Category: conferences

Total 78 Posts

OSCON 2009: Antifeatures

I picked “Antifeatures” from a crowded slate of good-looking sessions strictly because one of the speaker’s blogs, Revealing Errorsh/t Tom., which aims to teach basic computer science through images of basic computer science failing spectacularly, is awesome, and feels like a close cousin to What Can You Do With This?.

Benjamin Mako Hill didn’t waste time getting to his premise: our usual cost-benefit analysis w/r/t free and open source software is flawed. We often evaluate software on the basis of cost per feature. On that analysis, Hill’s 1990 Linux machine fared poorly, supporting 20 CD drives, none of which was stocked by the computer hardware store. Hill suggested, instead, we evaluate software and hardware not simply on which has more features but also on which has fewer antifeatures.

An antifeature, as defined by Hill, is like a feature. It requires effort and adds functionality. It’s just functionality that no consumer would ever want. It’s functionality a consumer might pay to have removed, if possible.

Bugs are not antifeatures, just mistakes. The lack of a feature is not an antifeature, it’s just a missing feature. Poor implementation is not an antifeature.

He set up four classes of anti-feature, disclaiming that they aren’t definitive and they aren’t without overlap:

1. Features that extract money from the user.

  • You had to pay to get your name unlisted from the phone book. Arguably, it required more work and more material for phone companies to include you than remove you, but your phone number was worth money to them packaged in lists for advertisers, so they passed that cost on to youI’m referring to phone books in the past tense here and I don’t really know why except I can’t remember how many years it’s been since I used one..
  • Spyware / Adware. The basic version of Kazaa (and also the video encoding software, DIVX) came bundled with spyware called Gator which reported your browsing activity, recorded parts of your credit card number, and issued pop-up advertising. Byte-for-byte, Kazaa contained ten times more spyware than useful software. Both DIVX and Kazaa charged users for professional versions of their software which were functionally the same except they didn’t include the spyware/adware.
  • Sony VAIO. A terrible laptop, pre-installed with so much junk trial software, VAIO applications, and games that some users reported crashes from the very first boot. Sony offered a “Fresh Start” program at point-of-sale which would remove all of the junk to (in their words) “conserve memory and processing power while maximizing overall system performance right from the start.” For fifty freaking dollars. Stay classy, Sony.

2. Features that segment the market.

  • Airline price points. Airlines have priced their tickets in order to determine exactly how much a given customer is willing to pay and then extract it.
  • Microsoft Windows NT Workstation / Server. Microsoft made several appearance throughout Hill’s talk. In this case, Workstation cost $200 and Server cost $800.00 but were byte-for-byte the same except that Workstation contained code that would update one bit in the registry to limit incoming TCP/IP connections to ten.
  • Six editions of Microsoft Vista. Vista Starter only allows three GUI applications open at a time. Hill wondered here (paraphrased) how the engineer tasked with writing this antifeature code lives with herself.
  • The Canon G7. The Canon G6 could shoot RAW files. The Canon G7 had better hardware and better software but could only shoot JPEG. This was to push customers toward Canon’s line of digital SLRs. Users had no trouble hacking the firmware and re-installing that functionality. In a world where there is any freedom for users, Hill noted, antifeatures are very hard to sustain.

3. Features that create or extend monopolies.

  • Panasonic Lumix. A firmware upgrade “protected” you from third-party batteries, turning off power management in order to leave users scratching their heads wondering why third-party batteries are so terrible.
  • Dr. Dos 6.0. Microsoft detected its installation and threw up all kinds of error messages, leading users to believe it was incompatible with their systems. This one went to court.
  • The battery authentication market. Hill said that this is the dystopian future: software companies waging war on each other inside your computer, a war where the user is collateral damage.

4. Features that protect copyright.

  • SimCity’s list of high scores. Printed with extremely low contrast in order to deter photocopying. This measure also made it impossible to read.
  • Dongles that prevent software copying. Hill: “How do they sleep at night? No one has ever purchased some software and said [disappointed] ‘I got the new version of the software and there wasn’t a dongle!'”. Billions of dollars are poured into making computers less good at what they’re naturally good at.
  • DVDs & DRM. “Basically every kind of antifeature.” Encryption, watermarking, region control, licensing, can’t skip the first track, etc.

He concluded that freedom has practical benefits. It may not be a methodology toward faster software, better software, that crashes less. But even when proprietary software works perfectly, exactly the way proprietary software engineers want them to, they don’t work perfectly the way the users want them to.

OSCON 2009: What Are Your Session Dealbreakers?

The 4:30PM Wednesday slot was packed at OSCON. I’m talking about three sessions I was either “eager” or “very eager” to attend at a conference where 95% of the conference titles were outright inscrutable. (eg. “Sun GlassFish (OpenSolaris) Web Stack – The Next Generation Open Web Infrastructure” – see what I mean?) One session concerned graphic design. Another risk models. The third session listed as “Antifeatures,” a title which was tough to resist in its own right.

I told myself I’d pick one and sit through the first five minutes. If, at that point, it had met certain criteria, I’d bail on it for one of the other sessions.

If you are loathe to leave a session under any circumstances, consider yourself exempt from this writing prompt. Otherwise, if you value your time and you vote with your feet, how do you judge a session by its first five minutes? Again, it’s possible the session turned into a winner exactly six minutes in. Under these constraints, though, we don’t have the luxury of patience.

I’ll post my own criteria to the comments shortly.

OSCON 2009: My Ignite Presentation

I was constrained by twenty slides at fifteen seconds apiece for a lean five minutes to talk about whateva to a crowd of open source software-types. I talked about a) teaching, b) why my first two years were miserable, c) the difference between teaching math and teaching citizenry, and d) what excites me lately.

[Click through to view embedded content.]

If any other pecha kucha survivors want to commiserate over the format, which required (for me) 400% more rehearsal and 90% less slidework than I’m accustomed to, the comments are all yours.

BTW: A couple of people have asked for a YouTube embed. I tried, but the audio stopped tracking with the video. Here, instead, is a link to a high-quality file.

OSCON 2009: Overview

I attended O’Reilly’s Open Source Convention last week in San Jose, which wasn’t my usual scene. I gave a brief, five-minute talk Tuesday night (which I’ll post shortly) and then spent Wednesday and Thursday wandering between sessions, feeling alternately like I was a) bobbing along the surface of something awesome or b) submerged several hundred feet beneath that same surface, all depending on the complexity of the content. I stuck to sessions on design, usability, business-building, and data visualization. Those were pretty great but OSCON punished me, without exception, anytime I decided to get adventurous, like the session on e-mail protocols that I would have understood just as well had it been delivered in Croatian.

Scattered notes:

  • It’s nice to know that, even at 27 years old, there are still things I don’t know, that there are at least a few things I have left to learn.
  • Everyone was exceptionally warm and welcoming, even people who, upon subsequent Google background-checking, turned out to be something of rockstars within the community. I wonder if this is particular to the open source ethos or if, more generally, I just go to all the wrong conferences.
  • Seriously, Valerie Aurora did everything but sound out the syllables in Li-nux Ker-nel for me. Such is the patience with noobs here.
  • There were no handouts whatsoever at this conference. I asked a few people about this and they looked at me like I was high.
  • Keynotes were fifteen minutes long. Sessions, forty-five.
  • The gender balance here is inverted at education conferences.
  • Everyone had an opinion or an anecdote to share about teaching. It was easy to lure someone into conversation by asking her to elaborate on why her ninth-grade science teacher was so good or bad. I had these conversations all throughout the conference, all throughout the convention center. I can’t imagine water management engineers enjoy this sort of ready social icebreaker so chalk one up for teaching.
  • Open Source Hero I: Clay Johnson, director of Sunlight Labs, who aims to make meaning out of the deluge of data from data.gov.
  • Open Source Hero II: Michael Driscoll, who makes awesome visualizations of huge data sets using the statistical analysis software R. Check out his six-dimensional analysis of baseball pitches.

Asilomar: Closing Remarks

From my comment at Jon Becker’s blog:

I’m having a difficult time determining why I walked away from three days at ILC so deflated while three days at CMC-North has me bouncing off walls.

It may be an issue of interest, true – I care more about math ed than tech – but give the same presentation outline on (eg.) “Wildlife of the Serengeti” to any randomly selected presenter at CMC-North and any randomly selected presenter at ILC, and my money’s on the CMC presenter every time to put up something lively, engaging, visual, and audience enfolding, scaled up by thirty extra minutes no less.

There is a lack of substantive criticism in the edtechno-blogosphere, I think, which is mirrored in these presentations, where people hop up to the front of the class for an unfocused, but definitely emotional, show-and-tell, and few people either care enough or have enough temerity to suggest that higher standards should apply. The feedback mechanism is, by and large, overly polluted by emotion.

By contrast, the crowds at CMC-North are vicious, though constructively so. I appreciate this and I know that overall session quality has risen to the occasion. As educators, the stakes are too high and the time constraints too stringent to settle for anything less than our best efforts, even if hearing that we shouldn’t lecture from bulleted slides for an hour is painful.

Jon’s response is entirely on point and he invokes my new favorite word, which I am trying to wedge as much as possible into my daily correspondence.