Re: text in idraw giving errors in recent versions of ghostscript

[Scott J. <sc...@ac...>]

>
>
>>>Specificially, if I create a new file with idraw 1.0.6 and put the
>>>text string "foo" in it, the gv the file using ghostscript-6.52-9.4 on
>>>Redhat 7.3, it gives the following error message:
>>>
>>>Error: /rangecheckGNU Ghostscript 6.52: Unrecoverable error, exit code 1
>>>      
>>>
To me this implies a bug in Ghostscript (or underlying library), not an 
error in PostScript.

>>>Current file position is 10438
>>>
>>>
>>>Position 10438 is the first character of the "End" just after
>>>the "] Text" line in the file.
>>>

It seems position 10438 is on the line before the "Begin %I Text"/

>>>
>>>This problem didn't happen on Redhat Linux 7.2, but it persists even
>>>if I downgrade ghostview in RH 7.3 to the 7.2 version.  
>>>
You would also need to downgrade ghostscript to the 7.2 version to see 
if the problem
is specific to ghostscript-6.52.  Did you do that?

>>>
>>>
>>>1. Has anyone else observed this bug?
>>>

Not that I know of.

>>>
>>>2. Is this a ghostscript or idraw problem?
>>>(Is idraw wrong and ghostscript is now more conservative in this release, or 
>>>did ghostscript regress.)
>>>

The test file you forwarded looks identical to what I generate on RH 
6.2.  I would say the
problem stems from some change to ghostcript (or underlying libraries). 
 Whether it is a bug
or a tightening up on the accepted PostScript language remains to be seen.

>>>
>>>I hpoe it's fixable because I have lots of old idraw files that I like
>>>to ghostview!
>>>

First figure out if you downgraded ghostscript at the same time you 
downgraded ghostview
(from the RH 7.3 version to the RH 7.2 version).  If so, I think you 
have a RH 7.3 specific problem
with ghostscript and you could report that to the maintainers.  If not, 
try that downgrade, and if
there is still a problem you still have a RH 7.3 specific problem. 
 Otherwise, you might have
isolated a problem interpreting idraw EPS with the ghostscript 6.52. 
 Which would be worth reporting to the ghostscript maintainers as well.

Scott Johnston

p.s.  After writing all this I vaguely recall some conflict between 
recent ghostscript and the idraw EPS format, but I can't locate any 
documentation on this.  Pinging the ghostscript maintainers (Raph 
Levien) might stir up what that was (or was it just a hallucinated memory).