random($foo)

After a quick jaunt to SD and back (I spent by far more time on the road than anywhere else) I’m trying to catch up with email, work, backfilling of SXSW notes (even short summaries and links are taking a surprisingly long time), and trying to focus coding some interesting things tonight. The SXSW notes are dated against the panel times (hmm, think about Time Zones) so they should make reverse chronological sense.

I’ll be trying to ingest etech content/commentary over the next few days. Apparently there were lots of good bits there this year. I should try hard to make it there properly next year.

My current obsession btw, here’s are some pre/post SXSW entries. I should keep the results of the water % numbers too, I’m sure it’d explain Wednesday’s entry:

“Improved focus can be achieved through activities such as meditation, yoga, and turning off instant messaging”

— Ulrich Mayr, University of Oregon

This panel wasn’t as exciting as I hoped it’d be, but it was the end of a long week and I was too tired to really care that much.

After this panel, I crashed for about an hour or so, missing out on some Hilton-bar schmoozing, but I think redeemed myself with the partying until 7AM.

Bruce Sterling’s always entertaining. This year, it was all about the fabricators: the physical being the new digital, within the context of sustainability. Really good stuff. (did IP concerns get touched here?) Anyway, just read the notes.

• Auscilate’s mostly complete transcript
• Lawver Wiki notes
• Clickable Culture notes
• The Actual is the New Virtual: Alex and Bruce Keynote SXSW – Alex posts a summary
• The Kind of Future Fabbing Suggests – Alex writes further on the topic matter

This was in the day-stage this year, and quickly became Sitting Here vs. Getting Food.

OK, not that funny, but worth noting that two blocks down from the Convention Center (betwen 4/5th on the Brazos side?) there’s a Taco Shack in the first floor of the Frost Bank Tower that has some great Mexican fast food.

OK, as you were.

I really enjoyed this panel (Jonas’ and Danah’s presentations were particularly interesting), and I liked that they actually brought up how retarded the ‘lets throw the data out there’ mentality is. Obviously there’s a real dichotomy going on w/ decentralization. How do you mine what’s already there to create useful services, and two, how do you deal with all that’s out there? At some point, I should go back and fill out some notes with all the thoughts triggered from the various stuff tossed out, but for now I’m tired of typing.

• Auscillate’s transcript
• jibber jabber’s notes
• Clickable Culture’s notes

Hmm. A lot of really smart people were up there talking, but I’m not sure I really got into the swing of the point they were making. Well, really that they were all making different ones maybe. I didn’t really take that much out of it I guess. I got really activisted out last year, so I may just still be recovering.

On the bright side, Zach had swung down and I got to say hi, see what they were up to w/ CivicSpace. I love picking his brain because this Drupal-based thing I’m getting down and dirty with is similar on so many levels (unfortunately, the community-scaling, which is what I really want to do, probably won’t be til post-1.0 and will undoubtedly require re-arch…)

• SxSwDelib – official wiki notes
• Full Circle Online Interaction Blog – liveblogging from the panel
• Lawver Wiki: Deliberative Democracy and Interactive Technology

The Tuesday morning SXSW panel was entitled: Web Design 2010: What Will the Web Look Like When It Turns 20?. After SubEthaEdit mishappery, the panel started off talking about “web apps” and gave some real WTF answers (ah, designers pontificating on development trends) so… I had to step out. Maybe I overreacted, but Molly had the same reaction in the hallway when she stepped in for a different 5 minutes, so I’m assuming that it wasn’t just me (todo: find transcript)

On the bright side, Dave had started a blog to do panel prep, and has opened it up for discussion. This is great. I wish more panels would have done that (or that all panels had this automatically, and that these sites were open for discussion beforehand and could also be used to help schedule rooms based on participation… oh wait, there I go again. Anyway, since I got kinda carried away in my posts (ahh, the inner pundit), here’s how I would have answered if I were on the panel (reproduced from comments for easy reading):

1. Are Web Apps going to happen?
What are they going to look like? Desktop apps? Something else?
How are we going to be doing them

We’ve very obviously been seeing the expansion of traditional web apps w/ the recent round of AJAX-based apps. Will web apps be able to expand beyond their core markets in the communications and shared media space? I think so, especially as the idea of the personal infocloud encompasses more and more facets. The need for data sharing and sync will outweigh the cost barriers of rewriting more and more applications. The subsequent improvements in ease of update and stable revenue stream will end up being better for almost everyone (B&M software shops excepted).

The biggest question IMO is whether the push for subscription-based net-delivered thick client apps (think .NET) wins over more webby variations. This may largely depend on the type of application (ie, Photoshop would fit easily in the former but not the latter), but also largely on strategic and technical decisions that won’t be easy to predict…

2. What will web designers be doing in 2010?
What sort of devices will be most commonly used to access the web?
What technologies will enable this?

“International Data Corp predicts that by the year
2004, there will be close to 1.3 billion web-enabled cellular phones
globally. And Motorola predicts that by that same year, more consumers
will be accessing the Internet from a wireless device, than a wired
one.”

src: http://www.thefeature.com/article?articleid=10825

3. Will IE6 still be significant to the work you will be doing?

[Jeremy,] I agree that CSS implementations will continue to be a problem regardless of which version of IE (see the latest IE7/CSS support hubbub), but I think that one of two things will happen: 1) people will stop testing/caring as IE6 support drops below a certain level, or more likely (I’ve begun doing this for various personal sites, it’s just a matter of time it becomes production ready), something like Dean Edwards’ IE7 library will be used to ‘fix’ the IE issues.

As far as scripting goes, I don’t think IE6 will play a major factor again, if it breaks things and goes under a certain percentage. This will be doubly so as XAML, FLEX, and other RIA front-ends fight it out.

4. In 2010, these technologies will be important for web developers:

(vote for, vote against, abstain)

CSS3
SVG
Flash
XML/XSLT (more generally than XHTML)
XHTML
XHTML2
XForms
XAML/XUL

[Mark, presuming that Macromedia is still around…] Anyway, one issues w/ the current incarnation of Flash Lite is that it’s the functional equivalent of Flash 4 (+ some XML), which means that it’s not a 1:1 transfer. I don’t see Flash going away on the desktop regardless, but their footing in the mobile (and to a much greater degree) in the enterprise space is quite precarious (well, nonexistant in the latter).

CSS3 – yes, although IE won’t support most of it properly so it will be of limited use. Also, there will be 5 bajillion selectors, but still sorely lacking in descriptors, especially for relations and dynamic calculations so JavaScript and large amounts of structural tags will still be required layouts that could be done ‘the bad way’ in 1998.

SVG – yes, especially if Flash flounders, but even if that doesn’t happen, SVG will ride the AJAX wave (especially once SVG rendering makes it into Mozilla core). Having DOM-based structure is a beautiful thing.

Flash – see top

XML/XSLT (more generally than XHTML) – an increasing need for transforms will push XSLT into the web design world so that designers everywhere will be able to enjoy how bass-ackwards and almost entirely retarded XSLT is. On the bright side XSLT 2.0 will have regular expressions (hallef’inlujah) although most web designers’ eyes will glaze over. XML will continue to dominate everywhere as the format for everything. Much fun will be had as people continue using it COMPLETELY INAPPROPRIATELY EVERYWHERE.

XHTML – yes, and it’ll need to validate too if you don’t want your XSLT parser to barf.

XHTML2 – haha, no.

XForms – despite Eris’ nice try in the Flash vs HTML panel, I will throw my hat firmly in the “XForms sucks and should burn in RFC hell and die” camp. XForms accomplishes the standards trifecta of reinventing the wheel, solving problems no one asked it to, and simultaneously not solving the the fundamental problems people were having with its predecessor. I’ve sat through lengthy presentations and even lengthier discussions on this and really, I’ll never get that time of my life back again. Luckily for the rest of you, XForms has this much traction -> . and we’ll get a much saner WebForms 2.0 courtesy of WHAT group (fingers crossed). Banks will love XForms though. (never underestimate the power of enterprise software)

XAML/XUL – Yes, although it’s a bit unclear how much these will encroach into web-space. Regardless, these XML-based widget sets w/ custom rendering engines (to differentiate from stuff like Dojo and to include things like Flex and Laszlo) will probably take over intranets, giving all the maintenance and development benefits of the web, and all the richness that comes with being able to draw lines and pulldowns that overlap properly among other things (XAML/Avalon will be downright scary).

5. Will industrialization drive the independent web developer out of existence, with the economies of scale of big development houses and cookie cutter design making this role economically unviable?

No, independents will just have to adapt to provide services (and service) that big development house, cookie cutter design can’t. I think web developers will have an easier time than some, but as Daniel Pink would put it, it’s just a function of the transition to the conceptual age. (That, BTW was a kick-ass presentation, and done in a way that I even remember most of it – guess that levity, brevity, and repetition thing really works).

6. Which philosophy will ultimately win on the web: openness (open source, open formats) or closed proprietary systems?

From an technical perspective, due to the nature of the data interchange, “openness” will increasingly be a simple fact of life in the coming years.

From an economic perspective, due to the nature of market efficiencies, “openness” will increasingly be a simple fact of life in the coming years.

I don’t think there’s much of a choice barring complete and total hegemony (whether this is legislative or otherwise).

There will be closed networks, protocols, stacks, but these will by definition not reside on the “Web” but by and large in the darknets.

I was a bit disappointed here. Everyone mentioned going through hundreds of blogs, but then discussion was either about the “philosophy” of customizing your blog (whuzzuh??) or a tortuous demo of not new stuff (seriously, how to make a blog roll?). My ideal? Lets spend some time talking about the philosophy. Dunstan made some great points about just noticing the annoying things on other people’s sites to improve your own, but then, lets go through like 20 or 30 blogs and point out the neat things or the different ways to go about things.

I guess what I wanted just wasn’t what they were doing. Still, if I were doing a trick-out your blog panel, I’d definitely try to highlight the “this is so fuckin cool” feature list. Here’s a few off-the-top-of-the-head picks:

• Kottke.org – yeah, I’m lame, but like Paul, I’m totally hot on the single timeline… It’s so the future.
• Erik Benson – the moraleometer rocks, as does the whole erikkgregator thing. Again, it’s the future
• Subtraction – beautifully designed, and quite functional. I like the category map, the star rankings, the org, all of it (damn, Khoi was at SXSW?)
• unfoldedorigami – this was the first site I saw doing LiveSearch
• urlgreyhot – Michael Angeles’ site has undergone a redesign from the one I fell in love with, but this and all the Drupal sites represent a different type of blogging. (see also: Jonas’ blog or Zack’s blog (AIM comments, insane, in all senses of the word). There’s lots of potential (especially w/ how nodes can be re-ordered and grouped — more about this soon)
• FTrain – probably not as interesting from a Geegaws point of view, it’s all about the geekage (ontology building baby)
• Fiftyfoureleven.com – this is a new redesign, and I’m loving the layout, very well thought out structure. Similar in the direction that Dave or Scrivs (layout since changed)
• On the plain clever design front, Andy Budd, Shaun Inman, Dave, Ryan, and of course Dunstan have all been doing neat things w/ comment numbering, real-time previews, etc. etc.
• bigempty – and now for something completely different

So the idea is to get a bunch of Flash and DHTML experts to (re)implement apps in their respective technologies. In theory, wacky hijinx should ensue.

So, let me start with the good. Seeing the implementations are definitely interesting. Dunstan takes charge on converting a Flash mountain resort map/travel site into DHTML, Kevin does an implementation of GMail in Flash (Flex), each side gives a JetBlue redesign a go, and then each team does an ‘only’ version of Flickr (the joke here being that Flickr tried a Flash only version and switched off for good reason, right?)

Judging was done w/ a not-very accurate flash-based sound-meter. Obviously things were tilted rather unfairly against the Flash guys just because of the room composition. What actually really sucked about the setup was that there was no time to substantively discuss pros/cons, or even to correct gross inaccuracies. A lot of this is based on the fact that most of it isn’t “what can we do that they can’t do”, but what makes it easier. Also, many of the diferences aren’t ‘sexy’ in the traditional sense (searchability, structure, etc.)

Some notes:

• Dunstan had a tough row to hoe with the maps. It’s not totally true that you can’t do zoomable maps (see Google Maps, after all), but it is true that it’s a complete PITA. Also, as far as a more usable/accessible version, you definitely would design it differently, and I’m a bit disappointed that Dunstan left it there w/o even sketching something out (probably a time thing)
• Kevin showed off some nice drag and drop w/ his GMail reimplementation (although DHTML can do that too, there have even been good photo album implementations). The central attachment filing is nice, although AFAIK, Flash can’t do desktop drag and drop to the desktop w/o Central?
• Jaxon does some great stuff witht the JetBlue redesign, having an Expose-style zoom map of where you are in the process. I can think of even wilder ways to go with that, that simply can’t be easily done in HTML.
• Eris (whom everyone I met has said is incredibly talented, and who has a beautiful site) inexplicably did the “HTML” version in XForms. I’ve ranted on XForms before…
• For the Flickr redesign, Vera added speech bubbles into which is also just plain bizarre because Flickr’s notes are already in Flash (and DHTML can do scaling speech bubbles trivially anyway, div w/ contained elements for curves))
• Chris showed off lots of math foo, calculating offsets/bounding to replicate Flickr annotations. Very trick.
• Simon asks a great question in regards to Greasemonkey — Flash just doesn’t allow a user to do any of those sorts of injections. This is a good and bad thing, actually and leads to…

So, I sort of got a bit agitated at various parts throughout this panel because of the general FUD, and I started making a list of Flash/DHTML pros/cons. Now, a lot of things just have to do with ease, but there are some things where they’re just (potentially different), which can be both good or bad. Alex added some things, and then I lost the file, so here’s just what I can pull off the top of my head

+ Significantly better vector, animation
+ Multimedia: native sound! video!
+ Better compositing
+ Richer built-in widgets (Flex)
+ Significatnly easier compatibility
+ Works or doesn't
+ Higher market penetration
~ Harder to externally modify

+ Potentially webbier
+ Browser navigation
+ Source viewable
+ RESTful
+ Bookmarkable
+ Progressive Enhancement
+ Structured (important, esp for search)
+ Standards Interaction (DOM, XML)
+ Easier to modify
~ "access" to variety of plugins
~ Easier to externally modify

That being said, the DHTML webbiness is definitely potential (the Google apps so far for example don’t do that well on those counts) and that tendency may even be a liability (think about the positives and negatives of easy external modification [not that you can trust any client-side data 100%, but still]). Not, by access to plugins, that can be bad because it’s usually limited and buggy, but also good in the sense that these can overcome DHTML’s inherent weaknesses (the best DHTML sound engine I’ve found is Scott Schiller’s SoundManager, which pings to Flash to play sound).

Has someone done a comprehensive pros/cons on the various RIA options? Seems obvious, but I haven’t seen one around (feel free to send URLs now :).

Notes:

• Joe Clark live-blogged this pretty comprehensively