RE: [cgmo-webcgm] Fwd: Re: WebCGM RGB/sRGB question

[Lofton Henderson <lofton@rockynet.com>]

At 03:37 PM 7/7/2005 -0600, Robert Orosz wrote:
>[...]
>I can't find sRGB in the Register of Graphical Items either.  By definition,
>that means that sRGB is not registered, contrary to what PPF section T.16.19
>says in the WebCGM 2.0 draft.

Groan!

This is going to be a tough one to sort out with ISO, because all of these 
registration ballots are vintage 1998.    On the positive side, I have 
found a ZIP file (dated 1999) with several of the ISO-format, multi-page 
.DOC files for the registrations.  (These were in the ZIP file with four 
ESCAPEs, and LZW compression, both of which *did* make it into the register.)

RGBa:  6
sRGB:  7
sRGBa:  8

The other good news is that this does not impact WebCGM as a valid profile 
(under CGM:1999 Rules for Profiles):  valid profiles according to CGM:1999 
"Rules for Profiles" require that such extensions either be registered, or 
profile-defined (in an unambiguous manner).  For WebCGM 2.0 First Release, 
we can solve the problem by posting the documents and making them the 
normative references for the these 3 items, plus PNG, on which SC24 
Registration apparently dropped the ball.

Note that PNG is not in the registry either.  ISO said last summer that it 
was not entered because of some delay that was requested by the Japanese 
delegation, but that such delay was apparently resolved (as of summer 2004).

PNG (compression): 9

I looked today and PNG is still is not in the Register.

Dave, add it to the AI list for me to sort this out with ISO.

Regards,
-Lofton.

>-----Original Message-----
>From: Cruikshank, David W [mailto:david.w.cruikshank@boeing.com]
>Sent: Thursday, July 07, 2005 3:31 PM
>To: Benoit Bezaire; cgmo-webcgm@lists.oasis-open.org
>Subject: RE: [cgmo-webcgm] Fwd: Re: WebCGM RGB/sRGB question
>
>
>So my question as a cgm generator:
>
>It says here (and in the WebCGM profile) that RGBa, sRGB, and sRGBa are
>registered.  Unfortunately, I can't find them in the ISO International
>Register of Graphical Items
>(http://jitc.fhu.disa.mil/nitf/graph_reg/graph_reg.html).  I now don't know
>how to encode these values in the Colour Table element, or direct colour
>setting attributes.  Is the profile underspecified for these?  Or am I
>misssing something?
>
>thx...Dave
>
>-----Original Message-----
>From: Benoit Bezaire [mailto:benoit@itedo.com]
>Sent: Thursday, July 07, 2005 12:44 PM
>To: cgmo-webcgm@lists.oasis-open.org
>Subject: [cgmo-webcgm] Fwd: Re: WebCGM RGB/sRGB question
>
>
>This is a forwarded message
>From: Chris Lilley <chris@w3.org>
>To: Benoit Bezaire <benoit@itedo.com>
>Date: Thursday, July 7, 2005, 3:17:59 PM
>Subject: WebCGM RGB/sRGB question
>
>===8<==============Original message text===============
>On Thursday, July 7, 2005, 8:14:03 PM, Benoit wrote:
>
>BB> Hi Chris,
>
>BB>   The WebCGM working group would like to ask you a question regarding
>BB>   color spaces within WebCGM documents.
>
>BB>   The allowed color spaces is currently under review for WebCGM 2.0.
>
>BB>   In WebCGM 1.0
>BB>
>BB> (http://www.w3.org/TR/REC-WebCGM/REC-02-CGM-Concepts.html#webcgm_2_5_3)
>BB>   the spec says:
>BB>   "The full range of standard CGM:1999 color models is limited in
>BB>   WebCGM Edition 1.0. The ATA GREXCHANGE RGB model is included, as
>BB>   well as the models: RGB-alpha; the colorimetric RGB space of the
>BB>   Web, sRGB; and sRGB-alpha. The latter three are registered in the
>BB>   ISO Register of Graphical Items."
>
>BB>   The group feels that too many color spaces are allowed and would
>BB>   like to reduce the set.
>
>Okay.
>
>BB> The intent is to deprecated sRGB and
>BB>   sRGB-alpha (and only keep RGB and RGB-alpha). After more carefully
>BB>   reviewing how colors are used in these technical markets, it may
>BB>   after all make sense.
>
>BB>   The bottom line is that no tools (within the group) have color
>BB>   management or some sort of color calibration method.
>
>The *entire point* of sRGB is of course that, while being colorimetricaly
>defined, and thus convertible  to other color spaces, it is specifically
>designed to use the transfer curve and primaries and illumination
>characteristics of a typical office environment and thus be displayable
>without any color calibration.
>
>BB>  They simply
>BB>   read in the R, G, B, data and blit them on the screen without any
>BB>   processing within those two steps. All applications basically work
>BB>   with uncalibrated R, G, B data. Since applications do not make the
>BB>   distinction between RGB and sRGB, the group's intent is to
>BB>   deprecated sRGB and sRGB-alpha...
>
>BB>   The group is however concerned, of negative feedback from an
>BB>   organization like W3C regarding such a decision.
>
>Yes, it would frankly be a really bad idea. Deprecate RGB and RGBA instead.
>
>
>
>
>--
>  Chris Lilley                    mailto:chris@w3.org
>  Chair, W3C SVG Working Group
>  W3C Graphics Activity Lead
>
>
>===8<===========End of original message text===========
>
>As I expected, Chris doesn't like the idea of deprecating sRGB and
>sRGBA (and I actually agree with him, we are trying to deprecate the
>wrong color spaces).
>
>--
>Best regards,
>  Benoit                            mailto:benoit@itedo.com