Re: wrong enumeration of adjustment statuses (request '.ad')

[G. Branden Robinson]

Hi Bjarni,

Wrong enumeration?  You are mistaken.

At 2024-12-19T23:04:23+0000, Bjarni Ingi Gislason wrote:
>   Still current:
> 
> adjustment types:   0 1 3 5 (left both center right)
> no-adjustment modes: 0 2 4
> 
>   So enumeration '0' is the same for two different statuses but should
>   be different

That's not AT&T or Heirloom Doctools troff-compatible.[1]

You observed yourself almost 6 months ago that 'all the different
implementations (also "plan9") ... have the same bug.'[2]

> like
> adjustment types:   0 3 5 7 (left both center right)
> no-adjustment modes: 2 4 6

Changing that is a wishlist item.  But rather than put lipstick on an
orc, I'd rather get us away from the `ad` request entirely, except for
backward compatibility, and even then I want to change some corner-case
behavior except in AT&T compatibility mode, as you should recall.[2]

Here is my proposal:

https://savannah.gnu.org/bugs/?65954

>   See more in bug #59795 (has a patch)

Resolved as fixed for nearly four years and I see no case here for
revisiting that status.

> (and #55579).

Resolved as a duplicate of #55579.

These bugs are exhibits of you desiring a feature change in groff, not
of you discovering a defect,[3] let alone one that is unresolved.

Regards,
Branden

[1]

$ cat EXPERIMENTS/bjarni-adjustment.roff
.tm \n(.j
.na
.tm \n(.j
.ad
.tm \n(.j
.ad l
.tm \n(.j
.ad
.tm \n(.j
.ad c
.tm \n(.j
.na
.tm \n(.j
.ad
.tm \n(.j
.ad l
.tm \n(.j
.ad
.tm \n(.j
.ad r
.tm \n(.j
.na
.tm \n(.j
.ad
.tm \n(.j
.ad l
.tm \n(.j
.ad b
.tm \n(.j
.na
.tm \n(.j
$ echo $(~/heirloom/bin/nroff < EXPERIMENTS/bjarni-adjustment.roff 2>&1)
1 0 1 0 1 3 2 3 0 1 5 4 5 0 1 0
$ echo $(DWBHOME=~/dwb ~/dwb/bin/nroff < EXPERIMENTS/bjarni-adjustment.roff 
2>&1)
1 0 1 0 1 3 2 3 0 1 5 4 5 0 1 0

[2] https://lists.gnu.org/archive/html/groff/2024-06/msg00056.html
    https://lists.gnu.org/archive/html/groff/2024-06/msg00057.html

[3] Although I have said here and elsewhere that the form of `ad` that
    landed in Seventh Edition Unix was a big goose egg.  And I expect I
    will continue to say so.

signature.asc
Description: PGP signature