Date before 1900 error - but where?
Hi all I could really do with some help searching for a problem. In the data http://www.cave-registry.org.uk/svn/CheddarCatchment/ (downloadable svn repository) Trying to run any of the thconfig (well I've tried most) files leads to this error configuration file: LongWoodValley.thcfg reading ... done reading source files ... done preprocessing database ... done therion: error -- automatic declination calculation before 1900 not supported, please specify declination explicitly writing xtherion file ... done This is not a useful error. This data set use to work, I've produced lots of surveys from it. It seems to indicate a date in the error message. Although grep has shown up a couple that I needed to fix for other reasons, none have shown dates before 1900. So I have no idea how to track down this error What else should I be looking for? Running therion with -d does not give any further information Help! thanks Andrew
Hi, I recently had the same issue. The reason here was, that I had added a small separate centerline to just tell the system, that there is line-of-sight between two stations. Turned out that this of course was also a normal valid centerline, and I had no date nor declination added. Maybe that is oh some help. Our human brains are sometimes filtering out information at the wrong place, so we dont recognize such issues as issue.
Am 19.02.2025 um 19:29 schrieb Andrew Atkinson <andrew@wotcc.org.uk>:
Hi all
I could really do with some help searching for a problem.
In the data
http://www.cave-registry.org.uk/svn/CheddarCatchment/
(downloadable svn repository)
Trying to run any of the thconfig (well I've tried most) files leads to this error
configuration file: LongWoodValley.thcfg reading ... done reading source files ... done preprocessing database ... done therion: error -- automatic declination calculation before 1900 not supported, please specify declination explicitly writing xtherion file ... done
This is not a useful error. This data set use to work, I've produced lots of surveys from it.
It seems to indicate a date in the error message. Although grep has shown up a couple that I needed to fix for other reasons, none have shown dates before 1900. So I have no idea how to track down this error
What else should I be looking for? Running therion with -d does not give any further information
Help!
thanks
Andrew _______________________________________________ Therion mailing list Therion@speleo.sk https://mailman.speleo.sk/listinfo/therion
Reservoir\CeilingAround\CeilingAround.th date 2013.04.12-13 You meant: date 2013.04.12-2013.04.13
Thank you very much That must be a change in Therion (Survex) as that file has not been edited since 2013.05.11 and it has been used many many times since then. I should probably file this as a bug, well the lack of reporting an error. How did you find it, and so quickly? I need that sort of skill set thanks Andrew On 19/02/2025 19:11, Tarquin Wilton-Jones via Therion wrote:
Reservoir\CeilingAround\CeilingAround.th
date 2013.04.12-13
You meant:
date 2013.04.12-2013.04.13 _______________________________________________ Therion mailing list Therion@speleo.sk https://mailman.speleo.sk/listinfo/therion
How did you find it, and so quickly? I need that sort of skill set
Same basic way I debug CSS. Make a copy of the codebase. Delete 50% of the inputs (caves). Did it go away? If not, try deleting the other half instead. Repeat with smaller and smaller chunks until you get which change made it go away. Don't delete anything that looks like it contains fixed points that everything else relies on. A few caves rely on .3d files which are not on SVN (eg. GB, Rhino). I deleted those first, and fortunately it wasn't in any of those (I would not have been able to debug otherwise). Now I know the cave. Delete contents of all maps and all joins, to remove dependencies. Now find inputs that are only used by one equate. Delete both. Repeat. Eventually, there was a loop of equates. Delete one to break the loop. Now continue deleting file+equate until deleting one input causes it to show a different error or no error (other than "could not open file for input). Now I know the file. The mistake was pretty obvious, but if there were lots of centrelines, I would just have deleted their data in reverse order. As it was, it was trying to work with "13" as a year 0013. I wonder if this broke your loop closure... The change in behaviour is a largely good thing. Personally, I would dearly love a warning of "date range is backwards", or "date looks insane", since if you managed to get some pre-Roman survey data, it would be very impressive indeed, and probably deserves a special config option "mindate 0001.01.01". However, the error message was not at all useful as a pointer, and needs to be corrected.
On 19/02/2025 20:02, Tarquin Wilton-Jones via Therion wrote:
A few caves rely on .3d files which are not on SVN (eg. GB, Rhino). I deleted those first, and fortunately it wasn't in any of those (I would not have been able to debug otherwise).
I have a script that processes the svx files that are needed, I should probably add that to the repository to save people doing it manually. The svx files are slowly being removed, all historic stuff. I was going through the same process, but obviously slower than you thanks again Andrew
On Wed, Feb 19, 2025 at 07:39:17PM +0000, Andrew Atkinson wrote:
That must be a change in Therion (Survex) as that file has not been edited since 2013.05.11 and it has been used many many times since the
Nothing to do with Survex here. Indeed Survex handles the equivalent situation with a clear warning and error which highlight exactly how this is interpreted and what's wrong with it: baddaterange.svx:1:18: warning: Assuming 2 digit year is 1913 *date 2013.04.12-13 ^~ baddaterange.svx:1:7: error: End of date range is before the start *date 2013.04.12-13 ^~~~~~~~~~~~~ Cheers, Olly
participants (4)
-
Andrew Atkinson -
Benedikt Hallinger -
Olly Betts -
Tarquin Wilton-Jones