Just some random thoughts of yet another someone who went to OCAD

When people saw our Blind Reading installation, they often ask, “How does it work?” The fact that the piece is an art installation and not a design artefact must not be obvious. Even the speeches alluded to results of design research (when there is, as far as I can tell, only one product design presentation in the whole exhibition). Even the official blog of the exhibition sounds as if they got it wrong and thought our art piece was a design artefact. But at least the installation held up for the evening. (The realization that, other than the incorrect wiring, the root cause of our problem was mechanical in nature helped immensely.) People did ask, we tried to explain, and I hope we got our points across. I still fear Wednesday: I have never been on this side of the table at an artist’s talk before, and I have not gone to enough artist’s talks, but my consolation is that I will probably not be the one doing most of the talking.

There were people inside working on their artwork, and I was able to get in. Hopefully this will not take more than 8 hours…

One thing that we might have done wrong is to have connected a 3V stepper to the Adafruit motor shield and using a 5V supply. But on further investigation this might not be as simple as that. According to the data sheet, the 57BYGH420 is a unipolar stepper. However, a page Google found suggests that 6-wire steppers can be connected either as a unipolar or bipolar motor. (It also tells me that our motor is a NEMA 23.) Indeed, according to Stepper Motor Connection Options on piclist.com this is indeed the case, and to be precise 6-wires can be connected either as unipolars or a “bipolar-series.” What’s interesting is that what Angela did to the two centre taps was in fact correct (whereas Adafruit’s documentation is not): To configure a 6-wire stepper as a bipolar, the two centre taps should “not [be] connected to anything.” That is probably why when I tried connecting the centre taps to GND (as per Adafruit’s instructions), the motor stalled. It is also worth noting that the page says that if we connect a 6-wire as a bipolar, we get twice the resistance than as rated and 4 times the inductance, and require 70.7% the current and 141.4% the voltage. Which would mean the 57BYGH420 as a bipolar would be actually a 4.23V motor at 1.414A. So it in fact needs a supply that can deliver approximately 5.43V. Our 5V supply is low, but not by so much. A 6V, but not a 9V, will probably do better.

They know I am on OCAD’s beta program and not on a paid account, but that is still not a reason to ignore my suggestions. Their user interface has a steeper learning curve than Behance (and to be blunt a few places are outright non-obvious), and if they keep throwing their strengths away I’m not sure how they plan to stay in business.

---

Date: Sun, 17 Mar 2013 00:23:37 -0400
Subject: Using 4ormat as a writing portfolio (please reply, as promised in the Help page)
From: Ambrose LI <ambrose.li@gmail.com>
To: info@4ormat.com Hello, in the “Help” section you promised that “a member of the 4ormat team will get back to you with an answer.” After more than a month, I am still waiting for a response to my previous suggestion. Last month I suggested a way to make it easier for those of us who have chosen to use 4ormat as a writing portfolio. Not only have you ignored my suggestion (and without giving me an explanation), you have just made changes that make it even harder for people to use 4ormat as a writing portfolio. I find this utterly incomprehensible. It was very difficult to use Behance as a writing portfolio, but things have improved on their side. When I first signed up with 4ormat I thought you guys had an advantage over them because despite their improvements things were still way easier (and much more flexible) here. But instead of building on that strength, you guys are throwing it into the trash. When a classmate told me two days ago that he had just created a (graphic design) portfolio himself, from scratch, using WordPress, I could not understand why he would do that when a ready-made solution like 4ormat exists. But if sites like 4ormat keep changing for the worse, I suppose his decision was justified. -- cheers, -ambrose <http://port.ambroseli.ca>

So apparently last night my Mac didn’t die; it’s just that the blinds were moving so slowly that the AJAX frontend thought the installation was not running. But right now something really isn’t right: The last status message the web backend received was from 4:32pm. That was when I headed out to Active Surplus to buy some pulleys (which I never got to try installing because my professor was using the room when I came back. Sigh). So, judging from that timestamp, I most likely had just forgotten to restart the python script. Or maybe the Mac really overheated this time, since the ssh process seems nowhere to be found either…

So I have been standing here, outside 240, filming the installation for the past half hour, and I just want to say that ths experience is entirely different from the experience I had inside, let alone the virtual experience on our web site. The tweaked installation runs incredibly slow, almost imperceptibly slow. On the site that would be unbearable, but outside, on the actual official viewing spot, it is calming. Well, the site is not even working. My old Mac must have overheated… Sigh…

Our professor was suggesting that we could document our experience of having a geographically separated team designing our gizmo for our installation. But I have this feeling that our biggest problem isn’t really our geography, it’s the simple lack of access. Our biggest progress was made approximately during the last one and a half weeks, yet we saw several days where Angela was in town but we could not do any work on our installation either because the room was in use all day, or simply because the office separating the elevator lobby and our classroom was locked and no one was available to open that door for us. We literally wasted at least four days because we physically could not get into our own classroom (in a building we were told would have access 24 hours a day, 365 days a year). It’s not the remote aspect; it’s the “last mile”—or, in our case, literally, the last inch (that is, the thickness of a locked door). So I could not even mount a set of information panels I freshly made yesterday because when I got back (from the cold, outside, in Butterfield Park, because students in our program apparently have no access to proper, ventilated studio spaces, and for obvious reasons I was not going to use rubber cement or spray fixative in the only studio space I actually have access to—I don’t want to literally blow up the studio we all love) the door was already locked. In hindsight, during our group meeting yesterday we could have easily defended our (unreasonable, I admit) slowness—we could not make any progress not because Angela was not in town, nor because I could not understand her code until I got my own Arduino, nor because of the steep learning curve of figuring out how to drive the stepper, nor even because the wheel was slipping and we could not make it not slip; we could not make progress simply because we could not get into the classroom to work on the installation. This has to be the stupidest reason ever why work could not be done, but it’s true, and it’s sad. And it has absolutely nothing to do with us allegedly treating this “real-world commitment” as mere “schoolwork.” And I fear this will repeat in our next installation, because 49 is another space we simply cannot get in.

Yesterday I finally remembered. I was talking to the professor a few days ago and thought there was a third point that I forgot, and indeed, there was a third point that I forgot. So here it is: We keep talking about agile, but agile values “working code.” And one way people doing agile keep their code working was to use TDD (test driven development) or BDD (behaviour driven development) techniques. They iterate quickly, but they don’t iterate in a vaccuum. Before they code, they write the test first. (Or at least that’s my understanding after taking that Coursera course.) To do agile, we have to first have our success criteria—a fluid set that changes over time, of course—set down. Success criteria, however, are sorely missing in our case: We just don’t have them. We know something is wrong, but we haven’t really defined what we mean by right. So no matter how short our iterations are, we still can’t be doing agile; if there’s a word for what we’re trying to do, we can probably call it hacking. And yes, this came out of that project too.

Unified English Braille distinguishes between many things that Unicode doesn’t distinguish. For example, UEB distinguishes between the seconds (of an angle) sign, the inch sign, and the double prime, whereas Unicode use the double prime for all three. How, then, can an automatic translator from Unicode to UEB be designed? Wouldn’t this require artificial intelligence?

Ever since I found out that it’s impractical to read PDF files on a laptop on the train, I have been thinking that’s the use of tablets and smart phones. So I tried taking a tablet and see how well it would work out. In short, it works, but not too great. To make it clear, the reading part was absolutely fine. I was able to read at the bus stop, on the bus, and on the train. No problem at all. The problems were with the highlighting. I couldn’t highlight individual words. When I tried highlighting phrases, the highlighting would often spill over to neighbouring words too (often starting from the left edge of the page). Perhaps it’s an Adobe Reader problem, but the highlighting was extremely imprecise. The bigger problem, however, was when I took the train back home after the Tyler Bruler lecture: When there’s not a lot of space around you, the sleeve of your jacket would often be the thing that touches the screen. It was very difficult to click (often the screen would zoom in instead). It was even more difficult to highlight (even if I managed to highlight something as soon as my sleeve touched the screen the highlighting would go completely berserk. I had to be so careful touching the screen that even the trackpad on a laptop was easier to use. So, if even someone wearing a winter jacket had so much difficulty using a tablet, I can’t imagine touch screens being the solution to all problems.