Apple comments on the W3C's proposal
Andrew Glasgow
Summary: Apple supports "royalty-free" use of patents in Web standards.
Cuss and discuss.
--
| Andrew Glasgow <amg39(at)cornell.edu> |
| SCSI is *NOT* magic. There are *fundamental technical |
| reasons* why it is necessary to sacrifice a young goat |
| to your SCSI chain now and then. -- John Woods |
David Venn-Brown
news:amg39.REMOVETHIS-C6...@newsstand.cit.cornell.edu...
> http://www.apple.com/about/w3c/
>
> Summary: Apple supports "royalty-free" use of patents in Web standards.
>
> Cuss and discuss.
It's a bit heavy reading - have I got this right? Apple thinks that Web
standards should be free for everyone to use without having to make payment?
It lost me a bit.
---
David.
dav...@ozemail.com.au fat...@hotmail.com
AVG wuz here.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.286 / Virus Database: 152 - Release Date: 10/9/2001
Simon Brooke
> http://www.apple.com/about/w3c/
>
> Summary: Apple supports "royalty-free" use of patents in Web standards.
>
> Cuss and discuss.
Good. That leaves Microsoft isolated as the only one of the authors of
this document who has not yet publicly retracted. It's also worth
noting that Pruce Perens and Eben Moglen have both accepted positions
on the W3C Patent Policy Board. It's not clear yet whether these are
voting positions, but they probably are not.
The battle is not, however, in my opinion, won, and won't be won until
membership of W3C is open to ordinary people at a reasonable rate;
until then the W3C must remain a rich corporates club and will
inevitably continue to advance the rich corporates' interest.
More discussion here:
<URL: http://www.jasmine.org.uk/~simon/bookshelf/papers/rand-response.html >
--
si...@jasmine.org.uk (Simon Brooke) http://www.jasmine.org.uk/~simon/
'You cannot put "The Internet" into the Recycle Bin.'
Arjun Ray
| The battle is not, however, in my opinion, won,
What battle? If it's to preserve the prerogative to ballyhoo the
hothouse products of the W3C as "standards", then you have found the
enemy...
| and won't be won until membership of W3C is open to ordinary people
| at a reasonable rate; until then the W3C must remain a rich corporates
| club
It has always been that, and nothing more.
The W3C is an industry consortium.
The W3C is not a standards body.
What part of "not" is so difficult to understand?
| and will inevitably continue to advance the rich corporates' interest.
Maybe, maybe not. A reasonable and non-discrininatory fee of $5000/yr
will allow you to have your say, too.
--
Standardization as a marketing tool is evil. - Eliot Kimber
Peter Wilson
news:878zed8...@gododdin.internal.jasmine.org.uk...
> Andrew Glasgow <amg39.RE...@cornell.edu.INVALID> writes:
>
> The battle is not, however, in my opinion, won, and won't be won until
> membership of W3C is open to ordinary people at a reasonable rate;
> until then the W3C must remain a rich corporates club and will
> inevitably continue to advance the rich corporates' interest.
I'm not too familiar with the workings of the W3C - a few years ago I was a
regular attendee of IETF meetings on behalf of my employer at the time. The
IETF had no membership fee - just pay yout $100 at the conference.
But - there are a number of reasons why this doesn't actually help - these
may or may not apply to W3C.
1. Progress...
At the IETF the bulk of the work and decisions were made at the meetings -
not on mailing lists. Although anyone could have a say - in practice it was
only those that could afford to be at the meetings in persons that had any
real influence.
2. Funding...
Large organisations have large pots of money - if they choose to use it. I
remember a time at the IETF when a certain organisation seemed to make a
concerted effort to 'buy' the IETF - and basically succeeded. The way they
did this was to bascially recruit all the influencial people at the IETF -
technologists, editors, chairpeople. It was quite amusing to watch the email
addresses change as they were recruited.
3. Weight of numbers
Related to the previous 2 points. Even should a small organisation consider
it worthwhile to attend these meetings - they will be in the company of -
say - 20 Microsoft people arguing for a (usually) over complex standard.
That can be qujite daunting - being a lone voice!
4. Implementation cost.
This approach is not in the interests of consumers. This standardisation
process seldomely results in a good practical technology - but rather a
bloated - over complex technology that give the likes of Microsoft the
advantage of sheer man-hours in development. They will get to that
technology first!
As far as I can work out - after a large number of years in this game - the
only really useful standards come from an initial small group of enthusiasts
producing a well honed technology for a specific task. That technology will
survive pretty well unscathed by the standards organisations for a few years
at least. In the mean time - a whole raft of pretty unnecessary and overly
complex technologies will evolve around that simple core (XML - the simple
adequate - elegant technology, XSLT - the complex over-engineered,
inefficient, sub functional result of a committee). (XPath 1.0 - good basic
technology solving a real problem. XPath 2.0 wish list, result of bandwagon
standards).
The organisation of the W3C and the membership fee, the patent argument is
not the real barrier - it's happened with every standardisation body in the
last 50 years. The difference now is that there is an Open Source movement
that wants to compete.
So - is there a solution? I think - in practice - the only chance the rest
of the world (read Open Source) stands to influence the standardisation
process is by getting organised. The Open Source community needs to create a
single front to the large organisations - push for some funding from those
organisations that benefit from Open Source and appear as a single entity -
Not - for example - Microsoft, Apple, IBM and 20 uncoordinated small
outfits - instead - Mircosoft, Apple, IBM and OpenSource.
It's similar to lots of little freedom fighters in a very loose coalition -
wanting to go in the same direction - but with no real leadership - going up
against a despotic regime. Doesn't matter how much they winge - doesn't
change anything. They need leadership and someone outside to provide them
with the means to fight.
Pete
--
Peter Wilson - Whitebeam architect (www.whitebeam.org) -
Combining Apache, JavaScript and XML solving real world problems.
-------
Simon Brooke
> Simon Brooke <si...@jasmine.org.uk> wrote:
>
> | The battle is not, however, in my opinion, won,
>
> What battle? If it's to preserve the prerogative to ballyhoo the
> hothouse products of the W3C as "standards", then you have found the
> enemy...
We need standards, because without them it's hard to create robustly
interoperable systems. If we are to have standards, we have to have a
mechanism for establishing them which has some degree of perceived
legitimacy - which people are prepared to buy into and respect.
The separation out of the W3C from the IETF has always seemed to me an
unfortunate move; for the IETF to once again take on the role of
standards setting body for the Web would seem to me a good thing. But,
from what I know of the personalities and history involved, that seems
to me not wholly probable.
So we're left with two alternatives: to increase W3C's legitimacy, or
to replace it.
Replacing it is not a trivial thing to do. A large collaborative decision
making process of any kind takes a lot of energy to actually make it
work, and a standards setting body for the Web is necessarily
large. That means organisation and a secretariat, and a lot of work
getting it set up and getting the necessary buy-in from as many as
possible parts of the Web constituency.
W3C already has the organisation and the secretariat, and it already
has buy in from a wide section of the constituency (although that's
been shaply damaged by the RAND debate). It seems to me that it is
less difficult to reform W3C than to replace it.
I'm still of the opinion that we should be prepared to replace it, if
necessary.
> | and won't be won until membership of W3C is open to ordinary people
> | at a reasonable rate; until then the W3C must remain a rich corporates
> | club
>
> It has always been that, and nothing more.
>
> The W3C is an industry consortium.
>
> The W3C is not a standards body.
That's entirely a semantic argument, akin to and as valid as arguments
about angels and pinheads. The W3C produces recommendations, and, up
until now, most of the technically competent parts of the community
have attempted to comply with those recommendations. If it walks like
a duck and it quacks like a duck, as far as I'm concerned it's a
duck. The question is, is it a good duck or a bad one?
Arjun Ray
| Arjun Ray <ar...@nmds.com.invalid> writes:
|> Simon Brooke <si...@jasmine.org.uk> wrote:
| We need standards, because without them it's hard to create robustly
| interoperable systems.
I disagree. "Need" - as in necessity - would apply were "robustly
interoperable systems" impossible otherwise. The more appropriate
verb is "want" - as in desirability or preference. And, in fact, the
people "wanting" will have to be the creators and maintainers of those
systems.
There are people who perceive value in interoperability, and sadly,
there are people who do not. If life weren't difficult, standards
would be as easy as wanting them. There is no quick fix.
| If we are to have standards, we have to have a mechanism for
| establishing them which has some degree of perceived legitimacy -
| which people are prepared to buy into and respect.
There's much more to it. Popular acclamation is not enough, because
*what* is "bought into" and "respected" has to survive the vagaries of
mob rule, vested interest and disinformation. Political constitutions
have the same problem: the concept of "sovereign authority" - a source
of impartial dispassionate judgment - works only when embodied in an
institution whose essential function is only to approve, and *not* to
propound or to promulgate. This is an essential feature of all true
standards bodies: they are the people doing the approving, and not the
proposing; and what really commands respect is that the approval can't
be suborned or bought. The approvers have no agenda - except perhaps
to make it as rigorous as possible for proposers to have to prove the
worthiness of their submissions - and play no favorites - anyone can
make a case.
Of course, it's just as easy to cheapen the meaning of "standard" -
all it takes today is people osmoting their feelgoods du jour from a
fourth estate that wouldn't recognize a clue even if it were on the
business end of a 2x4 and certainly not when there are deadlines to
meet and ad revenues to coddle - into something that is "given" to you
from some plausible or convenient source, the more big names dropped
in the process the better. There are a lot of people buying into and
respecting precisely that, but what they're getting are not standards,
but dicta from an emperor or dogma from a pope.
| The separation out of the W3C from the IETF has always seemed to me
| an unfortunate move; for the IETF to once again take on the role of
| standards setting body for the Web would seem to me a good thing.
| But, from what I know of the personalities and history involved,
| that seems to me not wholly probable.
The real reasons are that the IETF is much harder to convince, and
that the IETF can't be obliged to approve. They take the concept of
"consensus" very seriously: it is by no means out of the ordinary for
an IETF working group to dissolve without producing "deliverables".
This is entirely unacceptable to an agenda for hothouse "development"
of imprimaturs.
| So we're left with two alternatives: to increase W3C's legitimacy,
| or to replace it.
Only if you still want something "easier" than the IETF. Or the ISO.
The W3C's "legitimacy" is a PR elaboration of personal reputations.
There is no quick fix.
| Replacing it is not a trivial thing to do. A large collaborative
| decision making process of any kind takes a lot of energy to actually
| make it work, and a standards setting body for the Web is necessarily
| large. That means organisation and a secretariat, and a lot of work
| getting it set up and getting the necessary buy-in from as many as
| possible parts of the Web constituency.
Which "web constituency" other than the W3C has a problem with the
IETF?
| It seems to me that it is less difficult to reform W3C than to
| replace it.
Read "to make it what we want of it rather than what it is".
| I'm still of the opinion that we should be prepared to replace it,
| if necessary.
It became necessary quite a while back. I'd say, early '98 at the
latest though the signs were there close to a year before that.
|> The W3C is an industry consortium.
|>
|> The W3C is not a standards body.
|
| That's entirely a semantic argument, akin to and as valid as arguments
| about angels and pinheads.
Sorry, but it happens to be documented fact. The W3C as a "standards
body" is a PR-fanned fantasy.
| The W3C produces recommendations, and, up until now, most of the
| technically competent parts of the community have attempted to comply
| with those recommendations.
To their detriment. Most of the stuff emanating from the W3C is just
plain and simple bad engineering.
| If it walks like a duck and it quacks like a duck, as far as I'm
| concerned it's a duck. The question is, is it a good duck or a bad one?
Too late. You already said being a duck to you was good enough for
you.
Nick Kew
> Arjun Ray <ar...@nmds.com.invalid> writes:
>
> > Simon Brooke <si...@jasmine.org.uk> wrote:
> >
> > | The battle is not, however, in my opinion, won,
> >
> > What battle? If it's to preserve the prerogative to ballyhoo the
> > hothouse products of the W3C as "standards", then you have found the
> > enemy...
>
> We need standards, because without them it's hard to create robustly
> interoperable systems. If we are to have standards, we have to have a
> mechanism for establishing them which has some degree of perceived
> legitimacy - which people are prepared to buy into and respect.
Hear, hear!
We have a range of standards from fully open (best - typically
IETF-sponsored) through to totally proprietary. Lots of people
still regard MSWORD as a standard, and just accept getting shafted
every couple of years as it undergoes incompatible changes.
On that scale, W3C looks as good as we can reasonably hope for.
> So we're left with two alternatives: to increase W3C's legitimacy, or
> to replace it.
M$ have been trying to replace it de-facto since they first jumped on
the bandwagon and broke HTTP (which was of course pre-Halloween).
For the rest of the world to abandon W3C now could risk giving them
a clear run to the M$ Global Network.
> been shaply damaged by the RAND debate). It seems to me that it is
> less difficult to reform W3C than to replace it.
Better still, work with W3C to contribute whatever you can to
improving the Web.
> I'm still of the opinion that we should be prepared to replace it, if
> necessary.
Who's going to promote a White Paper for reform?
> > | and won't be won until membership of W3C is open to ordinary people
> > | at a reasonable rate; until then the W3C must remain a rich corporates
> > | club
It is a rich-mans club, but tolerates people as well. I feel more
welcome at W3C than I would expect to be at any of my local rich mans
clubs devoted to golf, horses, boats, or whatever. Indeed, my limited
contact with it has been a broadly similar experience to what Peter
Wilson describes elsewhere on this thread with IETF.
> --
> si...@jasmine.org.uk (Simon Brooke) http://www.jasmine.org.uk/~simon/
Oops - why has gnus not snipped your .sig? Must have it setup incorrectly.
> 'You cannot put "The Internet" into the Recycle Bin.'
What a shame!
--
Nick Kew
Site Valet - the essential service for anyone with a website.
<URL:http://valet.webthing.com/>
Simon Brooke
> Simon Brooke <si...@jasmine.org.uk> wrote:
>
> | Replacing it is not a trivial thing to do. A large collaborative
> | decision making process of any kind takes a lot of energy to actually
> | make it work, and a standards setting body for the Web is necessarily
> | large. That means organisation and a secretariat, and a lot of work
> | getting it set up and getting the necessary buy-in from as many as
> | possible parts of the Web constituency.
>
> Which "web constituency" other than the W3C has a problem with the
> IETF?
Boot's on the other foot. From conversations I have had with them, I
think key people in the IETF would be very wary of doing anything
which could be seen as 'stepping on W3C turf'. Some people got pretty
badly burned in the flamewars which led to the founding of W3C, and
don't want to go through that again.
Personally, I still see the IETF as the most legitimate available
body, but real politics is I don't think they want to play.
> | It seems to me that it is less difficult to reform W3C than to
> | replace it.
>
> Read "to make it what we want of it rather than what it is".
Fine, that's a reasonable summation.
Simon Brooke
> Simon Brooke <si...@jasmine.org.uk> writes:
>
> > I'm still of the opinion that we should be prepared to replace it, if
> > necessary.
>
> Who's going to promote a White Paper for reform?
You and me? Actually, there is quite a collective working on building
something out of this. I'm trying to keep in with the conversation,
but BeRo and Bruce Perens seem to be taking most of the lead.
> > > | and won't be won until membership of W3C is open to ordinary people
> > > | at a reasonable rate; until then the W3C must remain a rich corporates
> > > | club
>
> It is a rich-mans club, but tolerates people as well. I feel more
> welcome at W3C than I would expect to be at any of my local rich mans
> clubs devoted to golf, horses, boats, or whatever. Indeed, my limited
> contact with it has been a broadly similar experience to what Peter
> Wilson describes elsewhere on this thread with IETF.
Yeah, but, I pay my US$25 a year (plus a little more to sponsor third
world members) and that makes me a fully paid up voting member of
ISOC, which is IETF's parent body. I can't get into that position with
W3C for money I could possibly afford.
> > 'You cannot put "The Internet" into the Recycle Bin.'
Genuine Microsoft error message. I kid you not. I just loved the
honesty of it, and the keen sense of regret... oh how they wished they
could.
--
si...@jasmine.org.uk (Simon Brooke) http://www.jasmine.org.uk/~simon/
'You cannot put "The Internet" into the Recycle Bin.'
Andrew Glasgow
Arjun Ray <ar...@nmds.com.invalid> wrote:
> Which "web constituency" other than the W3C has a problem with the
> IETF?
I dunno. My perspective (which I admit is based on limited experience)
on standards organizations of the type that you mean -- that approve
existing ideas as standards, rather than creating new ideas -- is that
if we relied on them for our standards, the only standard HTML today
would be IETF HTML, i.e. HTML 2.0 -- and no one would use it.
Juan R. Pozo
> Arjun Ray writes:
>
> > The W3C is an industry consortium.
> >
> > The W3C is not a standards body.
>
> That's entirely a semantic argument, akin to and as valid
> as arguments about angels and pinheads.
This is from <http://www.ietf.org/rfc/rfc1602.txt>:
<<In general, an Internet Standard is a specification that
is stable and well-understood, is technically competent, has
multiple, independent, and interoperable implementations
with substantial operational experience, enjoys significant
public support, and is recognizably useful in some or all
parts of the Internet.>>
<<In summary, the goals for the Internet standards process
are: * technical excellence;
* prior implementation and testing;
* clear, short, and easily understandable
documentation;
* openness and fairness; and
* timeliness.>>
Something doesn't fit here...
To me, the W3C is ok, as long as you understand it's the
industry. It's better than nothing, I suppose. The WAI is a
great initiative (though it's very sad to see how they have
to claudicate on so many things in view of the poor
implementation that W3C members themselves make of W3C
Recommendations).
Perhaps you were talking about de facto standards, but then
could you please specify which de facto standard you were
talking about? (HTML as in IE, NS, Opera...? CSS as in MS,
NS, Opera...? Etc. And which version of each?).
I see all this randy thing as sheer blackmail, and if we
keep believing that their "standards" are "necessary", they
could even get away with it. It's absurd.
> If it walks like a duck and it quacks like a duck, as far
> as I'm concerned it's a duck. The question is, is it a
> good duck or a bad one?
It must be Scrooge McDuck.
--
Juanra || http://html.conclase.net/
Arjun Ray
| Arjun Ray <ar...@nmds.com.invalid> writes:
|> Simon Brooke <si...@jasmine.org.uk> wrote:
|> Which "web constituency" other than the W3C has a problem with the
|> IETF?
|
| Boot's on the other foot. From conversations I have had with them, I
| think key people in the IETF would be very wary of doing anything
| which could be seen as 'stepping on W3C turf'.
Depends. It has been a mixed bag anyway. The IETF still has custody
of HTTP and a few other things that could be characterized as "W3C
turf". The spectacular disaster, of course, was HTML (RFC 1866), from
which the IETF got nothing but lots of egg in face. Then there's new
stuff - e.g. topic maps. XTM uses XML and XLink: does that make it
"W3C turf"? I'd say the OASIS folks have a different opinion;-)
| Some people got pretty badly burned in the flamewars which led to
| the founding of W3C, and don't want to go through that again.
I thought the flamewars started after the W3C was founded. I'm
probably missing the reference.
| Personally, I still see the IETF as the most legitimate available
| body,
Agreed.
| but real politics is I don't think they want to play.
Again, depends. Stay away from "languages", keep it to protocols, and
the IETF may show interest, I think. The goldmine yet to be tapped is
the DNS infrastructure.
|>| It seems to me that it is less difficult to reform W3C than to
|>| replace it.
|>
|> Read "to make it what we want of it rather than what it is".
|
| Fine, that's a reasonable summation.
Ain't gonna happen. People pay $5000/yr, and companies $50000/yr, to
keep it the way it is.
Arjun Ray
| My perspective (which I admit is based on limited experience) on
| standards organizations of the type that you mean -- that approve
| existing ideas as standards, rather than creating new ideas -- is
| that if we relied on them for our standards, the only standard HTML
| today would be IETF HTML, i.e. HTML 2.0 -- and no one would use it.
Actually, in the case of HTML, the IETF goofed. Had they been more
forceful (and, unfortunately, politically more blunt) in insisting on
evidence of genuine consensus, RFC 1866 would never have happened,
because the WG would have been obliged to dissolve. At least now the
IETF has taken ("official") HTML off the Standards Track.
The reason why no one would use HTML 2.0 is that no one ever used it.
The IETF would have happily endorsed a spec for Tag Soup, if it came
to that, because such a spec would have had the virtue of being honest
to the facts.
Arjun Ray
| I see all this randy thing as sheer blackmail, and if we
| keep believing that their "standards" are "necessary", they
| could even get away with it. It's absurd.
No, it's merely the piper being paid. Wishful thinking says that
never happens, which is why, of course, it's wishful thinking.
Woolly Mittens
> never happens, which is why, of course, it's wishful thinking.
>
You can't charge people for what they've been using free for years.
No matter how many lawyers you send after them... it's regarded as
extortion, no matter how legal.
Simon Brooke
> Perhaps you were talking about de facto standards, but then
> could you please specify which de facto standard you were
> talking about? (HTML as in IE, NS, Opera...? CSS as in MS,
> NS, Opera...? Etc. And which version of each?).
>
> I see all this randy thing as sheer blackmail, and if we
> keep believing that their "standards" are "necessary", they
> could even get away with it. It's absurd.
OK, but if no-one is there to develop and set standards, whose variant
'HTML' do we code to? Or do we create mutiple versions of every site
for every different proprietary browser that might visit it?
The fact is that if we don't have a standards body of some kind, then
we have either Microsoft, or a non-interoperable Web, and both those
outcomes seem to me substantially worse.
Andrew Glasgow
Arjun Ray <ar...@nmds.com.invalid> wrote:
So why on earth would that be a good thing?
In your insistence that "tag soup" is "real" HTML because it's what
everyone uses, you're ignoring that "standard" HTML is substantially
better than tag soup -- and with your scenario we would never have had
it. At least with the W3C we have an alterative to tag soup and
proprietary crapola.
Alan J. Flavell
[Arjun...]
> > The IETF would have happily endorsed a spec for Tag Soup, if it came
> > to that, because such a spec would have had the virtue of being honest
> > to the facts.
>
> So why on earth would that be a good thing?
Are you reading a different thread from me, or what?
> In your insistence that "tag soup" is "real" HTML because it's what
> everyone uses, you're ignoring that "standard" HTML is substantially
> better than tag soup
"The more I learn about reality, the less I believe it possible" -me
> -- and with your scenario we would never have had
> it.
_Whose_ scenario? Can't you recognise a temporary "devil's advocate"
when you see one?
I'm sure Arjun was merely documenting what he had observed. I really
don't believe he was expressing his ideal solution!
--
realism: HTML 4.01 Strict
evangelism: XHTML 1.0 Strict
madness: XHTML 1.1 as application/xhtml+xml
- Bjoern Hoehrmann
Arjun Ray
| In article <j8gnstorvjt4srgoa...@4ax.com>,
| Arjun Ray <ar...@nmds.com.invalid> wrote:
| > Andrew Glasgow <amg39.RE...@cornell.edu.INVALID> wrote:
|> The reason why no one would use HTML 2.0 is that no one ever
|> used it.
|>
|> The IETF would have happily endorsed a spec for Tag Soup, if it
|> came to that, because such a spec would have had the virtue of
|> being honest to the facts.
|
| So why on earth would that be a good thing?
It's important to distinguish facts and advocacy, and to remember that
specifications are inherently also historical records. Documents that
whitewash, distort or falsify history are poor material for standards.
A Tag Soup spec delivered in late '94 or early '95 would have been:
able to show a basis in prior implementation and testing; clear, short
and understandable in the documentation; open and fair; and timely.
It probably would not have met the criterion of technical excellence,
but that would have impacted only the eligibility for the Standards
Track. BFD.
| In your insistence that "tag soup" is "real" HTML because it's what
| everyone uses, you're ignoring that "standard" HTML is substantially
| better than tag soup -- and with your scenario we would never have
| had it.
I try to keep facts and advocacy separate; but as it happens, I don't
view "standard" HTML as substantially better. I can't imagine a worse
advertisement of Generalized Markup than the caricature of SGML that
"official HTML" has always been.
| At least with the W3C we have an alterative [...]
We don't need the W3C for alternatives; with the W3C in the way, we
may never have them.
| to tag soup and proprietary crapola.
Tag Soup need not be proprietary crapola, nor is the W3C a guard
against that. People blinded by agendas make the best cats' paws.
Simon Brooke
> Andrew Glasgow <amg39.RE...@cornell.edu.INVALID> wrote:
> | At least with the W3C we have an alterative [...]
>
> We don't need the W3C for alternatives; with the W3C in the way, we
> may never have them.
Is what you are saying that you think the best outcome is for IETF to
resume responsibility for those areas which W3C is currently claiming?
Andrew Glasgow
Arjun Ray <ar...@nmds.com.invalid> wrote:
> Andrew Glasgow <amg39.RE...@cornell.edu.INVALID> wrote:
> | In article <j8gnstorvjt4srgoa...@4ax.com>,
> | Arjun Ray <ar...@nmds.com.invalid> wrote:
> | > Andrew Glasgow <amg39.RE...@cornell.edu.INVALID> wrote:
>
> |> The reason why no one would use HTML 2.0 is that no one ever
> |> used it.
> |>
> |> The IETF would have happily endorsed a spec for Tag Soup, if it
> |> came to that, because such a spec would have had the virtue of
> |> being honest to the facts.
> |
> | So why on earth would that be a good thing?
>
> It's important to distinguish facts and advocacy, and to remember that
> specifications are inherently also historical records. Documents that
> whitewash, distort or falsify history are poor material for standards.
>
> A Tag Soup spec delivered in late '94 or early '95 would have been:
> able to show a basis in prior implementation and testing; clear, short
> and understandable in the documentation; open and fair; and timely.
And would suck ass in all the ways that tag soup did, does, and will.
> It probably would not have met the criterion of technical excellence,
> but that would have impacted only the eligibility for the Standards
> Track. BFD.
I dunno about you, but I'd rather have a
> | In your insistence that "tag soup" is "real" HTML because it's what
> | everyone uses, you're ignoring that "standard" HTML is substantially
> | better than tag soup -- and with your scenario we would never have
> | had it.
>
> I try to keep facts and advocacy separate; but as it happens, I don't
> view "standard" HTML as substantially better. I can't imagine a worse
> advertisement of Generalized Markup than the caricature of SGML that
> "official HTML" has always been.
To hell with SGML. I don't care whether HTML is a "caricature" of
anything, as long as it works as markup for the WWW. SGML is useless as
a markup language for the WWW unless you run it through a translator to
generate HTML (or some other web-relevant markup language).
> | At least with the W3C we have an alterative [...]
>
> We don't need the W3C for alternatives; with the W3C in the way, we
> may never have them.
The W3C isn't in the way. It's Microsoft who mainly is the source of
resistance to change in open-standard internet markup (while of course
changing their own crap at the drop of a
> | to tag soup and proprietary crapola.
>
> Tag Soup need not be proprietary crapola, nor is the W3C a guard
> against that. People blinded by agendas make the best cats' paws.
Whether or not tag soup is composed of proprietary crap or standard
tags, it's a bad thing. Even proprietary things are fine as long as
they don't suck (not that that ever happens) and don't exclude
non-proprietary things (not that that ever happens either).
Andrew Glasgow
<Pine.LNX.4.30.011016...@lxplus023.cern.ch>,
"Alan J. Flavell" <fla...@mail.cern.ch> wrote:
> On Oct 16, Andrew Glasgow inscribed on the eternal scroll:
>
> [Arjun...]
> > > The IETF would have happily endorsed a spec for Tag Soup, if it came
> > > to that, because such a spec would have had the virtue of being honest
> > > to the facts.
> >
> > So why on earth would that be a good thing?
>
> Are you reading a different thread from me, or what?
>
> > In your insistence that "tag soup" is "real" HTML because it's what
> > everyone uses, you're ignoring that "standard" HTML is substantially
> > better than tag soup
>
> "The more I learn about reality, the less I believe it possible" -me
>
> > -- and with your scenario we would never have had
> > it.
>
> _Whose_ scenario? Can't you recognise a temporary "devil's advocate"
> when you see one?
>
> I'm sure Arjun was merely documenting what he had observed. I really
> don't believe he was expressing his ideal solution!
I think his ideal scenario would be for everyone to stop using HTML and
use SGML instead.
Jan Roland Eriksson
wrote:
[...]
>We don't need the W3C for alternatives; with the W3C in the way, we
>may never have them.
That probably sums up the whole situation...
My take is a "clear distress" of seeing just about every type of
document processing that can be moved to client side, actually being
moved there.
Every damned "publisher" expects me to have all of his predefined
"gadgets" available, if not? it's me who are stupid, not him.
--
Rex [who at this point may be "colored" from having been
through an internal "fight" with my employer's so
called "IT" section]
Jan Roland Eriksson
wrote:
>Arjun Ray <ar...@nmds.com.invalid> writes:
>> We don't need the W3C for alternatives; with the W3C in the way, we
>> may never have them.
>Is what you are saying that you think the best outcome is for IETF to
>resume responsibility for those areas which W3C is currently claiming?
Some kind of open organization, IETF may be one possibility, there are
other possibilities of course.
The main thing is, as Arjun said earlier, that a _standard_ is the
result of a decision made by some body that do not have any other
stake in it then just to acknowledge a consensus when they can see
that there is one that can be acknowledged.
Please remember that Craig Mundies words at the university of NY did
not appear out of his own brain only, they where sanctioned in house
MS from some board room decisions. All official MS activities are
sanctioned from board room meetings, even the default install setup of
MSOE in fact.
A few other "big players" inside W3C do exactly the same of course,
and since they are what they are on the platform they currently
possess, go figure...
--
Rex
Jan Roland Eriksson
<amg39.RE...@cornell.edu.INVALID> wrote:
>In article
><Pine.LNX.4.30.011016...@lxplus023.cern.ch>,
> "Alan J. Flavell" <fla...@mail.cern.ch> wrote:
>> "The more I learn about reality, the less I believe it possible" -me
[...]
>> I'm sure Arjun was merely documenting what he had observed. I really
>> don't believe he was expressing his ideal solution!
>I think his ideal scenario would be for everyone to stop using HTML and
>use SGML instead.
Then you have understood less than I have, and I'm still an amateur...
...what does that make out of the situation?
--
Rex
Arjun Ray
| Arjun Ray <ar...@nmds.com.invalid> writes:
|> We don't need the W3C for alternatives; with the W3C in the way,
|> we may never have them.
|
| Is what you are saying that you think the best outcome is for IETF
| to resume responsibility for those areas which W3C is currently
| claiming?
Basically, yes. But not without qualification.
Much of W3C "activity" is sheer busy work: hothouse "development",
originating not in some genuine community seeking to formalize a
consensus, but in the minds of the pope and his staff of cardinals.
They have anointed themselves "architects" and "coordinators": verily
they say "let this be so", and verily it is so, whereupon an Activity
is proclaimed, a Working Group convened, and deliverables enjoined.
The IETF shouldn't take over. It should merely stand ready - as it
always has - to accept genuine proposals, which are likely to be only
a very small fraction of all that frenzied "activity". The rest is
just the typical blather of an industry consortium.
Arjun Ray
| In article <t12pstcedbttqh1d6...@4ax.com>,
| Arjun Ray <ar...@nmds.com.invalid> wrote:
|> Andrew Glasgow <amg39.RE...@cornell.edu.INVALID> wrote:
|>| In article <j8gnstorvjt4srgoa...@4ax.com>,
|>| Arjun Ray <ar...@nmds.com.invalid> wrote:
|>|> Andrew Glasgow <amg39.RE...@cornell.edu.INVALID> wrote:
|> A Tag Soup spec delivered in late '94 or early '95 would have been:
|> able to show a basis in prior implementation and testing; clear, short
|> and understandable in the documentation; open and fair; and timely.
|
| And would suck ass in all the ways that tag soup did, does, and will.
Really? Would you care to explain why all sorts of people went ape
over Mosaic and Netscape? Look at item (e) here:
http://www.webhistory.org/www.lists/www-talk.1993q3/0668.html
[Yes, the "missing years" of the www-talk and www-html lists at
lists.w3.org can be found at www.webhistory.org .]
That was August 93. In January 94, there was RFC 1563 (note that
there were I-Ds before this date).
http://www.faqs.org/rfcs/rfc1563.html
Read item 7 in Appendix B. Now add inline images and simple links,
and you have what Netscape 0.9 effectively had in October 94 to no end
of hosannas and hallelujahs from all around.
Suck ass? Hindsight is easy, ain't it?
If you want to see what people might have been *missing*, see
http://www.viola.org/violaIntro.html
The Aug 94 date only reflects editorial changes of a document that was
available in Mar 94 when Viola 3.0 was released. Here is a directory
of screen dumps - note the earlier date:
ftp://ftp.cuhk.edu.hk/.1/packages/info-systems/www/viola/screenDumps/
Yep: collapsible lists, tables, multicolumn layouts, text flow around
images, scripted applets, you name it. But noooooo, Mosaic/Netscape -
which had *none* of these, months and months *later* - was the kewlest
thing since sliced bread, verily a "standard" for the herd stampeding
over their chosen bleeding edge.
Suck ass? Now, why did people justify MarcA's assertion by *choosing*
to suck ass?
History is embarassing, ain't it?
|> as it happens, I don't view "standard" HTML as substantially better.
|> I can't imagine a worse advertisement of Generalized Markup than the
|> caricature of SGML that "official HTML" has always been.
|
| To hell with SGML. I don't care whether HTML is a "caricature" of
| anything, as long as it works as markup for the WWW.
What works as markup for the WWW is Tag Soup.
|> Tag Soup need not be proprietary crapola, nor is the W3C a guard
|> against that. People blinded by agendas make the best cats' paws.
|
| Whether or not tag soup is composed of proprietary crap or standard
| tags, it's a bad thing.
Only in that it's a limited paradigm - a dead end, no future. But!
"Hard work pays off later. Laziness pays off immediately". Tag Soup
is *easy*. Not just for Joe HomePager, but much more importantly, for
the horde of instant expert wannabes.
Vanity, thy name is Web.
Arjun Ray
| If you want to see what people might have been *missing*, see
|
| http://www.viola.org/violaIntro.html
|
| The Aug 94 date only reflects editorial changes of a document that was
| available in Mar 94 when Viola 3.0 was released. Here is a directory
| of screen dumps - note the earlier date:
|
| ftp://ftp.cuhk.edu.hk/.1/packages/info-systems/www/viola/screenDumps/
|
| Yep: collapsible lists, tables, multicolumn layouts, text flow around
| images, scripted applets, you name it.
I didn't make the Mar 94 date up.
ftp://ftp.xcf.berkeley.edu/pub/ht/projects/viola/README
The file viola940323.tar.gz is bundled inside this:
ftp://ftp.cuhk.edu.hk/pub/www/unix/viola/viola.tar.gz
Technologically, there has been very little progress, if any at all,
in seven and a half years. You want to blame Tag Soup? Go ahead,
it's still alive and kicking.
Just keep the facts and advocacy separate. (It is advocacy to
"support" the W3C.)
Simon Brooke
> Andrew Glasgow <amg39.RE...@cornell.edu.INVALID> wrote:
> | In article <t12pstcedbttqh1d6...@4ax.com>,
> | Arjun Ray <ar...@nmds.com.invalid> wrote:
> |> as it happens, I don't view "standard" HTML as substantially better.
> |> I can't imagine a worse advertisement of Generalized Markup than the
> |> caricature of SGML that "official HTML" has always been.
> |
> | To hell with SGML. I don't care whether HTML is a "caricature" of
> | anything, as long as it works as markup for the WWW.
>
> What works as markup for the WWW is Tag Soup.
Now *that's* a charicature.
Tag Soup does not work well for anything, not even markup for the
Web. On the contrary, well structured markup works better for the Web
than tag soup, because the parsing of it does not depend on
unpredictable error recovery.
This is not to pretend, of course, that there is not an Augean Stables
out there.
> Vanity, thy name is Web.
Verily.
bit-b...@maney.org
So, if I am understanding you properly, you are suggesting that what
should happen is that the W3C (and any other consortium or individual)
should submit it's RECs to the IETF and then the IETF should evaluate
those RECs using it's normal methods towards producing IETF standards?
fpsm
--
| Fredrich P. Maney maney at maney dot org |
| Do NOT send me HTML formatted E-mail or copies of netnews posts! |
| Address in header is a spamtrap. Use one in signature for replies. |
| Please review http://www.maney.org/fred/site/uce/ before emailing. |
Arjun Ray
| So, if I am understanding you properly, you are suggesting that what
| should happen is that the W3C (and any other consortium or individual)
| should submit it's RECs to the IETF and then the IETF should evaluate
| those RECs using it's normal methods towards producing IETF standards?
Yes. I-Ds like any other. A document being a W3C Recommendation (or
any variant thereof) confers no special status whatsoever as far as
the IETF is concerned. No favorites. All the usual rules apply - of
which I'd consider open mailing lists[1] one of the most important.
[1] As the means to have *all* transactions out in the open, not only
for people to review, but also to track during the development - few
people can come to a spec cold and immediately grasp the implications:
you have to "live with it" for a while. This is one the reasons why
the W3C's policy of "public review" is bogus, btw.
Andrew Glasgow
I don't think that's what he's actually advocating here. He's made what
he is advocating clear in another post; the end of the W3C as a
"standards" body and continuing in its role as developer of new
recommendations, with the IETF taking on its original role as a
standards body.
My impression is that his ideal, pie-in-the-sky scenario would be for
HTML to die the gory death that (in his opinion) it richly deserves is
based on the rest of his posting history.
I'm not saying he's a troll or anything; his posts are always
informative and often interesting. But they are very rarely of any help
to anyone who wants to actually write HTML for the WWW. No offense
intended.
Arjun Ray
| My impression is that his ideal, pie-in-the-sky scenario would be for
| HTML to die the gory death that (in his opinion) it richly deserves
"Official" HTML, yes. It's simplistic in what it could cater to, and
hopelessly convoluted with all its grandfathering of adhockery. By
continuing to exist (as something the unwary or the innocent could be
inveigled into taking seriously), it occupies "mindshare" inertially
and thus inhibits alternatives.
Tag Soup, that is text/html, has its own dynamic.
| his posts are [...] very rarely of any help to anyone who wants to
| actually write HTML for the WWW. No offense intended.
None taken:-)
Arjun Ray
| Arjun Ray <ar...@nmds.com.invalid> writes:
| > Andrew Glasgow <amg39.RE...@cornell.edu.INVALID> wrote:
|>| I don't care whether HTML is a "caricature" of anything, as
|>| long as it works as markup for the WWW.
|>
|> What works as markup for the WWW is Tag Soup.
|
| Now *that's* a charicature.
Nope.
| Tag Soup does not work well for anything, not even markup for the
| Web.
Working well wasn't the issue, just working was.
| On the contrary, well structured markup works better for the Web
| than tag soup, because the parsing of it does not depend on
| unpredictable error recovery.
Neither does the parsing of Tag Soup need error recovery, nor does
"error recovery" in any form describe the behavior of common HTML
user-agents which, in the unlikely event you hadn't noticed, are tag
soup processors. The unpredictability stems from the fact that tag
soup "commands" or toggles aren't fully orthogonal. There is more
than a little context dependence that different implementations treat
in different ways.
Interestingly enough, "well-structured" markup isn't necessarily proof
against this. This is well structured:
<ul><dd><h1><dl><li>Blah</li></dl></h1></dd></ul>
but the result is ah, unpredictable. The point of *valid* markup (and
not merely "well-structured") is to enforce contextual constraints,
but you didn't say that, and the people who would care seem to be a
set of measure zero on the Web.
| This is not to pretend, of course, that there is not an Augean Stables
| out there.
That is a caricature.
Rijk van Geijtenbeek
> This is well structured:
>
> <ul><dd><h1><dl><li>Blah</li></dl></h1></dd></ul>
>
>but the result is ah, unpredictable.
Interesting experiment: try this in various browsers and see the nesting
the browser allows. How do the left- and top-margins add up, etc. Opera
and Mozilla close some elements (not as much as they should*), deciding
that the occurrence of some tag implies the closure of a previous
element.
<style type="text/css">
<!--
ul {border: thin dotted yellow;}
dd {border: thin dotted red;}
h1 {border: thin solid yellow;}
dl {border: thin solid red;}
li {border: thin dotted blue;}
-->
</style>
<ul><dd><h1><dl><li>Blah</li></dl></h1></dd></ul>
Of course, slapping a Doctype on top makes Mozilla render it differently
;-)
This tells us Tagsoup may be alive and well, but valid markup is
rendered more consistently the same across browsers.
* Opera tries to do the right thing w.r.t. contextual constraints,
unless it means breaking a lot of 'Netploder-enhanced' pages. So <b> is
always closed at the start of the next block-level element, but <font>
can contain block-level elements. Missing 'boldness' isn't a big deal,
but a missing font color can be disastrous...
Opera offers no choice between 'strict' and 'quirk' display based on
Doctypes, it is just trying to be usable with the code that's out there.
--
If you don't like having choices | Rijk van Geijtenbeek
made for you, you should start | Documentation & QA
making your own. - Neal Stephenson | mailto:ri...@opera.com
Arjun Ray
| Of course, slapping a Doctype on top makes Mozilla render it
| differently ;-)
Great. Now all we need is for a Mozilla-wonk to come around and tell
us that this would not be yet another example of tag soup processing.
("But but but but... Isn't the doctype a tag?!?!?")
| This tells us Tagsoup may be alive and well,
Is.
| but valid markup is rendered more consistently the same across
| browsers.
I'm not sure how this follows. We could expect it, inasmuch as
"official" HTML is a retrofit, but that wasn't the point. It is
possible for tag soup to be "well-structured", which doesn't make it
any less tag soup.
| * Opera tries to do the right thing w.r.t. contextual constraints,
| unless it means breaking a lot of 'Netploder-enhanced' pages.
I know the feeling...
| Opera offers no choice between 'strict' and 'quirk' display based
| on Doctypes, it is just trying to be usable with the code that's
| out there.
Sure. The problem is the comforting notion that the "code out there"
is causeless. It isn't.
Andrew Glasgow
Arjun Ray <ar...@nmds.com.invalid> wrote:
> Andrew Glasgow <amg39.RE...@cornell.edu.INVALID> wrote:
>
>
> | My impression is that his ideal, pie-in-the-sky scenario would be for
> | HTML to die the gory death that (in his opinion) it richly deserves
>
> "Official" HTML, yes. It's simplistic in what it could cater to, and
> hopelessly convoluted with all its grandfathering of adhockery. By
> continuing to exist (as something the unwary or the innocent could be
> inveigled into taking seriously), it occupies "mindshare" inertially
> and thus inhibits alternatives.
True, but what else are we going to use to author for the WWW? There's
no choice at the moment. Any substantially accepted "standard", whether
de facto or de jure, inhibits alternatives.
Arjun Ray
| In article <3hosstsajm8pa1tiq...@4ax.com>,
| Arjun Ray <ar...@nmds.com.invalid> wrote:
|> By continuing to exist (as something the unwary or the innocent
|> could be inveigled into taking seriously), ["Official" HTML]
|> occupies "mindshare" inertially and thus inhibits alternatives.
|
| True, but what else are we going to use to author for the WWW?
Think of HTML as an output-only format. Generate it, but don't use it
to store your real assets (your imformation.)
| Any substantially accepted "standard", whether de facto or de jure,
| inhibits alternatives.
Indeed, well said. Standards are never costless. By the same token,
we should be wary of accepting anything as a "standard" until the
costs of doing so are properly accounted for. There is no quick fix,
certainly not some frenized extrusion of the W3C.
--
Standardization as a marketing tool is evil. - Eliot Kimber
Isofarro
wrote:
>Andrew Glasgow <amg39.RE...@cornell.edu.INVALID> wrote:
>| In article <3hosstsajm8pa1tiq...@4ax.com>,
>| Arjun Ray <ar...@nmds.com.invalid> wrote:
>
>|> By continuing to exist (as something the unwary or the innocent
>|> could be inveigled into taking seriously), ["Official" HTML]
>|> occupies "mindshare" inertially and thus inhibits alternatives.
>|
>| True, but what else are we going to use to author for the WWW?
>
>Think of HTML as an output-only format. Generate it, but don't use it
>to store your real assets (your imformation.)
I'm in agreement with you there. HTML is a means to an end - the end
being communicating with the user using a browser.
I sometimes keep wondering where the translation between the real
assets and the HTML should happen.
Should it happen off-line, as in preprocessing,
or should it happen on the fly - thus generating depending on the
client's preferences.
At the moment, the client preferences are a handful of data formats:
HTML, XML, RSS, WAP, Plain Text - so preprocessing has its benefits,
especially on mainly static content - such as a FAQ.
Here the "duplication" of information shouldn't matter, as long as the
generator tools do its job correctly.
I'd guess if the output formats rises to a substantial number, then
the tendancy to offer the content dynamically would then be a better
solution. Although, dynamic delivery tools should have a very good
caching mechanism.
Iso.
--
AnyBrowser Campaign: http://www.anybrowser.org/campaign/
alt.html FAQ: http://www.html-faq.com & http://www.alt-html.org
alt.html QuickStart: http://rock13.com/webhelp/usenet/newbie.txt
Web Dev Wiki: http://www.allmyfaqs.com/
Nick Kew
at the keyboard of Isofarro <spam...@spamdetector.co.uk> wrote:
> I'd guess if the output formats rises to a substantial number, then
> the tendancy to offer the content dynamically would then be a better
> solution. Although, dynamic delivery tools should have a very good
> caching mechanism.
<spectacles appearance="rose-tinted">
Yep. On todays web, that's where XML is heading, with XML+XSLT=whatever
the client wants.
</>
I'm busy on a set of stylesheets for Site Valet, with the aim of offering
the choice of HTML, XML, EARL or plain text for every reporting function.
With Jim's help, we've already deployed one application that shifts
the whole of the presentational thing over to the Client.
--
Nick Kew
Site Valet - the essential service for anyone with a website.
<URL:http://valet.webthing.com/>