MS *CREATED* XML, and the W3C adopted it.
Daeron
- Erik Funkenbusch Aug 27 2004
- msg: <65gvvdn9...@funkenbusch.com>
------
er,, Erik - don't you mean they innovated it ;) Do you mind providing
some citations for the above statement and I don't mean quotes from
Ballmer. Sure they submitted some specs to the W3C working group. But so
did a lot of other people. How does that ever translate into *CREATED*.
Is this what you are refering to. "This document is a NOTE made
available by the World Wide Web Consortium for discussion only?"
http://www.w3.org/TR/NOTE-XSL-970910 Aug 27 1997
ps: Why is it the current version of Open Office supports XML as its
native file format and Word2000/6 doesn't ?
------
XML's Development start around 1996.
The motivation was in part due to the success of HTML and in part due to
the crushing need to simplify SGML.
Jon Bosak from Sun Microsystems was the prime motivator for the creation
of XML and inspired by Yuri Rubinsky's[1] "SGML on the Web"
http://www.sims.berkeley.edu/academics/courses/is290-8/s04/lectures/1/slide2.html
So in 1996, discussions began which focused on how to define a markup
language with the power and extensibility of SGML but with the
simplicity of HTML. The World Wide Web Consortium (W3C) decided to
sponsor a group of SGML gurus including Jon Bosak from Sun.
http://wdvl.internet.com/Authoring/Languages/XML/Tutorials/Intro/history.html
KLH
Erik Funkenbusch
> "Funny, considering that MS *CREATED* XML, and the W3C adopted it."
> - Erik Funkenbusch Aug 27 2004
> - msg: <65gvvdn9...@funkenbusch.com>
> ------
>
> er,, Erik - don't you mean they innovated it ;) Do you mind providing
> some citations for the above statement and I don't mean quotes from
> Ballmer. Sure they submitted some specs to the W3C working group. But so
> did a lot of other people. How does that ever translate into *CREATED*.
Jean Paoli, a MS employee, is considered the Co-creator of XML, along with
Jon Bosak from Sun. They did the majority of work on it.
> Is this what you are refering to. "This document is a NOTE made
> available by the World Wide Web Consortium for discussion only?"
> http://www.w3.org/TR/NOTE-XSL-970910 Aug 27 1997
No, that's XSL, which as you note, MS employees consisted of nearly half
the members. This was the initial work on XSL which MS and Inso submitted
to the W3C.
> ps: Why is it the current version of Open Office supports XML as its
> native file format and Word2000/6 doesn't ?
Apart from the fact that XML didn't even exist when Word 6 was created?
Apart from the fact that XML was hardly in use when Word 2000 was created
(1997 and shipped in 1999)?
> ------
>
> XML's Development start around 1996.
Word 6 was released in 1994. Word 2000 began development in 97 and shipped
in 1999. XML took a long time to become something in common use.
> So in 1996, discussions began which focused on how to define a markup
> language with the power and extensibility of SGML but with the
> simplicity of HTML. The World Wide Web Consortium (W3C) decided to
> sponsor a group of SGML gurus including Jon Bosak from Sun.
Also including Jean Paoli, of Microsoft.
KLH
http://www.xmlconference.org/xmlusa/2002/plenaries_paoli.asp
http://www.atnewyork.com/news/article.php/2193151
http://www.wininsider.com/news/?4647
Seems that Jean Paoli started work at Microsoft in 1996 but had been
involved with XML (SGML) since 1985. How does this translate into
Microsoft inventing XML? Care to elaborate? Seems that he was involved
in the standards creation for XML 1.0, not inventing it.
The technology involved seems to have started out at IBM in the later
part of the 1960s. Did Microsoft exist in 1969? I do not see any
reference to Jean Paoli at the site below.
Daeron
> Erik Funkenbusch wrote:
>> On Sat, 28 Aug 2004 19:35:40 +0100, Daeron wrote:
<snip>
>>> ps: Why is it the current version of Open Office supports XML as
>>> its native file format and Word2000/6 doesn't ?
>> Apart from the fact that XML didn't even exist when Word 6 was
>> created? Apart from the fact that XML was hardly in use when Word
>> 2000 was created (1997 and shipped in 1999)?
You know exactly what I meant: Word 6, Word 95, Word 96, Word
98/2000/XP/2003 or the yet to be released Word 2006. All equally
incompatible with the previous version. All to prevent anyone else
getting into *their* word processing market. All causing multiple
compatible problems for the poor schmuck of a consumer.
The XML file format has been the around since April 2001 This is now
2004 and Word2000 and four, five or six still hasn't got there. How come
? Is because then they couldn't play file format shuffle with the other
software developers ?
http://www.conectiva.com.br/cpub/pt/incConectiva/cases/cases2/005,030,62,188,917.html
http://snipurl.com/8qfn
Erik Funkenbusch
> KLH wrote:
>
>> Erik Funkenbusch wrote:
>
>>> On Sat, 28 Aug 2004 19:35:40 +0100, Daeron wrote:
>
> <snip>
>
>>>> ps: Why is it the current version of Open Office supports XML as
>>>> its native file format and Word2000/6 doesn't ?
>
>>> Apart from the fact that XML didn't even exist when Word 6 was
>>> created? Apart from the fact that XML was hardly in use when Word
>>> 2000 was created (1997 and shipped in 1999)?
>
> You know exactly what I meant: Word 6, Word 95, Word 96, Word
> 98/2000/XP/2003 or the yet to be released Word 2006. All equally
> incompatible with the previous version. All to prevent anyone else
> getting into *their* word processing market. All causing multiple
> compatible problems for the poor schmuck of a consumer.
Not completely true. For example, Word 6 and Word 95 had the same file
format, no changes. Word 97 and Word 2000 had the same format, no changes.
Excel has had the same compatibile format since Excel 5.
> The XML file format has been the around since April 2001 This is now
> 2004 and Word2000 and four, five or six still hasn't got there. How come
> ? Is because then they couldn't play file format shuffle with the other
> software developers ?
Word 2003 has full suport of XML file formats. Word XP had partial XML
support. Why do you insist that it doesn't?
iAn
> Not completely true. For example, Word 6 and Word 95 had the same file
> format, no changes. Word 97 and Word 2000 had the same format, no
> changes. Excel has had the same compatibile format since Excel 5.
Can you open a Word 2000 file with Word 97?
--
A fatal exception 0E has occurred at 0028:C000BD1D in VXD VMM(01) +
0000AD1D. The current application will be terminated.
Linønut
> Erik Funkenbusch wrote:
>
>
>> Not completely true. For example, Word 6 and Word 95 had the same file
>> format, no changes. Word 97 and Word 2000 had the same format, no
>> changes. Excel has had the same compatibile format since Excel 5.
>
> Can you open a Word 2000 file with Word 97?
Or a Word 97 document with Word 95?
Look, Donald Knuth had his TeX "word processor" fully debugged years ago.
Microsoft's "word processor" is still a buggy, frustrating tool.
--
[X] Check here to always trust content from Linųnut
Baruch
>
> Look, Donald Knuth had his TeX "word processor" fully debugged years ago.
> Microsoft's "word processor" is still a buggy, frustrating tool.
>
I doubt whether any non-trivial piece of software can ever be known to be
"fully debugged". About all we can say is that certain programs haven't
shown bugs, *yet*. Some programs take years to show up a bug; we just
can't tell.
Word is not particularly "buggy", in my opinion. I've used it at work for
about ten years, in its several incarnations. The frustrating part isn't
in using it; it's in learning the new quirks. For some reason, Microsoft
keeps changing the UI so that various features are moved around,
accelerators are redefined, etc. But it is a decent, if bloated, word
processor.
Linønut
> Linųnut wrote:
>
>> Look, Donald Knuth had his TeX "word processor" fully debugged years ago.
>> Microsoft's "word processor" is still a buggy, frustrating tool.
>>
> I doubt whether any non-trivial piece of software can ever be known to be
> "fully debugged". About all we can say is that certain programs haven't
> shown bugs, *yet*. Some programs take years to show up a bug; we just
> can't tell.
For years, Knuth has offered a bounty on bugs in TeX. I hear he hasn't paid
out lately.
> Word is not particularly "buggy", in my opinion. I've used it at work for
> about ten years, in its several incarnations. The frustrating part isn't
> in using it; it's in learning the new quirks. For some reason, Microsoft
> keeps changing the UI so that various features are moved around,
> accelerators are redefined, etc. But it is a decent, if bloated, word
> processor.
When it works. Every so often, it will crap out and corrupt a document. More
often, it will reapply undesired formatting (numbered lists seem particularly
susceptible).
It also has a number of annoying quirks that MS would say are features, but
they reduce the usability of Word.
Word is overly complex and buggy. If you don't think so, you don't use it
enough.
Baruch
>
> For years, Knuth has offered a bounty on bugs in TeX. I hear he hasn't
> paid out lately.
>
That only shows that no one has uncovered a bug, not that none exist. One
reason no one has collected on the bounty may be that there *are* no bugs;
but there's no way to prove that. That's all I'm saying - it's not
possible to prove that software is free of bugs.
>> Word is not particularly "buggy", in my opinion. I've used it at work
>
> When it works. Every so often, it will crap out and corrupt a document.
I have not had this happen to me in over ten years. It *did* happen to me
right when I was learning the program. It appears to happen to people who
try to save documents to floppies instead of the HD, for some reason.
> More often, it will reapply undesired formatting (numbered lists seem
> particularly susceptible).
>
No doubt; it can be cranky, but all you really need to do is make changes to
your normal.dot file (and hope it doesn't get too bloated, at which time it
will explode, taking all your macros and formatting with it).
> It also has a number of annoying quirks that MS would say are features,
> but they reduce the usability of Word.
>
> Word is overly complex and buggy. If you don't think so, you don't use it
> enough.
>
This is a nonsensical statement. It can be countered simply by my claim:
"Word is OK. If you don't think so, then you don't use it enough." One
could argue that if you're having problems with any software, then you
haven't used it enough. However, such statements are meaningless.
Robert A. Heinlein used to say, "If you pray hard enough, you can make water
run uphill. How hard do you have to pray? Why, hard enough to make water
run uphill". His point is that you can make any sort of claim that if you
X enough, then Y. If Y doesn't ensue, then of course you haven't X'ed
enough.
In this case, your statement is also false.
I use Word to earn my livelihood, and I have done so for about six years
now. By this I mean that my job entails me typing stuff in Word, almost
exclusively. Prior to that, I used Word in my job, but it was only part of
what I did. In other words, I spend eight hours a day, forty hours a week,
about 2000 hours a year, using Word now, and have done so for about 10,000
hours worth of learning that program and using it. Possibly even 12,000,
but who's counting? Even allowing for breaks, goofing off, surfing the net
on my boss's time, etc., I have put in several thousand hours of time using
the program for my livelihood.
Additionally, until I got a Linux box running, I used Word exclusively at
home for any word processing tasks, including keeping a diary and writing
various letters. This added hundreds of hours more to my use of Word, as
well as allowed me to experiment with many of the features that are not
commonly used or needed in an office setting.
Yes, Word has bugs in it. So does the Linux kernel. There are a large
number of misfeatures and undocumented features in Word. It tries to do
too much - to be a spreadsheet, HTML editor, database, presentation
manager, equation editor, and so on. It does none of these tasks well.
But it's a good word processor.
Of course, one could argue that it *did* take me a very long time to learn
its quirks, and because of that, it's not a good program. In other words,
it's so bad that only someone who spends thousands of hours learning it
could even make it work without it crashing. I might agree with that line
of reasoning - in fact, I'd say it's close to the truth. Software should
work without having to spend that long figuring out its idiosyncrasies.
Linønut
> Linųnut wrote:
>
>> For years, Knuth has offered a bounty on bugs in TeX. I hear he hasn't
>> paid out lately.
>>
> That only shows that no one has uncovered a bug, not that none exist. One
> reason no one has collected on the bounty may be that there *are* no bugs;
> but there's no way to prove that. That's all I'm saying - it's not
> possible to prove that software is free of bugs.
Knuth believes that TeX is free of bugs. Sure, there may be a few lurking, but
when you search hard and don't find any more, there probably aren't any.
Contrast that with Microsoft products (and, indeed, some OSS items), where you
can't work for more than a few hours without finding another bug.
>> When it works. Every so often, it will crap out and corrupt a document.
>
> I have not had this happen to me in over ten years. It *did* happen to me
> right when I was learning the program. It appears to happen to people who
> try to save documents to floppies instead of the HD, for some reason.
I'm chuckling here. Right now, we're using Word XP to create 200- to 500-page
documents, 12 Mb worth of text with embedded pictures and Visio diagrams.
It is extremely irritating to modify a style so that the hanging indents line
up and then, when you add some new numbered items in a new section of numbered
items, all numbered items throughout the whole document go back to having the
same badly-hanging indent. And that indent shouldn't even be part of the style
anymore.
At least Word XP is less buggy than Word 2000 (a nadir in Word stabily, I
feel). That version of Word would frequently corrupt a large document, so that
Word itself would often (but not always) hang when you opened it. If you could
open it without hanging, that was your chance to cut the document into pieces,
save the piece, and then reassemble them later.
Hardly what I would call a "productivity" app.
>> Word is overly complex and buggy. If you don't think so, you don't use it
>> enough.
>>
> This is a nonsensical statement. It can be countered simply by my claim:
> "Word is OK. If you don't think so, then you don't use it enough." One
> could argue that if you're having problems with any software, then you
> haven't used it enough. However, such statements are meaningless.
Not meaningless. I'm saying that Word may be fine for small documents, but it
is too buggy to use for large documents.
> Robert A. Heinlein used to say, "If you pray hard enough, you can make water
> run uphill. How hard do you have to pray? Why, hard enough to make water
> run uphill". His point is that you can make any sort of claim that if you
> X enough, then Y. If Y doesn't ensue, then of course you haven't X'ed
> enough.
What does that have to do with my story? I'm telling you about water that I
saw run uphill.
> In this case, your statement is also false.
o. It is me relating my experiences with Word.
>
> I use Word to earn my livelihood, and I have done so for about six years
> now. By this I mean that my job entails me typing stuff in Word, almost
> exclusively. Prior to that, I used Word in my job, but it was only part of
> what I did. In other words, I spend eight hours a day, forty hours a week,
> about 2000 hours a year, using Word now, and have done so for about 10,000
> hours worth of learning that program and using it. Possibly even 12,000,
> but who's counting? Even allowing for breaks, goofing off, surfing the net
> on my boss's time, etc., I have put in several thousand hours of time using
> the program for my livelihood.
So what? How long are your documents? What range of formatting do they use?
Do you embed objects in your documents? Do they have tables of contents and
figures? Do they "include" other documents. Do they employ sections and
header/footers?
> Additionally, until I got a Linux box running, I used Word exclusively at
> home for any word processing tasks, including keeping a diary and writing
> various letters. This added hundreds of hours more to my use of Word, as
> well as allowed me to experiment with many of the features that are not
> commonly used or needed in an office setting.
>
> But it's a good word processor.
It's not even a good word processor, unless you can get all your readers to
commit to using Word.
> Of course, one could argue that it *did* take me a very long time to learn
> its quirks, and because of that, it's not a good program. In other words,
> it's so bad that only someone who spends thousands of hours learning it
> could even make it work without it crashing. I might agree with that line
> of reasoning - in fact, I'd say it's close to the truth. Software should
> work without having to spend that long figuring out its idiosyncrasies.
Yeah, but you just said it is a good word processor. Make up your mind.
Erik Funkenbusch
> Not meaningless. I'm saying that Word may be fine for small documents, but it
> is too buggy to use for large documents.
I would agree. Word is not really designed for huge documents, though MS
has made some effort to address those problems. In fact, I question
whether *any* WYSIWYG word processor could be designed to handle such
documents without severely restricting the capabilities.
Large documents take extreme amounts of memory, and applying all your
styles and formatting to the entire document when you make changes is going
to be frustrating to say the least.
Also, the larger the document, the greater the chance that you're going to
get into some untested (or less tested) permutation of control paths.
The Ghost In The Machine
<linøn...@bone.com>
wrote
on Mon, 30 Aug 2004 11:12:28 -0500
<fo2dnS4Wz9J...@comcast.com>:
But it's a *profitable*, buggy, frustrating tool. (For Microsoft,
certainly.)
Which makes me wonder where the businessman's prorities are, at times.
ObQuack: Quork.
--
#191, ewi...@earthlink.net
It's still legal to go .sigless.
Baruch
>
> Knuth believes that TeX is free of bugs. Sure, there may be a few
> lurking, but when you search hard and don't find any more, there probably
> aren't any. Contrast that with Microsoft products (and, indeed, some OSS
> items), where you can't work for more than a few hours without finding
> another bug.
>
program, either. That's one of those things you just never know about
until one bites you. As far as Microsoft - come on! They have many
thousands of bugs, including many they know about and have not yet bothered
to patch. Compared to them, TeX is absolutely perfect, and was since v
0.0.01 or so...
>> right when I was learning the program. It appears to happen to people
>> who try to save documents to floppies instead of the HD, for some reason.
>
> I'm chuckling here. Right now, we're using Word XP to create 200- to
> 500-page documents, 12 Mb worth of text with embedded pictures and Visio
> diagrams.
>
Ah, you're doing *that* stuff, are you? OK, I retract my previous statement
- using Word like that is just begging for corrupted files. I had really
been thinking of how I normally use Word - as a word processor. Not as -
what, desktop publishing? I don't even know what you're trying to do...
But definitely, for your sort of document, Word is horribly inappropriate.
> It is extremely irritating to modify a style so that the hanging indents
> line up and then, when you add some new numbered items in a new section of
> numbered items, all numbered items throughout the whole document go back
> to having the
> same badly-hanging indent. And that indent shouldn't even be part of the
> style anymore.
>
It's a *feature*, don't you understand? *FEATURE*...
> At least Word XP is less buggy than Word 2000 (a nadir in Word stabily, I
> feel). That version of Word would frequently corrupt a large document, so
> that
I forgot about that... it's all coming back to me now. I must have blocked
it out of my mind...
Word 2000 was unbelievable. I am in awe that such a product could be
inflicted on an unsuspecting public. It was so bad, I reverted to Word 97,
rather than struggle with 2000. Yup, Word is buggy, all right.
> Word itself would often (but not always) hang when you opened it. If you
> could open it without hanging, that was your chance to cut the document
> into pieces, save the piece, and then reassemble them later.
>
> Hardly what I would call a "productivity" app.
>
Yeah, I agree... like I said, I was thinking lots smaller docs. Something
like you're working on is a recipe for disaster. No wonder you're not
happy with Word...
>> haven't used it enough. However, such statements are meaningless.
>
> Not meaningless. I'm saying that Word may be fine for small documents,
> but it is too buggy to use for large documents.
>
OK - first, I *do* agree with you, as far as Word being buggy - especially
for larger documents. However, IMNSHO, statements like "if you don't agree
you haven't used it enough" are still meaningless, since it's just a way to
invalidate one who disagrees with you, without providing some basis for the
claim.
>
> What does that have to do with my story? I'm telling you about water that
> I saw run uphill.
>
No, you're telling me that if I haven't seen it, I wasn't looking for it.
>> In this case, your statement is also false.
>
> o. It is me relating my experiences with Word.
No, you were telling me that if I disagree with you, I don't know what I'm
talking about. *That* statement is false. I wasn't questioning your
experience with Word. Your experiences are valid, and I am in no position
to question them.
What I was questioning - and saying is false - was your claim that if I
disagree, it is due to some deficiency in my knowledge or experience.
The reason I disagreed with you was not a result of my ignorance or
inexperience with Word. It had to do with using it for different purposes
than you do. Now that I understand what you're doing with it, I totally
agree with you - using Word for that is like kicking a dead whale down the
beach.
>>
>> I use Word to earn my livelihood, and I have done so for about six years
>> now. By this I mean that my job entails me typing stuff in Word, almost
>> exclusively. Prior to that, I used Word in my job, but it was only part
>> of
>> what I did. In other words, I spend eight hours a day, forty hours a
>> week, about 2000 hours a year, using Word now, and have done so for about
>> 10,000
>> hours worth of learning that program and using it. Possibly even 12,000,
>> but who's counting? Even allowing for breaks, goofing off, surfing the
>> net on my boss's time, etc., I have put in several thousand hours of time
>> using the program for my livelihood.
>
> So what? How long are your documents? What range of formatting do they
> use?
> Do you embed objects in your documents? Do they have tables of contents
> and
> figures? Do they "include" other documents. Do they employ sections and
> header/footers?
>
All of the above, except for the size of the document. In my experience,
Word blows up when the document gets bigger, even if you don't have a lot
of embedded objects or tables in the doc.
>>
>> But it's a good word processor.
>
> It's not even a good word processor, unless you can get all your readers
> to commit to using Word.
>
Wait - I don't understand this sentence here... For me, the end product of
my using Word is a physical document, something printed out on paper and
sent through the mail (except for the occasional e-mailed item). Do you
mean to say here that you're using it for online use, where others will be
reading your product using a computer? In that case, then yes, you'd have
to get everyone on board, because a Word document just isn't going to be
portable. In that case, it would be much better to use something else.
>> Of course, one could argue that it *did* take me a very long time to
>> learn
>> its quirks, and because of that, it's not a good program. In other
>> words, it's so bad that only someone who spends thousands of hours
>> learning it
>> could even make it work without it crashing. I might agree with that
>> line
>> of reasoning - in fact, I'd say it's close to the truth. Software should
>> work without having to spend that long figuring out its idiosyncrasies.
>
> Yeah, but you just said it is a good word processor. Make up your mind.
>
Right. I said it is a good word processor. Here I was saying how someone
could argue otherwise - not that I accept that argument, but I recognize
it. That's all.
Let me clarify: Word is an OK word processor, for relatively small
documents that don't have too much in the way of embedded objects or
subdocuments. Once you get over a few hundred pages, or start getting
fancy with tables, images, subdocuments, etc., then Word deteriorates and
becomes unusable.
Jim Richardson
Hash: SHA1
On Tue, 31 Aug 2004 17:54:27 -0500,
Erik Funkenbusch <er...@despam-funkenbusch.com> wrote:
Thank heavens for LaTeX.
Having just typeset a 120,000 word manuscript, I am *so* glad I didn't
have to do it in word or some other wysiwy(might)g "word processor"
LaTeX rules, all praise Knuth.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.5 (GNU/Linux)
iD8DBQFBNZBQd90bcYOAWPYRAnhbAJsHVEXhKwQXDl+aBA3K1Cf+YemXNgCfdZge
y9lVg+lt4FSw4RxDfO2rn8k=
=aH0O
-----END PGP SIGNATURE-----
--
Jim Richardson http://www.eskimo.com/~warlock
Never be in the company of anyone with whom you would not want to die.
-- Fremen Saying
LEE Sau Dan
Baruch> Linønut wrote:
>> For years, Knuth has offered a bounty on bugs in TeX. I hear
>> he hasn't paid out lately.
>>
Baruch> That only shows that no one has uncovered a bug, not that
Baruch> none exist.
In case you don't know:
1) no bugs for TeX has been reported in the last 10 years (or more?).
2) TeX is not a forgotten piece of software. It gets used daily by
many man people around the whole world (often via LaTeX, AMS-LaTeX,
etc.) Different users use and combine pretty different parts of
TeX. That can certainly be considered "rigorous testing".
Now, explain to me how these are possible? Indeed, many people
*believe* (yes, it is just a belief; no mathematical proof) that TeX
is already bug-free, based on (1) and (2) above.
Baruch> One reason no one has collected on the bounty may be
Baruch> that there *are* no bugs; but there's no way to prove
Baruch> that. That's all I'm saying - it's not possible to prove
Baruch> that software is free of bugs.
It's possible to believe it.
--
Lee Sau Dan 李守敦 ~{@nJX6X~}
E-mail: dan...@informatik.uni-freiburg.de
Home page: http://www.informatik.uni-freiburg.de/~danlee
Jesse F. Hughes
> That only shows that no one has uncovered a bug, not that none exist. One
> reason no one has collected on the bounty may be that there *are* no bugs;
> but there's no way to prove that. That's all I'm saying - it's not
> possible to prove that software is free of bugs.
All those folks working in software verification will be shocked.
It is not at all difficult to give a description of what a program
like "Hello World" is supposed to do and then to prove that it does
it.
As the specification and program increase in complexity, things get
more difficult, of course. Verifying something on the order of TeX is
not feasible with the present tools and methods, but nonetheless, your
statement is naive. It is sometimes possible to prove that software
is free of bugs (such proofs might depend on the compiler doing what
it is supposed to do, of course).
--
Jesse F. Hughes
"That's the base tautological space where by tautological space I mean
a region of truth." -- James S. Harris does philosophy of mathematics.
JSH is a renaissance man.
Baruch
> >>
> Baruch> That only shows that no one has uncovered a bug, not that
> Baruch> none exist.
>
> In case you don't know:
>
> 1) no bugs for TeX has been reported in the last 10 years (or more?).
> 2) TeX is not a forgotten piece of software. It gets used daily by
> many man people around the whole world (often via LaTeX, AMS-LaTeX,
> etc.) Different users use and combine pretty different parts of
> TeX. That can certainly be considered "rigorous testing".
>
> Now, explain to me how these are possible? Indeed, many people
> *believe* (yes, it is just a belief; no mathematical proof) that TeX
> is already bug-free, based on (1) and (2) above.
>
It probably *is* bug-free. My point wasn't that it has bugs, but only that
we can't prove it doesn't have any.
>
> Baruch> One reason no one has collected on the bounty may be
> Baruch> that there *are* no bugs; but there's no way to prove
> Baruch> that. That's all I'm saying - it's not possible to prove
> Baruch> that software is free of bugs.
>
> It's possible to believe it.
>
Absolutely. It's possible to believe in six impossible things before
breakfast... and it's certainly possible (likely) that TeX is bug-free.
Kelsey Bjarnason
On Wed, 01 Sep 2004 14:30:58 +0200, Jesse F. Hughes wrote:
> Baruch <baru...@N0sbcglobal.net$PAM> writes:
>
>> That only shows that no one has uncovered a bug, not that none exist. One
>> reason no one has collected on the bounty may be that there *are* no bugs;
>> but there's no way to prove that. That's all I'm saying - it's not
>> possible to prove that software is free of bugs.
>
> All those folks working in software verification will be shocked.
>
> It is not at all difficult to give a description of what a program
> like "Hello World" is supposed to do and then to prove that it does
> it.
You'd be surprised.
Not too long ago, there was a fairly in-depth discussion in comp.lang.c
over the observation that a program which produced *any* output whatsoever
apparently did so in an implementation-defined manner - thus even the
simplest "hello world" program may not behave as expected.
Then there's issues such as how the program is invoked. For example, if
one does something like hello > file.dat, one might expect the file to be
created and end up with the contents "hello world". Fine, great,
marvellous - but what if the medium is full, or read-only, or you lack
privileges to modify the file, or it's on a network share point and the
network goes down, etc, etc, etc.
All you can do here is assert that from a standpoint of the logic of
program flow, in ideal conditions, it _should_ produce an output of
"hello, world" then exit, hopefully with a "success" status.
Once you start adding in conditionals, however, verification becomes
extremely difficult. Indeed, to validate applications will behave
correctly in terms of desiging a general algorithmic approach to such
requires solving the halting problem - do have fun with that.
Jesse F. Hughes
> [snips]
>
> On Wed, 01 Sep 2004 14:30:58 +0200, Jesse F. Hughes wrote:
>
>> Baruch <baru...@N0sbcglobal.net$PAM> writes:
>>
>>> That only shows that no one has uncovered a bug, not that none exist. One
>>> reason no one has collected on the bounty may be that there *are* no bugs;
>>> but there's no way to prove that. That's all I'm saying - it's not
>>> possible to prove that software is free of bugs.
>>
>> All those folks working in software verification will be shocked.
>>
>> It is not at all difficult to give a description of what a program
>> like "Hello World" is supposed to do and then to prove that it does
>> it.
>
> You'd be surprised.
>
> Not too long ago, there was a fairly in-depth discussion in comp.lang.c
> over the observation that a program which produced *any* output whatsoever
> apparently did so in an implementation-defined manner - thus even the
> simplest "hello world" program may not behave as expected.
Well, C is a language much avoided by everyone that does program
verification, of course, but I don't think that fact is too relevant
to what you say below.
> Then there's issues such as how the program is invoked. For example, if
> one does something like hello > file.dat, one might expect the file to be
> created and end up with the contents "hello world". Fine, great,
> marvellous - but what if the medium is full, or read-only, or you lack
> privileges to modify the file, or it's on a network share point and the
> network goes down, etc, etc, etc.
>
> All you can do here is assert that from a standpoint of the logic of
> program flow, in ideal conditions, it _should_ produce an output of
> "hello, world" then exit, hopefully with a "success" status.
>
> Once you start adding in conditionals, however, verification becomes
> extremely difficult. Indeed, to validate applications will behave
> correctly in terms of desiging a general algorithmic approach to such
> requires solving the halting problem - do have fun with that.
I'll grant you that program verification requires certain assumptions,
but some of the issues you mention are simply outside the scope of
verification. Take piping the output to a file. The verifier doesn't
care what happens in that case and he doesn't care how the OS reacts
when the medium is not writeable. He has only been asked to confirm
that the correct output has been sent to stdout relative to environs
which satisfy the defined semantics of the language.
You can't ask him to ensure that the program works when executed while
the computer is in freefall, milliseconds from crashing into the earth
at some God-awful velocity, after all. But its behavior in those
cases aren't called bugs. An OS that isn't doing what it ought to (or
a compiler) also isn't a bug in the *program*.
You may call these assumptions "ideal conditions", but they seem quite
reasonable attempts to separate the bugs in the program from
undesirable environmental affects that have nothing much to do with
whether the program is written correctly.
--
Jesse Hughes
"I often told you of the dangers of hubris, and most importantly of
all, I TOLD you that I wanted to change the institution of mathematics
worldwide." -- James Harris, on the evils of pride
Arjun Ray
> Seems that Jean Paoli started work at Microsoft in 1996 but had been
> involved with XML (SGML) since 1985.
Yes, Jean Paoli came to Microsoft from Grif.
> Seems that he was involved in the standards creation for XML 1.0, not
> inventing it.
No more involved than a whole bunch of others.
It started in Summer 96, with Jon Bosak canvassing everyone he knew in the
SGML community to join a working group/mailing list where radical changes
in SGML, based on Yuri Rubinsky's "SGML on the Web", could be discussed.
Things got off the ground when Jon found a home for the list at the W3C in
late August early September. (No, the W3C did *not* originate the XML
initiative. They played host, and subsequently took over.)
The extremely aggressive goal was to produce a draft document in time for
SGML 96 in Boston (the annual GCA convention in mid-November.) The real
hard work in the beginning was mostly done by Michael Sperberg-McQueen.
He put up a huge questionnaire on Tim Bray's textuality.con site, that
went through the entire ISO 8879 document (and the later TCs), section by
section, enumerating each and every provision of the SGML standard, with
summaries of older proposals and appropriate questions in multiple-choice
format. You can see it here:
http://www.textuality.com/sgml-erb/dd-1996-0002.html
Michael and Tim then collated the answers, and summarized them into an
initial set of issues. The basic idea was that the Editorial Board would
vote on these issues and others as they came up, drawing on feedback from
the Working Group on the mailing list. This is how the deadline was met.
You can see the entire process in action by browsing this archive:
http://lists.w3.org/Archives/Public/w3c-sgml-wg/
Fair warning: a lot of posts will be inscrutable if you don't know SGML.
But, by the same token, it really isn't possible to grasp fully why XML is
what it is (and what serious problems it has - a lot of mistakes in that
first cut never got fixed once "W3C Process" took over) without the SGML
background.
As for Jean Paoli and Microsoft, I would estimate their constructive
contribution as indistinguishable from zero.