#OSGEO IRC Log - 2010-02-01

For logs after Feb 3, 2007, all times are GMT-8. Prior logs are GMT-9.
Back to Logs
00:13:36 CIA-84: osgeo: alfonx * r3031 /livedvd/gisvm/trunk/doc/descriptions/geopublishing_licence.html: Added licence information for geopublisher and atlasstyler
00:13:38 sigq: osgeofeed: Changeset [3031]: Added licence information for geopublisher and atlasstyler <http://trac.osgeo.org/osgeo/changeset/3031>
00:59:55 sigq: osgeofeed: Changeset [3032]: First version of install script, definition and description. <http://trac.osgeo.org/osgeo/changeset/3032>
01:00:15 CIA-84: osgeo: jldominguez * r3032 /livedvd/gisvm/trunk/ ( 3 files in 2 dirs ): First version of install script, definition and description.
01:51:57 sigq: osgeofeed: Ticket #519 ( task closed ): livedvd: new way to set desktop background <http://trac.osgeo.org/osgeo/ticket/519#comment:3>
03:04:32 CIA-84: osgeo: hamish * r3033 /livedvd/gisvm/trunk/bin/main.sh: fix hog list units
03:04:33 sigq: osgeofeed: Changeset [3033]: fix hog list units <http://trac.osgeo.org/osgeo/changeset/3033>
03:05:54 CIA-84: osgeo: hamish * r3034 /livedvd/gisvm/trunk/bin/main.sh: typo
03:05:55 sigq: osgeofeed: Changeset [3034]: typo <http://trac.osgeo.org/osgeo/changeset/3034>
03:31:06 sigq: osgeofeed: Changeset [3035]: repair patch <http://trac.osgeo.org/osgeo/changeset/3035>
03:31:39 CIA-84: osgeo: hamish * r3035 /livedvd/gisvm/trunk/app-data/gpsdrive/gpsdrive_fix_deps.patch: repair patch
03:35:33 sigq: osgeofeed: Ticket #519 ( task reopened ): livedvd: new way to set desktop background <http://trac.osgeo.org/osgeo/ticket/519#comment:4>
03:45:34 sigq: osgeofeed: Changeset [3036]: group java installers together <http://trac.osgeo.org/osgeo/changeset/3036>
03:45:49 CIA-84: osgeo: hamish * r3036 /livedvd/gisvm/trunk/bin/main.sh: group java installers together
04:09:14 CIA-84: osgeo: alfonx * r3037 /livedvd/gisvm/trunk/bin/install_geopublishing.sh: added "--no-install-recommends" to apt-get... if we want openoffice on the dvd it should not be because of a geopublisher dependency, but be installed explicitly.
04:27:16 sigq: osgeofeed: Ticket #528 ( task created ): Barcelona Local Chapter mailing list <http://trac.osgeo.org/osgeo/ticket/528>
05:03:01 : -lindbohm.freenode.net- [freenode-info] channel trolls and no channel staff around to help? please check with freenode support: http://freenode.net/faq.shtml#gettinghelp
05:32:28 sigq: osgeofeed: Ticket #527 ( task closed ): New quebec-comite and quebec-conf lists <http://trac.osgeo.org/osgeo/ticket/527#comment:2>
06:14:52 CIA-84: osgeo: pieterdg * r3038 /livedvd/gisvm/trunk/bin/main.sh: Incorrect filename was called for the geomajas install file ( geomajas_install instead of geomajas_install.sh )
06:20:20 FrankW: aboudreault: go ahead with the Barcelona mailing list. No need to wait for objections.
06:20:46 FrankW: Perhaps you could take care of http://trac.osgeo.org/osgeo/ticket/527 as well.
06:20:47 sigq: Title: #527 ( New quebec-comite and quebec-conf lists ) - OSGeo - Trac ( at trac.osgeo.org )
06:37:37 sigq: osgeofeed: Changeset [3039]: added academic track page and opening registrations <http://trac.osgeo.org/osgeo/changeset/3039>
06:38:26 CIA-84: osgeo: lbecchi * r3039 /foss4g/2010/website/ ( 4 files ): added academic track page and opening registrations
07:10:31 TylerM: happy monday! : )
07:11:20 seven: Morning!
07:12:58 TylerM: heya seven - what's on your radar this week?
07:15:47 seven: Work through all the todo's on the board list.
07:15:52 seven: They accumulated a bit...
07:16:12 seven: Then focus on sponsorship acquisition and alternative funding schemes.
07:16:25 seven: Frank suggested to move discussion ot the public Wiki.
07:16:53 seven: Can you add the achievements of our last funding board meeting to a new WIki page?
07:17:15 TylerM: You mean to make some categories for further discussion?
07:24:57 seven: We need the list of things that OSGeo might want to fund itself from. Including FOSS4G down to merchandising.
07:25:25 seven: And also tricky bits like offering "education resources" , see 501 3( c ) application.
07:25:38 seven: I am not sure whether they fit in with our agenda too well.
07:26:14 seven: I noticed that when I answered the questionaire.
07:31:56 TylerM: seven: will do
07:32:12 seven: Cool. Thank you.
08:06:42 StormTide: anyone know of a way to dump a shapefile of points into a csv?
08:07:00 StormTide: like lat/long/attribute1...
08:10:24 nhv: ogr2ogr http://www.gdal.org/ogr2ogr.html
08:10:25 sigq: Title: GDAL: ogr2ogr ( at www.gdal.org )
08:11:11 nhv: http://www.gdal.org/ogr/drv_csv.html
08:11:12 sigq: Title: Comma Separated Value ( .csv ) ( at www.gdal.org )
08:11:42 StormTide: aha excellent, thanks
08:14:18 seven: This is the coolest Geospatial One-Stop shop I know. :- )
08:18:26 TylerM: StormTide: let us know if you get stuck.. many of us do ogr2ogr commands in our sleep ;- )
08:18:47 TylerM: and some in the room even write the program :- ) hehe
08:19:12 StormTide: TylerM, thanks, will do, seems straightforward sofar
08:19:37 StormTide: im projecting to wgs84 at the shpfile creation time, so i dont think there'll be any issues with the co-ords being right
08:20:02 TylerM: you can do that transformation at the same time with ogr2ogr as well, if need be
08:22:36 StormTide: TylerM, hrm, ok maybe this is a lil more complicated
08:22:45 StormTide: TylerM, the shpfile i want is actually attributed areas
08:22:54 StormTide: i want to export the centroid as point in the csv
08:24:43 nhv: what are you using to create the shapefiles
08:24:59 StormTide: quantumgis...
08:25:06 StormTide: comes from a s57 enc from noaa
08:25:09 StormTide: originally that is
08:26:00 nhv: should have realized that given your nick
08:26:24 nhv: easiest ways I know of getting that is to either
08:26:40 nhv: use postgis or script a little python
08:27:05 StormTide: can ogr dump to postgis?
08:27:50 StormTide: er nm, can use shp2pgsql
08:28:19 nhv: or ogr2ogr
08:28:25 nhv: can do it
08:33:54 StormTide: nhv, when i try PGL:dname=mydb ... i get a postgresql driver doesnt currently support database creation error... but that database exists and is intialized with postgres
08:33:58 StormTide: with postgis even
08:35:21 StormTide: nm, i see the format of the connection string is changed to PG:"spec"
09:02:52 TylerM: seven: done basic braindump http://wiki.osgeo.org/wiki/Fundraising_2010
09:02:54 sigq: Title: Fundraising 2010 - OSGeo Wiki ( at wiki.osgeo.org )
09:03:34 TylerM: hope its what you were after
09:17:41 : * TylerM might drown under foss4g email pile soon ;- )
09:21:19 m_k: hi, does anyone know what happened to mapserver workshop? http://2009.foss4g.org/workshops/#workshop_02
09:21:21 sigq: Title: FOSS4G 2009 Conference, Sydney 20-23 October 2009 ( at 2009.foss4g.org )
09:21:29 m_k: the link goes to an empty dir http://download.osgeo.org/osgeo/foss4g/2009/SPREP/0Tue/Parkside%20GO3/0900/
09:21:30 sigq: Title: Index of /osgeo/foss4g/2009/SPREP/0Tue/Parkside GO3/0900 ( at download.osgeo.org )
09:22:27 TylerM: good question, not sure
09:46:36 nhv: m_k have you seen this http://mapserver.org/tutorial/index.html
09:46:37 sigq: Title: MapServer Tutorial MapServer 5.6.1 documentation ( at mapserver.org )
10:27:17 m_k: nhv: yes, I use mapserver docs page quite often - I was just wondering if the workshop was somehow different ; )
10:28:08 nhv: well the guy who gave the workshop is the guy who wrote the tutorial :- )
10:28:43 m_k: ah : )
10:35:09 CIA-84: osgeo: jldominguez * r3040 /livedvd/gisvm/trunk/ ( 4 files in 2 dirs ): Minor changes. Should work now. Doc needs love.
10:35:10 sigq: osgeofeed: Changeset [3040]: Minor changes. Should work now. Doc needs love. <http://trac.osgeo.org/osgeo/changeset/3040>
10:36:20 TylerM: oh heh - sorry m_k - I was the copresenter for that workshop : )
10:36:26 TylerM: didn't realise which one you were after : )
10:36:29 : * m_k imported openstreetmap data into postgis and made osm wms in mapserver last week
10:36:29 m_k: very cool results
10:36:40 TylerM: good work! : )
10:37:54 TylerM: not sure where the docs on that site went to - but they are almost identical to ones used previously http://2007.foss4g.org/workshops/W-09/
10:37:55 sigq: Title: FOSS4G 2007 : W-09: Getting Started with MapServer ( at 2007.foss4g.org )
10:38:23 TylerM: Perry's tutorial stuff is half the workshop, then part #2 is different.
10:41:41 m_k: thanks
10:53:22 m_k: BTW we ( the company I work for ) have more than 100 databases in postgis and over 50( ? ) different .map files in mapserver ; )
10:53:33 m_k: on a single machine
10:55:46 TylerM: m_k: excellent : ) I'd love to see it all in action sometime : )
10:58:46 m_k: http://halinow.serverwms.pl/ we do many services like this one ( top 3 layers are from our wms )
10:58:47 sigq: Title: Serwis Mapowy Gminy Halinów ( at halinow.serverwms.pl )
10:58:54 m_k: it's in polish, but... ; )
11:00:59 m_k: http://klembow.e-mapa.net/ this has osm layer
11:01:00 sigq: Title: Serwis Mapowy Gminy Klembów ( at klembow.e-mapa.net )
17:38:47 wildintellect: FrankW, the short of my incoming email, slight tweaks to the quotes and we will be under budget no problem
17:39:24 FrankW: excellent.
17:46:32 wildintellect: I'm wondering what you think of the idea of using Telescience as an offsite backup
17:46:59 FrankW: Hmm, I'm disappointed that we aren't using OCS for managing presentations for FOSS4G this year.
17:47:24 wildintellect: like a weekly rsync of the backups kept on one of the new servers
17:48:03 FrankW: I think that is quite reasonable.
17:48:42 wildintellect: how many snapshots would we expect to keep
17:49:03 wildintellect: from reading the peer1 setup it's not clear to me if they only keep the previous day or x number of previous days
17:49:18 FrankW: Are you referring to the peer1 tivoli system?
17:50:10 wildintellect: I can only assume based on http://wiki.osgeo.org/wiki/SAC:Backups
17:50:12 sigq: Title: SAC:Backups - OSGeo Wiki ( at wiki.osgeo.org )
17:50:33 FrankW: There are different backup details for different services.
17:51:08 FrankW: The tivoli backups are the "Daily PEER1 Backup" on that page.
17:51:22 FrankW: They are currently disabled, and they were opaque at the best of times.
17:51:41 FrankW: More critical is the dumps/compression/rsync backups we do service by service to osgeo2.
17:51:55 FrankW: That is what I believe we would want to replicate in some appropriate fashion.
17:52:04 FrankW: ( ( IMHO )
17:52:09 wildintellect: what I'm trying to wrap my head around is how many backups would we want archived
17:52:23 wildintellect: ie how far back in time might we need to go to rescue something
17:52:44 FrankW: I'll do a bit of digging to better understand what we do now.
17:53:19 wildintellect: thanks
17:53:44 FrankW: For svn our primary backup is an svnmirror. SVN has full history internally though we could get screwed if the database was corrupted and the corrupt version propogated
17:54:37 FrankW: For the mailing lists we essentially keep just one backup copy - the most recent rsync of the whole tree.
17:55:13 FrankW: likewise for 1GB of /var/www/html
17:55:23 FrankW: and /etc
17:55:51 FrankW: Hmm, i'm not seeing any backups for mysql ( backing drupal, etc ).
17:56:59 FrankW: It looks like we keep roughly two weeks of daily Trac snapshows, about 400MB for each day.
17:57:23 wildintellect: wow that will be real easy on the new setup
17:57:36 FrankW: I'm a bit weirded out that I can't see any mysql backups.
17:57:42 wildintellect: I was ballpark reserving 1TB for backup space
17:57:45 FrankW: But that isn't my area.
17:58:03 wildintellect: page says they should be in /var/lib/mysql-zrm/daily
17:58:43 FrankW: /home/back/mysql-zrm/daily on osgeo2 ( the backed up copy ) is empty.
17:59:07 FrankW: There may be saved copies on osgeo1, but if that server were pooched, it does not look like we could recover the content.
18:00:00 FrankW: In the future we might also want to be backing up entire VMs rather than just rsyncing selected data.
18:00:04 FrankW: At least within OSL.
18:00:19 wildintellect: right, I still need to have that discussion with them
18:00:45 wildintellect: live backup of kvm does seem possible with something called lvm snapshots according to my research
18:00:53 FrankW: I would note that we have been *plagued* by performance problems of our services while backups are running due to io load.
18:01:03 FrankW: I can imagine that would also be an issue backing up whole VMs.
18:01:30 FrankW: I'm a bit worried that with virtualization it will just be harder to understand the issue when we have io contention.
18:01:36 wildintellect: the biggest problem with backup of VMs is finding a way that doesn't require turning them off
18:01:56 wildintellect: actually vm should solve the io issue
18:02:09 FrankW: Really? Is there some sort of magic involved?
18:02:15 wildintellect: by partitioning the actions, certain practices are capped
18:02:34 wildintellect: if you give a vm 1 core and 16GB of ram, that's all it's allowed to use
18:02:49 FrankW: That doesn't address io contention.
18:03:11 wildintellect: you talking about at the cpu level or data flow in and out of the servers
18:03:22 FrankW: I am talking about IO to disk.
18:03:39 wildintellect: ah, well for that I'm proposing raid 6
18:03:44 FrankW: For instance, a big SVN operation might so whack the disks that everything else starts bogging down seriously.
18:04:10 FrankW: And we were seeing the backup daemons, database dumps, etc really whacking the disk and bogging things down.
18:04:26 wildintellect: do we have raid on the current machines?
18:04:28 FrankW: I don't understand what magic raid 6 provides other than somewhat increased io performance.
18:04:34 FrankW: I frankly don't know.
18:04:40 wildintellect: raid 6 is striped
18:04:47 wildintellect: like raid 5
18:05:11 bwoodall: I think it is 6 = extra parity
18:05:27 bwoodall: so you can loose one then one drive and recover
18:05:29 wildintellect: yes it's RAID 5 with the ability to lose 2 drives instead of 1
18:05:37 bwoodall: so you can loose more then one drive and recover
18:05:46 wildintellect: yes you can loose 2
18:05:56 bwoodall: howdy WI
18:06:14 wildintellect: also we opted for 10K rpm SAS drives
18:06:26 bwoodall: : )
18:06:33 wildintellect: should be much faster than SATA
18:06:46 wildintellect: I think they are rated at 6 GB/s +
18:06:56 FrankW: I don't see how this prevents io overload of one operation ( like backups ) causing other stuff to bog down.
18:07:20 wildintellect: might depend on how we separate them
18:07:35 FrankW: possibly.
18:08:00 wildintellect: I'm thinking trac/svn will be on a different server from the backup
18:08:09 FrankW: Anyways, in my opinion that has been our biggest headache in the past, and with several VMs running on one set of physical disks, it seems like things could get even more opaque.
18:08:27 FrankW: It's the backing up of the live server( s ) that kills, not writing to the destination server.
18:08:35 wildintellect: I'm estimating about 2-3 vms per physical machine right now
18:08:55 FrankW: By the way, I'm agog at the suggested amounts of RAM.
18:09:20 FrankW: If nothing else, we should be able to do god-like disk caching of actually used stuff.
18:09:27 wildintellect: yes
18:10:11 FrankW: do you have much experience working with VMs in high performance situations?
18:10:23 wildintellect: no, only light use so far
18:10:31 FrankW: I don't ( usually them essentially for development/hobbiest purposes )
18:10:51 FrankW: I'm wondering if the OSL guys have more experience in that regard.
18:11:02 wildintellect: they seem to
18:11:17 wildintellect: they did say they prefer kvm and run the base host on gentoo
18:11:25 wildintellect: which to me says they know what they're doing
18:11:31 FrankW: I understand some of the VM systems support cheap checkpointing. Essentially a way of cheaply capturing the running state of the service at a point in time that would be useful for backups.
18:11:52 wildintellect: yes thats similar to the lvm snapshot I was talking about
18:12:10 FrankW: Well, unfortunately I don't know of anyone on SAC who has mentioned any experience with kvm or gentoo, but hopefully it will be fine.
18:12:20 wildintellect: OSL would manage that part
18:12:29 wildintellect: we don't have to touch kvm at all
18:12:43 wildintellect: except to load our images for our vms
18:12:46 FrankW: Do you have an idea on when we could issue an order for the hardware, and how long it would take to fulfill?
18:13:06 wildintellect: dell is pretty quick I have a timeline in my spreadsheet
18:13:16 wildintellect: I think we can order in the next 1-2 weeks
18:13:18 FrankW: I'm keen to get this in place. I'm very pleased that you are stepping in since I know hobu was having trouble finding the time to followup.
18:13:26 wildintellect: so 1 month before setup starts
18:13:35 FrankW: ok, I can live with that.
18:13:59 wildintellect: we do have some important planning of resources to figure out before then
18:14:22 wildintellect: what are we doing with the existing hardware?
18:15:38 FrankW: Is that a specific question?
18:16:05 FrankW: I did look over a table of the various services and it seems reasonably accurate and comprehensive.
18:16:52 wildintellect: did you look at my ideas sheet
18:17:09 FrankW: Yes, I think that was the table I was looking at. Can you push out the url for me again?
18:17:30 : * FrankW would prefer that this material be collected in the SAC wiki rather than google docs.
18:18:21 wildintellect: I plan to make it a wiki page once it's done, I am using calculations though for some stuff
18:18:51 wildintellect: https://spreadsheets.google.com/ccc?key=0AifEHeRRT620dHF0ZDExSXFzSVlSQWxVcXRqUnBOQ1E&hl=en
18:18:53 sigq: Title: Welcome to Google Docs ( at spreadsheets.google.com )
18:19:30 wildintellect: well that and much of the information is from the wiki pages but collected in a way that I can grasp the quantities - table editing in wikis is a nightmare
18:20:25 FrankW: I can understand that, though there doesn't seem to be a lot of quantative values.
18:21:07 wildintellect: well, turns out I can't find those for the current systems
18:21:24 FrankW: What quantities are you looking for?
18:21:27 wildintellect: not without getting someone to log in and run some commands to find out cpu/ram etc
18:21:44 FrankW: Would you like an account and root access on osgeo1?
18:21:56 wildintellect: df -h, /proc/cpuinfo, amount of ram from dmidecode
18:22:27 wildintellect: hmm maybe ram from top would suffice
18:23:00 FrankW: [warmerdam@osgeo1 ~]$ df -h
18:23:01 FrankW: Filesystem Size Used Avail Use% Mounted on
18:23:01 FrankW: /dev/sda2 224G 108G 106G 51% /
18:23:01 FrankW: /dev/sda1 122M 104M 12M 90% /boot
18:24:18 FrankW: cpuinfo is 4x the following:
18:24:18 FrankW: http://osgeo.pastebin.ca/1775138
18:24:19 sigq: Title: pastebin - osgeo1 cpuinfo - post number 1775138 ( at osgeo.pastebin.ca )
18:24:42 FrankW: Mem: 2074860k total, 1927532k used, 147328k free, 22896k buffers
18:24:43 FrankW: Swap: 2040244k total, 388832k used, 1651412k free, 867980k cached
18:24:46 FrankW: Anything else?
18:25:14 wildintellect: osgeo 1 has 4 cpus that are all dual core?
18:25:45 wildintellect: same info for osgeo 2 would be it
18:26:07 FrankW: I presume that means one dual core cpu with each cpu further segmented into two virtual cpus per hypertreading, but I'm not absolutely positive.
18:26:13 FrankW: osgeo2 is an exact clone of the osgeo1 hardware.
18:26:21 wildintellect: ah ok
18:26:28 wildintellect: wow only 2 GB of RAM
18:26:33 StormTide: any of you guys familiar with route-me at all?
18:26:41 FrankW: wildintellect: yes, that kills us.
18:26:49 FrankW: But upgrades are punatively expensive with peer1.
18:26:55 wildintellect: no kidding
18:26:58 FrankW: StormTide: not me.
18:27:21 FrankW: Though I must confess I've been quite satisfied with some aspects of their services.
18:28:31 FrankW: By the way, for "service segmentation" purposes, I wonder if we might consider making the LDAP server it's own small VM as our "security core". I'd hate to have it compromised by other sloppy software services.
18:28:49 wildintellect: I had thought about that
18:29:11 wildintellect: For now I have it lumped with the Mail Service
18:29:21 wildintellect: as those both seemed low on resource needs
18:29:30 FrankW: We do move quite a bit of mail!
18:29:35 FrankW: Did you mean mailman?
18:29:53 wildintellect: I mean both mailman and the @osgeo.org email routing
18:29:59 FrankW: ok
18:30:03 wildintellect: so mailman and postfix was it
18:30:20 FrankW: I'm frequently surprised how busy the mailman list delivery stuff is even when I shut down apache and all it's related services.
18:48:52 wildintellect: hmm, I managed to theoretically allocate all the new space already, lol
19:01:21 TylerM: FrankW: no clue on the backup costs.. I'm thinking a couple hundred bucks maybe..
19:01:30 TylerM: can't seem to find those old emails. maybe their online system
19:01:41 wildintellect: might not matter if we migrate off in 1-2 months
19:01:42 TylerM: says it but I never log in : )
19:02:17 TylerM: hi Alex.. probably right
19:02:26 TylerM: never had a company harass me to reduce my costs before ;- )
19:02:34 TylerM: hobu: yeah i got a call too
19:03:09 wildintellect: lol, well from what I understand peer1 is odd
19:03:30 wildintellect: it's probably costing them in some way to handle the backup so they'd rather not
19:03:47 TylerM: we keep killing their backup app.. heh
19:03:50 wildintellect: like how the power company gives you incentives to not use electricity
19:03:56 TylerM: or at least it's off for a long time now : )
19:03:59 TylerM: heh yeah
19:05:55 TylerM: FrankW: re: OCS - sorry, I had to give up propping it up on my own. without someone else helping out or taking responsibility it was too much to adapt to new demands each year.
19:06:20 TylerM: every system seemed to neglect one small but important-to-current-organisers function ;- )
19:06:45 FrankW: TylerM: Was there a discussion about OCS with the 2010 LOC?
19:06:53 TylerM: you know when I'm hacking php that we're all in trouble ; ) heh I wrote patches for OCS in 2008 to handle a few of the LOC requirements.
19:07:11 TylerM: Only my telling Lorenzo what I just said above : )
19:08:40 TylerM: They'd have to take it through a few test runs first to ensure that it meets their needs.. that was the trouble previously :[
19:09:10 TylerM: having the mgt company running their own system just means asking for something.. an approach that works better for some LOCs than others ; )
19:09:38 TylerM: bbl
19:10:38 FrankW: Given that I wasn't prepared to be the one supporting OCS it is likely best I just back off.
19:19:47 CIA-84: osgeo: hamish * r3041 /livedvd/gisvm/trunk/doc/descriptions/ ( 6 files ): software versions
19:21:17 sigq: osgeofeed: Changeset [3041]: software versions <http://trac.osgeo.org/osgeo/changeset/3041>
19:36:16 CIA-84: osgeo: hamish * r3042 /livedvd/gisvm/trunk/doc/descriptions/ ( 5 files ): svn props
19:37:20 sigq: osgeofeed: Changeset [3042]: svn props <http://trac.osgeo.org/osgeo/changeset/3042>
19:49:07 CIA-84: osgeo: hamish * r3043 /livedvd/gisvm/trunk/README.txt: current structure
19:49:10 sigq: osgeofeed: Changeset [3043]: current structure <http://trac.osgeo.org/osgeo/changeset/3043>
23:23:55 CIA-84: osgeo: hamish * r3044 /livedvd/gisvm/trunk/bin/install_ugvsigmobile.sh: svn prop
23:23:59 sigq: osgeofeed: Changeset [3044]: svn prop <http://trac.osgeo.org/osgeo/changeset/3044>