I experienced serious technical difficulties when giving this
presentation. I was under the mis-impression that my
talk
was at 2:45 when it in fact was scheduled for 2:00. So, I was
in the midst of preparing my presentation for upload — always
have a viable backup — when I was
fetched to come give the talk.
I fumbled for a few minutes trying to get the video to display,
and despite the fact that I have been successful as far away as
Sri
Lanka, I had no success in Atlanta. My clip on-mike
didn’t pick up my voice sufficiently, so I ended up using a
hand mike. Trying to generate the HTML to copy to a USB
device while giving a presentation while holding a mike - well, I
simply didn’t have enough hands.
So I presented to a backdrop of a blank screen, complimented
occasionally by some illegible scribles I made on a flip chart.
Judging by the audience, I did get the impression that most
people were following along, and there was even some apparent
disagreement on some of my points - though I suspect that mostly we
would have ended up at the same place. Example: it apparently
is deemed somewhat unfortunate but expected that some parsers would
accept xml:stylesheet whereas others would flat out reject the
entire document. At one point there was even an audience to
audience disagreement as to whether the XML spec clearly states
that attribute order is irrelevant.
Normally, I like creating the environment for such discussions
as it means that I have successfully engaged the audience. I
just wish I could have more fully participated myself.
Note: while the URLs for the individual pages of this
presentation are stable, I may end up tweaking the script that
generated them over the next few days. I see to have hit
scalability issues in S5, so I backed off to a hybrid approach for
generating the slides. I will admit to a preference for every slide
having a unique URI, and for the address bar in the browser to
reflect the URI of the current page, so this is goodness
anyway.
With respect the “no ungetc in Java” comment in relation to the BOM, I present to you PushbackReader (since JDK 1.1); that may or may not have helped you solve your problem, but it’s one of those things few people seem to know aobut and which comes in handy here and there.
As you’ve pointed out in different terms, XML is even harder than comedy. But one of the reasons for the massive success of HTML was the fact that early browsers accepted tag soup, enabling just about anybody, not just those with the attention span to read an implement a spec, to create a web page. Looking over REXML, as you’ve documented yourself, the urge to go Postel on purported XML you find in the wild looks like it’s hard to resist.
With respect to the general issue of liberality, it depends on the context; a parser used in a SOAP stack deployed in a crucial application has every right to be nitpicky. When you’re talking about an RSS parser, I think Mark Pilgrim got it right and made something that’s extremely liberal.
Seems Sam Ruby’s presentation suffered a bit in the spectacle. Unfortunately, I’m no stranger to presentation set-up problems. I’ve also been lucky enough to have patient audiences. Maybe conference organizers will someday factor Linux A/V support...
I enjoyed reading the slides. Sadly there seem to be some quirks at the moment. Slides 107, 117, and 119 missed the formatting train. There’s a possible time warp and detour around 129, 130, and 131.
Talking about well-formedness is fine, except when you are talking about a part of the process of acquiring an XML document, parsing it, and presenting the results, where well-formedness either doesn’t apply, or doesn’t apply in the way you say it...
You did an admirable job, Sam, in spite of A/V difficulties. Interestingly, with the same laptop and same distribution, I had no projection difficulties. But I’m still using the open source ATI driver.
Unfortunately, I’m not exactly sure what point you were trying to make. That you have to understand how the various bits of XML fit together? Sure, you do. That some specifications could be clearer? Sure, name one that couldn’t. That some software has bugs? Sure, but that’s a surprise?
For example, you make a fuss about the fact that “--” can’t appear in a comment. Why is that more significant than the fact that “*/” can’t appear in a C or Java block comment?
You observe that Klingon is unlikely to be a supported encoding? Would it be better if only a single encoding was allowed? Sure, better for interoperability, I guess, but not better for those Klingon native speakers whose lives are made easier by the presence of a proper encoding (does Klingon really have an encoding, I wonder?)
There are Unicode code points whose glyphs are nearly (or exactly) indistinguishable from other glyphs. Were that not the case, there’d be one less thing to be aware of, I suppose, but that’s just the way the world is. Different communities have reused glyphs.
I guess the takeaway is that XML isn’t quite as simple as it appears on first glance. But what is?
Norman: thanks. I too am using the vanilla ATI driver now — my guess is that somehow I was flustered enough that I made a rookie mistake; screen resolution mismatch or some-such.
Overall, it I had any general point, it would be that abstractions leak, and that XML would benefit from the creation of a BCP document.
Filling in a few details from before. The XQuery stuff demoed by Jason Hunter is from MarkLogic; there’s a demo/trial/community license available, which allows you to run XQuery over 50MB datasets. Their developer site is also to be...
So in one of my classes, I have to do a group presentation. We decided to go for the minimalistic approach, which led to having many more slides in our presentation than normal presentations: 188 slides, which translates to 33.6 megabytes. It seems...
RDF Calendar, GRDDL, Microformats, and all that at XML2005 in Atlanta
My talk was: @ 2:45 on Weds : Semantic Web Calendaring: RDF Calendar, hCalendar, and GRDDL I spent a lot of time preparing my slides, Semantic Web Data Integration with hCalendar and GRDDL, but at show-time, there were still too many. I had to...
phil ringnalda » <style> in an age of HTML fragments
phil ringnalda » <style> in an age of HTML fragments. I love how everything is broken. And, I am surprised at how many aggregators get Atom feeds wrong....
Courtesy of Camping, Why’s latest stroke of genius (he’s the creator of Hoodwink.d as well), comes this lovely little graphic. Ever curious if you are an XML Geek? If that made you laugh, then yes. If you can also understand Sam...
[link] 이 슬라이드는 그냥 화면 아무데나 클릭하면 화면이 넘어갑니다. trax님의 글을 보면 <?xml version=“1.0″ encoding=“euc-kr”?> 이 잘못된 것처럼 설명되어있지만 제가 XML 스펙을 읽어본 바로는 잘못되었다는 점을 찾지 못하겠습니다. 물론 feedvalidator.org 에서는...
The extract method is basically done. I’m sure it could be improved a bit more, but it seems to be fairly effective. I added a few extra features beyond the original URI class’s capabilities, such as supplying a base uri to resolve...
Over the past couple of years, the Business Process Modeling Notation (BPMN) has established itself as the de-facto standard for process modeling. Nevertheless, an impedance mismatch exists between BPMN and BPEL, one of its target execution...
Die Menge der Blogs ist schier unüberschaubar — die Weblog-Suchmaschine Technorati verfolgte bereits letztes Jahr mehr als 57 Millionen. Jede Auswahl daraus ist damit ausgesprochen subjektiv und von den Interessen des Betrachters geprägt...
I love it when a plan comes together. Matt Terenzio : “Why we wouldn’t use XMPP as the basis for a decentralized microblogging platform?” XMPP is getting on the scale radar - at last. Dave Winer has picked up on the idea , despite rejecting it...