Ben’s posthaven

"Contrary to the foolish notion that syntax is immaterial, people optimize the way they express themselves, and so express themselves differently with different syntaxes." - Erik Naggum

This comes today, just as I'm rediscovering some of the worst offenders of Python's syntax. For example, accessing an instance and class variable "var" are "self.var" (5 chars) & "self.__class__.var" (15 chars!), respectively in Python and "@var" (1 char) & "@@var" (2 chars) in Ruby. The alternative of referencing through the class-name is hackish too as the behavior can't be changed by the inheritor! It's enough to make me want to avoid using class variables! "".join(list) and global functions are other examples counter-intuitive syntax.

Now, I've plenty of love still for Python, I attended PyCon 2006 and am happy to see the new "with" syntax comming in Python 2.5. I just would like to se Python 3000 do better.

I channelled my earlier frustration about the Blanton Museum into some social activism. Specifically, creating a Facebook group about the whole debacle. The result: "Herzog & de Meuron can make my Blanton any day!".

Spurred on by creating it, I collected a good amount of quotes surrounding the whole issue. I've put up a good deal, just from the print media available on-line, not even including architecture magazines! Anyway, I'll be putting up everything I've found there. For you, I have a collection of quotes:

But for all the time, money, and hotheadedness, the final incarnation of the museum is strikingly unremarkable, [...] The new Blanton is just a building, not a masterpiece. - Texas Monthly, 4/2006

Larry Speck on his Resignation:

The departure of [H & dM] was not a simple matter of personal difference, They were treated badly and were put in positions [...] that no top-flight architect should tolerate. - Larry Speck, Dean of the School of Architecture, in his resignation letter

The Regents:

I have a big problem with flat roofs, I've never seen an undulating roof, I am glad that they have [resigned] so that we can [...] select an American architecture firm that [...] understand[s] the cultural significance of the project. - Regents Rita Clements and Tony Sanchez, on rejecting the first and second H & dM designs, and in response to the firm's resignation

About Herzog & de Mueron:

One is hard put to think of any architects in history that have addressed [...] architecture with greater imagination and virtuosity - On H & dM's recieving the Pritzker Prize, the "Nobel Prize" of architecture (2001) The strength of our buildings is the immediate, visceral impact they have on a visitor. - Jacques Herzog

About the New Blanton:

[The museum] certainly won't go down in the history books. - Kevin Alter, associate dean for graduate programs in the School of Architecture (2001) For a building that's going to house famous artwork and because of its location on campus, I think the museum needed to be more prominent in design. - Tray Toungate, Vice President for the Architecture and Planning Council (2001)

I was reading about "Template metaprogramming" on Wikipedia tonight this morning and I noticed this line which got me chuckling.

The syntax and idioms of template metaprogramming are esoteric compared to conventional C++ programming, and advanced, or even most non-trivial, template metaprogramming can be very difficult to understand.

Maybe it's just late early, but lines like that get me thinking of the person who must have written it. I can see him hunched over the keyboard now, his little mustache twitching as he types, his glasses illuminated by the screen.

I'm feeling rather sad at the moment for my soon-to-be alma mater, The University of Texas at Austin. 1998: Herzog & de Meuron are chosen, museum to be "an architectural landmark":

"Herzog & de Meuron is known for creating highly imaginative buildings that stand in harmony with their purpose, materials, and site," remarked President Faulkner. "The new Blanton Museum will be not only an architectural landmark for The University and the city of Austin, but a building that will engage students, faculty, and visitors from all backgrounds with the world's artistic traditions."

1999: Architects resign in frustration w/ the Board of Regents. Dean of School of Architecture resigns in protest:

"The tortured saga of the Blanton Museum of Art has a new and painful chapter. The latest involves Lawrence Speck, dean of the UT School of Architecture for the past six years, who resigned that position on Monday, November 22, in protest of the departure of prestigious Swiss firm Herzog & de Meuron Arkitekten from the Blanton project."

2002: Why did this happen?:

"I didn't ever see any [Herzog] designs that I felt would fit with the architecture of the campus," said Rita Clements, UT System regent and chairwoman of the System's Facilities Planning and Construction Committee. However, neither the campus' most noteworthy buildings nor those buildings surrounding the future museum are built in the campus master plan's style, said Kevin Alter, associate dean for graduate programs in the School of Architecture. The campus plan leaves room for architectural innovation for special buildings, Alter said. Alter expressed disappointment that the regents took a conservative approach to the museum."

2006: Regents choose a design for the Blanton museum which is "strikingly unremarkable"

"But for all the time, money, and hotheadedness, the final incarnation of the museum is strikingly unremarkable. Yes, its Texas granite base, epe wood overhangs, and sixteen-foot-deep arcades are lovely, but they're, well, decidedly traditional. The new Blanton is just a building, not a masterpiece."

Now how could that be? Let's review...

In response to a Slashdot posting on "The Best of Web 2.0", I visited some of the honored sites. As I'm prone to do when I'm curious about something, on one particular site I went to browse the source and was rewarded with the following:

interested in joining meebo? we have full time and internship oppportunities available including front-end software engineer (JavaScript/DHTML/Ajax/C/C++), back-end software engineer (C/C++/networking/databases), and visual designer. email secretjobs at meebo dot com with your resume and cover letter. thanks!

I'm not particularly interested in the position, so I place the offer here. Enjoy! In a side-note, it's interesting to watch the url, where you can clearly see Meebo's load-balancing at work. Load http://meebo.com and you'll be redirected to some variant of http://www\n\n.meebo.com, where \n represents a number (in reg-ex). It seems they're airing their load-balancing for the world to see. I sort of appreciate the open-ness in a way. On a second note, what the heck is up with compressing the html into a single line? Is this to save on bandwidth or just obfuscate the code? If you happen to know, I'm once again curious Oh and my apologies to Meebo for airing their secrets...

I'm now using RSS to feed photos from my Flickr account to the side-bar of this blog. I take pride in my photos, particularly since they're taken using my Sony-Ericsson K-750i 2MP camera-phone. Currently the ones showing are from my trips to The Modern and the Nasher Sculpture Center, both in Dallas. Check them out!

Ever since I picked up Ruby on Rails, a Ruby-language web framework for a project of mine, I've been curious about what web frameworks were out there for Python, another of my favored languages. A few of them out there are Django and TurboGears. Another that happened to catch my attention was the web.py a framework by Aaron Swartz whom I recently met at the first (of hopefully many yet to come) Startup school in Boston. Anyway, I was browsing the source of 0.13 of Aaron's framework and noticed something amiss. Specifically, in storify, Aaron was testing isinstance(k, list) on a value k which he had just used to index into an array. Now in Python arrays are mutable, and arrays and dictionaries can only be indexed by immutable values. That's ints, strings, but not lists. Seeing that something was amiss, I sent him a quick e-mail, and now I'm in his changelog entry for version 0.132. I'm reproducing the relevant line here for my future self: Fix bug with storify when it received multiple inputs (tx Ben Woosley). In other news, Apple just announced it was recognizing contributors to their WebCore HTML Layout engine, with gifts of computers and free tickets to Apple's WWDC. They list none-to-trivial contributions to on a blog posting on David Hyatt's blog Surfin' Safari. One could call these events arguments both for the model of open source, and for code reviews as well. But then my evidence is anecdotal, so you're welcome to your own interpretation.

I'm now running the latest beta of Internet Explorer 7 alongside Firefox, Opera, and (even) Amaya. I've run up against many of IE flaws, so I'm glad to say that quite a few have been fixed in the latest release. I'm even limiting my special, conditional comments-enable css hacks for IE down to version 6 and earlier. Now, it's not perfect, the IE team still has a lot of work ahead of hit, but after years of stagnation it's nice to see some progress. In addition, I'm actually pretty happy with the user interface. It's a nice departure from the past, and while I can tell there's still some "old IE" in there, I'm happy with the direction it's going. I'm not dropping Firefox anytime soon, but I'm really glad to see some fresh change in the works at Microsoft. Now if we could only get some nice application/xhtml+xml and image/svg+xml implementations out of Redmond...

I'm faced with an interesting decision for my last semester of college. I'm required to do a senior design project, which might seem like an excellent chance to do whatever I want, but unfortunately the UT assumes ownership of any IP developed through the project. This means I can't or rather, won't, use any of my pet projects. As it would happen, I'm left with two very different options. The first is to take on a very bold, challenging project with a good friend of mine. The subject has just recently emerged into the main-stream and promises to be very important for at least the next five years. It's enticing, but possibly would be too much to handle well, in addition to my own projects, work and classwork. So I'd have to drop something else to accomodate it... The second option is to work with my current employer, SigmaTel, on a project which is useful and will be interesting in the sense that it's my code that I "own" that I've written and designed from scratch. The problem is interesting but lacks the gee-whiz factor of the former project. Of course there's a big bonus in the fact that I'd be getting paid to do it for SigmaTel. And, in the end, combining work and my Senior Design project will make for a much more livable semester, and may even give me more time for my other exploits. So, in review, I'm selecting a semi-mundane, paid-for project over a bold, daring (but possibly over-reaching), unpaid project with a hot new technology. So am I selling out? Should someone with bold dreams throw caution to the wind and take on the big problems, no matter what?

One cool aspect of the public nature of the web, is that you can watch some of the most effective people up close. For an aspiring hacker, this leads naturally to measuring oneself against their accomplishments, or to be more exact, comparing what you've done to what they'd done at your age, or vice versa. I've done this, with some mixed results. For example. Aaron Swartz is a number of years younger than I, and he has, among other things, just released his first web framework. On the other hand, I've five years left before I have to achieve the like of Ruby on Rails creator David Heinemeier Hansson. Now one could ask whether I actually expect to do so. I honestly don't know what I'll do these next few years, but I do know I won't do much if I don't aim to do much. That said, there's a refreshing familiarity to the thoughts I read in Hansson's "Loud Thinking," particularly the early years. What more indication do you need? ;-)