bug#14802: 24.3.50; (elisp) Multiple Terminals - what is a terminal attribute?

bug#14802: 24.3.50; (elisp) Multiple Terminals - what is a terminal attribute?

[Drew Adams]

The term "terminal attribute" is mentioned here only, and with no
explanation.  What is it?  Is it the same thing as a terminal parameter?
The wording would suggest that one of the terminal attributes is a list
of terminal parameters, but things are not very clear at all.

In GNU Emacs 24.3.50.1 (i686-pc-mingw32)
 of 2013-07-01 on LEG570
Bzr revision: 113246 lekktu@gmail.com-20130701165437-ea20s94hqwp3ttaj
Windowing system distributor `Microsoft Corp.', version 6.1.7601
Configured using:
 `configure --prefix=/c/usr --enable-checking CFLAGS='-O0 -g3'
 CPPFLAGS='-DGLYPH_DEBUG=1 -I/c/usr/include''

bug#14802: 24.3.50; (elisp) Multiple Terminals - what is a terminal attribute?

[Lars Ingebrigtsen]

Drew Adams <drew.adams@oracle.com> writes:

> The term "terminal attribute" is mentioned here only, and with no
> explanation.  What is it?  Is it the same thing as a terminal parameter?
> The wording would suggest that one of the terminal attributes is a list
> of terminal parameters, but things are not very clear at all.

I can't find that term at all in the manual, so this seems like it's
fixed now.

-- 
(domestic pets only, the antidote for overdose, milk.)
  bloggy blog http://lars.ingebrigtsen.no/

bug#14802: 24.3.50; (elisp) Multiple Terminals - what is a terminal attribute?

[Drew Adams]

> I can't find that term at all in the manual, so this seems like it's
> fixed now.

Well, you have to read. ;-)  The undefined concept is introduced in the
node I cited.

    "Each terminal object has the following attributes"

What does that mean?  What kind of an Emacs thingie is a terminal
attribute?

bug#14802: 24.3.50; (elisp) Multiple Terminals - what is a terminal attribute?

[Lars Ingebrigtsen]

Drew Adams <drew.adams@oracle.com> writes:

>> I can't find that term at all in the manual, so this seems like it's
>> fixed now.
>
> Well, you have to read. ;-)  The undefined concept is introduced in the
> node I cited.
>
>     "Each terminal object has the following attributes"
>
> What does that mean?  What kind of an Emacs thingie is a terminal
> attribute?

The definition is indexed under "terminal"...

-- 
(domestic pets only, the antidote for overdose, milk.)
  bloggy blog http://lars.ingebrigtsen.no/

bug#14802: 24.3.50; (elisp) Multiple Terminals - what is a terminal attribute?

[Eli Zaretskii]

> Date: Sun, 9 Feb 2014 17:27:57 -0800 (PST)
> From: Drew Adams <drew.adams@oracle.com>
> Cc: 14802@debbugs.gnu.org
> 
> What kind of an Emacs thingie is a terminal attribute?

It's an attribute of a terminal object.

bug#14802: 24.3.50; (elisp) Multiple Terminals - what is a terminal attribute?

[Drew Adams]

> > What kind of an Emacs thingie is a terminal attribute?
> 
> It's an attribute of a terminal object.

OK, that gives it a name.  Now how about saying what that
is (in the manual)?

Either that or don't introduce the term - it is just vapor
otherwise, and confusing.  Users can wonder what they're missing,
and where to find it.

If the text said "characteristics" or "qualities" or similar,
users wouldn't wonder so much.  But "attribute" typically means
something particular in software.  If what is meant here is the
ordinary English use of the word then I suggest picking a less
technical-sounding equivalent.

If what is meant really is something technical that is known
as a terminal attribute (as one user, I really don't know),
then that term either needs explaining or users should be
sent elsewhere for info about it.

bug#14802: 24.3.50; (elisp) Multiple Terminals - what is a terminal attribute?

[Eli Zaretskii]

> Date: Sun, 9 Feb 2014 20:03:52 -0800 (PST)
> From: Drew Adams <drew.adams@oracle.com>
> Cc: larsi@gnus.org, 14802@debbugs.gnu.org
> 
> > > What kind of an Emacs thingie is a terminal attribute?
> > 
> > It's an attribute of a terminal object.
> 
> OK, that gives it a name.  Now how about saying what that
> is (in the manual)?
> 
> Either that or don't introduce the term - it is just vapor
> otherwise, and confusing.  Users can wonder what they're missing,
> and where to find it.

I really don't understand the purpose of this bug report, because this
very node is the answer to all your questions and gripes.  It
includes:

  . a full list of the attributes of a terminal object (directly below
    the single sentence you cited in your report)

  . documentation of functions that access those attributes, and

  . references to other nodes where related features and issues are
    described

What else is missing?

> If the text said "characteristics" or "qualities" or similar,
> users wouldn't wonder so much.  But "attribute" typically means
> something particular in software.  If what is meant here is the
> ordinary English use of the word then I suggest picking a less
> technical-sounding equivalent.
> 
> If what is meant really is something technical that is known
> as a terminal attribute (as one user, I really don't know),
> then that term either needs explaining or users should be
> sent elsewhere for info about it.

I see nothing wrong with the word "attribute".  In this very manual,
we have:

  . text attributes
  . face attributes
  . file attributes
  . package attributes
  . process attributes

What's wrong with having "terminal object attributes"?

bug#14802: 24.3.50; (elisp) Multiple Terminals - what is a terminal attribute?

[Drew Adams]

> I really don't understand the purpose of this bug report, because
> this very node is the answer to all your questions and gripes.  It
> includes:
> 
>   . a full list of the attributes of a terminal object (directly
>     below the single sentence you cited in your report)
>   . documentation of functions that access those attributes, and
>   . references to other nodes where related features and issues are
>     described

So you _are_ saying that "terminal attribute" is a term and concept
being introduced, and we are not just using the English word
"attribute" in a general sense.  Great.

In that case, please put it in quotes, as we usually do for a term
introduction.  And, as this bug report requests, say something
about what it is, if you can - what the term means.  Listing values
is not really a substitute for that.

> What else is missing?

See above.

> > If what is meant really is something technical that is known
> > as a terminal attribute (as one user, I really don't know),
> > then that term either needs explaining or users should be
> > sent elsewhere for info about it.
> 
> I see nothing wrong with the word "attribute".  In this very manual,
> we have:
> 
>   . text attributes
>   . face attributes
>   . file attributes
>   . package attributes
>   . process attributes
> 
> What's wrong with having "terminal object attributes"?

Nothing.  Say what you mean by the term, as we do for each of
the others you cited.  And index it.  Compare what you see by
following `i face attribute': The term is quoted and described:

  "Face attributes" determine the visual appearance of a face...

Or follow `i package attribute', to see:

  Each package (whether simple or multi-file) has certain
  "attributes":

(That one is a minimal description - it just introduces particular
attributes without saying what is meant by the term.  But at least
it introduces the term explicitly.)

BTW, "attributes of text" is in the index, but "text attribute"
should also be added.  And the indexed node is `Text Properties'.
The term "text attribute" should be cited there as meaning the
same as "text property", which is the term described.

Also, I said "each" above, but that's not quite right:

* The term "file attribute" also should be introduced explicitly,
  in node `File Attributes'.  E.g.:

    ...detailed information about a file, including the owner and
    group numbers, the number of names, the inode number, the size,
    and the times of access and modification.

  Add something like this: `These characteristics are referred to as
  "file attributes".'

* "Process attribute" is also not introduced explicitly, in node
  `System Processes'.  Also, although function `process-attributes'
  is indexed, `process attribute' is not.

bug#14802: 24.3.50; (elisp) Multiple Terminals - what is a terminal attribute?

[Eli Zaretskii]

> Date: Mon, 10 Feb 2014 11:27:58 -0800 (PST)
> From: Drew Adams <drew.adams@oracle.com>
> Cc: larsi@gnus.org, 14802@debbugs.gnu.org
> 
> > I really don't understand the purpose of this bug report, because
> > this very node is the answer to all your questions and gripes.  It
> > includes:
> > 
> >   . a full list of the attributes of a terminal object (directly
> >     below the single sentence you cited in your report)
> >   . documentation of functions that access those attributes, and
> >   . references to other nodes where related features and issues are
> >     described
> 
> So you _are_ saying that "terminal attribute" is a term and concept
> being introduced, and we are not just using the English word
> "attribute" in a general sense.

It's not a term or concept, no.  It's a word.

> In that case, please put it in quotes, as we usually do for a term
> introduction.

It's not a term.

> And, as this bug report requests, say something about what it is, if
> you can - what the term means.  Listing values is not really a
> substitute for that.

Yes, it is.  Because it's not a term.

> > I see nothing wrong with the word "attribute".  In this very manual,
> > we have:
> > 
> >   . text attributes
> >   . face attributes
> >   . file attributes
> >   . package attributes
> >   . process attributes
> > 
> > What's wrong with having "terminal object attributes"?
> 
> Nothing.  Say what you mean by the term, as we do for each of
> the others you cited.

We don't.

> And index it.

It's not a term, so no need to index it.

> Or follow `i package attribute', to see:
> 
>   Each package (whether simple or multi-file) has certain
>   "attributes":

Which is exactly what we say about terminal attributes.

> (That one is a minimal description - it just introduces particular
> attributes without saying what is meant by the term.  But at least
> it introduces the term explicitly.)

It's not a term.

> BTW, "attributes of text" is in the index, but "text attribute"
> should also be added.

Yes, we should also index "text", and perhaps also "index", because
otherwise who knows what these mean?

> * The term "file attribute" also should be introduced explicitly,
>   in node `File Attributes'.  E.g.:
> 
>     ...detailed information about a file, including the owner and
>     group numbers, the number of names, the inode number, the size,
>     and the times of access and modification.
> 
>   Add something like this: `These characteristics are referred to as
>   "file attributes".'
> 
> * "Process attribute" is also not introduced explicitly, in node
>   `System Processes'.  Also, although function `process-attributes'
>   is indexed, `process attribute' is not.

Like I said: terminal attributes are not special.

Closing, as there's nothing more to be said here, and no bug anywhere
in sight.