peirce@gumby.cc.wmich.edu (Leonard J. Peirce) (01/12/89)
I retrieved a copy of the fig program from the sun archive and built it. (3/60 running 4.0). Fig works great. However, the f2ps program core dumps whenever I try to use it. Does anyone out there have a working copy? Or doesn't it work under 4.0? I retrieved the patch for fig and applied it but it didn't help. The patch file also tried to patch a file called fig2tex.c. If anyone has a copy of this, I'd be REAL interested in getting a copy... Leonard J. Peirce Internet: peirce@gumby.cc.wmich.edu Western Michigan University peirce@gw.wmich.edu Academic Computer Center Voice: (616) 387-5469 Kalamazoo, MI 49008
ken@cs.rochester.edu (Ken Yap) (01/18/89)
> I retrieved the patch for fig and applied it but it didn't help. The > patch file also tried to patch a file called fig2tex.c. If anyone has a > copy of this, I'd be REAL interested in getting a copy... Sorry, have no clues about your bug, but fig2tex is part of the Transfig package and is in svax.cs.cornell.edu:fig/transfig.tar.Z. Be warned that it is actually a fig to pictex filter.
will%robots.oxford.ac.uk@nss.cs.ucl.ac.uk (Will Dickson) (02/08/89)
We have also had some trouble with fig, some of which have been fixed in
fig-fs (along with the addition of some new facilities). The problem
mentioned by peirce@gumby.cc.wmich.edu also stayed hidden until we
upgraded to 4.0: it is a question of initialising all the pointers of the
structure obj in read_objects in read.c (fix below). This has been fixed
in fig-fs. Below is a summary of all the mods I applied to fig-fs to make
it work. The first stops fig from crashing out when points are deleted
from splines; the other prevents infinite looping for zero height
ellipses.
******************************
Script started on Fri Jan 13 11:50:46 1989
miranda:1% diff deletept.c deletept.c.OLD
100,102c100
< /* JWD fix 25-10-88 */
< /* if (n <= 1) {*/
< if (n <= 2) {
---
> if (n <= 1) {
107,114c105,108
< /* JWD fix 25-10-88 */
< /* else if (n <= 2) {*/ /* it must be an interpolated spline */
< /* put_msg("An interpolated spline may have less than 3 points");
< */
< else if (n <= 3) {
< put_msg("An interpolated spline cannot have less than 3 points");
< return;
< }
---
> else if (n <= 2) { /* it must be an interpolated spline */
> put_msg("An interpolated spline may have less than 3 points");
> return;
> }
miranda:2% diff ellipse.c ellipse.c.OLD
116,124d115
<
< /* JWD mod 25-10-88 */
< if ((x == fix_x) || (y == fix_y))
< {
< put_msg("cannot create flat ellipse !");
< return;
< }
< /* end JWD mod */
<
miranda:3% exit
script done on Fri Jan 13 11:53:05 1989
******************************
The fix for read.c (fig only, not fig-fs) is:
******************************
Script started on Fri Jan 13 11:56:42 1989
miranda:1% diff read.c read.c.oldOLD
114,125d113
< /* Modification by JWD to initialise pointers... not done otherwise. */
< /* 5-JAN-89 JWD */
<
< obj->lines = NULL;
< obj->ellipses = NULL;
< obj->splines = NULL;
< obj->texts = NULL;
< obj->arcs = NULL;
< obj->compounds = NULL;
<
< /* End of JWD mod 5-JAN-89 */
<
miranda:2% exit
script done on Fri Jan 13 11:57:00 1989
******************************
The patch mentioned by peirce@gumby.cc.wmich.edu appears to be intended
for the fig-fs distribution (which includes fig2tex etc...).
There is still another bug somewhere in fig-fs which causes a crash very
occasionally, but only after very prolonged sessions and not in repeatable
circumstances. If anyone has found anything like this, please let us
know!
Will Dickson,
Robotics Research Group,
19 Parks Road,
Oxford OX1 3PJ,
ENGLAND. JANET: will@uk.ac.ox.robots