[Gs-devel] "relpos" in MDRC (refine MDRC patch)

[mpsuzuki at hiroshima-u.ac.jp]

Dear Mr. Melichev,

Just a report. I've found "relpos" type behaviour in the
official CPSI in PS printers of OKI (2015, 3010), and
in Adobe Distiller 4.0. Using an artificial specification

	1 begincidrange
	<2121> <7e21> 1000
	endcidrange

makes full-width spaces for all <**22> characters.

>In this case you can say that your method is CPSI-compatible.
>Possibly it is bug-to-bug compatible.
>A proof may be taken only from specs.

Hmmm, to insist "relpos" as officially designed behaviour,
I should find explicit proof by Adobe (in specs).

>My point is : 'relpos' to be removed because
>(1) it is not needed for correct processing of known CMaps, and
>(2) it is not a documented feature.

I agree. As you called "bug-to-bug compatible", if it's bug,
Adobe will fix in future CPSI, so we should not emulate it,
even if older CPSI had such feature.

I wrote the "relpos" code before detailed experiments on official
CPSI. When I was reading the PostScript reference manual v3 roughly,
I supposed Adobe's CPSI will calculate CID so. I must reconsider.

Best Wishes,
mpsuzuki