--- Log opened Sun Jul 03 00:00:37 2011 00:16 -!- woglinde [~heinold@f052067045.adsl.alicedsl.de] has quit [Quit: zapp] 00:43 -!- Benja1972 [~sergei@227.213.10.109.rev.sfr.net] has quit [Ping timeout: 246 seconds] 01:32 -!- antenna1 [~ion@dslb-094-217-196-207.pools.arcor-ip.net] has quit [Quit: Leaving.] 03:07 -!- noradtux [~quassel@f054018147.adsl.alicedsl.de] has quit [Read error: Operation timed out] 03:09 -!- noradtux [~quassel@g224053115.adsl.alicedsl.de] has joined #navit 03:35 -!- cp15` [ayhuji@p57B1CFE0.dip0.t-ipconnect.de] has joined #navit 03:35 -!- cp15 [zibldl@p57B1CF09.dip0.t-ipconnect.de] has quit [Disconnected by services] 03:35 -!- cp15` is now known as cp15 03:36 -!- mode/#navit [+o cp15] by ChanServ 04:19 -!- laroche [~laroche@dslb-188-104-188-081.pools.arcor-ip.net] has quit [Ping timeout: 276 seconds] 04:20 -!- laroche [~laroche@dslb-178-002-188-042.pools.arcor-ip.net] has joined #navit 04:54 -!- chollya [~chollya@2a02:7200:0:3::150] has quit [Ping timeout: 255 seconds] 05:04 -!- chollya [~chollya@2a02:7200:0:3::150] has joined #navit 05:25 -!- chollya [~chollya@2a02:7200:0:3::150] has quit [Ping timeout: 255 seconds] 05:29 -!- chollya [~chollya@2a02:7200:0:3::150] has joined #navit 05:46 -!- ColdFyre [~lenny@c-98-248-33-157.hsd1.ca.comcast.net] has quit [Ping timeout: 240 seconds] 05:48 -!- sera [~sera@108-55.62-81.cust.bluewin.ch] has joined #navit 05:50 -!- sera_ [~sera@212-207.203-62.cust.bluewin.ch] has quit [Ping timeout: 240 seconds] 05:56 -!- ColdFyre [~lenny@c-98-248-33-157.hsd1.ca.comcast.net] has joined #navit 06:28 -!- Zaba [~zaba@about/goats/billygoat/zaba] has quit [Read error: Operation timed out] 06:38 -!- Zaba [~zaba@about/goats/billygoat/zaba] has joined #navit 07:42 -!- drlizau [~liz@billiau.net] has quit [Remote host closed the connection] 07:54 -!- drlizau [~liz@billiau.net] has joined #navit 08:58 -!- goldfndr` [~goldfndr@h-72-244-200-128.sttnwaho.dynamic.covad.net] has quit [Read error: Connection reset by peer] 09:00 -!- goldfndr` [~goldfndr@h-67-100-127-124.sttnwaho.dynamic.covad.net] has joined #navit 09:09 -!- goldfndr` [~goldfndr@h-67-100-127-124.sttnwaho.dynamic.covad.net] has quit [Read error: Connection reset by peer] 09:10 -!- goldfndr` [~goldfndr@h-72-244-206-51.sttnwaho.dynamic.covad.net] has joined #navit 09:23 -!- chollya [~chollya@2a02:7200:0:3::150] has quit [Ping timeout: 255 seconds] 09:23 -!- chollya [~chollya@2a02:7200:0:3::150] has joined #navit 10:21 -!- Amorphous [jan@unaffiliated/amorphous] has quit [Read error: Connection reset by peer] 10:29 -!- antenna1 [~antenna1@dslb-178-010-087-152.pools.arcor-ip.net] has joined #navit 10:35 -!- antenna1 [~antenna1@dslb-178-010-087-152.pools.arcor-ip.net] has left #navit [] 10:40 -!- Amorphous [jan@unaffiliated/amorphous] has joined #navit 10:57 -!- antenna1 [~ion@dslb-178-010-087-152.pools.arcor-ip.net] has joined #navit 11:30 -!- antenna1 [~ion@dslb-178-010-087-152.pools.arcor-ip.net] has left #navit [] 11:38 -!- khetzal [~quetzal@2001:470:1f13:6e7:6ef0:49ff:fee6:8e93] has joined #navit 11:47 -!- BrainDamage [BrainDamag@i.love.tiltshellz.org] has joined #navit 11:48 < BrainDamage> hello, sorry if I'm asking something probably obvious, but I couldn't find anything on the wiki, nor searching in the program itself, how do I toggle on/off avoidance of toll routes & highway on the android port? 11:50 < drlizau> i'm not sure its obvious - i've not heard anyone ask before 18:29 -!- NavitLogger [~number6@140.203.205.173] has joined #navit 18:29 -!- Topic for #navit: The open source vector based navigation program with routing engine. See http://wiki.navit-project.org. Please be patient if you don't get a response immediately. Please use http://navit.pastebin.com for your pastes! Channel is logged : http://irclogs.navit.ie/ Follow @NavitProject on Twitter. 18:29 -!- Topic set by KaZeR [~kazer@4.170.67.86.rev.sfr.net] [Mon May 23 18:10:05 2011] 18:29 [Users #navit] 18:29 [@ChanServ ] [ CIA-69 ] [ laroche ] [ ricky26 ] 18:29 [@cp15 ] [ ColdFyre ] [ latouche ] [ sebastian-w] 18:29 [@Horwitz ] [ curious ] [ marienz ] [ sera ] 18:29 [@KaZeR ] [ dandor ] [ merlijn ] [ tg ] 18:29 [ Amorphous ] [ dandor_logger] [ Mineque ] [ ThomasAH ] 18:29 [ antenna1 ] [ davem ] [ NavitLogger] [ toer ] 18:29 [ apmon ] [ Earthwings ] [ niekie ] [ tryagain ] 18:29 [ BrainDamage] [ goldfndr` ] [ noradtux ] [ udovdh ] 18:29 [ Brinky ] [ hape ] [ number6 ] [ UukGoblin ] 18:29 [ brot ] [ kenguest_ ] [ Number6_ ] [ Zaba ] 18:29 [ bzed ] [ khetzal ] [ pini ] 18:29 [ chollya ] [ korrosa ] [ playya_ ] 18:29 -!- Irssi: #navit: Total of 46 nicks [4 ops, 0 halfops, 0 voices, 42 normal] 18:29 -!- Channel #navit created Sun Nov 26 06:47:57 2006 18:29 -!- Irssi: Join to #navit was synced in 4 secs 18:29 <@cp15> Oohh... They fixed it 18:29 <@cp15> Quite fast if it really was a fibre break 18:30 < number6> Just got the nagios email \o/ 18:30 -!- number6 [~number6@zoidberg.geoghegan.me] has quit [Quit: leaving] 18:31 -!- Number6_ is now known as Number6 18:31 < NavitLogger> What did I miss? 18:32 < korrosa> ME! 18:32 * NavitLogger cuddles korrosa 18:33 < korrosa> I've closed the two toll road tickets in trac, with the information provided by cp15 earlier in irc (for the benefit of loggers) 18:33 < korrosa> Thanks NavitLogger! 18:34 <@cp15> korrosa, do you want to take care about the tickets? 18:34 < korrosa> You're really trying hard to get me to commit more, aren't you!! :) What would taking care of the tickets mean? 18:35 <@cp15> Well, everything someone else cares about I don't have to 18:36 < korrosa> Lol! Yeah, I'll give it a go! 18:36 < Number6> Taking care of the tickets entails bringing them down to the lake, with a baseball bat and concrete boots 18:36 < korrosa> Just got this: "maptool:read_item:itembin_buffer.c:37 assertion failed:(ib->len+1)*4 < sizeof(buffer)" followed by "Aborted". Just trying again to see if it's a disk space issue again... 18:37 < korrosa> Number6: Now that's something I'm familiar with... 18:37 <@cp15> I would say: Closing them when appropriate, ask the opener for more information if needed, eliminate duplicates 18:37 < korrosa> Maptool issue was after "PROGRESS: Phase 4: generating tiles" 18:37 <@cp15> Most likely some of the temp file was corrupt 18:38 < korrosa> cp15: Ah, OK - yeah, I've been doing that anyway from time to time. I really miss the CIA plugin though for trac - there's no way for me to keep up-to-date with tickets! 18:40 < tryagain> http://trac.navit-project.org/ticket/859 reopened, patch added. 18:40 <@cp15> Ok, and since you will be one of the main users of it: What bug tracker would you prefer? Staying with trac, moving to mantis, redmine, bugzilla... 18:41 < Number6> I have no qualms with trac. Mantis is nice, but I'm not too sure about using it for Navit 18:42 < korrosa> I'll have to have a look at those and let you know. Only other experience is with bugzilla. Whatever we use, we MUST be able to see changes to existing tickets somehow - even if it is through an RSS feed... 18:43 < Number6> I miss CIA-69's ticketing posts 18:43 < tryagain> korrosa you may use http://trac.navit-project.org/query?status=assigned&status=closed&status=new&status=reopened&col=id&col=summary&col=status&col=type&col=priority&col=milestone&col=component&col=changetime&desc=1&order=changetime 18:43 < tryagain> but CIA notifies were quite useful 18:44 < tryagain> kazer said he's going to fix them 18:44 < korrosa> Ah, and use the rss for that, I guess. That'll do for now! 18:44 < korrosa> Thanks! 18:44 < korrosa> tryagain: I still have maptool patched with the previous patches from 859 - will the new patch play nice? 18:48 < tryagain> korrosa you'll need to rollback previous patch. New one places icons in the center of area, no other difference should be seen. Also map size is a little less than before 'coz extra coordinates being skipped. 18:48 < korrosa> Yeah, the patch did apply, but certainly not cleanly! I'l start afresh! 18:57 < tryagain> btw my patch didn't tested with protobuf and sql input, but designed to support both. 18:58 < korrosa> Well, I only use pbf so I'll let you know! 19:01 <@cp15> tryagain, looks good so far 19:02 <@cp15> A few questions 19:04 <@cp15> In resolve_ways, if it is called for multiple slices, how do you prevent writing out the result multiple times? 19:10 < korrosa> First maptool test: "Floating point exception" then quit. Line before was "OSM Warning:http://www.openstreetmap.org/browse/way/38330476 Expected polygon, but only two points defined 19:10 < korrosa> " 19:13 < korrosa> Trying again with a bz2 file 19:16 -!- mase76 [~mase76@p5DD3A7E6.dip.t-dialin.net] has joined #navit 19:18 -!- mase76 [~mase76@p5DD3A7E6.dip.t-dialin.net] has quit [Client Quit] 19:19 -!- chollya [~chollya@2a02:7200:0:3::150] has quit [Ping timeout: 264 seconds] 19:19 < Number6> gone 19:23 -!- chollya [~chollya@2a02:7200:0:3::150] has joined #navit 19:29 <@cp15> tryagain, in osm_end_way, could you move the new item_bin_add_string to the if (outwaypoi) branch so polygons and lines won't have the same redundant attributes? 19:33 <@cp15> Your method of calculating the centroid seems to be a bit complicated. Just take the average of all coordinates (substract the first coordinate to prevent an overflow when building the sum) 19:35 <@cp15> If you correct these things, disable it by default and make a switch to enable it (for me to try it out and see what happes) I will commit it 19:35 <@cp15> When I have tested it on the planet we can remove the switch and turn it on permanently 19:38 < korrosa> Using a bz2 file also results in a floating point exception. 19:38 <@cp15> Ah, and I see another problem: You feed ways now multiple times to resolve_ways. That will lead to a way being split up at every node if it is a street 19:45 <@cp15> korrosa, can you enable coredumps (ulimit -c unlimited) and then check the core with gdb: gdb maptool core 19:46 < korrosa> Um, maybe?! What did all that mean? 19:47 <@cp15> ulimit -c unlimited: Enable core dumps 19:47 <@cp15> Then run maptool again 19:48 <@cp15> When it crashes it will create a file named "core" or "core.pid" depending on your os 19:48 <@cp15> That is a dump of the process memory 19:48 <@cp15> You can load this into gdb and see what happened when maptool crashed 19:50 < korrosa> OK, you'll have to wait a few minutes as I rolled-back the 859 patch and am trying again, just to check the problem isn't somewhere else... 20:00 < korrosa> Well, those recent changes to maptool means that 1.1GB of free space isn't enough to process a 311MB pbf file using maptool!! 20:01 -!- rd [~rd@p57B497B6.dip0.t-ipconnect.de] has joined #navit 20:16 < rd> cp15, thanks for confirming. I opened a bug report http://trac.navit-project.org/ticket/892 20:18 < korrosa> rd: highway=pedestrian is mapped to "street_pedestrian" in Navit. Is this not in your active layout? 20:19 < korrosa> Also, it seems your ticket is an exact duplicate of http://trac.navit-project.org/ticket/591? 20:32 < rd> korrosa, that for pointing out the duplicate 20:34 < rd> korrosa, I am running r4538, my map is from March. 20:34 < korrosa> No problem! It's also similar to http://trac.navit-project.org/ticket/847. 20:35 < rd> I am wondering why 591 is fixed since free months. 20:35 < rd> My map is from March, but cp15 did confirm that the problem is still there. 20:38 < korrosa> cp15: Does the toll road fix work with toll="yes"? It seems that my no_toll profile still routes over toll roads such as http://www.openstreetmap.org/browse/way/3578235 20:40 <@cp15> korrosa, if you click on the road, go to the globe, then select the street type and select view attributes, is there something with toll listed? 20:41 < korrosa> cp15: No, see http://i.imgur.com/L4omh.png 20:41 <@cp15> Then there is an error... I will check 20:42 <@cp15> Whats that car icon in the bottom right? 20:44 < korrosa> A super secret!! No, that's an icon to show the routing status of Navit in the OSD, but it sometimes shows above the internal menu for some reason. I suppose it's a bug, but it doesn't bother me! 20:44 < korrosa> See http://wiki.navit-project.org/index.php/OSD_Layouts#Netbook_Layout_2 20:46 <@cp15> korrosa, hmm... For me there is an AF_TOLL flag 20:48 < korrosa> Whaaaat?! navit -v gives me navit 0.5.0 4578M 20:52 <@cp15> Why not? 20:54 <@cp15> Your layout is pretty impressive... Sometimes I am impressed what is possible 20:55 < korrosa> Thanks :) 20:56 < korrosa> I'm just installing a clean version of Navit, and will process a new map. I'll let you know if I get the toll flag in the attributes or not... 20:57 <@cp15> For a quick test you can do the following: 20:57 <@cp15> Download http://www.openstreetmap.org/api/0.6/way/3578235/full and save it as full.osm 20:57 <@cp15> run maptool -D You should get: ... flags="0xfff00400:|AF_TOLL|AF_ALL" ... as output 21:00 < korrosa> Yup, I do. I'll check my patched version. Two seconds... 21:02 < korrosa> Ah, OK, for some reason the patched version doesn't. I'll investigate... 21:03 <@cp15> I assume you patched an older version 21:03 <@cp15> The patch won't apply on the newest version cleanly... At least it won't compile 21:04 < korrosa> Yeah, I screwed up. I overwrote my patched version with a clean version of maptool, but that clean source was old svn, not the latest. 21:05 < korrosa> Yup, OK, works now. 21:06 <@cp15> The AF_TOLL flag or even the routing? 21:06 < korrosa> The AF_TOLL flag - I've got a netbook, so a new map takes time!! Patience! 21:11 < rd> korrosa, I updated my map data, still the search does not yet work. Seems that the fix in 591 is incomplete (?) 21:11 < tryagain> cp15 will work on your suggestions. But before I'd like to clarify with centroid algo: just summing points as you suggest will move point from the center. For example, we have a triangle with one rounded corner. Say that corner is presented with 4 short line segments, giving 4 more points. So result will be moved towards rounded corner. 21:12 < korrosa> rd: If cp15 says so, then I'd certainly presume the same. I'd recommend closing your ticket as a duplicate of #591 (I can do this for you), and you re-opening 591 and saying that it doesn't work. 21:16 < tryagain> cp15 I call resolve_ways right after buffer is saved to disk, and later it should not be used... Though will look what really happens, especially if slices=0... 21:17 <@cp15> Well, suppose you have two slices, one with nodes 1-999, second with 1000-1999 21:18 < korrosa> antenna1: Can you point me to where it says Navit uses OSM uses max_speed_flag for routing? I presumed it just used route_weight... 21:19 <@cp15> A way with nodes 10 and 20 will be resolved with the first slice and put to the output 21:19 <@cp15> But it will be read again on the second slice (something might still have unresolved nodes) and it will be detected as resolved again 21:20 <@cp15> Unless I am missing something 21:20 <@cp15> Yes, I do... 21:20 < tryagain> How can nodes 10 and 20 be on the second slice? 21:21 <@cp15> They aren't, but you are using the same input, or am I wrong? 21:21 <@cp15> Have to look at your patch again 21:22 < tryagain> Count will be wrong, but buffer will be reloaded after that. 21:22 < tryagain> Well, I'll need to do one extra load_buffer. Or disable reference count change when outfile is defined in resolve_ways. 21:23 < tryagain> Problem is that first slice is not re-loaded from file between steps 1 and 2. 21:23 <@cp15> Ah, ok I see... Since ib->clen will be 2 it won't process a node again 21:23 < tryagain> steps=phases 21:24 <@cp15> Yes, since it is already in memory 21:25 < tryagain> So I have to do one of following: reload it or disable reference count change when outfile is defined in resolve_ways. 21:26 <@cp15> I would say: Make reference counting an aditional parameter and switch it off when processing the way pois 21:26 < tryagain> ok 21:27 <@cp15> And maybe you could move the way-to-poi conversion out of resolve_ways 21:28 <@cp15> cycle through a temporary file until all slices are processed, then run the result through a ways_to_poi function 21:28 < tryagain> I thought of it. But then there would be two very similar functions without any chance to reuse code. Don't mind? 21:29 <@cp15> Hmm? Which functions? 21:29 <@cp15> My intention is to keep resolve_ways reuseable... Right now it will convert ways from the input to nodes in the output 21:31 < tryagain> Only if way on input is has type < type_line... For real ways it will work as should. 21:32 <@cp15> Ah... Right, your input is already POIs with multiple coordinates 21:33 <@cp15> You seem to know maptool now better than me 21:33 < tryagain> :) 21:36 < korrosa> cp15: Routing avoiding that toll road works brilliantly now! 21:37 -!- dandor [~fodoragi@91-82-68-226.pool.invitel.hu] has quit [Remote host closed the connection] 21:37 < rd> korrosa, I would appreciate, if you can reopen 591 and reference the data in 892 there. Need to leave for today. 21:37 < tryagain> cp15 So just add one more argument to resolve_ways to switch reference counting on/off? 21:39 <@cp15> Yes 21:39 < tryagain> What with centroid algo? Do you like to test simplier approach? I thought we have IO bottleneck, not CPU one for maptool. 21:40 <@cp15> Keep your algo. But maybe you can find out with korrosa to find out where the floating point exception comes from 21:40 -!- drlizau [~liz@billiau.net] has joined #navit 21:41 < tryagain> Will try. If that is not his HDD space ;) 21:42 <@cp15> Might be, I see no floating point operations in your code 21:42 < korrosa> tryagain, cp15 - it may very well be my HDD space. I'll re-patch the up-to-date maptool and see what happens. 21:43 <@cp15> Ok 21:43 < korrosa> cp15: Back to tolls - is there a quick, dirty way to have toll roads indicated in the map layout? 21:43 < tryagain> Division by zero gives something similar. But I hope I get rid of it in any situation. 21:47 -!- khetzal [~quetzal@2001:470:1f13:6e7:6ef0:49ff:fee6:8e93] has quit [Remote host closed the connection] 21:50 < korrosa> tryagain: Does your patch take into account the latest additions to maptool concerning o5m? The patch has 1 chunk which fails: http://pastebin.com/LR6PC0e5 21:53 < korrosa> Hmmm, when trying to compile after manually patching I get: osm_o5m.c:307: error: too few arguments to function ‘osm_end_way’ 21:56 < korrosa> tryagain: Can you provide a patch against latest svn? 21:56 <@cp15> change osm_end_way(out_ways); to osm_end_way(out_ways, NULL); in osm_o5m.c line 307 for a quick fix 21:57 < tryagain> Will fix it later. My working copy now is not current svn. 21:57 < tryagain> Now bed is calling... 21:58 -!- tryagain [3e21ef88@gateway/web/freenode/ip.62.33.239.136] has quit [Quit: Page closed] 22:00 < korrosa> OK, make worked, now re-compiling the map... 22:03 -!- rd [~rd@p57B497B6.dip0.t-ipconnect.de] has quit [Remote host closed the connection] 22:17 -!- mase76 [~mase76@p5DD3A7E6.dip.t-dialin.net] has joined #navit 22:17 < korrosa> OK, another floating point exception, and I have more than enough space!! I'll get the core dump 22:19 -!- mase76 [~mase76@p5DD3A7E6.dip.t-dialin.net] has quit [Client Quit] 22:19 < korrosa> cp15: I'm in the gdb terminal. What do I do now? 22:19 -!- mase76 [~mase76@p5DD3A7E6.dip.t-dialin.net] has joined #navit 22:19 < korrosa> Last few lines: Program terminated with signal 8, Arithmetic exception. 22:19 < korrosa> #0 0x080b27db in __divdi3 () 22:20 <@cp15> What comes after #0 22:20 -!- mase76 [~mase76@p5DD3A7E6.dip.t-dialin.net] has quit [Client Quit] 22:20 -!- mase76 [~mase76@p5DD3A7E6.dip.t-dialin.net] has joined #navit 22:20 -!- mase76 [~mase76@p5DD3A7E6.dip.t-dialin.net] has quit [Client Quit] 22:21 < korrosa> cp15: This is the terminal: http://i.imgur.com/EsV3O.png 22:22 <@cp15> Please type in "bt" 22:22 < korrosa> cp15: http://pastebin.com/QfapiJNz 22:23 <@cp15> Interesting... 22:23 < antenna1> korrosa: I may be wrong here, but I thought it depends on maxspeed handling in vehicleprofile what speed attribute is used. 22:23 -!- drlizau [~liz@billiau.net] has quit [Remote host closed the connection] 22:24 <@cp15> Heh... Wow... You seem to have found a polygon which has an area of 1 22:25 <@cp15> korrosa, can you please change geom.c line 99 from 22:25 <@cp15> if(area!=0) { 22:25 <@cp15> to 22:25 < korrosa> antenna1: Ah yes, I see! 22:25 < korrosa> cp15: I aim to please! 22:25 <@cp15> if(area>1) { 22:26 <@cp15> I will read the log tomorrow 22:27 <@cp15> Good night 22:27 < korrosa> OK, done, re-building 22:27 < korrosa> Good night. 22:30 < antenna1> korrosa: nevertheless, the point is that the road with the higher speed limit is not necessarily the fastest way to go, due to traffic lights. 22:30 < korrosa> antenna1: Definitely agree! 22:31 < antenna1> korrosa: if you are going to test tunnel extrapolation, please log your gps with gpspipe, so cp is able to follow your route, if it doesn't work. 22:31 < antenna1> I always forget about it ;) 22:33 < korrosa> I tend to use gpxlogger - is gpspipe better? 22:35 < korrosa> Oh, OK, gpspipe is for the raw nmea. Makes sense! 22:45 -!- Benja1972 [~sergei@227.213.10.109.rev.sfr.net] has joined #navit 23:03 < korrosa> OK, cp15, tryagain: the fix cp15 suggested has produced a map which works fine. No floating point exceptions. The file size has grown by 3MB over an unpatched maptool... 23:05 < korrosa> Well, 2.3MB from 185.7MB to 188MB, so a ~1.2% increase... seems OK! 23:05 < korrosa> Night all 23:05 -!- korrosa [~netbook@2.96.82.9] has quit [Quit: Leaving] 23:49 -!- antenna1 [~ion@dslb-178-010-087-152.pools.arcor-ip.net] has left #navit [] --- Log closed Mon Jul 04 00:00:37 2011