Description of problem: When I tried to print out the plain text file, it failed and it seems that gs8 shows an error when PS file from mpage is given. Version-Release number of selected component (if applicable): ghostscript-8.15-0.rc3.7 mpage-2.5.4-5 How reproducible: always Steps to Reproduce: 1.LANG=ja_JP.UTF-8 mpage -bA4 -l -o -m18lr36bt -P- /usr/share/printconf/tests/testpage.euc | gs 2. 3. Actual results: /invalidrestore error appears Expected results: no error should happens. Additional info: I noticed this problem when I was testing with s-c-p's test printing, but the printer didn't output anything at all. then tried with lpr, but the dialog said an error happened or so. So I've changed LogLevel to debug and see what happened. it also shows '*** buffer overflow detected ***' message and Backtrace stuff. although I didn't see this message with mpage | gs on the terminal, will attach this error_log as well. BTW a PS file which was generated by mpage on FC5 can be printable on FC4, gs7. so this is an regression bug, I think. P.S. Once I stored the output from mpage to the file, and when I give that file to gs, no error happened. I mean, I did "mpage .... testpage.euc > mpage.ps" and gs mpage.ps. However "cat mpage.ps | gs" still shows the error.
Created attachment 117649 [details] error_log
Please try 8.15-0.rc4.1.
Oh, wait -- there are two bugs here. The overflow one is fixed, but the /invalidrestore one is not.
An easy work-around is 'gs /dev/fd/0' though. So I'd like to know if 8.15-0.rc4.1 allows you to print again after all.
Well, yes, lpr testpage.euc works fine. but mpage | gs still doesn't.
Well, I'm not really inclined to fix this. There is an easy work-around, which the printing system uses.