My thought’s on Martin’s advice…

 

>Try to create small scraps (max 100 m?). And try to create new .th2 file for 
>each one scrap. This will reduce opening time much!

 

Small scraps are good, and fast to load and render, but one scrap per file 
would seem to be excessive unless you are dealing with extra large scraps.

By limiting to one scrap per file, you are removing the significant advantage 
of automatic scrap joins that Therion enforces when points and line points from 
adjacent scraps are snapped to the same location.  And in addition you will 
have many more ‘input drawing.th2’ statements and ‘join scrap1 scrap2’ 
statements.

 

I am looking at a small cave I surveyed two weeks ago, just under 130m in 
length.  Parts of it are a little bit complex, and I am drawing at 1:250 
instead of my usual 1:500 or 1:1000, so scraps have been broken into smaller 
pieces than would be typical and so I have more of them.  It is much more 
detailed than my typical effort.

*       I made 3 PocketTopo files, because the plan and elevation complexities 
made it expedient to start new files (in other cases 500-1000m per PocketTopo 
would be a typical upper limit).
*       That led to 3 survey.th files, 3 plan.th2 files and 3 elev.th2 files.
*       Between the plan and elevation projections, the drawing files have 32 
scraps (5-6 scraps per file, 4 of them are cross sections), and total 2800 
lines (292-790 lines per file).
*       By maximising scraps per file, I estimate I have saved myself 24 input 
statements, and 20 - 25 join statements, and in addition, I am mostly drawing 
with the context of adjacent scraps in full view.

 

If this was part of a large cave drawn at 1:1000, there might have been less 
than one third the number of scraps (still a lot, but required to avoid single 
scraps passing over or under themselves).  The general point is, in a large 
cave system, one scrap per file would seem to create a lot of unnecessary work 
and input/join statements.

 

And I doubt the number of scraps per file is what actually causes the reduction 
in performance.

I would expect it is more related to the number of entities or number of lines 
within the th2 file, and the size of any xvi or image files loaded.  From 
memory the rendering starts to become noticeably slower at 1000-1500 lines per 
th2 file, and intolerable at 3000 lines per th2 file (assuming a minimal xvi 
image).  This will vary from computer to computer, and your tolerance to 
waiting.

 

My rules of thumb are;

*       Minimise the size of survey trip files (limits size of and delay due to 
xvi and sketch files)  I find 200 - 600m is good - same as Martins advice, we 
just have different thresholds.
*       Make as many scraps as you want or need in each th2 file, and avoid 
more than one th2 file per survey trip file if you can help it – differs from 
Martins advice – almost opposite!
*       Once your th2 file reaches about 1500 lines, start considering another 
th2 file. 

Having said that, I don’t think there are any absolute limits, it’s just a 
matter of preference.

 

Bruce

_______________________________________________
Therion mailing list
[email protected]
https://mailman.speleo.sk/listinfo/therion

Reply via email to