Re: [groff] 05/14: tbl(1): Say decimal "separator", not "point".

[G. Branden Robinson]

Dave noted in Savannah #61371[1] that there hadn't been much follow-up
on this.

> However, Keith's claim that "decimal separator" is "invalid
> terminology" is contradicted by Wikipedia's extensive, and extensively
> annotated, entry for the term
> <http://en.wikipedia.org/wiki/Decimal_separator>, which also says that
> "radix point" is the more general term used when bases other than 10
> are under consideration.

However, as I think Ralph might remind us, Wikipedia is not
authoritative.

At 2021-11-01T22:48:21+0000, Keith Marshall wrote:
> On 01/11/2021 13:19, G. Branden Robinson wrote:
> > gbranden pushed a commit to branch master
> > in repository groff.
> > 
> > commit a0ec5ffd258b9f54daa46b88471ec837e8213ad1
> > Author: Bjarni Ingi Gislason <bjarniig@rhi.hi.is>
> > AuthorDate: Sun Oct 31 00:42:09 2021 +0000
> > 
> >     tbl(1): Say decimal "separator", not "point".
> 
> Surely, the correct English-language terminology is "radix point",

There is little in English language usage that is apodictic.

> (and where the radix is "decimal", this becomes "decimal point").  It
> is irrelevant whether the prevailing convention, of the user's locale,
> is to represent the radix point by a comma, or a period, we do *not*
> refer to a "decimal comma", or a "decimal period", (or a "decimal
> dot"); the correct terminology is "decimal point".

tbl(1) itself doesn't recognize any number base other than decimal, and
overgeneralizing could mislead the reader into an inference that it
does.

> This change introduces invalid terminology, and should be reverted.

I don't agree.  POSIX uses the terms "decimal delimiter" and "radix
character"[2].  On the commercial side, Oracle[3] and Microsoft[4] both
use the term "decimal separator" in technical documentation.  On the
more general international standardization front, apparently the term
"decimal marker" is preferred[5], at least by the Bureau International
des Poids et Mesures (BIPM).

Given that the table option name to which the commit refers is
"decimalpoint", which goes back to 1970s AT&T tbl, has no synonym in any
tbl implementation I'm aware of, and therefore would be inadvisable to
withdraw in a short time frame, the reader's view of your approved
terminology is not obscured.

In a more purely mathematical context, I prefer the term "radix point",
myself.

Anyone have access to authoritative sources that might weigh on the
issue?

Regards,
Branden

[1] https://savannah.gnu.org/bugs/?61371
[2] https://pubs.opengroup.org/onlinepubs/9699919799/basedefs/V1_chap07.html
[3] in "Solaris Internationalization Overview"
    https://docs.oracle.com/cd/E19455-01/806-0169/overview-9/index.html
[4] in a document hierarchy with a root named "Globalization overview"
    https://docs.microsoft.com/en-us/globalization/locale/number-formatting
[5] 
https://web.archive.org/web/20210121220851/https://www.bipm.org/utils/en/pdf/Resol22CGPM-EN.pdf

signature.asc
Description: PGP signature