Whim

Query re: Michael Jackson's "Problem Frames" - any listserv or similar?

== Updated: See "Comment" below ==

My e-mail inquiry.

Query re: Michael Jackson's "Problem Frames" - any listserv or similar?
From: William Reilly
To: jblaine@san.rr.com

Greetings, Mr. Blaine,

I came upon your 2002 review of Michael Jackson's "Problem Frames," in Software Quality Professional:
http://www.asq.org/pub/sqp/past/vol4_issue2/resource.html#problemframes

...and thought I might ask if you are familiar with any sort of forum, listserv, wiki, etc., in which people are discussing Problem Frames.

I've recently begun this book,
http://www.amazon.com/exec/obidos/tg/detail/-/020159627X/
Problem Frames
Michael Jackson. Addison Wesley, 2001.

...led to it by recently finishing Ben Kovitz's, which is based in part on Jackson's problem frames work:
http://www.amazon.com/exec/obidos/tg/detail/-/1884777597/
Practical Software Requirements: A manual of content and style.
Manning, 1999.

Thank you for your time in considering my request.

Best regards,
William Reilly
william@reilly2001.info
http://reilly2001.info
Somerville, Massachusetts

DocBook As SpecBook?

My e-mail to experts at Adobe and Sun (on the OASIS Technical Committee for ebXML), as well as my posting to the docbook-apps listserv:
http://lists.oasis-open.org/archives/docbook-apps/200401/msg00301.html
===================
Fwd: [docbook-apps] DocBook for authoring Technical Requirements, Specifications?
From: William Reilly
To: mxxxx@ adobe.com, dxxxxxxxxxx@ Sun.com

Hello, Matt MacKenzie & Doug Bunting,

I'm very interested in (possibly) using DocBook not to "Doc"ument but to
Specify a system.

Looking for a "SpecBook," or "ReqBook" ;^) essentially.

Or, to learn that DocBook itself might be a good choice...

--- WHY I'M WRITING YOU ------------------
From the DocBook WIKI I learned that:
"OASIS ebXML Message Services TC has decided to use DocBook for version 3.x
of its specification"
http://www.docbook.org/wiki/moin.cgi/WhoUsesDocBook

I then found your names from this OASIS page:
http://www.oasis-open.org/committees/documents.php?wg_abbrev=ebxml-msg

Looks like V2 Spec. "formatting was based upon the Internet Society's Standard
RFC format, converted to Microsoft Word 2000."
http://www.oasis-open.org/committees/ebxml-msg/documents/ebMS_v2_0rev_c.pdf

And I found the ebXML DTD and XML used in Oct. 2002:
http://www.oasis-open.org/committees/ebxml-msg/documents/requirements.xml
http://www.oasis-open.org/committees/ebxml-msg/documents/requirements.dtd

--- MY BASIC QUESTION -----------------
My question would be a wide open one:
- Are you using DocBook to write requirements, or specifications?
- How is that going?
- Have you needed to extend DocBook (to get your "reqDetails" kind of tags,
etc.)?

Any info, comments, or sample files you feel you could share would be most
welcome.

I'm at a not-for-profit educational organization, http://www.cast.org -
creating improved online test-taking for learning-impaired and print-disabled
kids.

--- THANK YOU -----------------
Many thanks for your time...
Best regards,
William Reilly
william@ reilly2001.info
Somerville, Massachusetts
http://reilly2001.info
=========================
P.S. This page (WhatIsDocBookUsedFor) lists everything _except_ specifying
systems!)
As I note (below, in my post to docbook-apps mailing list):
"Perhaps it's that the overhead of authoring in a markup language is seen
as detrimental to the shorter life of these "upstream" documents, as compared
to downstream documentation efforts, which ought to endure long after a
system is built and in use."
http://www.docbook.org/wiki/moin.cgi/WhatIsDocBookUsedFor
---------------------------------------------
# help systems
# Web sites
# books
# reference pages
# FAQs
# white papers
# training courseware
# articles
# API documentation
# reports
# functional specifications
# "how to" guides and other procedural documentation
# presentations
---------------------------------------------
---------- Forwarded Message ----------

Subject: [docbook-apps] DocBook for authoring Technical Requirements,
Specifications?
Date: Thursday 29 January 2004 11:47
From: William Reilly
To: docbook-apps@lists.oasis-open.org

Greetings.

Searching Google, OASIS, and "Requirements" community sites hasn't yet shown
me whether there are XML vocabularies generally in use for creating system
requirements and technical specifications documents.

DocBook occurred to me, despite its having been designed to "doc"ument a
system more than specify one.

Any "SpecBook" or "ReqBook" out there?

Or is going the MS-Word or OpenOffice .DOC template route the only way these
days? (or Rational ReqPro, LiveSpecs, and similar expensive commercial
products)

Perhaps it's that the overhead of authoring in a markup language is seen as
detrimental to the shorter life of these "upstream" documents, as compared to
downstream documentation efforts, which ought to endure long after a system
is built and in use.

Many thanks for suggestions, comments.
William Reilly
Somerville, Massachusetts, U.S.A.
william@reilly2001.info

To unsubscribe from this list, send a post to
docbook-apps-unsubscribe@lists.oasis-open.org, or visit
http://www.oasis-open.org/mlmanage/.

-------------------------------------------------------

FINALLY:
Dave Pawson was kind enough to respond with:

You might try .gov sites?
I guess they are still into full requirement specs etc?
Perhaps mix it with iso9000 and stuff as part of the search?

HTH DaveP
(ex mil.std :-)
==========================

XML Editors: Cladonia Exchanger vs. SyncRO oXygen

I wanted an inexpensive XML Editor IDE for Linux (ideally for Windows and as plug-in to Eclipse, and doubly ideally for one license price).

DISCLAIMER: Below, I've written not a review, but taken some semi-organized notes.

If anyone reading has comments, questions, clarifications, I would definitely like to hear about all.

Minor UPDATE in "Comments" below... [Feb. 13, 2004]

Cladonia Exchanger, v. 1.2 is $98, or $45 non-profit
SyncRO oXygenXML, v. 3.0 is $74, or $48 non-profit.
One license lets you install to Linux or Windows; oXygen can also work as Eclipse plug-in.

Click on the two screenshots below to see full-screen.

Exchanger

oXygen

30-day trials, more screenshots at:
http://www.cladonia.com
http://www.oxygenxml.com

(One more I've not (yet) looked into is MyEclipseIDE.)

Conclusion First:
Well, if I really had to buy right now, I think Cladonia's Exchanger would provide me more useful functionality than oXygen, but, there are as always some tradeoffs.

Read the first few bullets under "NOT SO GOOD, COMPARISON" and "GOOD COMPARISON" for each product below to get the gist pretty quickly.

I know it's always kind of deadening to go through such a lot of info.
Briefest High Level Points:
-=-=-=-=-=-=-=-=-=-
Exchanger is stronger in XPath; Schema viewing; and keeping main and outline views in synch (something I rank high). It also has some interesting other "services" (SOAP. SVG, etc. details below).
But - Exchanger cannot work too well with DTDs directly (I was disappointed). To get context-sensitive dropdown of valid elements (something I rank high), you can't use a DTD but must convert it to either a Schema or to the Exchanger proprietary concept of "named Type." Kind of odd; ah well.

oXygen creates v. nice HTML doc for your Schema; can work better with DTDs; can work as Eclipse plug-in; has better U/I generally.
-=-=-=-=-=-=-=-=-=-

N O T E S

==================
Cladonia Exchanger = $45 (non-profit; $98 reg.)
(Like oXygen, one license per user, not OS. Buy once and run on Linux, Windows. Nice.)
==================

GOOD, COMPARISON

MAJOR
- XPath
Superior in Exchanger. Larger textarea box, for starters (more of the long Paths visible), but more importantly, clicking around in either the document or the outline keeps that XPath textarea updated. This is good. (XML Spy only lets you "calculate XPath" when you enter its modal dialog devoted to that. Exchanger's is nicer in that it's integrated into working with the document.)
oXygen has just about no automatic connection to the XPath textarea.
Both do have a dropdown button for "memory" of reusable recently executed XPaths.
- "Schema" View
Pretty interesting: you open not the Schema per se, but an XML document, and ask for the view of the Schema document it's using. (You can then ask for the Schema of the Schema...). Provides good navigation among "Globals," "Refers," and "Substitutions". Useful. (And yes, if you want, you can always simply open a Schema as a file, but this mechanism lets you skip that.)
Generally, it's fair to say that Exchanger Schema support (viewing, esp.) is far superior to Oygen. The latter provides typical text view of the .XSD document, and the usual Outline available in side pane. But Exchanger provides far more (as described just above).
(Do note that Oxygen can produce nice HTML documentation of your Schema, but right in the IDE is kind of hard to beat.) Also, note that although Oxygen has an attractive "Tree Editor," that doesn't really bring anything esp. useful to looking at an .XSD.
On the other hand, Oxygen does offer a nice button to get a browser-link, re: XSD, out to the W3C Spec, for any validation errors encountered in your XSD (e.g. auto connect to URLs like: http://www.w3.org/TR/xmlschema-1/#src-resolve ).
One more important note: Apparently the Exchanger "Schema Viewer" is also a FREE stand-alone product (they call it "service") offering: http://www.cladonia.com/schemaviewer/ (v. 0.5)

- Synchronizing between Outline and Main Edit works both ways (click and it synchs other pane).
- Outline is a little less elegant looking, but the fact you can, in Preferences, turn on "Show Attributes" means that the Outline in Exchanger can be much more useful than its counterpart in oXygen, where it's more attractive at a glance (elegant, narrower), but as it doesn't carry any attribute info, you are presented with a long list of "unlabelled," as it were, elements. (e.g. Exchanger shows: 'chunk id="passage4"' on top of 'chunk id="passage5"' and so on, while oXygen shows: 'chunk' on top of 'chunk' on top of 'chunk' and so on. You get the idea.)

- Also important to note are Exchanger's XQuery 1.0 and XSLT 2.0 support, though I'm not using these things.

MINOR
- Exchanger has "Viewer" view, which complements the "Editor" view, by providing the '+/-' sign convention for collapse/expand (as seen in IE presenting raw XML). oXygen lacks this, and you realize its value when you have large areas of XML to range over, and it's nice to be able to "move" some out of the way like this.
- Outline has handy "Collapse All" and "Expand All" buttons.
- Search Again (F3) does NOT synchronize between Main and Outline.
- The "Format" (indent) button took care of removing all the hard linebreaks that were in the book.xml document. This could be quite useful in certain situations...

GOOD, SINGULAR
MAJOR
- Has "Go to Start Tag" and "Go to End Tag" buttons. Useful.
MINOR
- Has immediate, top of right-click menu, Find (word cursor's on)
- Has "Select Element Content" and "Select Element" buttons. Useful.
- Has "Convert Characters to Entities", "Convert Entities to Characters", and "Strip Tags from Selection" buttons. Interesting, sometimes useful, I suppose.

NOT SO GOOD, COMPARISON

MAJOR
- Cannot use DTD for context-sensitive elements
(offers auto-convert to Schema, or "Type")
- Looking at a DTD only works in "Editor" mode; no color-coding even. (Exchanger requires well-formed XML to do much else (e.g. "Viewer" mode, Outline, etc.). This is fair (oXygen too shows DTDs simply as text file, though you do get color-coding).)
(For careful reading of DTDs, use LiveDTD or DTDParse anyway - converters to useful hyperlinked HTML.
http://www.sagehill.net/livedtd/
http://sourceforge.net/projects/dtdparse/ (Whoops, my Perl modules lacking; sorry.)
/home/william/bin/livedtd-run
file:///home/william/schemata/dtd/ims_qtiasiv1p2p1/index.html
- Doesn't work/plugin to Eclipse
- No nifty "Schema Documentation" like Oxygen has.

MINOR
- Not getting Line Numbers (?) (though Preferences says I can set them. Hmmmm.)
( I submitted a Bug to SourceForge )
UPDATE: If you turn off "Soft Wrap Lines," then Line Numbers will work. They're working on making Line Numbers work regardless... (Nice responsive next-day e-mail to bug report :^)
- Doesn't provide "EOF/EOL" visible markers
- Bottom pane can't go away, and it's shortest height is too high
- Nothing re: "webdav" nor "ftp" in Help Search. Hmmm.
- Cannot 'Ctrl' click multiple files open (not a v. big deal)
- No "Close All Files" (not a big deal)

OTHER
- Also important to note (things I've not looked into): "Services" like SOAP; PIMnet; SVG; XHTML; Java Conduit. See: http://xngr.org/services.htm

- A "real" review (Oct. 2002) is at: http://www.nwfusion.com/newsletters/web/2002/01590348.html.

-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

==================
OxygenXML = $48 (non-profit; $74 reg.)
(Like Exchanger, one license per user, not OS. Buy once and run on Linux, Windows, & Eclipse. Nice.)
==================

GOOD, COMPARISON

MAJOR
- CAN use DTD for context-sensitive elements
In fact, can even provide you context-sensitive elements list dropdown with NO schema, dtd, relaxng. How? It has a "Learn Structure" feature which automatically kicks in to provide you this editing assist. Quite nice. (It's also supposed to have a "Save Structure" to create a new file, but, this was not working for me.)
- DOES work/plugin to Eclipse
- Wow. "Schema Documenter" produces a beautiful, information-rich HTML page of your XSD, including (collapsible) "Schema Component Representation" and "XML Instance Representation" (code snippets).
- DOES provide "EOF/EOL" visible markers

MINOR
- CAN 'Ctrl' click multiple files open
- Has "Open File at Cursor" e.g., if there's a filename or URI in your file, put the cursor on it, right-click, and this opens that file. Nice.
- WebDAV and FTP features

GOOD, SINGULAR
- Tree Editor. Quite nice, separate app (separate window). Exchanger has an "Outliner" which (in the same IDE) permits some "editing/writing without messing with tags" like this oXygen Tree Editor does, but the Tree Editor is far more impressive, far more mature.

NOT SO GOOD, COMPARISON
MAJOR
- XPath use not as good as Exchanger (details above).
- Schema viewing not as good as Exchanger (details above).
- Does NOT synchronize between Outline and Main Edit both ways - only one: click in Outline to get synched in Main Edit. Would be nice to click in Main and get Outline to synch. Ah well.
- Outline is more elegant looking, but less useful. (See note above under Exchanger.)

MINOR
- No "Viewer" view (as in Exchanger) ('+/-' sign convention for collapse/expand (as seen in IE presenting raw XML).The utility in this is when you have large areas of XML to range over, and it's nice to be able to "move" some out of the way like this. Ah well; in oXygen you just cursor up and down all the XML in plain text editor kind of mode.
- Outline does NOT have handy "Collapse All" and "Expand All" buttons.
- Search Again (F3) does NOT synchronize between Main and Outline.
- Left pane can't go away, and it's shortest width is too wide.
- Did NOT remove all the hard linebreaks that were in the book.xml document. (The "Format" (indent) button)
Maybe you could make the case it shouldn't, but, I still think this could be quite useful in certain situations...

==============
BOTH
==============

COMPARABLE - NOT SO GOOD
- Not a show-stopper, but neither have the nice visuals for XML Schema as found in more focussed (and more expensive) products like Tibco ($270) (formerly extensibility) Turbo XML (visual = http://www.tibco.com/images/solutions/products/extensibility/schema_editing_screenshot.gif) or Altova's (Windows only) XML Spy ($400) (visual = http://www.xmlspy.com/images/shots/schema_view_tree.gif) [They have a "Home" version for $50 that does get you the XML Schema editor :^), but no XPath, no XSLT debugger, and many other things not :^( ]
[News (to me): If you want Windows on Linux, go to CodeWeavers.com and for $60 get CrossOver Office. Then you can run XML Spy on Linux. Hmmmph.]

COMPARABLE - O.K.
- Transformation Scenarios. Look to be pretty comparable. Exchange has XQuery; Oxygen I think doesn't. Oxygen lets you append a header or footer (from a URI) to your transform output.
- Converting Schemata. At a high level, both seem to offer a lot vis-a-vis Converting Schemata, one type to another. Exchanger also has concept of named "Types," which permits even more along these lines. Oxygen tells us it's using James Clark's Trang; don't know about Exchanger. (as with so much of this, you'll have to see the documentation, website, or product to understand more. sorry...)
- Both have "Find All" or "Find in Files," which put clickable links to all occurrences of word in a 2nd pane. oXygen has somewhat better context provided (line text, file, and line/col number). Exchanger 2nd panel info not quite as elegant context provided (line; file (full path; can be good, can be too verbose), and line (no col) number).
Exchanger makes this feature available only if you create a Project. (In both products, it's not available if you're searching on one file only (then you just get hit highlight, with F3 to "find again.")
Exchanger 2nd panel info not quite as elegant context provided (line; file (full path; can be good, can be too verbose), and line (no col) number).

MINOR
- Both have buggy Java (Swing?) menubar dropdowns that don't permit another application's window to which you change the focus, to paint over the dropdown. You have to go back to Exchange or oXygen and click somewhere else to close the dropdown, then you can go back to your other window.
- Presentation of error messages on validation are comparable, but I think I prefer Exchange's (oXygen's is a bit slicker, ease of getting about in them, but you have to click more to read full message. Exchange's plainer, fuller, right there. A matter of taste, etc.)
- Presentation of Find In Files uses same pane and functionality. In this case I find I prefer oXygen's better controls; Exchanger lacks even a horizontal scrollbar to read overly wide info, or resizable columns.

MS-Word's Awful XML

My posting to the cms-list back on November 7, in response to a big ol' long thread (some 23 entries!) on the topic.

My own inquiry (pasted below, and available online at:
http://archivist.incutio.com/viewlist/cms-list/2012
elicited exactly zero replies!
Wm.

===============================
[cms-list] Re: Cutting and Pasting from Word?!
From: William Reilly
To: cms-list@cms-list.org

I'd be curious:

== Has anyone ever heard from any Microsoft employee / defender / apologist on
any public listserv (or other venue) speak up about the how and why their
rendition of HTML-izing has to be quite so bad as it is???

I mean, we know why, I suppose: round-tripping to MS-Word being the central
raison d'etre for working with HTML, according to Redmond, but still...

== Are the new "File | Save As..." options in Office 11/2003 making any of
this appreciably better?

William Reilly
Somerville, Massachusetts U.S.A.
=============================
On Friday 07 November 2003 peeter talvistu, wrote:

> It is hard to make it work 100% well, though, because Word's html is
> so - and let me be blunt - idiotic (thousands of <b> and <strong> tags for
> marking one paragraph bold for example).

--
http://cms-list.org/
please trim your posts.

Slash Post Posted to Blog. Ornery XSLT Question...

Will my personal information be public if I post here?

Original Post

---

Driving me a bit crazy.

Is the use of the [1] predicate really required?

I'm finding when I remove it, I'm still getting same (correct?) results.

Seems like the 'select' attribute on a key() function's returned node-set is giving me the same thing whether I explicitly indicate I want the first one '[1]' or if I don't explictly do that. (It even seemed to return same results when I explicitly asked for [2] second node. Hmmm. (See final example below.))

Is that the nature of a 'select' attribute on a node-set, to get the first one in node-set anyway? (Hmmmm.)

I expect in the end I'm wrong about all this, and I want to believe my textbook(s)!, yet, I've not been able (yet?) to demonstrate to myself the difference.

[Please note: I do note the explanation worked through in Beginning XSLT on pages 384-385 (TVGuide8.xsl and TVGuide9.xsl), regarding getting "only one" of the elements (<Series> elements in her example). But that still doesn't address what I'm bringing up here (see "Third Example" below).]

Please somebody show me. :^)

Thanks!
(Details below.)
William Reilly

Jeni Tennison's Beginning XSLT book Code Download:
http://jenitennison.com/xslt/5946.zip

FIRST EXAMPLE (of 3)
"Apply-Templates"

Beginning XSLT, p. 386 ff.
=== TVGuide10.xsl ======
<xsl:template match="TVGuide">
<xsl:apply-templates
select="Channel/Program
[generate-id() =
generate-id(key('programsByHour',
substring(Start, 12, 2))[1])]" />

=== Same File, My Edit to Remove '[1]'======
<xsl:template match="TVGuide">
<xsl:apply-templates
select="Channel/Program
[generate-id() =
generate-id(key('programsByHour',
substring(Start, 12, 2)))]" />

=== FileCompare the Two Outputs ("no diff") ======
C:\XSLT-Beginning\Chapter10>fc XSLOutput_10.html XSLOutput_10_no1.html
Comparing files XSLOutput_10.html and XSLOUTPUT_10_NO1.HTML
FC: no differences encountered
=========

SECOND EXAMPLE (of 3)
"For-Each"

http://cscie153.dce.harvard.edu/lecture_notes/7/sl ide14.html

http://cscie153.dce.harvard.edu/lecture_notes/7/ex amples3/group3.xsl
=== group3.xsl ======
<xsl:template name="members_by_state">
<xsl:for-each select="/ushousemembers/member
[generate-id() = generate-id(
key('membersbystate',state)[1]
)
]">
=========

=== Same File, My Edit to Remove '[1]' ======
<xsl:template name="members_by_state">
<xsl:for-each select="/ushousemembers/member
[generate-id() = generate-id(
key('membersbystate',state)
)
]">
==========

=== Unix "diff" on the Two Outputs ("no diff") ===
======
$ diff out_group.html out_group_no1.html
$
=========

THIRD EXAMPLE (of 3)
"xsl:if"

Beginning XSLT, p. 385
=== TVGuide9.xsl ======
<xsl:key name="programsBySeries" match="Program" use="Series" />

<xsl:template match="Program/Series" mode="EpisodeList">
<xsl:variable name="episodes" select="key('programsBySeries', .)" />
<xsl:variable name="firstEpisode" select="$episodes[1]" />
<xsl:if test="generate-id() =
****************************************
generate-id($firstEpisode)">
****************************************
<div>
<h3><a name="{.}" id="{.}"><xsl:value-of select="." /></a></h3>
<h4>Episodes</h4>
<ul>
<xsl:for-each select="$episodes">
===============================

=== Same File, My Edit to Change $firstEpisode to $episodes ========
<xsl:key name="programsBySeries" match="Program" use="Series" />

<xsl:template match="Program/Series" mode="EpisodeList">
<xsl:variable name="episodes" select="key('programsBySeries', .)" />
<xsl:variable name="firstEpisode" select="$episodes[1]" />
<xsl:if test="generate-id() =
****************************************
generate-id($episodes)">
****************************************
<div>
<h3><a name="{.}" id="{.}"><xsl:value-of select="." /></a></h3>
<h4>Episodes</h4>
<ul>
<xsl:for-each select="$episodes">
===========================

=== FileCompare the Two Outputs ("no diff") ======
C:\XSLT-Beginning\5946\Chapter10>fc XSLOutput_09_firstepisode.html XSLOutput_09_
episodes.html
Comparing files XSLOutput_09_firstepisode.html and XSLOUTPUT_09_EPISODES.HTML
FC: no differences encountered
===================================

"Extra Credit" Example ;^)
Same as Third Example, but One Small Edit:
Tried _Second_ Episode '[2]' vs. First.
(Still No Diff!)
=== TVGuide9.xsl ======
<xsl:key name="programsBySeries" match="Program" use="Series" />

<xsl:template match="Program/Series" mode="EpisodeList">
<xsl:variable name="episodes" select="key('programsBySeries', .)" />
<xsl:variable name="secondEpisode" select="$episodes[2]" />
<xsl:if test="generate-id() =
****************************************
generate-id($secondEpisode)">
****************************************
<div>
<h3><a name="{.}" id="{.}"><xsl:value-of select="." /></a></h3>
<h4>Episodes</h4>
<ul>
<xsl:for-each select="$episodes">
===============================

=== FileCompare the Two Outputs ("no diff") ======
C:\XSLT-Beginning\5946\Chapter10>fc XSLOutput_09_secondepisode.html XSLOutput_09
_episodes.html
Comparing files XSLOutput_09_secondepisode.html and XSLOUTPUT_09_EPISODES.HTML
FC: no differences encountered
===================================</ul>< /div></ul></div></ul></div >

FAVICON. ('Bout Time You Got Here)

Just figured out (been seeing these guys _quite_ a while now) how to get that tiny "/favicon.ico" into the Location bar.
_Quite_ exciting.
I think (?).

Of course, the 16x16 pixel, 16-colors-only icon I was able to put together is rather "clastic," as it were, a.k.a. presentationally challenged.
(icon-o-clastic, get it? oh oh, v. bad joke/pun).

Resources:

• Favicon.ico Info Page at "The Site Wizard."
  
• Free tool (Windows) From Axiomx.

PHP: O'Reilly Missed The Boat. This Reilly, Too?

This article on the lousy XML "Book Business" mentions a Tim O'Reilly comment vis-a-vis PHP vs. Perl. It's for this Reilly (William) yet another clue as to why a look into PHP may well be in order. Hmmm. Sounds good.
Time to catch a boat (one among several boats).
Wm.

XML.com: The XML Book Business [Oct. 29, 2003]

Hi Martin, (my e-mail to Martin)

Hmmm. Another quote encountered telling me there's something interesting going on with PHP, again, something I'd kind of totally ignored. Hmmmm...
Maybe I oughta pursue that chapter on putting PHP together with MySQL from that book I'd borrowed from the library, after all... (Need to find the time!)

Cheers,
William
P.S. You'll find this e-mail up in my blog :^)

Quote:
"in fact, Tim O'Reilly seemed to admit at a FOO Camp presentation that O'Reilly had missed the PHP boat, something others have been saying too."

"The XML Book Business"
http://www.xml.com/pub/a/2003/10/29/deviant.html

Longer excerpt:
"Indeed, the XML book business is fairly grim across the board, not just for O'Reilly.

That's an important point to think clearly about in my view. After all, O'Reilly as an organization is far from omniscient; ...>> in fact, Tim O'Reilly seemed to admit at a FOO Camp presentation that O'Reilly had missed the PHP boat, something others have been saying too.<<... In that case, of course, it was a tricky thing to see, especially for O'Reilly, which has put a ton of effort behind Perl. It particularly tried to pitch Perl as something other than a language for writing cgi-bin Web applications. That must have made PHP's success as a kind of Perl++ in that particular space hard to see coming. One of the most common ways in which people and organizations err is zigging when a zag would have worked better.

So [anyway] the first point is that everyone's XML book business is in the dumper. ..."

A Few Thoughts as I Get Close to Finishing This (Computer) Book...

I've just about finished reading:
Designing Data-Intensive Web Applications
which goes along with a methodology:
WebML ("UML for the Web")
and a CASE tool (which I've got a 9 month "academic" license to work with):
WebRatio

Herewith, just a couple of (very positive!) thoughts about it (as I penned them in an e-mail today)...

Believe it or not, this system actually provides me a tool (and a methodology, a way of approaching the problem) to build a Struts (or, for that matter, .NET) website. Fairly amazing. I hope to use it to actually build things, but more importantly gain better understanding about how Struts, MVC(2), etc. work.

As for the WebML methodology, it is pretty interesting. The web geek software dev guru
guy who wrote the preface for their book (Adam Bosworth, chief architect at
BEA WebLogic, and inventor of MS-Access, Quattro Pro database, etc.),
described what these Italian computer science professors and database "big
brains" types have come up with as "something very nearly as simple and elegant as
the relational calculus."

---- (Let's use the (new today!!!) Amazon "Search Inside The Book" feature, to give you a link right to the page I'm talking about (and let's use tinyurl.com, to make it more palatable/clickable)):
---- http://tinyurl.com/2fdqk (See the Page in the Book!)
Also at: http://snipurl.com/webml_book

Now, I scarcely know what "the relational calculus" really means, actually, but, I do know when something's supposed to be "simple and elegant," so that works for me.

At any rate, even I, non-Adam Bosworth that I am, can appreciate how this books' contents just flow: the ideas stack up and just keep making sense, chapter after chapter. You get a feeling you know where to put things, even pretty abstract things. Nice.

With the tool, just following along their tutorial, I'm able to create what is under the hood a somewhat involved Struts application (though it's older Struts - don't know the number exactly, 1.1 (?)). Interesting. (You can do .NET too.)

Don't just take my word for it. Here's another (glowing) opinion, seen on the WebML "Community" board: THIS book is probably one of the BEST technical books I've ever picked up... (and it gets better... ;^)

What I don't yet understand is why there are NO Reader Reviews up on Amazon for this title. Hmmmm. And in bookstores, I find it sort of mis-shelved, over with books with the word "Design" in them (near CSS and site architecture and such). So, I hope it's getting around and getting known. May have to log in to Amazon and put up a review of my own, when I'm better prepared.

From the WebRatio product page:
"COMPLEXITY MADE SIMPLE"
"THE UML OF THE WEB"
"DEVELOPMENT PROCESS RESHAPED"
"SIMPLICITY NOT SIMPLIFICATION"
and finally
"SIMPLICITY IS COMPLEXITY RESOLVED.
TO ACHIEVE SIMPLICITY WE MASTERED COMPLEXITY"

You can tell they're pretty excited. But, they have reason to be!

More to follow.
Wm.

-----------------
fwiw: here's that tiny URL in all its non-tiny glory:
http://www.amazon.com/gp/reader/1558608435/ref=sib_vae_pg_6?%5Fencoding=UTF8&keywords=relational%20calculus&p=S008&twc=1&checkSum=kvwd00gcHud0ts%2FwazcCG%2FCxsldcKHYBLYtQDZloBSo%3D#reader-link See The Page Inside The Book

Mid-Morning Break Blog Entry: Info Architecture...

O.K., it's coffee break number 1 here at the User Interface Conference here in Kendall Square, Cambridge, Mass. (I'm on a Press Pass, for CMS Review)
http://www.uiconf.com/8/fullday_merholz_young.html

From Construct to Structure:
Information Architecture from Mental Models
Peter Merholz & Indi Young, Adaptive Path

Kernel Idea being presented here is a "Two-Parter":

1) "Content Model"
Basically, the bottom-up model is your stuff (content), which is the stuff you
already know, work with, hang out with, and which you are usually doing kinda
badly re: putting up on the web (organize website by company org chart, etc.
etc.)

2) "Mental Model"
The new bit here is the top-down model, which stops pretty much at the top nav
of the site (O.K., maybe nav2 level), and is all about _not_ the web but what
(the h$%&#) the user is actually thinking as they try to "get something done."

In looking at some of the morning's slides showing this dichotomy I was kind of
thinking back to Digitas days, as, "where John Robinson (MModel) meets William Reilly (CModel)." John (Copy, Info Arch, Content guy)
would prob. disagree (figures he's doing both). Maybe he is.

Anyway, kind of interesting. Glad it's free though ($800/day around here. Yow.)
Wm.

MS-Outlook, Office Bashing (rants by others)

Scott Johnson blogged a useful link to Chris Pirillo on "Office 2003" and his, ahem, er, "singular expression of unhappiness" with it.

Turns out Pirillo was going on about Outlook, but, I'd say either way it's worth some ink and reading time to better dissect these new offerings from Redmond.

Chris P. didn't like Outlook's interface (at all. - "gack").

For my part, I am mostly interested in knowing just how good/bad is the XML support in Office, and have to admit to enjoying some bit of schadenfreude reading Taking the Pulse of XML Editing on xml.com (Oct. 1, 2003):

" Office 2003's much-reported XML support generates less fear on the part of vendors and...less interest on the part of managers."

"...every vendor who showed XML output from its tool showed XML that was vastly cleaner and more comprehensible than any XML output I've seen from Office 2003."

I'd seen something similar-ish (though not as strongly stated) from PCMag recently.

Here's simply hoping that the big gorilla "safe choice" tool (for IT, that is) doesn't become a de facto choice for XML authoring going forward.

-- Admittedly, it might be nice if things have improved to the extent that you could design a system that could take advantage of whatever MS-Word 2003 can produce for reasonably clean XML, for your more casual contributors.

But it would not be nice to learn that corporate America might be buying into the hype and coming to regard Office 2003 as the automatic choice ("nobody ever got fired choosing...") for the ubiquitous authoring tool for content in XML. Harrumph.

Wm.

CSS + XHTML with Eric Meyer: The "Real Deal"

Couple quick notes from highly useful daylong seminar.
Eric & Holly took a page all the way down and built it all the way back up, using the modern tools of XHTML 1.1 and CSS (1). Good time had by all!

Roadmap to Markup Success: Understanding Structure and Style - Fullday Seminar - Molly Holzschlag & Eric Meyer

Unfortunately, the key 21 pages of good stuff going over every detail of the original HTML page, the CSS, and the re-built XHTML page are available only on paper, not electronic.

At any rate, highly instructive to see the innards of the "transitional" re-make--a task awaiting many a website...

=== TIP ============
Nifty Presentation Layer Developer's Toolbar
http://www.placenamehere.com/pnhtoolbar/screengrab.png
Mozilla + Firebird + Netscape, all platforms
"XUL + JS"
April 2003 last updated

Here, links to actual code from earlier Eric Meyer presentation:
=== CODE ====================================
CSS for Navigation (in lieu of Javascript...)
Seybold S.F., Sept.
Slides:
http://complexspiral.com/events/archive/2003/seybold/cssnav.html
Resources: CSS, HTML:
http://complexspiral.com/events/archive/2003/seybold/cssnav/
Plain, but he builds up from here...
http://complexspiral.com/events/archive/2003/seybold/cssnav/list-ex.html
Kinda out there, not all working...
http://www.meyerweb.com/eric/css/edge/menus/demo.html

Other examples he cites:
http://www.serve.com/apg/workshop/cssMenu.html
http://www.surguy.net/menu/index.html
=======================================

Linux on the (notebook) desktop: joys or noise?

My first "Trackback" (!) :: (like, "I'm a first-time caller..." to your radio show)
-----------------------------------------------
Hey Gregor,

I liked your entry critiquing this 180+Kb of joys/noise(?) on "moving from Linux to Windows." That little piece sure generated a lot of Reply to This, whew.

Gregor J. Rothfuss :: Imagination is key to your dreams coming true

You and I had talked about this some.

My case? Windows user since long time; growing use of *nix last few years. But about four months ago, moved entirely from Windows to Linux on the (notebook) desktop, just for the Big Experiment of it all.
For what it's worth, I administer my site's hosting on FreeBSD, I ssh in to sundry *nix school/work systems, and at home it's been two dual-boot machines, (notebook + tower), so I can get as much Linux admin + user exposure as possible.
But it's been the total desktop user shiftover that has been decidely useful for my *nix skills, though yes there are desktop things that work better/easier back in the Windows side. I could see myself heading back there sometime (?), but, I couldn't tell you when. Probably still a _good_ long ways off... :^)

Anyway, to get here, I do have to admit I did go the easy route, and got the Linux guy at PCsForEveryone to install the dual-boot onto the notebook. Things looked a little too hairy on linux-on-laptops.com (and a lot of the contibutions being in German!). Red Hat install to desktop machines, O.K., but the notebook changes things.
In fact, even with the help, turns out:
1) I can't run WiFi native Centrino on the Linux side (no open code from Intel for Linux team to program to, I'm told); have to buy a PCMCIA card, etc., and even that's tricky, getting the right one, etc., I'm told.
2) The sound board is unhappy (no go) on the Linux side (that isn't killing me).
3) Battery management is there, sort of?, on Linux (dead easy automatic in Windows). Linux guy says I can "look it up on the web" how to get the (visible) icon to (actually) do something. (don't worry, I'm sort of chuckling at this -- what the hell, there is a hardware orange "idiot light" for me to see before I lose all my exciting lengthy blog entries before the power goes out on me kaput!)
4) The modem won't go on the Linux side at all, I'm told (that doesn't matter at all).

Random:
Slightly embarrassing, but took more doing than it ever should have to just get from installed 800x600 to 1024x768 resolution. That's 2 seconds in Windows.
The slightly deep voodoo of "how do I switch Workspaces without using the mouse?" also called for someone to show me where to go do it (then, it was easy, granted, but still).

Took me an also slightly embarrassing number of days to wean all off of MS PIM-ish tools onto the KDE set: e-mail; address book; organizer; notes; talk to my Palm Pilot, etc. etc. (and the Palm-2-Linux infrared is still a mystery, though it did sort of get half-way there? (useful, that)). The world of "documentation" on these sorts of things is still, ahem, a bit, eh, thin.

I finally got away from paying MS-Office (ouch, once) and use OpenOffice (which yes, I could be/should be/will be! using on Windows side too). Besides, I'm far too interested in OOs XML; and, who can afford to re-pay for Office-11, even if it does come with mo' better XML? (the which I'm reading is beginning to look less compelling, according to some in the know (some recent PCMag piece)).

And yes your lineup of Mozilla, Eclipse etc. is in line with things I do run on my Windows side (which I also do log in and run at times). Hooray for open source on the Microsoft platform. Of course.

Something Missing:
One nice (expensive; someone else paid for) tool on Windows only: Altova's XML Spy. Eclipse's XML Buddy hasn't been updated in a long while; I have to follow some links to others (off I.B.M. developerWorks, I believe). But XML Spy is quite strong and I've used it to good effect. (Likewise Corel's XMetal I've got; Win32 only. Ah well.)

As for the intellectual honesty and the crap filter and all, yep, I'd say there's a lot of knee-jerk-ness out there to be had, no question.
Thanks for your bit, on all the noise!

[Finally - I'd like to look further into that Mindmanager you mention. Interesting...]

How To Find A Cable Car In The Fog: Category-Based Navigation

Something else I found (in addition to the Center in a Box I'm now working on!), in my recent searching at Berkeley: "Category-Based Navigation" -- drilling down through the faceted data, in this case describing photographs of "life at Berkeley & environs."

(or "dim light" anyway...)

When I recently went looking for interesting doings at my grad school alma mater, U.C. Berkeley's School of Information Management and Systems (SIMS), I found not only Professor Glushko's "Document Engineering" initiative (with whom I've made contact, and will help with beta-testing one of their products soon), but also this quite interesting, Endeca-like (I'd seen their dog & pony show about a year ago?) drill-down navigation and search tool.

I find this really interesting, and would love to someday work on the metadata design for this sort of approach.

(See final link for your cable car in the fog...)

Wm.

http://bailando.sims.berkeley.edu/flamenco.html
= BAILANDO ("dance" in Spanish), or "Better Access to Information using
Language Analysis and New Displays and Organizations"
= FLAMENCO ("pink bird" in Spanish ;^) stands for "FLexible information Access
using MEtadata in Novel COmbinations"
(Python, MySQL & Lucene)

Commercial Implementations
"Several commercial implementations of ideas related to those we've
explored in Flamenco are now available. These include Endeca, Siderean
(formerly bpallen's teapot system), and DieselPoint."

SIMS 202: Information Organization and Retrieval
http://www.sims.berkeley.edu/academics/courses/is202/f02/PhotoProject.html
The Photo Project
"...outcome of this project will be a sensible and usable classification to
structure and describe photos, and a browser application to look at the
photos within that classificatory structure..."

Example drill-down search URL:
http://flamenco.sims.berkeley.edu/cgi-bin/flamenco/202photos/Flamenco?q=conditions:3/location:35&group=location
Conditions = "Lighting | Brightness | Dim"
Location = "Geographic Area | U.S. | California | Bay Area | San Francisco County"
Results: 7 pictures (one including cable car in the fog...)
====================================

MapQuest 3 Pointer: 44 Cedar Street in 2 Nearby Towns

Always entertaining to come upon ambiguities in information exchange.

A friend e-mailed saying we're "practically neighbors" when MapQuest helped him to see where I live (47 Kingston, Somerville (near Davis Square)). He let me know he lives at "44 Cedar Street, near Porter Square." And so I promptly turned to the same online tool to see where that might be, and typed in "44 Cedar Street, Cambridge, MA", and indeed found a very nearby address! But it wasn't really near Porter Square, unless you gave that a quite wide interpretation... Hmmmm.

Ah-hah, I wonder is it good old 44 Cedar Street, Somerville, MA? Ah, yes, clickity-click, type-type, and there it is, another nearby address (and near Porter Square). Interesting.

It was then I discovered that MapQuest can do something I'd never asked it to before: "Add Location":
You can put a bunch of addresses on one map, label them, assign icons (like pins, stars, and so forth), and get an enormous, quite custom, URL that you can then zing around in e-mail, blogs, and browsers to share the joy.
Fun!
================
My e-mail back to Lars:
MapQuest: Maps: map

I'm enclosing the enormous URL to a funny custom map (with not one but three
positions) I drew using MapQuest:
- my house
- your house
- and your ghostly Cambridge cousin's house
[that is, the "44 Cedar Street" in that other nearby, "Somerville wannabe"
town ;^) ]

=============
http://www.mapquest.com/maps/map.adp?mapdata=EifUonSP7c3NebxVtmnURk6gd9UfB7oWLkAmxk5YqGYn1XOtbF3ImklA8DcrGdnaAPAezR9hRk6Ar1eyGlQm2uCD%2fhiBqoXEJzZS1%2fnPeXY9wntTEmBQSuHLkragdC2wl1hY1j5Z0P8%2bb4NVxANJfybniK6liEFZGSOo%2f9rL%2fANrRLixeRKDp1NMEhzlgH4rmX566zOuExdr6ZmXlcJ1aK1LAqRwrLUz4EHiRH%2fRCjvGmPLTwkpIPJw3mienpHT4T5IDTThmk0%2fPSqv1Tviz%2fscKC3K8xcPWc%2fkKeYwU8V5ldNEir42Uifu6rjpcBMwBPeMM0lDMskjIP4Qm8NZrmFmAX1KeJ8aTpNopGzEiyRCnEfw3FdRTPmtZ1Knte1Kq%2f1m2ID%2fmDbyeKKYgCUFzRqGQ2DN0XIT1tavUnF1w7cnDIjeBpWK5KepkVYUnGfABhh0i7TwfGPRFnjUTEf%2fIwyfqCRYPmttadqNryWLG4LEA153ai6UUhBaa3%2bPClHRB%2bI%2fumBzBC0NvbLCCONH8XmJCSXLANiTCdk35yZvJ9T5DQy7nPviiEiAZuG34HDZFUvqWWarGOPUVhRclxMD%2bjiuqxZzUTun8JWAaZ67fE%2fJC3PsS8hJYJ1sSj9fCB2r%2bqx1%2bIUpRpsmUjQGMUxbT%2fZR2jyKAzAhY5IyYb7faBy%2fLVJDCSDPsQYzmVBxSQQgCR8cIRB85AMGlrqTh2fR2rddWjSJBLyfVEVyW0hp%2f1%2bipNBtgfofDrmktqaT9Wy7a%2fMM5FzCCKHsdPPGYk1a5j1YRt5u5ZKARJyaQNAHfiyCC6crIHd%2fjvF7KZZzj6yBbqKBtjl3bFSrhWGuHHJmHtDziIhj9z%2bqLZWK3SI012Wzfdb5mP8PCh1a4vSD25%2fw7dHDGBjUsklGL7m0cC%2fa%2bHBxUwr5BetQ3JcErDTlbRfuHwLJQHn6Dtv%2fX2BapSLKM0mMPuMN9jYA%3d
=============

It's funny because though you did mention "near Porter," you didn't say which city, and I first typed in Cambridge (!). Interestingly (as you'll see on the map, if you can get your browser to swallow the whole enormous long URL -- you may have to manually chop out any linebreaks created by e-mail sending & receiving), the 44 Cedar in Cambridge is even more of a "practically neighbors" location...

European Commission v. Open Source

Clicking this morning over to the homepage for Apache Cocoon Lenya (a Swiss-initiated, open source concern), my websurf was interrupted for 60 seconds by this announcment (read below):

[Gist: Patents v. Open; sign a petition (they say they've got quarter million); this news is already a little late ("August 27").]
================================
Update: More patent news (Microsoft's IE use of ActiveX):
Setback for Microsoft Ripples Through the World Wide Web

"The court ruling and its potential impact, according to Mr. Weitzner, points to the larger issue of the need to keep the basic software of the Web free of patent royalties."
``If you try to charge individual companies for patents on Web standards, you risk balkanizing the Web and breaking it,''

Plaintiff: Michael Doyle, the founder of Eolas Technologies in Chicago, while he was an adjunct professor at the University of California at San Francisco. Wants $1.47 a pop for Windows licenses sold...

Consortium of firms working to minimize damage to web browsing experience: Real Networks, Sun, Apple, Adobe, Macromedia, Microsoft.
========================================

Important notice

On September 1st the European Commission is going to vote a revised
version of the European Patentability rules. The proposed revision
contains a set of serious challenges to Open Source development since
regulation regarding software patents will be broadly extended and
might forbid independent development of innovative (Open Source and
not) software-based solutions.

The European Open Source community is very concerned about the upcoming
new regulation and has organized a demo protest for August 27, asking
Open Source supporting sites to change their home pages to let everyone
know what is going on at the European Parliament.
[ The Apache Software Foundation has decided to support this initiative, and this is why you
are seeing this page. ]

For further information please see
http://swpat.ffii.org and
http://petition.eurolinux.org.
[ You will be redirected automatically to the Cocoon homepage in 60 seconds
(or continue on to cocoon.apache.org/lenya/index2.html). ]

San Francisco, Content Management

San Francisco this week, for Content Management extravaganza:
Gilbane Conference on Content Management at Seybold San Francisco 2003 ("at-a-glance").

Logistics information below (hotel; phone numbers; etc.)

Some Shared Logistics Info:

CONTACT
Cell Phone = (617) 290-9689
wreilly@cmsreview.com
http://www.cmsreview.com/Editors/
william@reilly2001.info
http://reilly2001.info

AIR
I'll be getting in late Monday morning (6:00 A.M. out of Logan (!)), and
departing Thursday evening v. late (11:45 P.M.).

HOTEL
Staying at the Canterbury Hotel on Sutter Street
http://www.canterbury-hotel.com/
750 Sutter Street * San Francisco, CA 94109 * 415-474-6464 * Guest Fax
415-474-5856 * 800-227-4788

SCHEDULE
Tuesday evening I've an appt. with former U.C. Berkeley professor and a couple
of his students; otherwise (so far) free...
http://cde.berkeley.edu/
http://sims.berkeley.edu/
Tues. eve. event:
http://www.sdforum.org/p/calEvent.asp?CID=1151&mo=9&yr=2003

SSI 'XBitHack on' (WAS) messing up my CGI

Well, this is FIXED now, but I thought I'd document the hassle
(while I'm still just chortling away at how much fun it's been)
===========================
Draft (not sent) post to "Tech Help!" at my hosting svc. (hub.org)

Dear Sirs,

I'm having success at getting two things to work (good), BUT I am not having success at getting them both to work at the same time (not so good).

These be SSI and CGI.

Gory annotation notes in my Apache configuration file...

/usr/local/etc/apache/virtual_host/www.reilly2001.info.conf
<VirtualHost>
# WR_ 20030828. As noted at TOP of httpd.conf file...
# Adding SSI to root (home) dir.
# RH7.3 Bible, p. 805 etc.
# (The intent is _not_ to have SSI on all sub-directories!) (Here's hoping!)
# This is so Movable Type 'recent.html' can be SSI'-ed into 'index.html'

## WR_ 20030829 SUMMARY
# O.K., I had trouble trying to get SSI to work, without
# it also taking down my CGI (!).
# That is, for SSI, I had 'Options Includes' for
# the home (web root) dir, and then (to avoid the
# dreaded ".shtml" way) I set 'XBitHack' to 'on'.
# O.K., SSI kicked in (on the single page I needed: /index.html)
# but my Movable Type CGI complained:
# "Forbidden, you don't have permission to access..."
# (All CGI, actually. /cgi-bin/wiki/wiki.cgi did the same)
# I Googled a bit to see about 'XBitHack' affecting CGI, "Forbidden" etc.
# Best I found were these, but they still didn't address my need.
# http://httpd.apache.org/docs/howto/ssi.html.html#configuringyourservertoperm\
itssi
# http://www.apacheweek.com/features/ssi
# Then, the Apache FAQ has:
# http://www.surreyinstitute.org.uk/manual/misc/FAQ.html#forbidden
# which can only offer the general help of:
# "The Apache configuration has some access restrictions in place which forbid\ access to the files"
# Still, I realized it was prob. going to be this sort of .htaccess thing.
# So, I followed some ideas I'd seen in Apache books at the QuantumBooks store\ today...
# 1. Working only on my <Virtual_Host> file (rather than httpd.conf) was a goo\d idea... (I'd sorta figured that out...)
# 2. Understanding that yes, these <Directory> directives were for all sub-dir\s was important (!)
# 3. So putting a more exclusive <Files> directive, within the broader
# <Directory> directive seemed a good idea.
# 4. Finally, some mucking about with trying XBitHack on, then commented out, \then back on again, finally led me to getting it all to work.
# Hallelujah.
<Directory "/usr/local/www/reilly2001.info/www">
<Files "index.html">
Options Includes
</Files>
</Directory>
XBitHack on
</VirtualHost>

"Is Knowing XML Worth Anything Anymore?"

XML.com: Nobody Asked Me, But... [Aug. 27, 2003]

This article rings true, as I'm busy (re-)sizing up which piece(s) of the grand world of XML are for me (these days).

I've long now said (two years, maybe?), XML is like ASCII. Saying, "we do XML" is like saying, "our product does ASCII." Yes, of course it does; nobody asks that anymore! The question is _what_ are you doing with XML? (and the legitimate replies can be myriad; hence the "grand world...")

Wm.

by John Simpson

"A few years ago, I crammed my head with everything XML-related that I could find; it all used to fit -- if not in my head, at least between the covers of a 300- or 400-page book."

"At work nobody knew anything about XML, and non-XML project priorities couldn't wait." [...so I was a leading XML generalist...!]

"It's true that there are people -- many if not all of them participants on the XML-DEV list -- who are apparently profoundly smart about nearly everything going on in the XML world; many of them grew up with and are still fond of SGML, Smalltalk, Lisp, and awk, yet are equally at home discussing Java, XML Schema, Unicode, web services, Perl, the Campaign for Real Ale, and the handling characteristics of the Segway. Face it, though: you are not one of those people" (emphasis added)

"There's just too much going on, and it's going on in too many places at once, for you to get it all."

"Select a specialty. Go deep instead of broad."

Blocklist Blocks Self (and all of us!)

My e-mail to cms-list participant and "listmom"
========================================
Subj: Flakey Internet... Boston Globe article today on "blocklists"

Hi Chris (and greetings to listmom, whom I'd also written about this),

Thanks for your note back.

On Wednesday 27 August 2003 07:39 pm, Chris Harrington wrote:
> Yes, I got both of your emails. The Internet has been overall pretty
> flakey the past week.

This error message is what I got yesterday, trying to post to the list, and I noticed then the domain "osirusoft.com".

=== Undelivered Mail Returned to Sender ===================
<cms-list@cms-list.org%gt;: host ashiya.cms-list.org[64.81.246.22] said: 554
Service unavailable; [64.117.224.150] blocked using
dialups.relays.osirusoft.com (in reply to RCPT TO command)
=================================================

Then at my morning breakfast table, the Boston Globe spilled the beans on what'd happened (!). (Is this what they mean by "convergence"? ;^0) )

http://www.boston.com/news/nation/articles/2003/08/28/saboteurs_hit_spams_blockers/

=== Osirusoft.com excerpt ===============================
"The attackers have managed to drive one popular blocklist entirely offline. On Tuesday, Californian Joe Jared shut down his Osirusoft blocklist in an unexpected manner. Jared blocklisted all Internet addresses worldwide. As a result, businesses that relied on his list were suddenly unable to receive any e-mail at all, even legitimate e-mail.

"He said . . . I'm going to blacklist the world. And by golly, he did," said Jim Miller, network administrator at Simutronics Corp., a St. Charles, Mo., firm that formerly used the Osirusoft blocklist.

Jared expressed regret for the way he shut down his blocklist. "I thought there had to be a better way to do it," Jared said. "But there wasn't."

Jared said his blocklist server also hosted the website for his small business, which makes shoe inserts for people with foot problems. He couldn't shut down the blocklist server without also closing his business website, so he chose to make the blocklist unusable by blocking everything.

He said he'd spent weeks trying to fend off the denial of service attacks against his servers, but "they just beat the hell out of them . . . I just can't be attacked like that."

Jared isn't sure he'll ever run a blocklist again. "What I am going to do is take a vacation," he said. "I need one.""
=== /Osirusoft.com excerpt ===============================

william.reilly.name the 1st

Dear Hiawatha Bray (Boston Globe technology writer),

Subj: william@reilly.name -- What happens to the other William Reillys??

I today came upon the "firstname.lastname.name" phenomenon at http://norman.walsh.name (DocBook guru), and a Google-click or two later was at various ".name" registrars happy to tell me my name was still available (for $25 - 60...).

My question: sounds like a nice new appealing way to simplify web addresses, etc., but it is (obviously!) a land rush for the John Smiths (and William Reillys) of the world (maybe less so for Hiawathas! :^) )

> What, if anything, is/was the plan for how this would really play out??
ICANN or whomever runs this hasn't exactly made it clear what the 2nd (and 200th) William Reilly is supposed to do, to join this new world.
Are we just going to HotMail it? (William99.Reilly.name ?) I suppose so.

Anyway, I gave up trying to find a decent news article or opinion piece on this, because Google just leads me to the registrars, largely.

Do you have info on this? Interested in it as a story? Or maybe just a one-liner e-mail reply... (I know you are busy.)

THANK YOU for your time,
Best wishes,
William Reilly
william@reilly2001.info
-- as you can see, (back in 2001) I got an '.info' domain, still pretty unusual...

"Metadata Wins Again" (Norm Walsh)

Following some series of links from Simon St. Laurent's site (http://www.simonstl.com), I've been spending a few moments at Norm Walsh's new (to me) (ca. this spring, looks like?) "pseudo-blog" website, which is full of RDF and interesting ideas (par for the course for Norm "DocBook" Walsh).

This one I thought worth clipping:
"Metadata wins again"
http://norman.walsh.name/2003/08/21/metadata

"On more than one occasion, I've been saved by the fact that I can influence the publication through no more than the addition of metadata"

-- gives one example of being pleasantly surprised by auto-magically getting the lone .GIF on his site to be handled/processed by the metadata he thought he'd set up for .JPGs description only...

-- in another example "he [realizes that the] RSS feed, like everything else, just comes from the metadata!" and for the price of a quick snippet of RDF he gets the RSS listing he wants, without the onus of having to create a (real, non-metadata) placeholder article.

Concluding remark:
"This metadata stuff, it's got legs."
+++++++++++++++++++++++++

Wm.
P.S. "On the other hand..."
Then again I have to note Simon St. Laurent's just-the-other-day "rant" on "Why I Don't Like RDF"
http://simonstl.com/articles/RDF.html (Aug. 2003)
"URIs identify resources, which are sort of whatever you want or I want or whatever URI owners want if they can be bothered to communicate it. The understandings of the division of labor between server and browser regarding things like fragment identifiers have been tossed out in favor of nonsense about magic hashes (#) signalling a difference between representation and abstract ideas."

[re: XSLT as a help] "The XML world has frequently deluded itself into thinking that agreement-by-committee can solve its communications problems, but at the same time it has hedged its bets with transformation practices that let developers get from vocabulary A to vocabulary B just in case the big-picture designs [a la RDF] aren't quite what was needed."

"RDF has managed to inflict its assumptions on XML enough (largely through namespaces) that there are periodic efforts to convince XML users to put themselves in the RDF straitjacket. Rather than merely accepting the constraints of trees, these people suggest, developers would be wise to subject their content to the constraints of trees and graphs simultaneously"

He does link to his own (older) "What's right with RDF"
http://www.oreillynet.com/pub/wlg/2322 (Nov. 2002)
but that has some qualified statements of interest, too.
"RDF is excellent at addressing a particular set of problems. The Resource Description Framework's primary approach is description. XML often presents something (a document, a table) directly; RDF more typically presents a description of something, not the thing itself"