I noted the bit in the map defiition which says I can can include names of surveys in the map command so that the centrelines are drawn. I tried this with the terikan survey and therion just says 'error' in the compiler and doesn't say what. Running Therion standalone shows that it is segfaulting. You should be able to demonstrate this with the dataset I uploaded last night. Comment out the join soundrivercrab_s1 soundriver2_s1 line so it runs,then add the name of a survey (I tried 'river' and 'goon') to theend of the scrap list. Now it crashes. I noticed the bit about -projection being required 'if the map contains survey', but adding "-projection plan" to the map command didn't help. This looks like a bug. And is there some subtlty about adding surveys (centrelines) to maps? Wookey -- Aleph One Ltd, Bottisham, CAMBRIDGE, CB5 9BA, UK Tel +44 (0) 1223 811679 work: http://www.aleph1.co.uk/ play: http://www.chaos.org.uk/~wookey/
Wookey wrote:
This looks like a bug. And is there some subtlty about adding surveys (centrelines) to maps?
Yes, it is a bug. I have downloaded your data, made your changes and it really fails - no idea why, but I will fix it. If you would like to have map with centerline also, create a separate map with centerline only - it should work with it (at least on my machine it worked - see attached picture). Regards, S.
+++ Stacho Mudrak [05-01-11 10:08 +0100]:
Wookey wrote:
This looks like a bug. And is there some subtlty about adding surveys (centrelines) to maps?
Yes, it is a bug. I have downloaded your data, made your changes and it really fails - no idea why, but I will fix it.
If you would like to have map with centerline also, create a separate map with centerline only - it should work with it (at least on my machine it worked - see attached picture).
OK. Done that and if I get rid of all the scraps then I get an image which has complete lines for the surface surveys but station-marks for the underground stations. Is that expected? See http://wookware.org/centrelineonly.pdf However when I put the scraps back in then I only see station inside scraps,not any stations or lines on centrelines which don't have correspondging scraps. Any idea why? I'd really like to get this owrking as it's the easiest way to see which bits I haven't done yet. I have noticed a problem with the need for a 'dummy' survey outside a .3d file import - it stops you using the suggested index.th and index.th2 files for centreline and scrap info (because you need a 'survey dummy' in both files around everything and it complains about 'duplicate survey'). This will go away when we get rid of the need for the dummy survey (maybe you already did that - I can't find the mail right now (I'm using the 20050209 snapshot)). I've also realsied that I can't do what I wanted to do - import more than one .3d file. I wanted to import one per cave, but of course thinking about it, this isn't going to work because the co-ordinates will overlap. I need to generate one .3d file covering al the caves. (no problem except for the need to strip extra hierarchy layers. I wonder if this is going to cuase problems when I want to be able to produce both one large overview survey and various small single-cave surveys. I need to think about this... Still it would be good if it gave a useful error - it took me some time to work out what the problem was as it just gave a tex overflow error - which I can't reproduce right now due to getting a different error: It says: therion: error -- file import -- index.th [15] -- invalid survey name -- de0/.deception I think it is faiing to cope with the oddly-name "de0/" survey. It has this peculiar name because the data is imported from compass and it was expedient to use this name. Survex eats it happily and so does aven so therion should. As ever current dataset here: http://wookware.org/terikan20050217.tgz you need to change index.th at top level to use terikan/terikan.3d and deception/deception.3d to illustrate the above points (I changed it to a single overall dataset to check it worked - it does - very cool - except that I can't get centrelines to appear - indeed it doesn't seem to mattermuch what I put in the various 'map' options I seem to get 'all scraps mentioned in any map but no centrelines' whatever I do?). Wookey -- Aleph One Ltd, Bottisham, CAMBRIDGE, CB5 9BA, UK Tel +44 (0) 1223 811679 work: http://www.aleph1.co.uk/ play: http://www.chaos.org.uk/~wookey/
OK. Done that and if I get rid of all the scraps then I get an image which has complete lines for the surface surveys but station-marks for the underground stations. Is that expected?
Yes. This is only because of symbol definition of line survey:underground. This symbol just puts marks on the stations. It can be changed easily.
See http://wookware.org/centrelineonly.pdf
However when I put the scraps back in then I only see station inside scraps,not any stations or lines on centrelines which don't have correspondging scraps. Any idea why? I'd really like to get this owrking as it's the easiest way to see which bits I haven't done yet.
When therion has no scraps to export, it exports all centerline data. If it has some, it exports only 2D data and not centerline. If you wish to have also centerline data on the map, you need to do two things: 1. create some toplevel map, where you mix maps and surveys together, you would like to have on the output. 2. select and export this map
I have noticed a problem with the need for a 'dummy' survey outside a .3d file import - it stops you using the suggested index.th and index.th2 files for centreline and scrap info (because you need a 'survey dummy' in both files around everything and it complains about 'duplicate survey').
This will go away when we get rid of the need for the dummy survey (maybe you already did that - I can't find the mail right now (I'm using the 20050209 snapshot)).
dummy survey is still needed :( If you do not want to have it, use import inside terikan survey with option "-filter terikan"
Still it would be good if it gave a useful error - it took me some time to work out what the problem was as it just gave a tex overflow error - which I can't reproduce right now due to getting a different error:
It says: therion: error -- file import -- index.th [15] -- invalid survey name -- de0/.deception
I think it is faiing to cope with the oddly-name "de0/" survey. It has this peculiar name because the data is imported from compass and it was expedient to use this name. Survex eats it happily and so does aven so therion should.
OK - adding / as a special survey character should be OK - I think. I will enable it right now.
As ever current dataset here:
http://wookware.org/terikan20050217.tgz
you need to change index.th at top level to use terikan/terikan.3d and deception/deception.3d to illustrate the above points (I changed it to a single overall dataset to check it worked - it does - very cool - except that I can't get centrelines to appear - indeed it doesn't seem to mattermuch what I put in the various 'map' options I seem to get 'all scraps mentioned in any map but no centrelines' whatever I do?).
Once more - if you want to mix centerline with maps, you need to define a top level map, where you will mix maps and survey names together. When exporting this map - centerline should appear. If you wish to view whole centerline for underground, just redefine symbol in layout: code mpost let l_survey_cave = l_survey_surface_SKBB; endcode S. -=x=- Skontrolované antivírovým programom NOD32
OK. Done that and if I get rid of all the scraps then I get an image which has complete lines for the surface surveys but station-marks for the underground stations. Is that expected?
Second answer is NO :) Obviosly, there is another bug. I will try to fix it soon.
I have noticed a problem with the need for a 'dummy' survey outside a .3d file import - it stops you using the suggested index.th and index.th2 files for centreline and scrap info (because you need a 'survey dummy' in both files around everything and it complains about 'duplicate survey').
If previous bug will be fixed - no problem to remove need for dummy survey :)
I've also realsied that I can't do what I wanted to do - import more than one .3d file. I wanted to import one per cave, but of course thinking about it, this isn't going to work because the co-ordinates will overlap. I need to generate one .3d file covering al the caves.
May be - I can add some -shift option for import command. E.g. -shift 1200 1300 500 m will shift all stations with given vector. This is no problem and it can be probably very usefull in some cases.
Still it would be good if it gave a useful error - it took me some time to work out what the problem was as it just gave a tex overflow error - which I can't reproduce right now due to getting a different error:
I am curious to see this error - on my computer your data worked without errors.
It says: therion: error -- file import -- index.th [15] -- invalid survey name -- de0/.deception
I think it is faiing to cope with the oddly-name "de0/" survey. It has this peculiar name because the data is imported from compass and it was expedient to use this name. Survex eats it happily and so does aven so therion should.
This is already removed. Regards, S. -=x=- Skontrolované antivírovým programom NOD32
+++ Stacho Mudrak [05-02-17 12:27 +0100]:
May be - I can add some -shift option for import command. E.g. -shift 1200 1300 500 m will shift all stations with given vector. This is no problem and it can be probably very usefull in some cases.
Yes, I can imagine situations where this might be necessary. Put it on the todo list if it is easy. (I am now using one large .3d file for the whole mountain so it is not an issue for m).
Still it would be good if it gave a useful error - it took me some time to work out what the problem was as it just gave a tex overflow error - which I can't reproduce right now due to getting a different error:
I am curious to see this error - on my computer your data worked without errors.
Yes, sorry - the problem is that it takes a long time to write these mails as I try out the various options, and edit a lot of files, so at the end of the mail the dtaaset in in the status of the last test. I wrote this mail over 3 evenings and had this problem at the beginning but it went away whilst I changed other things :-( and now I can't remember the exact status that cuased it.
This is already removed.
cool - stargely the error does not appear when using the full-mountain benarat.3d file, only when using a small deception.3d file for that cave. Odd. Wookey -- Aleph One Ltd, Bottisham, CAMBRIDGE, CB5 9BA, UK Tel +44 (0) 1223 811679 work: http://www.aleph1.co.uk/ play: http://www.chaos.org.uk/~wookey/
+++ Stacho Mudrak [05-02-17 09:40 +0100]:
When therion has no scraps to export, it exports all centerline data. If it has some, it exports only 2D data and not centerline. If you wish to have also centerline data on the map, you need to do two things:
1. create some toplevel map, where you mix maps and surveys together, you would like to have on the output.
Yes- I have donethis - the centrelineplan map is all the centreline data. It is mixed with the terikanplan survey. I noticed that I could have a survey called 'terikan' and a map called 'terikan' and this mostly worked anyway, but I did get an error saying 'no projection for survey' (from memory) because it thought I meant the centreline data, not the map once. I can see whay this would happen. Not sure if we should do anything about that or not? Now that survey names can be referred to in maps (instead of scrap names) then maybe we need to enforce that the namespace does not overlap? This could be annoying becuase you may often call a map the same as the cave name, which is also a survey name?
2. select and export this map
OK - I think this is the thing I was not doing. Because therions magic 'automatic selection' mostly does what I want I do not have any selects in the dataset anymore. If I select the centrelineplan map, does that disable automatic selection of other maps so now I have to select all the maps I want to appear?
If you wish to view whole centerline for underground, just redefine symbol in layout:
code mpost let l_survey_cave = l_survey_surface_SKBB; endcode
I take it from your second message that this did not in fact work? Wookey -- Aleph One Ltd, Bottisham, CAMBRIDGE, CB5 9BA, UK Tel +44 (0) 1223 811679 work: http://www.aleph1.co.uk/ play: http://www.chaos.org.uk/~wookey/
At 11:58 +0000 17.2.2005, Wookey wrote: *******************************************
error saying 'no projection for survey'
map xxx -projection plan nnn endmap
I take it from your second message that this did not in fact work?
Wookey
check wiki - exaples MS. --
whay this would happen. Not sure if we should do anything about that or not?
We solved this problem very easily. To each map we add suffix, indicating that this is a map _m or .m . If there are more projections, you can add also projection type to the suffix (e.g. .mp, .me or .mx). The same we do with scraps. This way there are no conflicts and you see directly what is what.
OK - I think this is the thing I was not doing. Because therions magic 'automatic selection' mostly does what I want I do not have any selects in the dataset anymore. If I select the centrelineplan map, does that disable automatic selection of other maps so now I have to select all the maps I want to appear?
Yes - usually it's only one top-level map. But I can probably advance magic of therion automatic selection - that this will not be necessary.
code mpost let l_survey_cave = l_survey_surface_SKBB; endcode
I take it from your second message that this did not in fact work?
Yes and no. It will help, if you will use centerline data defined properly in therion. It will not, if you will import them - because of bug deep in therion data structure. S.
+++ Stacho Mudrak [05-02-17 13:31 +0100]:
whay this would happen. Not sure if we should do anything about that or not?
We solved this problem very easily. To each map we add suffix, indicating that this is a map _m or .m . If there are more projections, you can add also projection type to the suffix (e.g. .mp, .me or .mx). The same we do with scraps. This way there are no conflicts and you see directly what is what.
Yes that is sensible. The question is whether we should enforce this behaviour on people with warnings or error messages?
OK - I think this is the thing I was not doing. Because therions magic 'automatic selection' mostly does what I want I do not have any selects in the dataset anymore. If I select the centrelineplan map, does that disable automatic selection of other maps so now I have to select all the maps I want to appear?
Yes - usually it's only one top-level map.
But I can probably advance magic of therion automatic selection - that this will not be necessary.
Maybe. I've actually been finding the magic selection confusing over the last few days as I seem to get 'everything' even when I comment stuff out of map lists. I haven't exactly got to the bottom of it so I haven't sent mail about it (yet). The problem with 'automagic' software is that it can be very hard to work out what is going on when it is not doing what you want. Of course my dataset with many caves and multiple maps is more complicated than most so this probably isn't a problem (we should design to make the simple case easy - the complicated case is going to be complicated whatever we do). Let me do some more work on producing multiple maps from one dataset and then we can discuss the whole issue some more. I expect there will be a number of problems - like wanting to have labels and cross sections in different places on different maps - I don';t think we have that option at the moment.
code mpost let l_survey_cave = l_survey_surface_SKBB; endcode
I take it from your second message that this did not in fact work?
Yes and no. It will help, if you will use centerline data defined properly in therion. It will not, if you will import them - because of bug deep in therion data structure.
Right. understood. Wookey -- Aleph One Ltd, Bottisham, CAMBRIDGE, CB5 9BA, UK Tel +44 (0) 1223 811679 work: http://www.aleph1.co.uk/ play: http://www.chaos.org.uk/~wookey/
Maybe. I've actually been finding the magic selection confusing over the last few days as I seem to get 'everything' even when I comment stuff out of map lists. I haven't exactly got to the bottom of it so I haven't sent mail about it (yet).
Well, the only way how to get what you want is to select maps in configuration file.
then we can discuss the whole issue some more. I expect there will be a number of problems - like wanting to have labels and cross sections in different places on different maps - I don';t think we have that option at the moment.
:) ... I am looking forward. S.
participants (3)
-
Martin Sluka -
Stacho Mudrak -
Wookey