Fix POSTSCRIPTSTREAM to refine or default values for COLOR.

[MattHeffron]

(Don't require them to be FLOATP.)

This resolves #2176.

[rmkaplan]

This corrected the zeros, but now I see that the file EQUATIONEDITORPRGMERSGUIDE.TEDIT is also drawing a curve where the brush has color 7. Are integer colors like this some sort of index into a table of colors?

…

On May 31, 2025, at 5:38 PM, Matt Heffron ***@***.***> wrote: (Don't require them to be FLOATP.) This resolves #2176 <#2176>. You can view, comment on, or merge this pull request online at: #2177 Commit Summary 6a841f0 <6a841f0> Fix POSTSCRIPTSTREAM to accept SMALLP values 0 and 1 as COLOR. File Changes (3 files <https://github.com/Interlisp/medley/pull/2177/files>) M library/POSTSCRIPTSTREAM <https://github.com/Interlisp/medley/pull/2177/files#diff-836e34ca40efc69ae4673638214411ef21a8a53f765942dd5e407eaa6afe8ded> (114) M library/POSTSCRIPTSTREAM.LCOM <https://github.com/Interlisp/medley/pull/2177/files#diff-ff2743be5a3adf057b43978de681fc2651599edb78373ed0aa308d980eb429de> (0) M library/POSTSCRIPTSTREAM.TEDIT <https://github.com/Interlisp/medley/pull/2177/files#diff-08006b4611d5215594af331a55307c5d439d9dc6d8b551c94ac0ec28df4a5496> (6) Patch Links: https://github.com/Interlisp/medley/pull/2177.patch https://github.com/Interlisp/medley/pull/2177.diff — Reply to this email directly, view it on GitHub <#2177>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AQSTUJPKO4TTRWFMCABPIRD3BJDRRAVCNFSM6AAAAAB6KODAQ6VHI2DSMVQWIX3LMV43ASLTON2WKOZTGEYDMMRZGY4TGMI>. You are receiving this because your review was requested.

[nbriggs]

@rmkaplan there are a few things in can be, but one option is that it's something that is COLORNUMBERP, which is defined on library/LLCOLOR.

[MattHeffron]

Are integer colors like this some sort of index into a table of colors?

COLORNUMBERP checks against the color map for the screen, so I don't know what it would do for a B/W screen, nor for an imagestream that isn't to the screen.

This is the BRUSHCOLOR of a curve in the second embedded sketch. It appears that color 7 is specified in the sketch, but I don't see how to set the color of a BRUSH to use in the sketch. Is it actually a TEXTURE? So, I don't know how to find out where that "color" came from, and what it should do.
It looks like I might have to dig into SKETCH. How come TEDIT has no problem with that?

[rmkaplan]

Tedit doesn't know about colors in image objects, and it doesn't interpret colors in any other way. DSPCOLOR of a Tedit stream just puts the color-pointer (whatever it is) in the character-looks, and passes that to the back-end image stream through the image-stream's DSPCOLOR.

…

On Jun 1, 2025, at 11:43 AM, Matt Heffron ***@***.***> wrote: MattHeffron left a comment (Interlisp/medley#2177) <#2177 (comment)> Are integer colors like this some sort of index into a table of colors? COLORNUMBERP checks against the color map for the screen, so I don't know what it would do for a B/W screen, nor for an imagestream that isn't to the screen. This is the BRUSHCOLOR of a curve in the second embedded sketch. It appears that color 7 is specified in the sketch, but I don't see how to set the color of a BRUSH to use in the sketch. Is it actually a TEXTURE? So, I don't know how to find out where that "color" came from, and what it should do. It looks like I might have to dig into SKETCH. How come TEDIT has no problem with that? — Reply to this email directly, view it on GitHub <#2177 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AQSTUJOH5ZKZC2LS5L6F22D3BNCUXAVCNFSM6AAAAAB6KODAQ6VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDSMRXGY3TONJRHA>. You are receiving this because you were mentioned.

[MattHeffron]

I did a bunch of digging through the source code and discovered:

• When a drawing operation targets the display, the BRUSHCOLOR is ignored if the target is 1BPP ScreenBitMap.
  If the target is ColorScreenBitMap , then BRUSHCOLOR is set into a bitmap of the same BPP as ColorScreenBitMap with no further interpretation, leaving that to the use of the ColorScreenBitMap.
• In INTERPRESS the function \SETBRUSH.IP ignores the BRUSHCOLOR if it is a NUMBERP that's <= 0.
  (It doesn't do this safely. There are a couple of points there where the BRUSH function argument is used as a BRUSH record type, but it hasn't verified that it is that type.)

So, I propose that POSTSCRIPTSTREAM should behave similar to INTERPRESS and just ignore a BRUSHCOLOR if it is a SMALLP that's neither 0 nor 1, and use the current DSPCOLOR of the stream. If someone wants a specific color, or grayscale, they can use the color descriptions as documented in POSTSCRIPTSTREAM.TEDIT.
This seems the best alternative, both for consistency with INTERPRESS and simplicity.

@rmkaplan @hjellinek @masinter Any comments/objections?

[rmkaplan]

I tried it, and it worked fine on the problematic equation file. But the paragraph in the documentation beginning "Anywhere that a texture or color..." seems to only allow 0 and 1, should that be updated to describe the other pass-through possibilities ?

[rmkaplan]

LGTM