待辦事項 #41071

Add publicly known LCS_sRGB and LCS_WINDOWS_COLOR_SPACE

啟用日期: 2020-12-24 06:52 最後更新: 2021-04-12 17:39

回報者:
負責人:
類型:
狀態:
關閉
元件:
(無)
里程碑:
(無)
優先權:
5 - 中
嚴重程度:
5 - 中
處理結果:
Accepted
檔案:
1
Vote
Score: 0
No votes
0.0% (0/0)
0.0% (0/0)

細節

Here is a public source to add definitions of LCS_sRGB and LCS_WINDOWS_COLOR_SPACE

https://docs.microsoft.com/en-us/openspecs/windows_protocols/ms-wmf/eb4bbd50-b3ce-4917-895c-be31f214797f

Also here is public definitions of PROFILE_LINKED and PROFILE_EMBEDDED with and without LCS_ prefix

https://docs.microsoft.com/en-us/openspecs/windows_protocols/ms-wmf/3c289fe1-c42e-42f6-b125-4b5fc49a2b20

Please add to wingdi.h

Ticket History (3/17 Histories)

2020-12-24 06:52 Updated by: apodtele
  • New Ticket "Add publicly known LCS_sRGB and LCS_WINDOWS_COLOR_SPACE " created
2020-12-29 07:52 Updated by: keith
  • 類型 Update from Issues to Feature Request
評語

Thanks for these references. Please note that this should have been filed as a "feature request", not as an "issues" report; I've redesignated it accordingly.

Do you happen to know, for sure, what versions of Windows first supported these? If not, we'll have to accept the guidance of the "footnotes", which suggest that the first pair first appeared in Win95/WinNT4, while the latter pair didn't appear until Win98/Win2K.

2020-12-29 09:59 Updated by: apodtele
評語

I only know these must be defined as possible values of bV5CSType in BIITMAPV5HEADER (ca Win98) https://docs.microsoft.com/en-us/windows/win32/api/wingdi/ns-wingdi-bitmapv5header

However, VC2017 did not complain when I used LCS_sRGB in BITMAPV4HEADER (ca Win95).

2021-01-03 02:50 Updated by: apodtele
評語

To complete BITMAPV5HEADER constants, also add LCS_GM_ABS_COLORIMETRIC from https://docs.microsoft.com/en-us/openspecs/windows_protocols/ms-wmf/9fec0834-607d-427d-abd5-ab240fb0db38

There is also BI_CMYK, BI_CMYKRLE8, BI_CMYKRLE4 in https://docs.microsoft.com/en-us/openspecs/windows_protocols/ms-wmf/4e588f70-bd92-4a6f-b77f-35d0feaf7a57 and BI_ALPHABITFIELDS slightly off spec but searchable.

2021-02-07 02:58 Updated by: apodtele
評語

How can I possibly help to add the missing definitions? Would you like me to prepare a patch? I would like MinGW to support the BMP feature properly.

2021-02-07 03:22 Updated by: keith
評語

Reply To apodtele

How can I possibly help to add the missing definitions? Would you like me to prepare a patch?

A patch would certainly be welcome, (and helpful), although this is on my to-do list for the next release, in any event. If you do elect to submit a patch, please include a ChangeLog entry, (in the patch header ... not in diff format), and indicate how you would like the attribution to appear, in the ChangeLog file itself.

I would like MinGW to support the BMP feature properly.

As would I. Thanks for your interest, and support.

2021-02-08 03:28 Updated by: apodtele
  • File wingdi.diff (File ID: 5993) is attached
2021-02-08 03:32 Updated by: apodtele
評語

A diff is attached.

I also found multiple discrepancies with this block https://docs.microsoft.com/en-us/openspecs/windows_protocols/ms-wmf/55961a7b-f6f6-4f0b-a76b-b4cf652bc86b but I am not sure if you want to correct or just add them.

2021-02-08 06:26 Updated by: keith
  • File wingdi.diff (File ID: 5993) is deleted
2021-02-08 07:15 Updated by: keith
  • 負責人 Update from (無) to keith
評語

Reply To apodtele

A diff is attached.

Thanks. That is helpful ... although you did do exactly what I asked you not to, w.r.t. the ChangeLog entry, so I've taken the liberty of correcting it for you. I've also made some minor layout adjustments, so that this appears correct in the patch, as well as in the <wingdi.h> file itself.

I also found multiple discrepancies with this block https://docs.microsoft.com/en-us/openspecs/windows_protocols/ms-wmf/55961a7b-f6f6-4f0b-a76b-b4cf652bc86b but I am not sure if you want to correct or just add them.

If anything else is missing, then please add it; if it's already present, but incorrect, then please correct it.

Perhaps you could also clarify something for me: in the Microsoft docs, to which you referred me, I see no reference to PROFILE_LINKED, or to PROFILE_EMBEDDED, as aliases for LCS_PROFILE_LINKED, and LCS_PROFILE_EMBEDDED respectively. Whilst I don't mind leaving the aliases in place, can you please provide alternative references to justify doing so?

2021-02-08 11:29 Updated by: apodtele
評語

Re. PROFILE_* vs LCS_PROFILE_*

I just see both used in MS docs: BITMAPV5HEADER and BitmapV5Header Object. Go figure.

2021-02-14 01:57 Updated by: keith
評語

Reply To apodtele

Re. PROFILE_* vs LCS_PROFILE_*
I just see both used in MS docs: BITMAPV5HEADER and BitmapV5Header Object. Go figure.

Ambiguous, indeed. I guess it does make sense to support both forms, although officially, the LCS_prefixed form should be preferred.

Anyway, I committed this. Is there anything else you would like to add, within the scope of this ticket, or may I close it?

2021-02-14 02:25 Updated by: apodtele
評語

Thank you. Please close it.

2021-02-14 02:53 Updated by: keith
  • 狀態 Update from 開啟 to 關閉
  • 處理結果 Update from to Accepted
2021-04-12 00:21 Updated by: apodtele
評語

It would be prudent to release updated WSL before this summer. Many students use the break to try new tools or participate in GSoC.

2021-04-12 17:39 Updated by: keith
評語

Reply To apodtele

It would be prudent to release updated WSL before this summer ...

Hmm, yes ... it was on my TODO list, and I guess it was overdue. Done now: package files published yesterday, and mingw-get catalogue updated today. New release designated as composite of mingwrt-5.4.2, and w32api-5.4.2.

Attachment File List

編輯

Please login to add comment to this ticket » 登入