Closed hvoss49 closed 3 years ago
I tried implementing truncate
but I couldn't verify it with the test document since I don't have the pst-map3d
package.
replace pst-map??
by pst-geo
and it should work. Now lua:528: Unknown na me "arccos"
. is rthe problem
Now
lua:528: Unknown name "arccos"
. is rthe problem
This is an odd one. arccos
is not a standard operator in PostScript so it has to be defined somewhere. There is a definition in tx@geoDict
, but arccos
is used before tx@geoDict
is invoked so it shouldn't be able to find it. So I don't really know why GhostScript doesn't fail here...
I see ... with my Ghostscript (9.55.0):
GS<1>systemdict /arccos known ==
true
However, in pst-geo.tex
I load now tx@geoDict
for arccos
:
/AngleRayonsSoleil latitude_parallel 23.5 sin 360 jours mul 365.25 div sin mul tx@geoDict begin arccos end sub def
But then I get:
[Loading MPS to PDF converter (version 2006.09.02).]
) [1[string "-- luatex-core security and io overloads ....."]:51: bad argument
#1 to 'io_open' (string expected, got nil)
stack traceback:
[C]: in upvalue 'io_open'
[string "-- luatex-core security and io overloads ....."]:51: in function 'io.
open'
...ive/texmf-local/tex/generic/pstricks-lua/luapstricks.lua:2533: in function
<...ive/texmf-local/tex/generic/pstricks-lua/luapstricks.lua:2531>
I see ... with my Ghostscript (9.55.0):
GS<1>systemdict /arccos known == true
Oh, that explains it. I now found it as a documented extension for GhostScript. They are very easy to add so I might just do that.
But then I get:
[Loading MPS to PDF converter (version 2006.09.02).] ) [1[string "-- luatex-core security and io overloads ....."]:51: bad argument #1 to 'io_open' (string expected, got nil) stack traceback: [C]: in upvalue 'io_open' [string "-- luatex-core security and io overloads ....."]:51: in function 'io. open' ...ive/texmf-local/tex/generic/pstricks-lua/luapstricks.lua:2533: in function <...ive/texmf-local/tex/generic/pstricks-lua/luapstricks.lua:2531>
That's a terrible way of telling you: "File not found". The new version should feature a better error message (including the filename/path which wasn't found). It also fixes some other issues about the decompression.
On my system the new version seems to work except that the second globe isn't positioned correctly.
I found the issue causing the globe to shift: pst-geo
defines /MaillageSphere
which starts with gsave
but never executes grestore
. This messes up the nesting of the gsave
/grestore
groups, causing everything after the figure to be shifted (since the shift from the origin to the figure is never reverted)
Good catch! Fixed in svn.
The path setting is only for my system and must be modified if you want to run it.
expected result: