It seems I was wrong about the cause of the problem, the problem is more in Therion than in data entry. Looks like Therion 5.3.15 cannot handle survey dates in 2015 (as implied in the changes logs), and so 5.3.16 is required to use coordinate systems with surveys dated from 2015 onwards. That is OK, but what could be improved, is the reporting of the error to something like. magnetic declination not available for survey date and also to improve the variable \magdecl in this circumstance. Bruce _____ From: Bruce [mailto:bruce@tomo.co.nz] Sent: Monday, 19 January 2015 4:34 p.m. To: 'List for Therion users' Subject: Magnetic declination meltdown A random discovery with the new tex variable \magdecl = magnetic declination in degrees. If one uses a defined coordinate system and forgets to provide a survey date for all surveys, then there is some 'not very well trapped' undesirable behaviour. The log file produces something like this. "C:\Program Files\Therion\therion.exe: warning -- unable to determine magnetic declination for undated surveys output coordinate system: EPSG:2193 meridian convergence (deg): -0.0696 " which is appropriate behaviour, however it is followed by. " #################### end of metapost log file #################### converting scraps ... done making map ... done ######################## pdftex log file ######################### This is pdfTeX, Version 3.1415926-1.40.10 (Web2C 2009) (format=pdftex 2010.1.2) 18 JAN 2015 20:21 entering extended mode **data.tex (./data.tex (./th_enc.tex) (./th_texts.tex) (./th_resources.tex (c:/Program Files/Therion/texmf/tex/glyphtounicode.tex)) (./th_fontdef.tex{c:/P rogram Files/Therion/texmf/fonts/pdftex.map}) (./th_formdef.tex Normal \count register pool exhausted, switching to extended pool.) (./th_pagedef.tex Overfull \hbox (299.59392pt too wide) in paragraph at lines 12--12 []\thfb gridnorth -1261120821493519700000000000000000000000000000000000000000 00000000000000000000000000000000.00deg | \hbox(8.33333+2.33333)x330.05206, glue set - 1.0 .\hbox(0.0+0.0)x0.0 .\thfb g .\thfb r .\thfb i .\thfb d .etc. ) (./th_pages.tex) 253 numeric registers used out of 32768 [1] ) Here is how much of TeX's memory you used: 1706 strings out of 95617 .. " which is not so good, but at least it does produce a generally OK looking pdf (with my experimental debug like header layout). In this case it might be better to set \magdecl to a more meaningful error value. If it can take a string value then "unable to determine magnetic declination for undated surveys" is probably good. Bruce
