> I think I understand that using the global dataset would work, however, this > means that everything has to be tided into the global dataset from the > start. For various reasons this is not going to be possible (well not in the > near future.) I was also hoping to be able to send distinct sections out, > without sending the whole dataset out, eg just send the folder darkside out > and someone would be able to edit it. Also I like to try to have data as > simple to change as possible. For example when a cave has been misnamed I > could just change a couple of references and that would be it, I am thinking > of situations of when a small cave is linked to a large system and then is > just an area of that system. If the full global reference is used in the th2 > files this would not be possible.
I am not sure, whether we have understood. From my point of view, everything you described would work - without needs of knowing this global dataset structure. Assuming the most simple scenario: You have a centerline in survex. There is only one survey "entrance" and stations are named from 0 to 10. In therion, your dataset (entrance.th) will be like: import entrance.3d -surveys use survey entrance input entrance.th2 endsurvey ... assuming in entrance.th2 stations are without any prefix - named just 0,1,..,10. Now you have two such directories: "entrance" and "continuation" for example. In survex, you have a global file named "cave.3d" where stations from entrance dataset look like cave.entrance.1 and from second dataset cave.continuation.1 and cave.3d in in the upper level directory. Now the corresponding cave.th file in this directory will be like: import cave.3d -surveys use survey cave input entrance/entrance.th input continuation/continuation.th endsurvey And that's all. You are able to work on single peace without knowing global datastructure. You may send only directory continuation to somebody without any files from top level directory and it will work. He may edit it, send it back to you and it will work also. If he will rename this survey, you just change references in cave.th and that's all. I do not see any of problems you have described. > Also if I understand you correctly, I would have to edit every station name > in all the th2 file. Certainly not. It looks I missed something. I was wondering, why you use such a long station names... I need to check your source files once more, may be your problem is different. > >From a users point of view, a possible solution is to use some sort of > prefix name - endprefix that would use the survex naming structure ie add > name.* to all the stations. Scrap already has this possibility via -station-names option. You may add prefixes or suffixes if you like. OK, I will check your files more in the detail, may be I will understand. But today, thinking is a little bit difficult for me - I drunk something last night and I slept only 3 hours... Probably when I will wake up tomorrow, everything will be clear. S.
