- Therion - speleo.sk

FW: Scalebar meatpost labelling errors
by Bruce 07 Feb '10

07 Feb '10

Hi Thomas > that's strange, it should work at any scale. What is error cp-02 ? Cp-02 from therion, actually the metapost complains about the tmpx in the loop that calculates the position of the intermediate text labels. It is some sort of numeric overflow. As from the wiki it only runs for me if scale = 1:1000. If I constrain the loop to cycle only twice it runs for scale =1:500, 1:1000, 1:2000, 1:5000, but crashes similarly for 1:10000. If I play around with the values that are plotted it seems that there is quite a bit of rounding error - maybe we are putting the wrong type of number into the variables, or the intermediate calculation exceeds the variable range. I agree though, on the face of it the maths looks like it should work at any scale. In a few days I could cut and paste examples of what is produced, to illustrate, if necessary (I'm supposed to be drawing cave maps). Here is the code annotated a little, you can see how I have curtailed the loop somewhat to get it to work most of the time ... Bruce layout LayoutScalebar1 # Scalebar by Thomas Holder code metapost def s_scalebar (expr l, units, txt) = % l = value of scale-bar length % units = ?? % txt = string representing units begingroup interim warningcheck:=0; tmpl:=l / Scale * cm * units / 2; % tmpl = half plotted length of scale bar from central top insertion point tmpx:=l / Scale * cm * units / 5; tmph:=5bp; % bar height endgroup; pickup PenC; draw (-tmpl,0)--(tmpl,0)--(tmpl,-tmph)--(-tmpl,-tmph)--cycle; p:=(0,0)--(tmpx,0)--(tmpx,-tmph)--(0,-tmph)--cycle; for i:=-2.5 step 2 until 2: fill p shifted (i * tmpx,0); endfor; begingroup interim labeloffset:=3.5bp; for i:=0 step (l/5) until 0.3*l: % (l-1): % 0.9*l: alternative exit criteria tmpx:=(tmpl * i * 2 / l) - tmpl; % problem is here typically on third run thru this loop, unless scale = 1:1000 % tmpx:=(tmpl * i * 2 / l) - tmpl; % added () to original line label.top(thTEX(decimal (i)),(tmpx,0)); endfor; label.top(thTEX(decimal (l) & "\thinspace" & txt),(tmpl,0)); label.bot(thTEX( "Original Scale 1 : " & decimal (Scale*100)),(0,-tmph)); endgroup; enddef; endcode endlayout LayoutScalebar1

2 2

wiki metapost for station name size
by Thomas Holder 06 Feb '10

06 Feb '10

Hi, the proposed code in the wiki for "To make the debug station names smaller (Stacho Mudrak)" does not work. When looking at thpoint.cxx the "switch (this->type)" statement is followed by a "switch (this->scale)" statement which always clears out what was defined by \thstationname, \thlabel and \thremark. Thomas

1 0

FW: Hiding small objects at large scales
by Bruce 06 Feb '10

06 Feb '10

Thomas >the label size is not passed to metapost but is controlled with tex >macros like \thlargesize. On re-reading I think I misinterpreted Martin Budaj's comment (You can test even for user-defined attributes.) I was puzzled because there seemed to be no size or scale variable passed. I thought maybe it was a global. Thanks for the label code suggestion. I may well try it out because it is an improvement on my current solution. Is one point the smallest that can be passed to the likes of \def\thsmallsize{\size[1]} ? If we could pass 0.001 then the code is no more elegant, but multiline text would be caught. The solution I have been using for a year or so is code metapost fonts_setup(0.01,0.01,9,12,16); % xs s m l xl fonts in points and this works perfectly, in that the xs text is invisible. Trouble is in a file with lots of labels the pdf file size is a lot larger than it needs to be, and text searches in the pdf bring up text that is not visible. The obvious solution is to allow the use of zero values in fonts_setup(). As written at present it crashes, and I'm still too bumbly to be able to modify it. Thanks Bruce

2 1

symbol-hide group all failure
by Bruce 04 Feb '10

04 Feb '10

A bug perhaps. symbol-hide group all no longer hides point entities in 5.2.14 Has anyone else noticed this works or doesn't work for them? Bruce

1 0

FW: Clarification, Request for 2D output header title option
by Bruce 03 Feb '10

03 Feb '10

Referring to my previous post of last month... > If there are multiple 'selected' maps and 'source' surveys for a particular projection, then a blank output map 'title' is produced. And I suggested a possible solution might be to make a -layout-title as in; >export map -projection plan \ > -layout-title "test override title " \ > -output ./Outputs/FloraCavesPlanMap.pdf > Such a '-layout-title' should have properties like; > UseAlways, and > UseOnlyIfNoUniqueTitleAvailable > for maximum versatility. I have found a work around that would have the same effect as the 'UseAlways' property in the second example herein. http://therion.speleo.sk/wiki/doku.php?id=tex ie code tex-map \cavename={Terikan River Cave} endcode So now my question is; Is it possible to modify it something like the following pseudo-code, and what is the actual code that should be used? code tex-map If \cavename IS EMPTY OR AMBIGUOUS THEN \cavename={Terikan River Cave} ELSE \cavename = \cavename endcode Bruce

1 0

Paperless Cave Survey with DistoX, PocketTopo and Therion
by Bruce 30 Jan '10

30 Jan '10

You may recall that 6 to 9 months ago I was having difficulty with our new DistoX http://paperless.bheeb.ch/. It would not calibrate properly. This is just to let you know that it would seem that all the problems are resolved, and it is now our preferred surveying device. It took us a while, but we did our first completely paperless cave survey a couple of weeks ago (after a few above ground practice runs). The finished map was done by lunchtime the next day, with minimal effort compared to an equivalent 'paper survey'. A bit more time is spent sketching in the cave, but much more detail tends to be collected, and it would appear to be more accurate. There are a few rough edges to the software, so it pays to practice the whole process before doing a real survey. With luck these will be ironed out in due course. The problem with the distoX calibration was resolved by replacing the circuit board (Beat provided this service free of charge). There was a short term problem after this, because the compass behaviour was erratic - increasing counter-clockwise etc. This was due to the calibration being carried out in Switzerland where the earth's magnetic field is weaker than where we were using the device in New Zealand. So even though in theory a calibration is valid for use anywhere on earth, in practice this is only the case if the magnetic field is weaker than where the device was calibrated. Cheers Bruce

2 1

FW: show only fixed survey point on maps
by Bruce 30 Jan '10

30 Jan '10

I too would like this type of control over point station visibility. I'm sure it can be done with metapost code, by just redefining the temporary station symbol and survey line fragments as barely visible or invisible dots. There may be a hint as to how to do this in recent posts on splay lines and points. It's too low on my priority list at the moment to follow up, but I have been of a mind to do this for a year or two. A more satisfactory solution would be to include this functionality directly into Therion, along the lines of what Gilbert was trying to do. symbol-hide point station symbol-show point station:fixed Related to this, I suspect the behaviour of pdf outputs with respect to flag:<entrance/continuation/sink/spring/doline/dig> has changed. With recent releases text and symbols associated with stations that have these seem to 'print' by default. This is reasonable, however I have not figured out how to stop them printing. There should be an easy way to do it, but I have not found it yet. Ideas anyone?

1 0

show only fixed survey point on maps
by Gilbert Fernandes (mailinglist) 27 Jan '10

27 Jan '10

hello, I'm looking to display the fixed survey points on maps, but I didn't succeed. I try to set the symbol-hide / symbol-show command in the layout bloc, but if I put symbol-show point station I have all the station points displayed (not only the "subtype fixed" points) and if I put symbol-hide point station symbol-show point station:fixed I have no station points. I saw in the manual that we can use flags to show specific stations ( flag:<entrance/continuation/sink/spring/doline/dig> ) But can we use the subtype of the station points to do the same selection ? Thanks Gilbert

1 0

Preview above below, map-level, revise
by Bruce 16 Jan '10

16 Jan '10

Does anyone have an explanation for the what and how of 'map-level' as it relates to maps as described in this extract from Pg 31 of the Therion Book 5.2.14. preview <above/below> <other map id> will put the outline of the other map in the specified preview position relative to the current map. Preview is displayed only if the map is in the map-level level as specified by the select command. Use the revise command if you want to add maps from higher levels to the preview. I have only ever managed to create previews of maps in the very simplest of cases (except I'm not sure exactly what constitutes a simple case - the ones that don't work seem to be quite straight forward, and not much different to those that don't work ). I think it may relate to 'map-level'. I have asked on this forum about a year ago, but alas I still have not resolved this. I have not been able to comprehend the explanations in the wiki (which is similar to the Therion book extract above). I think the first step to working it out might be to understand what a map-level is. I think it is not referring to the layering of scraps or maps within a map. Maybe it is the level of the map in the hierarchy displayed in the XTherion Compiler window 'Map Structure' panel? If so how are they numbered? How to determine the map-level of any given map? Top level 0 or 1? Bottom level 0 or 1? How would I use 'revise' to change the map-level of a map to enable a preview to be displayed? My attempts to change something simple like a map -title have all failed, so there must be something fundamental that I'm not understanding. Any assistance appreciated. Bruce

1 0

Fw: Bug#565049: therion: ftbfs with gcc-4.5
by Wookey 16 Jan '10

16 Jan '10

Guys - someone is testing if debian packages build with gcc 4.5, and therion doesn't. Anyone got any ideas how to fix this, or do we contend that it is actually a compiler bug? ----- Forwarded message from Matthias Klose <doko(a)debian.org> ----- From: Matthias Klose <doko(a)debian.org> Date: Wed, 13 Jan 2010 01:06:03 +0000 To: maintonly(a)bugs.debian.org Subject: Bug#565049: therion: ftbfs with gcc-4.5 Reply-To: Matthias Klose <doko(a)debian.org>, 565049-maintonly(a)bugs.debian.org X-Spam-Status: No, score=-3.6 required=4.5 tests=BAYES_00,RCVD_IN_DNSWL_LOW autolearn=ham version=3.2.5 Package: therion Version: 5.2.8-1 Severity: normal Tags: sid User: debian-gcc(a)lists.debian.org Usertags: ftbfs-gcc-4.5 The package fails to build in a test rebuild on at least amd64 with gcc-4.5/g++-4.5, but succeeds to build with gcc-4.4/g++-4.4. For the compiler version used, see: http://lists.debian.org/debian-devel/2010/01/msg00230.html The full build log can be found at: http://people.debian.org/~lucas/logs/2010/01/08b/therion_5.2.8-1_lsid64b.bu… The last lines of the build log are cwat the end of this report. The reason of the build failure is likely in the package (compiler or linker message found in the build log), if this turns out as a gcc bug, please reassign to the gcc-4.5 package. If the build ends with a message "confused by earlier errors, bailing out", please ignore it for now; these compiler errors are filed separately. [...] c++ -c -Wall -DTHLINUX -O2 -o thscraplp.o thscraplp.cxx c++ -c -Wall -DTHLINUX -O2 -o thscrapen.o thscrapen.cxx c++ -c -Wall -DTHLINUX -O2 -o thpoint.o thpoint.cxx c++ -c -Wall -DTHLINUX -O2 -o thline.o thline.cxx c++ -c -Wall -DTHLINUX -O2 -o tharea.o tharea.cxx c++ -c -Wall -DTHLINUX -O2 -o thlegenddata.o thlegenddata.cxx make -C ./mpost make[2]: Entering directory `/build/user-therion_5.2.8-1-amd64-g4N5cQ/therion-5.2.8/build-tree/therion/mpost' perl genmpost.pl reading therion.mp reading thPoint reading thLine reading thArea reading thText reading thSpecial reading thTrans done make[2]: Leaving directory `/build/user-therion_5.2.8-1-amd64-g4N5cQ/therion-5.2.8/build-tree/therion/mpost' c++ -c -Wall -DTHLINUX -O2 -o thmpost.o thmpost.cxx c++ -c -Wall -DTHLINUX -O2 -o thsymbolsets.o thsymbolsets.cxx c++ -c -Wall -DTHLINUX -O2 -o thjoin.o thjoin.cxx c++ -c -Wall -DTHLINUX -O2 -o thmap.o thmap.cxx c++ -c -Wall -DTHLINUX -O2 -o thexpmap.o thexpmap.cxx In file included from thexpmap.cxx:75:0: thsvg.h:26:79: error: cannot call constructor 'legenddata::legenddata' directly thsvg.h:26:79: note: for a function-style cast, remove the redundant '::legenddata' make[1]: *** [thexpmap.o] Error 1 make: *** [build-stamp] Error 2 make[1]: Leaving directory `/build/user-therion_5.2.8-1-amd64-g4N5cQ/therion-5.2.8/build-tree/therion' dpkg-buildpackage: error: debian/rules build gave error exit status 2 ----- End forwarded message ----- Wookey -- Principal hats: iEndian - Balloonboard - Toby Churchill - Emdebian http://wookware.org/

3 5