Having been immersed in the wonder of c# and XML the past few days -- I 
am shocked -- SHOCKED -- at the low level of /mindshare/ when it comes 
to XML in the Linux Community.

XML is so much the man !

And yet, I rarely hear it as part of any of the various Linux 
*architectures*

Why not?

What is it about Advocates that make them fear XML ?

XML is data, interface and explanation all in one !

XML is the answer to so many data issues.

XML is the original /Open/ source technology -- it is human-readable and 
self explanatory -- it is the open source of data -- unlike the 
obscurities and arcanitude of the DBMS.

XML is Open Source.

Embrace it -- GNU and Linux Advocates -- I beseech thee !()!

Seeking Twine wrote:

> XML is so much the man !

You are a few years late. The XML hype took place already some time ago. :)

M.

Michael Vondung wrote:

> Seeking Twine wrote:
> 
>> XML is so much the man !
> 
> You are a few years late. The XML hype took place already some time ago.
> :)
> 
> M.

of course by that measure, all talk about linux is twice as /late/  

-- 
Valhallah !

Viking Bastards <deathtoValhalla@daeth.crem> writes:

> Michael Vondung wrote:
> 
> > Seeking Twine wrote:
> > 
> >> XML is so much the man !
> > 
> > You are a few years late. The XML hype took place already some time ago.
> > :)
> > 
> > M.
> 
> of course by that measure, all talk about linux is twice as /late/

Linux is a mostly uncertified UNIX implementation, so it actually is
totally retrocool again.  I have some (c) 1982 book here that is
totally suitable for Linux, give or take a few quaint file system
peculiarities like the boot procedire.  That is described as "Getting
UNIX started on a dead processor is sometimes riminiscent of black
magic rites: various secret incantations to be muttered; shaking a bag
of bones over the power supplies; appeasement of the gods and
sacrifice of a virgin -- you know the sort of thing."  It is obvious
that the book is concerned more with the procedures in UNIX Version 7:
old Linux users will probably rather remember the black cat and the
incense that are more typical with BSD-like variants.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Seeking Twine wrote:

> 
> Having been immersed in the wonder of c# and XML the past few days -- I
> am shocked -- SHOCKED -- at the low level of /mindshare/ when it comes
> to XML in the Linux Community.
> 
> XML is so much the man !
> 
> And yet, I rarely hear it as part of any of the various Linux
> *architectures*

XML is a file/data format.
> 
> Why not?

Because it isn't a big deal.

> 
> What is it about Advocates that make them fear XML ?

Nothing.
> 
> XML is data, interface and explanation all in one !

It is a text file format.
> 
> XML is the answer to so many data issues.

No one is really asking the questions, except,  of course, Micro$soft.

> 
> XML is the original /Open/ source technology -- it is human-readable and
> self explanatory -- it is the open source of data -- unlike the
> obscurities and arcanitude of the DBMS.

XML is a hugely inefficient mechanism to transport data. It has its places,
but it is not a universal cure all.

David Kastrup wrote:

> sacrifice of a virgin -- you know the sort of thing."  It is obvious
> that the book is concerned more with the procedures in UNIX Version 7:
> old Linux users will probably rather remember the black cat and the
> incense that are more typical with BSD-like variants.

Soon we'll be carrying out papers in jute bags and quoting from Abbie
Hoffmann's 'Steal This Book'



-- 
Valhallah !

> Having been immersed in the wonder of c# and XML the past few days -- I am
> shocked -- SHOCKED -- at the low level of /mindshare/ when it comes to XML
> in the Linux Community.

Well, XML is a technical detail.  Anybody who's competent enough will use
XML when it makes sense and won't bother mentioning it, just like she won't
mention that she used Emacs to edit the source code.


        Stefan

In article <x5wu39nbyb.fsf@lola.goethe.zz> (Tue, 18 May 2004 14:42:36
+0200), David Kastrup wrote:

> Linux is a mostly uncertified UNIX implementation [...]

ITYM POSIX.

> so it actually is totally retrocool again.

The Linux fad, if that's what it is, is as unexplainable as hula hoops,
platform shoes, and afro hairdoos.  (OTOH, nothing is as unexplainable as
disco music.)

> I have some (c) 1982 book here [...] That is described as "Getting
> UNIX started on a dead processor is sometimes riminiscent of black magic
> rites [...]

Like Unix and Windows, Linux certainly won't start on a dead processor; I
haven't tried black magic.

-- 
"[The SCO Group] is (and will remain) a scam stock with no assets --
 just a business model based on litigation."
-- http://www.itmanagersjournal.com/management/04/04/21/1529253.shtml

Hamilcar Barca wrote:

> The Linux fad, if that's what it is, is as unexplainable as hula hoops,
> platform shoes, and afro hairdoos.  (OTOH, nothing is as unexplainable as
> disco music.)

Platform shoes are worse. I was seriously shocked when they reappeared, 
and I seek comfort in knowing that soon the fashion style of the 
eighties will return. (The really depressing part is to experience all 
of this for the second time, since it shows so clearly that we've gotten 
old.)

M.

Stefan Monnier wrote:

> just like she won't
> mention that she used Emacs to edit the source code.

Or vi.

M.

Stefan Monnier <monnier@iro.umontreal.ca> writes:

> > Having been immersed in the wonder of c# and XML the past few days
> > -- I am shocked -- SHOCKED -- at the low level of /mindshare/ when
> > it comes to XML in the Linux Community.
> 
> Well, XML is a technical detail.  Anybody who's competent enough
> will use XML when it makes sense and won't bother mentioning it,
> just like she won't mention that she used Emacs to edit the source
> code.

XML is not a binary format.  That means that human readability is a
concern.  And coding conventions etc.

I am maintainer of AUCTeX.  Don't tell me your LaTeX editor is a
technical detail.  Good source code formatting is important, and it
does not fall off the trees.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Michael Vondung wrote:

> Hamilcar Barca wrote:
> 
>> The Linux fad, if that's what it is, is as unexplainable as hula hoops,
>> platform shoes, and afro hairdoos.  (OTOH, nothing is as unexplainable as
>> disco music.)
> 
> 
> Platform shoes are worse. I was seriously shocked when they reappeared, 
> and I seek comfort in knowing that soon the fashion style of the 
> eighties will return. (The really depressing part is to experience all 
> of this for the second time, since it shows so clearly that we've gotten 
> old.)
> 
> M.

Eventually the head of the snake eats the tail.  As always....

>> just like she won't mention that she used Emacs to edit the source code.
> Or vi.

It's different.  People who use Emacs don't mention it simply because it's
so obviously natural that it doesn't make sense to say it explicitly.
OTOH people using vi don't mention it simply out of shame.


        Stefan


PS: who said we couldn't have reasoned thoughtful arguments in g.m.d ?

> XML is not a binary format.  That means that human readability is a
> concern.  And coding conventions etc.

If readability were really an issue, I think XML would have been defined
differently.

Most actual uses of XML I've seen are pretty damn close to binary
(e.g. see Subversion).  Even if the XML is generated "by hand" it's most
often done via some specialized editor that hides a large part of the XML
textual representation and lets the user manage the conceptual tree instead.
I.e. for all the writer knows, the data representation could just as well be
binary or lisp-sexp and she wouldn't know the difference.


        Stefan

Stefan Monnier wrote:

>>XML is not a binary format.  That means that human readability is a
>>concern.  And coding conventions etc.
> 
> 
> If readability were really an issue, I think XML would have been defined
> differently.

http://www.w3.org/TR/REC-xml/#sec-origin-goals

The design goals for XML are:

XML shall be straightforwardly usable over the Internet.

XML shall support a wide variety of applications.

XML shall be compatible with SGML.

It shall be easy to write programs which process XML documents.

The number of optional features in XML is to be kept to the absolute 
minimum, ideally zero.

XML documents should be human-legible and reasonably clear.

The XML design should be prepared quickly.

The design of XML shall be formal and concise.

XML documents shall be easy to create.

Terseness in XML markup is of minimal importance.

Stefan Monnier wrote:

> Right.  readability is only mentioned as part of
> 
>    XML documents should be human-legible and reasonably clear.
> 
> and only after
> 
>    XML shall be compatible with SGML.

Yes, but come on.  We are talking about a document that in many ways can 
replace the use of SQL databases.   An XML document is far more readable 
than a binary database file -- wouldn't you say ?


> 
> so I doubt readability played a significant role in the design: most of the
> "not too bad" readability comes from SGML and so does the cruft that makes
> it not so hot readbility-wise.
> 
> 
>         Stefan

>>> XML is not a binary format.  That means that human readability is a
>>> concern.  And coding conventions etc.
>> If readability were really an issue, I think XML would have been defined
>> differently.

> The design goals for XML are:
[...]

Right.  readability is only mentioned as part of

   XML documents should be human-legible and reasonably clear.

and only after

   XML shall be compatible with SGML.

so I doubt readability played a significant role in the design: most of the
"not too bad" readability comes from SGML and so does the cruft that makes
it not so hot readbility-wise.


        Stefan

>> Right.  readability is only mentioned as part of
>> XML documents should be human-legible and reasonably clear.
>> and only after
>> XML shall be compatible with SGML.

> Yes, but come on.  We are talking about a document that in many ways can
> replace the use of SQL databases.   An XML document is far more readable
> than a binary database file -- wouldn't you say ?

I never said it's not readable.  Just that readability is not the main
concern.  Having a "self-describing" "text" representation is the main
concern.  Readability is pretty secondary.

And I'm not saying is a problem of XML, given that it's almost never
edited/read in its ASCII form anyway (other than for debugging and a few
weirdos like me who insist on editing their environment.plist and foo.xhtml
by hand in a text editor).


        Stefan

In article <2h17s5F7vnabU1@uni-berlin.de> (Wed, 19 May 2004 16:04:52
+0200), Michael Vondung wrote:

> Hamilcar Barca wrote:
> 
>> The Linux fad, if that's what it is, is as unexplainable as hula hoops,
>> platform shoes, and afro hairdoos.  (OTOH, nothing is as unexplainable
>> as disco music.)
> 
> Platform shoes are worse.

Platform shoes may be worse than 'fros and hula hoops but nothing is 
worse than disco unless it's Microsoft's windows.  At least disco 
isn't dangerous.

> I was seriously shocked when they reappeared

Retro is cool, I guess.

> since it shows so clearly that we've gotten old.)

I started to wonder when I heard a "classic rock" radio station playing
music that was released after I started college.  However, anything but
disco and windows...

The world rejoiced as Seeking Twine <any@twine.for.me> wrote:
> Having been immersed in the wonder of c# and XML the past few days --
> I am shocked -- SHOCKED -- at the low level of /mindshare/ when it
> comes to XML in the Linux Community.
>
> XML is so much the man !
>
> And yet, I rarely hear it as part of any of the various Linux
> *architectures*
>
> Why not?

It's used quite heavily in the GNOME and KDE "architectures," in the
application layer.  It is used quite widely in applications.

> What is it about Advocates that make them fear XML ?

I have no idea.

> XML is data, interface and explanation all in one !

So the theory goes.  It is also fair to regard it as a dumbed-down,
more-redundant, less-readable alternative to Lisp S-expressions.

> XML is the answer to so many data issues.

It is also the cause of many data issues, when people blindly try to
hammer data into XML form.

> XML is the original /Open/ source technology -- it is human-readable
> and self explanatory -- it is the open source of data -- unlike the
> obscurities and arcanitude of the DBMS.

XML is easy to make every bit as obscure as any technology you'd
regard as arcane.  It is NOT an answer to problems better solved using
an RDBMS.

> XML is Open Source.

Have you looked at the opaque set of standards?  I was a technical
reviewer on a [horrid] book on Web Services Security, which actually
amounted to a presentation of a bunch of "planned standards"
surrounding XML 'security.'  XML is NOT 'Open Source;' it usually
isn't even implemented correctly.

And by the time IBM, Microsoft, and such get through with adding
'standards' to add the functionality that is missing, they'll have
bloated things that are much more verbose, slow, bloated, and such
than the things they tried to replace.

> Embrace it -- GNU and Linux Advocates -- I beseech thee !()!
-- 
let name="cbbrowne" and tld="ntlug.org" in name ^ "@" ^ tld;;
http://cbbrowne.com/info/
I knew you weren't really interested.
-- Marvin the Paranoid Android

In an attempt to throw the authorities off his trail, Ministry Of Jute <allthings@jute.net> transmitted:
> Stefan Monnier wrote:
>
>> Right.  readability is only mentioned as part of
>>    XML documents should be human-legible and reasonably clear.
>> and only after
>>    XML shall be compatible with SGML.
>
> Yes, but come on.  We are talking about a document that in many ways
> can replace the use of SQL databases.  An XML document is far more
> readable than a binary database file -- wouldn't you say ?

No, I wouldn't.

An XML document is exactly analagous to a PostgreSQL 'pg_dump', which
is a serialized version of a database.  The 'pg_dump' is text...
-- 
output = ("cbbrowne" "@" "acm.org")
http://www.ntlug.org/~cbbrowne/multiplexor.html
Signs of a Klingon  Programmer - 5. "Indentation?! -  I will  show you
how to indent when I indent your skull!"

Christopher Browne wrote:

> 
> Have you looked at the opaque set of standards?  I was a technical
> reviewer on a [horrid] book on Web Services Security, which actually
> amounted to a presentation of a bunch of "planned standards"
> surrounding XML 'security.'  XML is NOT 'Open Source;' it usually
> isn't even implemented correctly.
> 

I was at a Microsoft seminar last year, I thinks something like "toolshed,"
in Boston. I didn't want to go and as expected it was a complete and total
marketing scam. Anyway, one presenter showed how "secure" .NET was. He
added a password field to the XML message and showed how it wouldn't work
without the password. Amongst the "oos" and "aaahs" in the audience, me an
the guy next to me we laughing at "Is this what Microsoft calls
'security?'"

> And by the time IBM, Microsoft, and such get through with adding
> 'standards' to add the functionality that is missing, they'll have
> bloated things that are much more verbose, slow, bloated, and such
> than the things they tried to replace.

I just finished doing a contract writing a server, it had to take multiple
sources of XML input and, after manipulation, produce one XML output.

XML is horrific because it is conceptually no different that ASCII. Yea,
sure, ASCII is a standard text format, but it doesn't help you interpret
the text. XML, is a standard data format, but it doesn't help you interpret
the data. The fact that you can put virtually anything in XML means that it
is virtually impossible to create a program which uses "XML" for anything
other than a data format. You have to create a program that accepts a set
of records, with "these" names, with "this" hierarchy, formatted in XML.

My client kept getting new data sources, each formatted differently, but
each "valid" XML. I tried over and over to explain, "I need to know the
names of the records, the hierarchy of the data, and how to map the various
data entities." Their response? "What's wrong? It's all XML, shouldn't it
just work?" sheesh.

> 
>> Embrace it -- GNU and Linux Advocates -- I beseech thee !()!
> --
> let name="cbbrowne" and tld="ntlug.org" in name ^ "@" ^ tld;;
> http://cbbrowne.com/info/
> I knew you weren't really interested.
> -- Marvin the Paranoid Android

>>>>> "mlw" == mlw  <mlw@nospam.no> writes:

  mlw> XML is horrific because it is conceptually no different that
  mlw> ASCII. Yea, sure, ASCII is a standard text format, but it
  mlw> doesn't help you interpret the text. XML, is a standard data
  mlw> format, but it doesn't help you interpret the data.


This is a little harsh. XML is something which enables you to define
and specify upfront a syntax. This is an advance from ASCII. At least
XML has standard tools for parsing (and rendering). 

The problem is that people think its more. "Look," they say "it's
self-describing". Well no, its not. 


  mlw> My client kept getting new data sources, each formatted
  mlw> differently, but each "valid" XML. I tried over and over to
  mlw> explain, "I need to know the names of the records, the
  mlw> hierarchy of the data, and how to map the various data
  mlw> entities." Their response? "What's wrong? It's all XML,
  mlw> shouldn't it just work?" 


And it will at one level "all work". At a syntactic level. 

Like many technologies in IT, there is nothing wrong with XML(1), its
just that the hype was so great that people think it can do things
which it can't. 


Phil


1) Actually there are somethings wrong with XML, I'm sure...

Phillip Lord wrote:

>>>>>> "mlw" == mlw  <mlw@nospam.no> writes:
> 
>   mlw> XML is horrific because it is conceptually no different that
>   mlw> ASCII. Yea, sure, ASCII is a standard text format, but it
>   mlw> doesn't help you interpret the text. XML, is a standard data
>   mlw> format, but it doesn't help you interpret the data.
> 
> 
> This is a little harsh. XML is something which enables you to define
> and specify upfront a syntax. This is an advance from ASCII. At least
> XML has standard tools for parsing (and rendering).

ASCII has many more "standard tools" for parsing and rendering. The function
calls printf and scanf work better and faster than *any* XML parser.

> 
> The problem is that people think its more. "Look," they say "it's
> self-describing". Well no, its not.

"Self describing?" It is called that, but it isn't at all. SOAP, based on
XML, "may" be self descriptive of the data element format, but very little
else.

You still need to know the schema of the data and its purpose to use it in
an application, as such, simple text is much more efficient to use. Yea,
XML, as a documented textual format, has some value, but it is by no means
even a "good" idea from a technical perspective.

> 
>   mlw> My client kept getting new data sources, each formatted
>   mlw> differently, but each "valid" XML. I tried over and over to
>   mlw> explain, "I need to know the names of the records, the
>   mlw> hierarchy of the data, and how to map the various data
>   mlw> entities." Their response? "What's wrong? It's all XML,
>   mlw> shouldn't it just work?"
> 
> 
> And it will at one level "all work". At a syntactic level.

Yea, no worries, yup goes through my XML parser just fine. Can't find
anything in the tree because I don't know what to look for.

> 
> Like many technologies in IT, there is nothing wrong with XML(1), its
> just that the hype was so great that people think it can do things
> which it can't.

Stupid people confuse and conflate different and separate technologies. XML
is a textual file format, nothing more, nothing less. The fact that it is a
"standard" whatever that means, i.e. ASCII is a "standard," has allowed
structures and protocols to evolve around it. These structures and
protocols have made some tasks, which were previously difficult, a bit
easier. Like all conveniences, there are tradeoffs. XML communication is
slow and bloated, across an application server, it can double transfer data
size, and inflicts a great deal of CPU overhead.

In a low utilization system, these are not measurable issues. In a high
speed system, it can mean the difference between a $100K data center and a
$1M data center.

> 
> 
> Phil
> 
> 
> 1) Actually there are somethings wrong with XML, I'm sure...