| | <<<Back 1 day (to 2020/06/21) | Fwd 1 day (to 2020/06/23) >>> | 20200622 |
| bosobject | What is the purpose of tag parameter in --writeobject-- ? Ghostscript scanner requires it to be 0. | 04:06.16 |
| chrisl | bosobject: writeobject is a Postscript operator, thus it is documented in the PLRM | 06:41.33 |
| bosobject | chrisl: It is documented but --writeobject-- easily creates files that Ghostscript scanner cannot read. | 07:00.59 |
| | I wonder how Adobe does it? | 07:03.59 |
| kens | Real worl duse of binary object sequences is vanishingly rare. Its possible you ahev found a bug in whcih case, as you know, you should report it. Chris' comment is correct, the tag operand is detailed (under printobject) in the PLRM. | 07:05.49 |
| bosobject | But does the reference implementation follow PLRM? I'd expect it to ignore the tag field, rather than consider it "unused" and require 0. | 07:08.57 |
| kens | The original Adobe PostScript interpreters (which reasonably haev to be considered the 'reference implementation') follow the specification very closely. What they did with binary object sequences is less clear given, as I said, that use of binary object sequences is about as common as the proverbial hens teeth | 07:10.29 |
| | I no longer have access to even a CPSI interpreter and I wouldnt' trust Distiller for this. | 07:10.51 |
| | page 165 of the 3rd edition PLRM defines byte 2 of the object in a binary object sequence as being unused and must have the value 0 | 07:16.28 |
| bosobject | I know but this but suspect that Adobe does not check it. What's the reason to generate unreadable objects? | 07:19.28 |
| kens | You could use it as a debugging aid, but set it to 0 for production | 07:20.00 |
| | <<<Back 1 day (to 2020/06/21) | Forward 1 day (to 2020/06/23)>>> | |
Log of #ghostscript at irc.freenode.net.