FAQ Topic - I have window.status="Moomin"; why doesn't the statusbar change? (2009-11-04)

Thomas 'PointedEars' Lahn wrote:

Dr J R Stockton wrote:

Garrett Smith posted:
The reason the FAQ is structured using semantic fragment identifiers has
been explained by me, at least twice, IIRC, and in discussion where you
were involved.

I cannot recall if you specifically responded to those.
An explanation and a good reason are not the same thing. Mark-up in
source has its uses; but it does not satisfy all needs. If it did, it
would not be necessary to have the existing two-level numbering.

One can look at the visible rendition of the FAQ and then write "See FAQ
4.26", for example; and for the whole of sections 4 to 12, that provides
sufficient resolution. In section 1.3 it is inadequate.

JFTR:

NO to any section numbering.
[...]

YES to section titles
[...]

Yes and yes as discussed at before and here again.

The want for numbering is perhaps coming from some sort of want to have
a bible-type of reference.

e.g. "see FAQ 12:20"

Which is fine for documents that are intended to final or unrevised.

That is not fine for a volatile document that is questioned and
criticized frequently; not fine for the FAQ.

Is there further objection to remove this "status bar" entry?
--
Garrett
comp.lang.javascript FAQ: http://jibbering.com/faq/

In comp.lang.javascript message <23746114.vaeNJFYEL5 (AT) PointedEars (DOT) de>,
Mon, 16 Nov 2009 12:23:04, Thomas 'PointedEars' Lahn
<PointedEars (AT) web (DOT) de> posted:

NO to any section numbering. Numbering requires efforts for tedious editing
(or programming of numbering) better spent elsewhere, and is not future-
proof: Any newly inserted question-answer pair could require renumbering,
thereby invalidating past numeric references. This would cause the FAQ to
have a tendency to be static instead of being structured dynamically in the
best way for the dedicated reader. (We have discussed this already.)

Numbering to the extent required is a trivial task, taking less time
than arguing about it.

Historical references to FAQ sections are of no practical importance;
the contents of sections can also change. And when referring to an aged
section, one should clearly give a link to the WayBack Machine to get
the correct version, which will have its appropriate number.

--
(c) John Stockton, Surrey, UK. ?@merlyn.demon.co.uk Turnpike v6.05 MIME.
Web <URL:http://www.merlyn.demon.co.uk/> - FAQish topics, acronyms, & links.
Proper <= 4-line sig. separator as above, a line exactly "-- " (SonOfRFC1036)
Do not Mail News to me. Before a reply, quote with ">" or "> " (SonOfRFC1036)

In comp.lang.javascript message <hdrsdj$fj0$1 (AT) news (DOT) eternal-
september.org>, Mon, 16 Nov 2009 07:51:13, Garrett Smith
<dhtmlkitchen (AT) gmail (DOT) com> posted:

Thomas 'PointedEars' Lahn wrote:
Dr J R Stockton wrote:

Garrett Smith posted:
The reason the FAQ is structured using semantic fragment identifiers has
been explained by me, at least twice, IIRC, and in discussion where you
were involved.

I cannot recall if you specifically responded to those.
An explanation and a good reason are not the same thing. Mark-up in
source has its uses; but it does not satisfy all needs. If it did, it
would not be necessary to have the existing two-level numbering.

One can look at the visible rendition of the FAQ and then write "See FAQ
4.26", for example; and for the whole of sections 4 to 12, that provides
sufficient resolution. In section 1.3 it is inadequate.
JFTR:
NO to any section numbering.
[...]

YES to section titles
[...]

Yes and yes as discussed at before and here again.

The want for numbering is perhaps coming from some sort of want to have
a bible-type of reference.

e.g. "see FAQ 12:20"

Which is fine for documents that are intended to final or unrevised.

Nonsense : if such a reference is given in News, which is designed as an
ephemeral medium. it is clearly intended to be seen as a reference to
the current FAQ, and will work as such.

If one is reading back in News, which few normal people actually do, and
finds a link-with-anchor to Jibbering, Jibbering will serve the
*current* version, which will not be what is cited. The only reasonably
safe long-term reference to a past version is to give a full WayBack
Machine link-and-anchor.

AFAIR, **ALL** previous CLJ FAQ maintainers have used a numbering system
for sections and subsections, to the general satisfaction in that
respect.

That is not fine for a volatile document that is questioned and
criticized frequently; not fine for the FAQ.

Is there further objection to remove this "status bar" entry?

Yes.

--
(c) John Stockton, Surrey, UK. ?@merlyn.demon.co.uk Turnpike v6.05 MIME.
Web <URL:http://www.merlyn.demon.co.uk/> - FAQish topics, acronyms, & links.
Proper <= 4-line sig. separator as above, a line exactly "-- " (SonOfRFC1036)
Do not Mail News to me. Before a reply, quote with ">" or "> " (SonOfRFC1036)

Garrett Smith wrote:

Thomas 'PointedEars' Lahn wrote:
Dr J R Stockton wrote:
JFTR:

NO to any section numbering.
[...]

YES to section titles
[...]

Yes and yes as discussed at before and here again.

The want for numbering is perhaps coming from some sort of want to have
a bible-type of reference.

e.g. "see FAQ 12:20"

Which is fine for documents that are intended to final or unrevised.

That is not fine for a volatile document that is questioned and
criticized frequently; not fine for the FAQ.

So are you going to remove the numbering in favor of a more flexible
approach?

Is there further objection to remove this "status bar" entry?

In its current form it should be disabled. It should not be removed because
the question might come up again.

However, if the following replacements would be made, it could as well stay
in: "event" should be "event-handler attribute value", "setTimeout" should
be "window.setTimeout", "15" should be "50".


PointedEars
--
Danny Goodman's books are out of date and teach practices that are
positively harmful for cross-browser scripting.
-- Richard Cornford, cljs, <cife6q$253$1$8300dec7 (AT) news (DOT) demon.co.uk> (2004)

Dr J R Stockton wrote:

Garrett Smith posted:
The want for numbering is perhaps coming from some sort of want to have
a bible-type of reference.

e.g. "see FAQ 12:20"

Which is fine for documents that are intended to final or unrevised.

Nonsense : if such a reference is given in News, which is designed as an
ephemeral medium. it is clearly intended to be seen as a reference to
the current FAQ, and will work as such.

If one is reading back in News, which few normal people actually do,

Wishful thinking. You underestimate the usefulness of a newsgroup archive.

and finds a link-with-anchor to Jibbering, Jibbering will serve the
*current* version, which will not be what is cited.

That is the problem that comes with numbering, a problem that can be avoided
when numbering is abandoned, and references are bound to a specific context
instead.

The only reasonably safe long-term reference to a past version is to give
a full WayBack Machine link-and-anchor.

Non sequitur. If the different to a past version only means that the
position of a question-answer pair has changed, a fragment identifier serves
better than a number, also with a Web archive.

AFAIR, **ALL** previous CLJ FAQ maintainers have used a numbering system
for sections and subsections, to the general satisfaction in that
respect.

So what? Times and opinions change; we are growing with our experiences.
Except for technical Specifications, e-books, and other seldom changed
content, you are not going to find many numbered lists on the Web anymore;
the OL element has become relatively rare, and CSS property support for
numbered lists has been waiting for wide implementation for several years
now. That is because numbered lists essentially create static reference
points; they do not meet the requirements of dynamic Web publishing. The
Web provides links and anchors to refer to content instead of section
numbers and page numbers as in the old days, before the Web.


PointedEars
--
Prototype.js was written by people who don't know javascript for people
who don't know javascript. People who don't know javascript are not
the best source of advice on designing systems that use javascript.
-- Richard Cornford, cljs, <f806at$ail$1$8300dec7 (AT) news (DOT) demon.co.uk>

In comp.lang.javascript message <5621194.EvYhyI6sBW (AT) PointedEars (DOT) de>,
Tue, 17 Nov 2009 01:10:43, Thomas 'PointedEars' Lahn
<PointedEars (AT) web (DOT) de> posted:

Dr J R Stockton wrote:

Garrett Smith posted:
The want for numbering is perhaps coming from some sort of want to have
a bible-type of reference.

e.g. "see FAQ 12:20"

Which is fine for documents that are intended to final or unrevised.

Nonsense : if such a reference is given in News, which is designed as an
ephemeral medium. it is clearly intended to be seen as a reference to
the current FAQ, and will work as such.

If one is reading back in News, which few normal people actually do,

Wishful thinking. You underestimate the usefulness of a newsgroup archive.

and finds a link-with-anchor to Jibbering, Jibbering will serve the
*current* version, which will not be what is cited.

That is the problem that comes with numbering, a problem that can be avoided
when numbering is abandoned, and references are bound to a specific context
instead.

You fail to realise that the content of the section now at Jibbering may
not be the content at the time that it was linked to.

It is highly likely that there exist, in the news archives, numerous
cases where someone has strongly criticised a section of the FAQ as it
then was. A link given at the time and archived will still point to
Jibbering, where the current version is. Very possibly the current
version will have changed and will no longer deserve that criticism.

The only way of making a long-term reliable reference is to refer to a
copy of the cited document unchanged since the time of citation.

The only reasonably safe long-term reference to a past version is to give
a full WayBack Machine link-and-anchor.

Non sequitur. If the different to a past version only means that the
position of a question-answer pair has changed, a fragment identifier serves
better than a number, also with a Web archive.

That is why I did not, in the part quoted, say number; I said anchor.

--
(c) John Stockton, Surrey, UK. ?@merlyn.demon.co.uk Turnpike v6.05 MIME.
Web <URL:http://www.merlyn.demon.co.uk/> - FAQish topics, acronyms, & links.
Proper <= 4-line sig. separator as above, a line exactly "-- " (SonOfRFC1036)
Do not Mail News to me. Before a reply, quote with ">" or "> " (SonOfRFC1036)

Dr J R Stockton wrote:

Thomas 'PointedEars' Lahn posted:
Dr J R Stockton wrote:
Garrett Smith posted:
The want for numbering is perhaps coming from some sort of want to have
a bible-type of reference.

e.g. "see FAQ 12:20"

Which is fine for documents that are intended to final or unrevised.

Nonsense : if such a reference is given in News, which is designed as an
ephemeral medium. it is clearly intended to be seen as a reference to
the current FAQ, and will work as such.

If one is reading back in News, which few normal people actually do,

Wishful thinking. You underestimate the usefulness of a newsgroup
archive.

and finds a link-with-anchor to Jibbering, Jibbering will serve the
*current* version, which will not be what is cited.

That is the problem that comes with numbering, a problem that can be
avoided when numbering is abandoned, and references are bound to a
specific context instead.

You fail to realise that the content of the section now at Jibbering may
not be the content at the time that it was linked to.

On the contrary, that the FAQ is a document that is frequently changed over
time is exactly my/our point. Two cases have to be considered:

A) Content being referred to that is no longer there at all

In this case, neither the fragment identifier, section title or
section number helps to find the content there.

There are no negative effects with a fragment identifier for which no
anchor can be found. The document will be displayed at the top (and the
observant user will recognize that this must be due to the fragment
identifier that refers to an anchor that not -- assuming the best case,
no longer -- exists).

The negative effect for a section number or section title being referred
to but can no longer be found is that the user will have to read the
section numbers from top to bottom until they discover there is no longer
such a section.

Using a Web archive is necessary in either case, and it matters little
whether the archived version uses section numbering or not because the
section can still be identified by its title (that should be given in
any case).

B) Content being referred to is placed at a different position than before

In this case, the fragment identifier still points to the correct
section; the Web browser will scroll down to the anchor that correlates
with the given fragment identifier.

The section number not only does not point to the intended section, it
may very well point to the *wrong* section, thereby creating confusion.
That is, regardless whether the reference uses the section number as
plain text, in the fragment identifier, or both. If only the section
number was given (worst case), the user will be forced to consult Web
archives to find the content being referred to, even though the content
referred to is still there in the original FAQ. Where is the logic in
that?

The only reasonably safe long-term reference to a past version is to
give a full WayBack Machine link-and-anchor.

Non sequitur. If the different to a past version only means that the
position of a question-answer pair has changed, a fragment identifier
serves better than a number, also with a Web archive.

That is why I did not, in the part quoted, say number; I said anchor.

But the anchor is the target of the fragment identifier, so whatever applies
to the section number must apply to the section number in the fragment
identifier and the anchor name, if and when section numbers occur in the
fragment identifier and anchor name. So, for the very same reasons that
section numbering is unwise to use in the FAQ text, it is bad for fragment
identifiers for FAQ sections and anchor names in the FAQ.


PointedEars
--
realism: HTML 4.01 Strict
evangelism: XHTML 1.0 Strict
madness: XHTML 1.1 as application/xhtml+xml
-- Bjoern Hoehrmann