" versus "
Greg Berigan
EKN...@tiny.computing.csbsju.edu (Liz Knuth) wrote:
>gber...@cse.unl.edu (Greg Berigan) wrote:
>>EKN...@tiny.computing.csbsju.edu (Liz Knuth) wrote:
>>> But you can't use double quotes nested inside double quotes,
>>> because the browser will think your string ends at the second
>>> quote.
>> Are you sure on this?
> Yes. Read it again.
Yes, you are quite right. Mea culpa.
>>> Special cases:
>>> Some bozos use the ampersand in their domain names or file names.
>> I don't think ampersand is even allowed in domain names. I've never
>> heard or seen of such a thing, even as part of the machine-name
>> portion of the fully qualified domain name (FQDN).
> Dunno for sure. I can think of two links on the library pages
> where I had to put the ampersand in hex. Now, it may be that it was
> a username or part of the path, but I thought that at least one was
> in a domain name.
In "FAQ" <http://www.internic.net/faq/new-reg.html> it is written:
| 3. What are the valid characters for a domain name and how long can it be?
|
| The only valid characters for a domain name are letters,
| numbers and a hyphen. Special characters like the underscore
| (_) or an exclamation mark (!) are NOT permitted. Domain names
| cannot contain spaces or begin or end with a hyphen.
>> Having to escape such ampersands for such domains would serve as a
>> major deterrant to anyone choosing to register one.
> Lots of us are using the tilde in URLs... Ask Jukka how safe that is.
I believe ~ has been officially moved from the unsafe to the safe
category. Unfortunately I don't have a reference for this, so feel free
to treat it as anecdotal. :-)
Let's take this to comp.infosystems.www.authoring.misc.
--
,=<#)-=# <http://incolor.inetnebr.com/wotw/> (The War of the Worlds)
,_--//--_,
_-~_-(####)-_~-_ "Did you see that Parkins boy's body in the tunnels?" "Just
(#>_--'~--~`--_<#) the photos. Worst thing I've ever seen; kid had no face."
Claus André Färber
> I believe ~ has been officially moved from the unsafe to the safe
> category. Unfortunately I don't have a reference for this, so feel free
> to treat it as anecdotal. :-)
RFC 2396, Appendix G.2, paragraph 5:
| The tilde "~" character was added to those in the "unreserved" set,
| since it is extensively used on the Internet in spite of the
| difficulty to transcribe it with some keyboards.
--
Claus Andre Faerber <http://www.muc.de/~cfaerber/> fax: +49-8061-3361
PGP: ID=1024/527CADCD FP=12 20 49 F3 E1 04 9E 9E 25 56 69 A5 C6 A0 C9 DC
Warren Steel
> Greg Berigan <gber...@cse.unl.edu> schrieb:
> > I believe ~ has been officially moved from the unsafe to the safe
> > category. Unfortunately I don't have a reference for this, so feel free
> > to treat it as anecdotal. :-)
> RFC 2396, Appendix G.2, paragraph 5:
> | The tilde "~" character was added to those in the "unreserved" set,
> | since it is extensively used on the Internet in spite of the
> | difficulty to transcribe it with some keyboards.
This refers, I believe, to "national keyboard"
systems used in several countries, which may lack
the tilde. My own logs reveal, however, that there
remain problems with the tilde in URLs. My error
logs show requests for _mudws, for -mudws, and other
constructs. Perhaps ~mudws was mistranscribed by
hand on a piece of paper, or misread from the same
piece of paper, or mistyped in a news-paper article.
Though I routinely use the tilde, I would not call
it safe on the Web.
--
Warren Steel mu...@olemiss.edu
Department of Music University of Mississippi
URL: http://www.mcsr.olemiss.edu/%7Emudws/
Jukka Korpela
wrote:
>Though I routinely use the tilde, I would not call
>it safe on the Web.
I routinely try to avoid the tilde, and I don't think it's safe at
all. It was a mistake to move it away from the unsafe set; luckily we
can still keep encoding tilde to stay on the safe side.
There's one point that needs to be emphasized in addition to the
points I make in http://www.hut.fi/u/jkorpela/tilde.html
The point is that in _any_ processing of data, tilde is problematic
due to its history. It was originally introduced to ASCII as a dual,
if not split-minded, character, and actually called "overline". (Don't
ask me for what purpose. Perhaps something similar to "underline").
The secondary purpose intended was to use it as the "tilde" diacritic
so that e.g. letter "n" followed by backspace and tilde would produce
letter n with tilde (the letter used in Spanish which can be denoted
in HTML by ñ). Such approach to creating letters with
diacritics is not very popular, luckily, but many programs still
reflect it, some way or another (e.g. so that the tilde key, which
might actually carry some other label, is "mute" - you need to press
space to get the ~ character - that's what just had to do to type
it!). The originally intended primary use as overline is probably
completely forgotten, partly because the glyph of the character is
tilde-like ("wave"), not a straight overline. But in the early days of
computing, when bits were expensive and bytes cost fortunes, people
took the tilde into all kinds of uses. Just because it was there, in
the keyboards and character code. I'll buy a beer to anyone who
explains at least quasi-rationally why the tilde, originally formed
from letter "n" written over a character e.g. to indicate that it is
nasalized, got the function of being negation operator or the "home
directory of..." meaning. Moreover, the ISO 646 standard (the
internationalized generalization of ASCII) specifies the code position
of tilde as being for national use, to be defined in national
standards. Thus, one should _expect_ inconsistencies. The tilde
character may appear, for example, as letter "u" with diaeresis (u
umlaut).
Well, I'd say, enough of this madness. Never use the tilde unless you
really must. In URLs, one can always escape it.
--
Yucca, http://www.hut.fi/u/jkorpela/ or http://yucca.hut.fi/yucca.html
Warren Steel
[ many good points and http://www.hut.fi/u/jkorpela/tilde.html ]
Thanks for the informative post, and the cited essay.
Liz Knuth
of comp.infosystems.www.authoring.html in these words:
> In comp.infosystems.www.authoring.html,
> EKN...@tiny.computing.csbsju.edu (Liz Knuth) wrote:
> >gber...@cse.unl.edu (Greg Berigan) wrote:
> >> I don't think ampersand is even allowed in domain names. I've
> >> never heard or seen of such a thing, even as part of the
> >> machine-name portion of the fully qualified domain name (FQDN).
> > Dunno for sure. I can think of two links on the library pages
> > where I had to put the ampersand in hex. Now, it may be that it was
> > a username or part of the path, but I thought that at least one was
> > in a domain name.
> In "FAQ" <http://www.internic.net/faq/new-reg.html> it is written:
> | 3. What are the valid characters for a domain name and how long can
> | it be?
> |
> | The only valid characters for a domain name are letters,
> | numbers and a hyphen.
I checked. The two URLs that required my putting the ampersand in
hex were: one path, one file name. No domain names. i.e.,
http://www.foobar.com/baz&bat/foo.html
and
http://www.foobar.com/baz&bat.html
So the deciding factor for how to get around ampersands in URLs is
whether or not the ampersand is being used as a field separator in a
query to a server side program. If yes, then & or & . If no,
%26 .
Liz
Vote NO on moderation of ciwah
--
Elizabeth T. Knuth "A cockroach in a tuxedo
ekn...@tiny.computing.csbsju.edu is still a cockroach."
http://www2.csbsju.edu/library/training/incen.html -Richard K. Bethell,
Internet-Centric Resources nanae, 6 Nov 97