Atom 1.0 Sample

<?xml version='1.0' encoding='UTF-8'?>
<feed xmlns=''
 <link href='./' />
 <link rel='self' href='' />
 <author><name>Tim Bray</name></author>
 <subtitle>ongoing fragmented essay by Tim Bray</subtitle>
 <rights>All content written by Tim Bray and photos by Tim Bray Copyright Tim Bray, some rights reserved, see /ongoing/misc/Copyright</rights>
 <generator uri='misc/Colophon'>Generated from XML source code using Perl, Expat, XML::Parser, Emacs, Mysql, and ImageMagick.  Industrial strength technology, baby.</generator>
 <entry xml:base='When/200x/2005/11/30/'>
  <title>On &#x201c;Beyond Java&#x201d;</title>
  <link href='On-Beyond-Java' />
  <category scheme='' term='Technology/Java' />
<category scheme='' term='Technology' />
<category scheme='' term='Java' />
<content type='xhtml'><div xmlns=''>
<p>I just got around to reading Bruce Tate’s
<a href="">Beyond Java</a>.  
I think that the
senior people in the Java groups at Sun, and all the other Java
powers, should read and think about it (and for that matter the CLR people
over at Microsoft).
The premise of the book is really nothing new: There are a lot
of problems out there for which smart, senior people are reporting that there
are languages and/or frameworks that produce solutions quicker and better than
Java.  <cite>Beyond Java</cite> assembles a lot of this testimony, 
claims that we’re at an inflection point, and goes on
to speculate about what comes next.  It argues interestingly by looking back
at the history of Java’s explosive rise from nowhere to world domination; but
at the end of the day I’m not sure the historical analogies are useful.
In other gripes, the book’s structure is a little messy, and
the kayaking anecdotes that introduce each chapter could have been
dropped without loss of value.
I also disagree with Tate’s argument that Swing and SWT are
useless, part of the problem 
not the solution. I hear loud complaints about <em>every</em> GUI-builder;
somewhat fewer about those in OS X; anyhow, as far as I know
none of the beyond-Java alternatives are rich-user-interface
champs. <i>[Thanks to Sam Ruby for pointing out that I’d misread Tate’s
argument on this, first time around.]</i>
Still, it’s a solid piece of work; see also
<a href="">Sam Ruby’s
take</a> and the discussion
<a href="">over at</a>.
The book got me thinking about two great big important complicated issues: the
future of 
the JVM, and the right way to build Web applications; but each of those deserves
its own essay.</p>
 <entry xml:base='When/200x/2005/11/30/'>
  <title>Office Politics and Profits</title>
  <link href='Office-Profits' />
  <category scheme='' term='Business/Microsoft' />
<category scheme='' term='Business' />
<category scheme='' term='Microsoft' />
<content type='xhtml'><div xmlns=''>
<p>In recent weeks I’ve been spending quite a bit of time talking to
journalists and analysts about the issues around office-document XML file
formats in general, and the Massachusetts dust-up in particular.
There’s one exchange that pops up in almost every one of these conversations,
and it goes something like this. <i>Journo</i>: “Now, you guys are taking all
these idealistic high-minded positions, but you know and I know that what we
have here is a battle for market share.”  <i>Tim</i>:
“That’s part of it, but we
think that our interests, and the customers’, are both best-served when
there’s no file-format lock-in and there’s a wide-open competitive market.”
Now it’s not <em>entirely</em> about
business, because governments have policy objectives, for example transparency
and freedom of information, that aren’t directly business-related.  But
indeed, there is a dollars-and-cents business dimension.  And to
help broaden the knowledge of those dollars and cents, I went and checked
<a href="">Investor Relations</a>
page to look up the Office-related numbers.  In the fiscal year that ended July
1st, they reported profit of $7.915B on $11.013B in revenue.
The trend continues: in the most recent quarter (ending last September), it
was $1.934B on $2.675B. 
Just FYI.</p>
 <entry xml:base='When/200x/2005/11/29/'>
  <title>Ultra Sysadmin</title>
  <link href='Ultra-Wrangling' />
  <category scheme='' term='Technology/Sun' />
<category scheme='' term='Technology' />
<category scheme='' term='Sun' />
  <category scheme='' term='Technology/Software' />
<category scheme='' term='Software' />
<summary type='xhtml'><div xmlns=''>I&#x2019;ve got this <a href='/ongoing/When/200x/2005/10/18/New-Computer'>Ultra 20</a> which among other things has a huge disk with room for lots of operating systems.  I want to fool around a bit with Linux and  Solaris Classic and GNU/Solaris, run some Bonnie numbers and also shake down my own stuff on all these different boxes.  I expected pain, and I&#x2019;ve already had some.  This post will be my diary of the things that go wrong, and right.</div></summary>
<content type='xhtml'><div xmlns=''>
<p>I’ve got this
<a href="/ongoing/When/200x/2005/10/18/New-Computer">Ultra 20</a> which among
other things has a huge disk with room for lots of operating systems.  I want
to fool around a bit with Linux and 
Solaris Classic and GNU/Solaris, run some Bonnie numbers and also shake down
my own stuff on all these different boxes.  I expected pain, and I’ve already
had some.  This post will be my diary of the things that go wrong, and
<p>As I review my first draft, it’s kind of grumpy.  Do bear in mind that
I’m coming out of the OS X world, which sets an awfully high
standard for fit, finish, and polish.
Another reason is that, while I’ve done my time as a professional sysadmin,
and anyone who’s used Unix-family systems for as long a I has built a few
from scratch, I’ve always found this sort of work profoundly irritating.  It’s
so easy to make dopey mistakes, and since OS installation is by definition a
slow process, each mistake costs time.</p>
<img src="IMG_4247.png" alt="Four Operating Systems on Sun Ultra 20 keyboard" />
<h2 id='p-1'>Kubuntu</h2>
<p>First thing I did was install Kubuntu; I’m out of touch with the
Gnome/KDE wars, but I used to like KDE.
I wanted to split my 250G disk into four slices for Linux, Solaris Classic,
GNU/Solaris, and a chunk to fiddle with filesystems.</p>
<p>The Kubuntu partition wrangler is pretty klunky - every time I carved out a
new chunk, as a side-effect it made another little mini-partition, so I ended
up with a total of 9, four big, one Linux-swap, four small. 
Clearly operator error, but it shouldn’t be this hard.</p>
<p>In fact, my overall impressions of Kubuntu weren’t that positive;
admittedly, it’s a work-in-progress.
First off, why don’t they either compress the ISO or offer a torrent?
Downloading it over my DSL link is pretty painful.</p>
<p>Adept, the package manager, is disappointing; if I wanted to work at the
<code>apt-get</code> level, then I know that I have to install and update and
upgrade and get things in the right order.  But when I’ve got a graphical tool
and I select “OpenSSH server” and click “Install”, well, you know, it ought to just
install it.</p>
<p>One Ubuntu-thing that’s nice is that there’s no obvious root account; it’s like
the Mac, a clever wrapper over <code>sudo</code> so that if you want to
install or configure something, you get prompted for your password.</p>
<p><a href="">Akregator</a> is advertised as
an “RSS/Atom feed reader”, but it couldn’t cope with 
<a href="">my Atom 1.0 feed</a>.
Then when I clicked on “Submit bug report”, everything went to hell, Konqueror
roached my X frame buffer somehow, static all over the place.
And Konqueror is kind of dumb too, if you open a page with lots of pictures,
it stupidly redraws after loading each.  Gotta get Firefox.</p>
<p>Going back to Akregator, the keystrokes for moving between articles seemed
really awkward, but then again I’m used to
<a href="">NetNewsWire</a>, which many people
think is the gold standard.
Hey, NNW is now part of NewsGator, and they’re a multi-platform play, right?
How ’bout a Linux reader, guys?</p>
<p>Here’s a real gripe: after I installed Java, I wanted to get it in my path,
and it turns out that <code>bash</code> on this particular Linux, contra its
man page, does <em>not</em> read <code>.profile</code> or
<code>.bash_profile</code>, all it reads is <code>.bashrc</code>.  Getting out
of sync with your man page is not particularly forgiveable.</p>
<h2 id='p-8'><s>Ku</s>Ubuntu!</h2>
<p>I eventually got myself all tangled up in the disk-partition mess, so
tried again with K-less Ubuntu.  I’m pretty sure that the disk chopper is the
same between the two, but I’d screwed it up once by this time and got it right
second time through; four nice, neat 62.5G partitions.</p>
<p>That aside, Ubuntu’s installation/management experience is, well,
It all Just Worked.  (Well OK, the Synaptic package manager made you click twice to
install which is once too many, but that’s a pretty minor gripe).
This is something that I’d feel pretty confident
<a href="/ongoing/When/200x/2005/08/27/New-Mac-for-Mom">giving to Mom</a> to
install, and I bet she’d do just fine.
Plus, it all looks nice; which is a pretty strong compliment coming from
someone who lives in OS X-land.</p>
<p>You know, if people at Microsoft aren’t losing sleep over the combination
of Ubuntu, the Gnome tools, the Mozilla tools, and OpenOffice, they should
be.  (Also see
<a href="">Sam
Ruby</a> on this topic).</p>
<h2 id='p-3'>Linux &amp; Java</h2>
<p>Ubuntu comes with the <code>gij</code> GNU bytecode stuff, which I’m
pretty sure won’t work with NetBeans, so I went off to and
grabbed the x64 download.   
It unpacked OK, but I wonder why it can’t have a nice graphical installer,
and why it won’t place itself in all the right directories under
<code>/usr/local</code> or some such?</p>
<h2 id='p-4'>Linux &amp; NetBeans</h2>
<p>Now this is how it ought to be.  You download the <code>.bin</code> (a
mere 50M), launch
it, it starts a nice InstallShield wizard, and puts everything
in the right place.  Then you run NetBeans it and It Just Works.</p>
<p>The Java guys should go get some installation hints from the NetBeans
<h2 id='p-5'>Solaris</h2>
<p>After running Solaris on my servers for a few months, I’m pretty well
hooked and won’t be going back to Linux any time soon.
But for a desktop box, the user interface and admin interfaces are just as
important—more important, actually, for lots of people—than threading
performance or filesystem 
magic.  And on the desktop, Solaris installation has some catching-up to do.</p>
<p>I want to run OpenSolaris, obviously, and the install page said that first
of all, I needed to start with a recent “Solaris Express—Community Release”,
so I bopped over there and discovered that I had to download four (not one,
not two, not three, but four) ISOs.  They were slightly compressed, but still,
this is a job for BitTorrent, n’est-ce pas?</p>
<p>The Solaris install startup disk partitioning was confusing, but Googling
led me to
<a href="">Rich
Burridge’s guide</a>, and it worked, although it really sounded like it was
threatening to blow away the Linux partition.
Confusingly, Solaris reported subtly-different sizes for the partitions than
the ones Ubuntu said it was writing.  Well, we all know that a
<a href="/ongoing/When/200x/2005/03/24/Gigabyte">gigabyte by any other
name</a>, isn’t.</p>
<h2 id='p-13'>Aargh</h2>
<p>After I had Solaris running, I wanted to build a ZFS and got confused at
the difference between 0-based controller numbering and 1-based partition
numbering, and built it right on top of my Ubuntu install.  Sob.</p> 
<h2 id='p-14'>Gripe Gripe Gripe</h2>
<p>A few irritating things about the Solaris install:  It doesn’t assume DHCP,
thus stalls for a long time stupidly trying to configure the network.
At several
points during the process, it goes catatonic; it’s actually doing something,
but it gives you no visual (or any other) indicator, it just sits there, while
you worry that Something Has Gone Wrong.
When you log in for the first time, it’s as
root, unlike Ubuntu which insists on creating a user account as a side-effect
of installing.  I suppose
this is OK, but nowhere in the JDS menus (that I could find) is a “create
user account” button.  Resorting to Google taught me that I had to type
<code>/usr/sadm/bin/smc</code> and then I was fine, but really.
The installation process doesn’t include a prompt for a node-name, so it comes
up as hostname “unknown”, and you have to do quite a bit of digging to find
out that you need to put the hostname in <code>/etc/nodename</code>, and woe
betide you if you run SMC before you do that because it stashes a bunch of
stuff in <code>~/.smc/&lt;current-hostname>/whatever</code> (or something like
that, I erased it in disgust).</p>
<p>The upside: given that you’re loading four CD’s worth of stuff, the Solaris
install gets done in a pretty big hurry without demanding much thought.  Also,
SMC is a decent program, did what I needed and got out of the 
<p>Anyhow, I have my Ubuntu, I have my recent Solaris with ZFS, they both run
great, I even know 11 different ways to refer to disk partitions.</p>
<h2 id='p-15'>Onward and Upward</h2>
<p>Now on to OpenSolaris.  Yow,
<a href="">this
looks kind of hairy</a>.  Think I’ll try the
<a href="">Nexenta alpha</a> first.</p>
<p>Hmm, the Nexenta CD boots and says <code>root login:</code>.
OK.  No README, no INSTALL.  Poking around leads me to
where there’s some things that look like installers.
I launched one and it started to look like the Solaris installer, I want to be
real sure I don’t roach any more partitions and I was tired, so I nuked it.
More later.</p>
 <entry xml:base='When/200x/2005/11/29/'>
  <title>How They Learn</title>
  <link href='How-They-Learn' />
  <category scheme='' term='The World/Family' />
<category scheme='' term='The World' />
<category scheme='' term='Family' />
  <category scheme='' term='Education' />
<category scheme='' term='Education' />
<content type='xhtml'><div xmlns=''>
<p>It’s amazing, contrary to all the clichés, how slowly little kids learn.
My little guy, now in first grade, has been learning to read for a year now
and still struggles with some obvious-seeming words; oh, those English vowels.
But slow is OK, because 
time is what he has, no end of it; and when you’re my age that time rushes by,
fast and unceasing like a spring flood.  This evening, reading the first
chapters of the first Harry
Potter at bedtime, 
he seemed to want a turn so I pointed him at a paragraph and
he hurried through it, the tale’s urgency carrying him over words he couldn’t
make out.   I remember being the same age doing the same thing, wondering what
some word meant, but not enough to stop, or even slow down much.</p>
 <entry xml:base='When/200x/2005/11/29/'>
  <link href='HBR1' />
  <category scheme='' term='Arts/Music' />
<category scheme='' term='Arts' />
<category scheme='' term='Music' />
  <category scheme='' term='Technology/Internet' />
<category scheme='' term='Technology' />
<category scheme='' term='Internet' />
<content type='xhtml'><div xmlns=''>
<p>I couple of years back I wrote about the advantages of using iTunes or XMMS 
to soak up
<a href="">Ambient
Internet Brain Goo</a>.  These last couple of days I’ve been learning how to
partition the disks and install multiple operating systems, some of ’em pretty
bleeding-edge, on my Ultra 20,
which involves quite a bit of waiting and irritation.  I owe my continued
sanity to
<a href="">HBR1</a>; as far as I’m concerned they’ve got
the best brain goo going.</p>
 <entry xml:base='When/200x/2005/11/29/'>
  <title>NetBeans Tip</title>
  <link href='Faster-NetBeans' />
  <category scheme='' term='Technology/Mac OS X' />
<category scheme='' term='Technology' />
<category scheme='' term='Mac OS X' />
  <category scheme='' term='Technology/Coding/Java' />
<category scheme='' term='Coding' />
<category scheme='' term='Java' />
<content type='xhtml'><div xmlns=''>
<p>I tried out Joshua Marinacci’s recent
<a href="">Netbeans on Mac Tip</a>
(different GC setting) and it worked like a charm.  I suspect that this ought
to be the default 
setting for NetBeans on most systems.  I talked to a NetBeans insider in
Prague and he mumbled about hard-to-reproduce weirdness with this setting on
previous NetBeans &amp; Java releases, but I haven’t seen a single hiccup.  
It turns out that on the AMD64 systems, lots of people are running more or
less everything Java in <code>-server</code> mode.  Hmmm.</p>
 <entry xml:base='When/200x/2005/11/28/'>
  <title>Liberals Fall</title>
  <link href='Canadian-Election' />
  <category scheme='' term='The World/Politics' />
<category scheme='' term='The World' />
<category scheme='' term='Politics' />
  <category scheme='' term='The World/Places/Canada' />
<category scheme='' term='Places' />
<category scheme='' term='Canada' />
<content type='xhtml'><div xmlns=''>
<p>As I (very safely)
<a href="/ongoing/When/200x/2005/11/08/Canadian-Politics">predicted</a> a
couple of weeks ago, the Canadian government 
<a href="">fell
today</a> and we’re looking at a January election.  You have to feel sorry for
the candidates who, most places in the country, will be slogging through the
snow and sub-zero temperatures; but not that sorry, it’s long past time we had
this little catharsis.
As I was scanning the coverage today I ran across the blog of
<a href="">Monte Solberg</a>, an Alberta
Tory  of whom I’d never heard, but who gives the political narrative in an
unforced, flowing, insider’s voice; he’s a natural.  I was a little
intrigued that it’s not obvious at all from his web-site which party he
represents.  On the other hand, it’s tough to see a guy this unaffected and
natural being real comfy associating with a plastic on-message droid like
Stephen Harper.</p>
 <entry xml:base='When/200x/2005/11/28/'>
  <link href='ODF-and-Atom' />
  <category scheme='' term='The World/Life Online/People' />
<category scheme='' term='The World' />
<category scheme='' term='Life Online' />
<category scheme='' term='People' />
  <category scheme='' term='Technology/XML' />
<category scheme='' term='Technology' />
<category scheme='' term='XML' />
  <category scheme='' term='Business/Microsoft' />
<category scheme='' term='Business' />
<category scheme='' term='Microsoft' />
<content type='xhtml'><div xmlns=''>
<p>It seems like my little
<a href="/ongoing/When/200x/2005/11/27/Office-XML">thought experiment</a> has
touched a nerve.
<a href="">Scoble</a>, 
<a href="">Dare
Obasanjo</a>, and
<a href="">Randy
Holloway</a> all push back, amazingly enough all making the same argument: how
can I be against duplication in office-document XML format while at the same
time being mixed up in the
<a href="">Atom
The argument is fallacious, but at least Robert and Randy made it in grown-up,
polite terms, leaving the childish name-calling to Dare.
Now, as for RSS and Atom:  When I came on the scene in 2003, RSS was already
hopelessly fragmented, and there was exactly zero chance of any of the 
thin-skinned proponents of the various versions deciding to make
nice with each other.  Atom is precisely an attempt to reduce the number of
vocabularies that implementors feel they have to support.
Turning to the office-document space: right now the world has
<a href="">exactly one</a>
finished, delivered, standardized, totally-unencumbered, multiply-implemented
XML-based office document format.  You are the guys who want
to introduce another, incompatible one.  And I think that’s OK; but
restrict your invention to the specialized Microsoft stuff that ODF can’t
do, and don’t re-invent the basics.  Why is this controversial?</p>
 <entry xml:base='When/200x/2005/11/28/'>
  <title>Seems Like Forever</title>
  <link href='Technorati' />
  <category scheme='' term='Technology/Web' />
<category scheme='' term='Technology' />
<category scheme='' term='Web' />
  <category scheme='' term='Technology/Syndication' />
<category scheme='' term='Syndication' />
<content type='xhtml'><div xmlns=''>
<p>But it’s only 
<a href="">Technorati’s third
I don’t remember when I first stumbled across them, but I actually
paid real money for a feed of pointers to my brand-new blog.
Nobody who hasn’t been behind the firewall at Technorati or one of their
competitors can grasp how pathologically hard it’s been to keep a service like
that up and running in the face of the continuing insane growth of the
blogosphere; they’ve had some tough times but it’s been a long time since they
weren’t there when I needed them.
Today, Nicholas Carr
<a href="">tries
to explain the big picture</a> that Technorati and their sector fit
into.  I don’t know, I think anyone who claims to understand what’s going on
is being awfully damn brave.   Interestingly, I’ve heard Dave Sifry make a
compelling big-picture pitch several times that’s as convincing as anything
I’ve read, and as far as I know he’s never actually written it down.  Dave?
<i>[Disclosure: I
<a href="">may have
a conflict of interest</a> with respect to Technorati.]</i></p>
 <entry xml:base='When/200x/2005/11/28/'>
  <title>The Saga Continues</title>
  <link href='Mass-Again' />
  <category scheme='' term='The World/Politics' />
<category scheme='' term='The World' />
<category scheme='' term='Politics' />
  <category scheme='' term='Technology/XML' />
<category scheme='' term='Technology' />
<category scheme='' term='XML' />
<content type='xhtml'><div xmlns=''>
<p>The Massachusetts Office XML File Formats saga, that is.  The latest news
is that the Microsoft announcements last week are
<a href="">playing
well</a> in Boston.
Commonwealth secretary Thomas Trimarco stated “we are optimistic that Office
Open XML will meet our new standards”, and
<a href="/ongoing/When/200x/2005/11/27/Office-XML">I’m optimistic
too</a>.  Obviously the key word is “will”, since we haven’t seen what’s
getting submitted to ECMA and nobody’s seen what will come out of ECMA.
Our own chief standards geek Carl Cargill wrote Mr. Trimarco a letter, which
you can read over at
<a href="">Piper Cole’s

 <entry xml:base='When/200x/2005/11/27/'>
  <title>Thought Experiments</title>
  <link href='Office-XML' />
  <category scheme='' term='Technology/XML' />
<category scheme='' term='Technology' />
<category scheme='' term='XML' />
  <category scheme='' term='Business/Microsoft' />
<category scheme='' term='Business' />
<category scheme='' term='Microsoft' />
<summary type='xhtml'><div xmlns=''>I see that Microsoft has posted a <a href=''>litigation covenant</a> on the OfficeXML formats (also read <a href=''>Brian Jones&#x2019; exegesis</a>). In response, there&#x2019;s a bunch of legal poking and prodding  <a href=''>here</a> and <a href=''>here</a>; I don&#x2019;t understand the legal arguments, and I don&#x2019;t think they&#x2019;re the interesting part of the story anyhow. So, let&#x2019;s do two thought experiments.  First, what if Microsoft really is doing the right thing?  Second, how can we avoid having two incompatible file formats? <i>[Update: There&#x2019;s been a lot of reaction to this piece, and I addressed some of those points <a href='/ongoing/When/200x/2005/11/28/ODF-and-Atom'>here</a>.]</i></div></summary>
<content type='xhtml'><div xmlns=''>
<p>I see that Microsoft has posted a
<a href="">litigation
covenant</a> on the OfficeXML formats (also read
<a href="">Brian
Jones’ exegesis</a>).
In response, there’s a bunch of legal poking and prodding 
<a href="">here</a> and
<a href="">here</a>;
I don’t understand the legal arguments, and I don’t think they’re the
interesting part of the story anyhow.
So, let’s do two thought experiments.  First, what if Microsoft really is
doing the right thing?  Second, how can we avoid having two incompatible
file formats?
<i>[Update: There’s been a lot of reaction to this piece, and I addressed some
of those points
<a href="/ongoing/When/200x/2005/11/28/ODF-and-Atom">here</a>.]</i></p>
<p>Clearly, there’s still work to be done on the legals; the current Microsoft
statement only covers the 2003 versions, and to see the schemas you have to
fight through a bunch of technical barriers and draconian legalese.
But unless Brian Jones is lying in his teeth, their
intent is to remove the legal encumbrances and let developers loose on
their file formats.  (If that’s <em>not</em> the case, we’ll find out pretty
soon, and in that case I think the Microsoft XML is pretty well doomed;
the world is simply not willing to live with legally-tainted file formats any
So, as a thought experiment, let’s assume the legal obstacles are gone; what
<p>To keep things short, let’s call OpenDocument Format 1.0 “ODF” and the
Office 12 XML File Formats “O12X”.</p>
<h2 id='p-1'>Alternatives</h2>
<p>In ODF we have a format that’s already a stable
OASIS standard and has multiple shipping implementations.
In O12X we have a format that will become a stable ECMA standard with one
shipping implementation sometime a year or two from now, depending on
software-development and standards-process timetables.
ODF is in the process
of working its way through ISO, and O12X will apparently be sent down that road too,
which should put ISO in an interesting situation.</p>
<p>On the technology side, the two formats are really more alike than they are
But, there are differences:  
O12X’s design center, Microsoft has said repeatedly,
is capturing the exact semantics of the billions of existing Microsoft Office
ODF’s design center is general-purpose reusability, and leveraging existing
standards like SVG and MathML and so on.</p>
<p>Which do you like better?  I know which one I’d pick.  But I think we’re
missing the point.</p>
<h2 id='p-2'>Why Are There Two?</h2>
<p>Almost all office documents are just paragraphs of text, with some
bold and some italics and some lists and some tables and some pictures.
Almost all
spreadsheets are numbers and labels, with some sums and averages and pivots
simple algebra.  Almost all presentations are lists of bullet points with occasional
<p>The capabilities of ODF and O12X are essentially identical for all this
basic stuff.  So why in the flaming hell does the world need two incompatible
formats to express it?  The answer, obviously, is, “it doesn’t”.</p>
<p>Microsoft wants there to be an office-document XML format
that covers their billions of legacy documents, and
they want it to be open.  Fair enough; I approve.  But why do we have to
re-invent all the basic stuff, and have two ways to express “This paragraph is
in 12-point Arial with 1.2em leading and ragged-right justification” or “K37
is the average of B37 through H37”?</p>
<p>The ideal outcome would be a common shared office-XML dialect for the
basics—and it should be ODF (or a subset), since that’s been designed and
debugged—then another extended vocabulary to support Microsoft features	,
whether they’re cool new whizzy features or mouldy old legacy features (XML
Namespaces are designed to support exactly this kind of thing).
That way, if you stayed with the basic stuff you’d never need to worry about
software lock-in; the difference between portable and proprietary would be
And, for the basic stuff that everybody uses, there’d be only one set of
<p>This outcome is technically feasible.
Who could possibly be against it?</p>
 <entry xml:base='When/200x/2004/12/12/'>
  <link href='BMS' />
  <category scheme='' term='Technology/Publishing' />
<category scheme='' term='Technology' />
<category scheme='' term='Publishing' />
  <category scheme='' term='Technology/Web' />
<category scheme='' term='Web' />
<summary type='xhtml'><div xmlns=''>Almost every Sunday I grab the week&#x2019;s <span class="o">ongoing</span> logfiles and update my numbers. I find it interesting and maybe others will too, so  this entry is now the charts&#x2019; permanent home. I&#x2019;ll update it most weeks, probably. <i>[Updated: 2005/11/27, and I added the RSS/Atom feed graph.  I&#x2019;m going to hang on a week before tossing in the Google Analytics numbers, because they&#x2019;re showing severe Thanksgiving-related distortion, also visible in the latter two graphs here.]</i></div></summary>
<content type='xhtml'><div xmlns=''>
<p>Almost every Sunday I grab the week’s <span class='o'>ongoing</span>
logfiles and update my numbers.
I find it interesting
and maybe others will too, so 
this entry is now the charts’ permanent home. I’ll update it most weeks,
<i>[Updated: 2005/11/27, and I added the RSS/Atom feed graph.  I’m going to
hang on a week before tossing in the Google Analytics numbers, because they’re
showing severe Thanksgiving-related distortion, also visible in the latter two
graphs here.]</i></p>
<img src="Browser-Market-Share.png" alt="Browser market shares at ‘ongoing’" />
<div class="caption"><p>Browsers visiting <span class='o'>ongoing</span>,
<img src="Browsers-via-search.png" alt="Browser market shares at ‘ongoing’, visitors via search engines" />
<div class="caption"><p>Browsers visiting <span class='o'>ongoing</span> via
search engines, percent.</p></div>
<img src="Search-Engines.png" alt="Search engine market shares at ‘ongoing’" />
<div class="caption"><p>Search referrals to <span
<img src="Feeds.png" alt="Rss and Atom feed fetches" />
<div class="caption"><p>Fetches of the RSS 2.0 and Atom 1.0 feeds.</p></div>

<p>The notes on usage and source code will return in coming weeks when I get
the cycles to rewrite this whole article.</p>
 <entry xml:base='When/200x/2005/11/26/'>
  <title>Fair is Fair</title>
  <link href='Fair-is-Fair' />
  <category scheme='' term='Technology/Web' />
<category scheme='' term='Technology' />
<category scheme='' term='Web' />
<content type='xhtml'><div xmlns=''>
<p>And borked is borked.  A few days back, I
<a href="/ongoing/When/200x/2005/11/21/Microsoft-XML-News">flamed away</a> at
<a href="">egregiously-broken</a>
markup over at MSDN.
Alex Barnett over at MSDN
<a href="">points out</a> that our own
<a href="">Sun Developer Network</a> site
<a href="">isn’t
valid either</a>.  Not as bad as MSDN (I noticed the problem there because it
misdisplays on Safari) but still borked.  It’s pathetic, just pathetic,
when two of the leading developers’ resources on the Web can’t get their shit
together and emit  valid web content.  Guys, go buy
<a href="">Zeldman</a> and
<a href="">Shea</a> and, like I said before,
get a clue.
<i>[Update: Make that “borked is borked is borked”; per James Snell, 
<a href=";entry=100822&amp;ca=drs-bl">IBM too</a>.
Let’s those of us working for the companies behind those sites all do some
more public shouting and private pleading, and maybe we can make ’em do the
right thing.]</i></p>
 <entry xml:base='When/200x/2005/11/25/'>
  <link href='OSI' />
  <category scheme='' term='Technology/Open Source' />
<category scheme='' term='Technology' />
<category scheme='' term='Open Source' />
<content type='xhtml'><div xmlns=''>
<p>That stands for “Open Source Innovation”, and Dana Blankenhorn
<a href=";part=rss&amp;tag=feed&amp;subj=zdblog">suggests
there isn’t any</a>.  Uh... Apache.  Emacs.  Vi.  Perl.  Python.  Ruby.  PHP.
Those were the result of twenty seconds thinking.  I don’t think software innovation
historically has correlated negatively or positively with open-source-ness.
In the future though, I think pretty well all software innovation will be either
open-source or inside a big server, because the business model for shipping
closed-source software as a product is just too 
twisted and weird.</p>
 <entry xml:base='When/200x/2005/11/25/'>
  <title>Vancouver Blogs</title>
  <link href='Vancouver-Blogs' />
  <category scheme='' term='The World/Places/Vancouver' />
<category scheme='' term='The World' />
<category scheme='' term='Places' />
<category scheme='' term='Vancouver' />
<content type='xhtml'><div xmlns=''>
<p>This notion of regional blogging is starting to get traction; right here in
Vancouver, which is a pretty wired place, we now have three different
pretenders to the “Vancouver City Blog” throne:
<a href="">Urban Vancouver</a>,
<a href="">The Vancouverite</a>, and
<a href="">Metroblogging Vancouver</a>, 
the latter part of the Metroblogging empire.  I’ve subscribed because, you
know, I live here; maybe this will make me a bit hipper and cooler, because
they’re always talking about events that I’m not clued-in enough to know
about; if we start going to a few, that would be a good sign.
A good thing happened already: one of the above linked to
<a href="">static photography</a>, a very nice
and thoughtful local photoblog, check it out.</p>
 <entry xml:base='When/200x/2005/11/25/'>
  <title>FSS: Fireworks, &#x2019;57-&#x2019;61</title>
  <link href='FSS' />
  <category scheme='' term='Arts/Photos' />
<category scheme='' term='Arts' />
<category scheme='' term='Photos' />
  <category scheme='' term='The World/Places/Oregon' />
<category scheme='' term='The World' />
<category scheme='' term='Places' />
<category scheme='' term='Oregon' />
<summary type='xhtml'><div xmlns=''>Friday Slide Scan #14 is four of Dad&#x2019;s fireworks pictures, all over forty years old.</div></summary>
<content type='xhtml'><div xmlns=''>
<p>Friday Slide Scan #14 is four of Dad’s fireworks pictures, all over forty
years old.</p>
<img src="A6001.png" alt="Fireworks in Corvallis, Oregon, 1957" />
<img src="A6006.png" alt="Fireworks in Corvallis, Oregon, 1957-1961" />
<img src="A6009.png" alt="Fireworks in Corvallis, Oregon, 1957-1961" />
<img src="A6012.png" alt="Fireworks in Corvallis, Oregon, 1957-1961" />
<p>The first is clearly dated 1957 and the others simply marked “Corvallis”.
That would be
<a href="">Corvallis,
Oregon</a>, where
<a href="">Dad</a> was getting his Ph.D. at
<a href="">OSU</a>.  
He graduated, and we moved to Lebanon, in 1962, so that’s an upper bound on
the dates.</p>
<p>The years may be in doubt, but the rest isn’t; these would all have been
taken on July 4<sup>th</sup> of whichever year.  I can remember (less than
seven years old, but I remember) watching Dad
brace the camera against a tree, pointing at the fireworks and holding the
button down.</p>
<p>Images in the Friday Slide Scans are from 35mm slides taken between 1953
and 2003 by (in rough chronological order)
<a href="">Bill Bray</a>, 
<a href="">Jean Bray</a>, Tim Bray, Cath
Bray, and
<a href="">Lauren Wood</a>; when I know
exactly who took one, I’ll say; in this case, Bill Bray. 
Most but not all of the slides were on Kodachrome; they were digitized using
a Nikon CoolScan 4000 ED scanner and cleaned up by a combination of the Nikon
scanning software and PhotoShop Elements.</p>
 <entry xml:base='When/200x/2005/11/24/'>
  <title>Roller Redux</title>
  <link href='Roller-Redux' />
  <category scheme='' term='Technology/Java' />
<category scheme='' term='Technology' />
<category scheme='' term='Java' />
  <category scheme='' term='Business/Blogging' />
<category scheme='' term='Business' />
<category scheme='' term='Blogging' />
<content type='xhtml'><div xmlns=''>
<p>Which is to say,
<a href="">Roller
2.0 has shipped</a>.  Lots of new stuff, and anything that can hold up under
<a href=""></a> load is pretty well
battle-tested.  I wonder if there are any other freely-downloadable blogging
engines (WordPress maybe?) getting as much full-time dedicated attention.</p>
 <entry xml:base='When/200x/2005/11/24/'>
  <title>Giving Thanks</title>
  <link href='Thanksgiving' />
  <category scheme='' term='The World/Places/America' />
<category scheme='' term='The World' />
<category scheme='' term='Places' />
<category scheme='' term='America' />
<content type='xhtml'><div xmlns=''>
<p>This is the day where those of us around the world who are in the
US-centric  high tech business, but not Americans, give thanks for the fact
that our phones are silent and for the certainty that no major high-tech
announcements or initiatives will occur.
Facetiousness aside, this is, as
<a href="">Dan 
Gillmor says</a>, America’s best holiday.  There’s a lot to be said for a
major civic celebration that has nothing to do with theologically-significant
births or deaths that happened millennia ago in the Middle East.
Here’s a charming,
<a href="">very personal, small-scale</a>
narrative about remembering—and, implicitly, giving thanks.</p>
 <entry xml:base='When/200x/2005/11/23/'>
  <title>Drawing Drama</title>
  <link href='Dramatic-Drawing' />
  <category scheme='' term='Arts' />
<category scheme='' term='Arts' />
<content type='xhtml'><div xmlns=''>
<p><a href="">David Weinberger</a> pointed out
<a href="">this</a> (warning, very
slightly not-work-safe) and I found it compelling, hypnotic; I have no idea
what the site,
<a href=""></a>, is about, but the the page called
“Flash Cards II” in English,
<a href="">Виртуальные открытки II - Вся коллекция</a>
in Russian, has lots and lots more
of these, and I’m not even sure David picked the best; they are elegant,
erotic, witty, dramatic; well worth a visit.
You probably want to turn the speed up to
8× or 16× for most of these or you’ll end up watching them all day.
I suppose this has long been prefigured in the work of sidewalk caricaturists
and the like, but still, I think it’s a new thing in the world.</p>
 <entry xml:base='When/200x/2005/11/22/'>
  <title>Feed Rates</title>
  <link href='Feed-Rates' />
  <category scheme='' term='Technology/Syndication' />
<category scheme='' term='Technology' />
<category scheme='' term='Syndication' />
  <category scheme='' term='Technology/Web' />
<category scheme='' term='Web' />
<summary type='xhtml'><div xmlns=''>I was glancing at my server logfiles, and was impressed, as always, by the huge number of feed-readers out there.  So I made a graph of how often the <span class="o">ongoing</span> feeds have been fetched so far this year, and the popularity of RSS vs. Atom 1.0.</div></summary>
<content type='xhtml'><div xmlns=''>
<p>I was glancing at my server logfiles, and was impressed, as always, by the
huge number of feed-readers out there.  So I made a graph of how often the
<span class='o'>ongoing</span> feeds have been fetched so far this year, and the popularity of
RSS vs. Atom 1.0.</p>
<img src="Feed-fetches.png" alt="Weekly fetches of the ongoing feed" />
<p>Note that the Y-axis bottoms out at 200K, so the graph overemphasizes
the growth relative to the total traffic.
Bottom line: 
During prime-time hours, the feeds are
getting fetched multiple times per second.</p>

Previous | Next | Top | Cafe con Leche | Cafe au Lait

Copyright 2005, 2006 Elliotte Rusty Harold
Last Modified December 1, 2005