a team heading to Dallah and a few other towns so we tried to map them fully.
[17:22:04] <pierzen> gives the link with arry appl.
[17:22:15] <pierzen> harry.
[17:22:29] <AndrewBuck> The towns are much more complicated to trace then we are used to, but they are manageable.
[17:22:43] <AndrewBuck> It takes quite a while to trace to that level of detail though.
[17:23:07] <AndrewBuck> That town took me approximately 6 hours I think.
[17:23:20] <scream> awesome!
[17:23:40] <scream> do we have an image of how it was at the beginning so to compare?
[17:23:51] <AndrewBuck> Well, it was blank.  :)
[17:23:57] <scream> totally?
[17:24:09] <AndrewBuck> yes, I started from just the grey outline.
[17:24:27] <AndrewBuck> This is how we prepare an area for walking papers.
[17:24:30] <pierzen> scream: this is our business :))
[17:24:45] <scream> ok guys, sorry to go out of the discussion
[17:24:48] <scream> but pierzen
[17:24:54] <AndrewBuck> So now your team can print out that page and write all the info you want on it, school health facility, etc.
[17:25:04] <scream> could i have a couple of screenshots
[17:25:16] <AndrewBuck> We will try to do similar for Mopti, but that might take longer.
[17:25:24] <AndrewBuck> scream: when will you be in mopti?
[17:25:31] <scream> of the best examples of before/after mapping with the time frame it took and how many people involved?
[17:25:40] <pierzen> Bing + OSM http://harrywood.co.uk/maps/bingosm.html?lat=15.10531&lon=-2.6385&zoom=15&bingopacity=16
[17:25:48] <AndrewBuck> If we have a fixed deadline, we can recruit some volunteers from the rest of OSM.
[17:26:11] <pierzen> A slider lets see OSM or Bing Satellite or both in the middle.
[17:26:15] <scream> I have my flight wednesday 6th
[17:26:18] <scream> march
[17:26:32] <scream> and will present Information Management and would like to present OSM
[17:26:39] <scream> so having 3-4 examples of images
[17:26:43] <scream> before/after
[17:26:48] <scream> to give a direct example would be super
[17:27:05] <pierzen> We will ask for that.
[17:27:13] <AndrewBuck> scream: that is in mopti on 6th march? or somewhere else?
[17:28:00] <scream> mopti the 7th will be presenting
[17:28:34] <AndrewBuck> scream: Ok, we will try to have good coverage by the 5th or 6th so you can print maps before you fly out.
[17:28:52] <scream> by the way the harrywood application is super cool
[17:29:14] <pierzen> these guys are amazing, a new tool everyday.
[17:29:22] <scream> it is true
[17:29:29] <scream> education are you still there?
[17:29:39] <education> still here
[17:30:06] <education> anything else I can be of assistance with?
[17:30:16] <education> or should we talk more in detail tomorrow?
[17:30:28] <AndrewBuck> education: what towns would be best for us to map in detail for you?
[17:30:31] <education> Guido and I
[17:30:42] <AndrewBuck> You suggested timbuctu I think.
[17:30:56] <education> yes, the towns in the north
[17:30:58] <education> Gao
[17:31:03] <education> timbouctou
[17:31:20] <education> the towns of mopti and segou
[17:31:45] <AndrewBuck> Mopti we will already do for scream, what would be your top priority and when willl your team be there?
[17:31:59] <AndrewBuck> We probably can't do all of those on a short timeline.
[17:32:34] <education> I would need to speak with some co-workers
[17:32:40] <education> and see what they think
[17:32:46] <AndrewBuck> ok, no need to know for sure right now.
[17:32:58] <education> I can come back to you tomorrow with a set list if you like
[17:33:04] <AndrewBuck> Try to get back to us with at least a few days to a week leadtime if possible.
[17:33:15] <education> for sure
[17:33:19] <AndrewBuck> if you are going somewhere the next day, we probably can't help.
[17:33:56] <education> yea, I would never be expecting that sort of timeline
[17:34:17] <AndrewBuck> Also, if anyone is going to have a team in one of these smaller towns it would be good to do a thorough survey with ither fieldpapers or walking papers which would take about 2 or 3 hours for a town like dallah.
[17:34:20] <education> any assistance you can bring in helping map schools in much appreciated
[17:34:22] <scream> @education you saw how amazing the mapping they are doing?
[17:34:43] <education> yea, it's great
[17:34:58] <pierzen> Segou we wait for imagery.
[17:34:58] <AndrewBuck> If you get us a thorogh survey of one town like that we can then work out the population density of these settlements and get good estimates for the others.
[17:35:24] <scream> @education
[17:35:25] <AndrewBuck> They seem to be pretty similar in their design, so extrapolation should work quite well.
[17:35:30] <scream> is konna a priority for you?
[17:36:07] <education> I think the areas with high IDP density are prioroties
[17:36:15] <pierzen> konna is quite done. let me check.
[17:36:31] <education> towns of Bamako, segou, mopti
[17:36:42] <scream> ok perfect
[17:36:53] <AndrewBuck> I don't think we have imagery for segou.
[17:37:00] <scream> so segou i am still waiting for an answer for the imagery....i am working on that
[17:37:03] <scream> bamako we have
[17:37:07] <pierzen> Map of Konna http://harrywood.co.uk/maps/bingosm.html?lat=15.10531&lon=-2.6385&zoom=15&bingopacity=16
[17:37:14] <AndrewBuck> in fact, no we don't I am looking at it now. The others we have though.
[17:37:25] <scream> and if you all agree as i will go next week and it is a priority for education we can go on that?
[17:37:46] <AndrewBuck> scream: we already have bamako imagery, which is reasonably recent I think so it might not help us there.
[17:37:47] <education> in the long term it would be interesting to map fllod zones in the south
[17:37:55] <education> and see what schools are located in these zones
[17:38:01] <AndrewBuck> Segou would be a prime target if you can get that though.
[17:38:22] <AndrewBuck> education: do you mean in the Niger Inner Delta?
[17:38:27] <pierzen> education: for flooding, other imagery would be needed.
[17:38:55] <AndrewBuck> scream: I understand, just letting you know in case there is any choice and you can only get one or the other.
[17:39:03] <scream> for flooding next week we go to the geographic institute for having historical data and talk to the civil protection
[17:39:13] <education> on great
[17:39:15] <education> ok
[17:39:17] <scream> they might have a database existing already
[17:41:09] <education> yes, so to recap for education the 2 priorities are areas of high IDP density in the south (towns of Bamako, segou, mopti) and schools in the north
[17:41:22] <PovAddict> augh too much backlog
[17:41:22] <PovAddict> wonder if I should even bother reading it
[17:41:33] <education> I understand that is a lot of schools
[17:41:36] <AndrewBuck> PovAddict: it will be logged.
[17:41:41] <pierzen> plus flooding.
[17:41:49] <education> yes
[17:42:00] <PovAddict> I probably won't bother once you put it on the wiki :)
[17:42:09] <education> but as you said, there is some background work that needs to be done on flooding first
[17:42:18] <AndrewBuck> for the flood area, is south of mopti, or north of mopti more important?
[17:42:49] <education> I think I would need to look at school density in these areas
[17:43:06] <education> to see which is more of a priority
[17:43:29] <AndrewBuck> ok, we will just "follow the schools" then and prioritize settlements with them.
[17:43:37] <pierzen> we might prioritize some sectors north and south.
[17:43:57] <education> and there are many regions prone to flooding, mopti, kayes where a lot of schools were affected last year
[17:44:13] <education> but we would need to look at general trends as guido suggested
[17:44:38] <pierzen> then this would become a way to prioritize.
[17:44:56] <AndrewBuck> yes, this kind of information is what we need.
[17:45:11] <AndrewBuck> It is far to big a place for us to just map blindly and hope we hit what you need.
[17:45:21] <education> for sure
[17:45:33] <AndrewBuck> It sounds like you work mostly in the larger towns. Is this correct?
[17:45:43] <pierzen> There are probalby maps of flooding that give a general view.
[17:46:09] <pierzen> like what pnorman found.
[17:47:14] <education> we would like to map schools everywhere, but areas of high population density make the most sense to start
[17:47:20] <pnorman> Imagery is on order, but may take 1-3 days for USGS to process it so I can download
[17:47:48] <pierzen> pnorman: do you cover large cities in the south.
[17:47:55] <AndrewBuck> Also, how much trouble do your teams have when travelling, are road maps a problem for your teams, or do you stick to the main roads and not have so much roblem with that?
[17:48:36] <pierzen> would the gps files usefull for you to download on a gps unit?
[17:48:45] <AndrewBuck> We have a routing system that can give turn by turn directions and/or export files to a handheld GPS for your teams to use for navigating.
[17:49:31] <pnorman> pierzen: I'm going as far south as Path 197 Row 50 which ends approximately at a line between Diena and Sokoura. I could go farther if desired
[17:51:40] <education> sorry, people but I need to jump off shortly
[17:51:50] <education> I want to thank you all for the support
[17:51:57] <AndrewBuck> education: no problem.
[17:51:59] <education> amazing work and much needed
[17:52:00] <pierzen> ok, we can continue by email. This was a great exchange.
[17:52:04] <AndrewBuck> I think we are about done anyway.
[17:52:18] <pierzen> you are doing a great job too.
[17:52:25] <pierzen> this is why we support you.
[17:52:30] <AndrewBuck> education: can you answer on the roads issue before you go?
[17:52:36] <education> I will follow up on 1. the USAID license, 2. priority zones of education to be mapped, 3. school pictures
[17:53:04] <education> sorry, I didn't know the roads question was for me
[17:53:10] <education> what was the question?
[17:53:34] <AndrewBuck> basically we want to know if we should focus on finiding more towns/villages or more on tracing the roads between the ones we have already found.
[17:53:57] <AndrewBuck> So what is the bigger problem, finding people, or getting to them.
[17:54:04] <AndrewBuck> ?
[17:54:16] <education> I think for us it is simply locating schools
[17:54:36] <education> educational authorities in the ground know how to get therer
[17:55:07] <AndrewBuck> ok, so wel will focus more on finding villages and trying to locate the schools in the files we have then.
[17:55:08] <education> so finding more towns and villages would be more helpful from our perspective
[17:55:15] <education> perfect
[17:55:16] <AndrewBuck> Excellent.
[17:55:19] <AndrewBuck> :)
[17:55:21] <pierzen> education: yes, this is for your cluster. But we have to take account of all clusters , you surely understand.
[17:55:22] <education> thanks again for everything
[17:55:34] <education> I do
[17:55:45] <education> just answering from my perspective
[17:55:58] <pierzen> it was a pleasure.
[17:56:01] <AndrewBuck> pierzen: yes, at least we know for sure what helps them the most though, the situation may be the same for others as well.
[17:56:02] <education> whatever you chose to do will be helpful in the long run
[17:56:17] <AndrewBuck> yes, we just want to do the priority stuff first if possible.
[17:56:33] <AndrewBuck> Thank you for your time, education.
[17:56:35] <pierzen> the discussion list will be good for that.
[17:56:35] <education> thanks for your time everyone
[17:56:43] <pierzen> thanks to you both.
[17:56:46] <education> will follow up via email
[17:57:24] <education> signing off now
[17:57:26] <education> thanks again!
[17:57:29] <AndrewBuck> Ok, time to adjourn then I think, unless anyone has anything else.
[17:57:43] <AndrewBuck> Goodnight education.
[17:57:49] <scream> thanks education
[17:57:51] <scream> and thanks to you all
[17:57:53] <education> bonne nuit
[17:57:58] <scream> we catch up and we keep updated
[17:58:00] <scream> via email
[17:58:02] <pierzen> Bonne nuit éducation.
[17:58:04] <scream> thanks a gain