Monday, December 8, 2008
Tuesday, December 2, 2008
Maybe it's a sign of the economic times?
Thanks to Jeff for continuing to inspire. I really should take his cue and start making some small apps...
I know what I'm thankful for tonight: the possibility of no more PM Harper come next Monday!
Saturday, November 1, 2008
For reference, here are Bertini's 5 False Myths (I think 'false' may be a redundant word here):
- FM1 - InfoVis is about data exploration
- FM2 - InfoVis is about discovery
- FM3 - InfoVis is about new visualization techniques
- FM4 - InfoVis is about vision
- FM5 - InfoVis is about the data
Well, this is a very interesting list. I agree with many of the points, especially the second class way our community seems to treat interaction. I think interaction design should play as central to good InfoVis as the visual design.
I'm not sure what I think about FM1 and FM2 -- I think some combination of discovery and exploration certainly is undertaken by real users using real infovis tools, especially in the bioinformatics and intelligence domains. I've seen this sort of exploration myself in ethnographic studies of scientists at work with ad hoc visualizations. But, yes, these are not the only reasons for doing InfoVis research.
Some work is moving forward on FM3 -- we are seeing more studies of how people see data and use visualization (e.g., Van Ham & Rogowitz at this year's InfoVis) rather than always creating "new" techniques. My own work from last year's conference was about combining existing techniques to leverage the benefits linking of multiple visualizations of related data.
Finally, I think FM5 is my favourite -- I'm a big proponent of the "human-in-the-loop" decision making model, often advocated in the CHI and CSCW communities. I think we need to create interactive experiences that aid task completion through a blending of the user's world knowledge and any new information present in the data to be visualized. We shouldn't make presenting the right answer from the data our goal, as we won't be able to do it without solving artificial intelligence in a real way. And if we could do that, we could just solve problems algorithmically and wouldn't need InfoVis. Bertini is right in that most tools do not take account of any prior knowledge, but there are some good examples from the VAST community where prior knowledge can be explicitly entered into the analysis process (e.g. i2's 'Analyst's Notebook' or IBM's Research's HARVEST project). My U of C colleague Torre Zuk has also done some analysis of how a physician's prior knowledge affects their decision making when presented with a visualization. This is certainly a challenging and fruitful area for more research.
Thanks Enrico the thoughtful posting!
Monday, September 15, 2008
As a result, I think I'll be sticking with the NDP. The Green Party argument that
it is "the largest killing of marine mammals anywhere on the planet" is vacuous. Just because something is big doesn't make it qualitatively different from something that is any other size. It is hypocritical to argue against the seal hunt unless you advocate against all animal harvest, and I don't think they do. That would be a more defensible position than singling out a particular industry because it has a bad reputation.
The perceived cruelty of the seal hunt is mostly just misrepresentative propaganda of the IFAW, showing a few sealers committing crimes for which they were charged. Governments are supposed to work with the facts, not the rumors and misinformed ideas of foreign countries and lobby groups. It seems the Greens wouldn't be up to that task after all. I also disagree that the local impact is minor. The annual seal hunt in Newfoundland provides a living for many rural families who work very hard to provide a product for which there is a market. Any party that doesn't support a sustained and managed sealing industry doesn't get my vote.
Their response, in full:
Thank you very much for your question regarding the commercial seal hunt in Newfoundland. Also, thank you for your support in getting Ms. May in the debates on October 1 and 2... it was a real showing of Canadian democracy at its best.
The Green Party of Canada has developed an extensive vision for Canada entitled Vision Green. If you are interested in our complete platform please visit www.greenparty.ca and click on "Issues".
The seal hunt is viewed by many people in Canada and abroad as an inhumane activity that is not ecologically sound nor sustainable. This is particularly true now that climate change has increased pup mortality in spring due to a lack of, and thin ice.
The annual seal hunt is the largest killing of marine mammals anywhere on the planet. Its enormity threatens Canada's overseas reputation for little local value.
The Green Party does not support a commercial seal hunt in Canada. We are not opposed to subsistence hunting by aboriginal peoples and local communities. However, we consider seal hunting, like whaling, to be a threat to the marine ecosystem. The loss of ice due to climate change threatens seal populations and exacerbates what many believe is an already unsustainable level of hunting:
Again, thank you for your concerns about the seal hunt and taking the time to email us. If you have any questions do not hesitate to email, call or visit our website.
Green Party Information Centre
Centre d'information du Parti Vert
Vote for Tomorrow
Votez pour l'avenir
Friday, September 12, 2008
"Dear Samsung Customer,
Occasionally we send email messages with news and special offers. If you would prefer not to receive such messages
in the future, simply remove yourself from our email list by selecting the option below and clicking the 'Update' button.
 Please remove me from your email list.
[X] On second thought, please continue to notify me of new products and promotions
 Please change my email address to: ________________
Yes, they pre-select an "I've changed my mind" button when you click "unsubscribe" in their email. If I hadn't been paying attention, which is entirely likely, I would have assumed it was the usual blah blah and just clicked "update", thinking I was getting out.
What I really don't understand is who Samsung thinks will actually click to unsubscribe and then change their mind. Probably no one, but everyone who sees this smarmy form will be slightly ticked off at the audacity of pre-selecting a form to do exactly the opposite of what you've already indicated as a preference. It's subtle, but now I'm annoyed at Samsung more than I was when they were just spamming me. Annoying potential customers is not a good business practice!
It got worse when I pressed "update" on the form:
"We're sorry to see you go. Please allow 7-10 business days for us to process your request."
7-10 days to process a single database command? Are they still using punch cards or something?
Tuesday, July 22, 2008
I understand they want to make their product sound appealing, but I'm so sick of seeing announcements that say things like new fees are being introduced to "serve you better", or features are being removed from a service to "address customer needs". Just call a spade a spade: features are removed, prices are raise, to make more money. The changes probably still wouldn't make me happy, but at least I could respect the company for respecting me.
The latest example of this comes from President's Choice Financial, Canada's big "no fees" bank. PCF services have been gradually reduced since it's inception -- I used to get free cookies and "P.C. Points" for my transactions. Now, I get ever increasing fees for "special services". The newest special service will be depositing cheques drawn on US banks (which I do often, given that I work in the US part of the year). What really gets me though is this sentence from their notice: "Please select the “view fee changes” button below for the upgraded costs." They need a dictionary: upgrade = to improve what was old or outdated (WordNet). I guess these fees are "upgrades" from their point of view, but certainly not the customers'.
Friday, June 27, 2008
So, when I took money out of the ATM and when I spent $5 at Starbucks, I was actually charged these fees. Stupid me for thinking that when there was no money available, the ATM wouldn't give me money (there was still a hold on my first deposit). At the very least, a warning message to the effect of "this transaction will incur a $35 fee" should have been presented. In Canada we get warnings of this sort all the time, e.g. when you use a different banks ATM and they disclose the fee.
I called Bank of America to decline this service recently, just in case. However, you cannot decline this "courtesy service". I doubt many people would view a $40 coffee as a courtesy, especially when they could just choose to use cash. Covering a rent cheque, ok, charging for ATM transactions, not ok.
There is hope -- the Federal Reserve is considering mandatory opt-out availability. Opt-in would be better, but this is a good step.
Wednesday, June 25, 2008
Anyway, I brought home my laptop tonight because I was still so frustrated with this stupid slow XML parsing problem and I couldn't put it away. I started running my test program when not connected to the internet, and it generated an exception saying it couldn't connect to www.w3.org. So, I was thinking, why the heck is it doing that? I'd already set the parser to be non-validating.
It turns out the parsers all fetch any external DTD that is referenced, even if the parser is non validating! So for every file, the header line was referencing the XHTML DTD and it was downloading it from an external site. So, if you have an abnormally slow XML parser, maybe this is why!
This feature disables that, in case you are ever parsing xhtml in the future and don't care about validation:
xmlReader.setFeature("http://apache.org/xml/features/nonvalidating/load-external-dtd" , false);
Now it parses 100 trials in 250ms total (down from about 55 seconds). And, I guess my program has stopped retrieving the same XHTML DTD from www.w3.org thousands of times an hour (sorry w3.org, it wasn't a DoS attack, I swear).
Phew! What a waste of time. I would never have thought to check for unexpected network connections. Glad my internet wasn't working at home.
3 days of work, help from colleagues, testing various parsers, running on various machines, profiling my code, ... for a 1 line fix that I found by accident. Sometimes I hate computers.
Now, time to move on with my super-cool-top-secret project. :)
Monday, June 23, 2008
Well, it's gotten to a really crazy level here. I was in the shopping mall the other day and I bought a water at a little convenience kiosk (the type of place that sells snacks, lotto tickets, and cigarettes). And, to my surprise, there was a "Tips appreciated" jar. Tips? For taking my money for the water I got for myself? What's the going rate for that? 20%? What then does the grocery cashier deserve, after handling so many products and packing them up carefully? It just makes no sense.
I can't imagine that the amount of tip money generated by this can make up for the shame of begging for tips when you actually offer no discernible level of customer service.
Macleans has a nice article about this, that I think is right on. Maybe I should send the author a tip.
Thursday, June 19, 2008
I recently received some mail from the US, and the postage was canceled with a skull and crossbones postmark. It looks rather creepy and unusual. I searched online but couldn't find any description of a special canceling stamp being used in Chicago, or any images of this stamp.
Does anyone know what's the story behind this? Is it just a normal thing? I've never seen anything like that before.
Thursday, June 12, 2008
I love working at IBM Research because there are cool things happening all over the place -- reminds me of being at a university lab in many ways. It's really fun, energizing, and, yes, even interns get paid to do it. One of those cool things, although officially a personal project supported by IBM, is Jonathan Feinberg's "Wordle", a beautiful, highly packed, word cloud, launched online today. Think tag clouds but more artistic.
In many ways, this Wordle is very similar to a visualization I have proposed, called an "author fingerprint visualization", in that it has a unique looking structure for each individual dataset and uses free text as input. From talking to Jonathan, however, it seems that his layout is random, and I would want something deterministic and stable, so that the same text would create the same or very similar images each time, like a real fingerprint! Also, Wordle uses simple word counts to determine sizes, whereas we have envisioned a series of measures based more on information content.
But, it's more beautiful than anything I could make, and the website, launched today, is just buzzing with activity. Check it out!
Monday, June 9, 2008
Incidentally, here in Cambridge, MA, they plant the trees with narrow metal gratings around the roots, so water can enter. They seem to be doing very well, and don't cause any disruption to the available walking space on the busy sidewalks, while still allowing for rainwater irrigation and some air exchange in the soil.
Friday, April 18, 2008
As much as I'm loathe to say this, Starbuck's is doing something innovative. They are collecting customer (and employee!) ideas of "doable" ideas to improve Starbuck's, and inviting everyone to vote on their favourites. Some are have already been implemented, others are coming. I've often complained that Starbucks doesn't ask if you want your coffee for here or to go, and they don't recycle -- both ideas are high on the list of votes. While this isn't the same as a corporation being responsive and responsible to the community (the shareholders still rule), it is a step in the right direction. And, a good PR stunt, as evidenced by this post. :)
My Starbucks Idea
Tuesday, March 25, 2008
Take this printer setup box for example:
Under "HP" printers, there are 20+ "pages" of printers listed. The scrolling resolution on such a small scroll bar with such a long list is awful. Given every printer name starts with "HP", you can't really hit the first letter or number of what you want either. Having installed the three printers I use maybe 6 times over the past months thanks to Vista printing issues, I've become all too familiar with this long scrolling operation, and now I've resorted to using the keyboard.
This window should be much bigger, giving a better overview allowing faster "jump to" the area of the list you want. Also, a bigger window would increase the resolution of the scroll bar. The availability of screen space is obvious in this screen shot, from one of my two monitors:
Windows certainly knows the resolution of my screen and could size this box accordingly, if they really don't want to allow me to have a resize handle.
I strongly believe it really is the little things that add up to ruin an interface experience.
I sent the below letter to the Prime Minister. I do not claim to be an expert on the region, or on what the state of Tibetan human rights during the reign of the monastic government, but I think what is happening now is truly abhorrent. Why do we ignore Tibet? Darfur? Congo? Guantanamo? Canada should be a country standing up for human rights.
Dear Prime Minister Harper,
I am writing today to express my concern for the people of Tibet. For many years I have wondered why the developed countries of the world, like Canada, have ignored the plight of that country, suffering under occupation by China. Recently, it is becoming increasingly clear that the occupation of Tibet by China is unjust and against the ideas of freedom and human rights that the people of Canada support and the privileges we enjoy. Why does Canada do nothing to support Tibetan human rights? If another country had invaded Canada, I’m sure we would want someone to come to our aid. Imagine losing our free press, our rights to assembly and free speech, and having yourself, the legitimate head of government, exiled to a neighbouring country. Of course, it’s unthinkable. The situation in Tibet is, to me, equally appalling. I understand that perhaps it is too late for a "free Tibet", but we must advocate for the fair treatment of Tibetans.
I’m appalled to see the treatment of the Tibetan people as they make peaceful protests against their oppressors. That the Chinese government blames His Holiness the Dalai Lama is just absurd and insulting. No one inspires more people to lead a peaceful path than him.
Please support the Dalai Lama’s call for a United Nations-lead investigation into Chinese human rights abuses in Tibet.
In addition, I call on you and all political leaders and other prominent Canadians to decline to attend the Beijing Olympic Games in protest. It is not right to celebrate the Olympic spirit in a country that is clearly not making sufficient progress towards supporting the ideals put forth by the Olympic organization.
Friday, March 14, 2008
"follow-on update releases will no longer be listed as separate items in the Windows "Add or Remove Programs" dialog"This is really great! When I did the update I had to remove 4 versions of JRE and 3 JDKs. Generally, I only use the latest one, but all the others are kept around, cluttering things up. With my OCD, this change will save me some time. :)
Thursday, March 13, 2008
I've had a vastly different experience with each. My laptop is always up-to-date, and the update program has never failed, performing bios flashes, software upgrades, and even installing some new features. It's not always present in the system tray, and does not send me stupid and useless messages. From the first day, the Dell update program has been sending inane popups like ads for a new wireless mouse, a notice about upcoming daylight savings time, and update notifications for hardware that I don't even have. I don't need such interruptions! I just want driver updates. The program has the "express service tag" of my machine -- doesn't that supply all the info on what hardware I have? Worst of all, whenever I click any of the messages that I might actually need, two annoying things happen: first, I get a popup asking me to inform Dell that I've read the message (which I decline); second, the support program reports and error, saying it can't find "Windows NW\applications\wordpad.exe". As you probably know from previous posts, I use Vista. Anyway, why does this program want to open WordPad? Perhaps it is opening some file type that has been registered to a non-existent version of WordPad? I don't know. But I also can't ask anyone. You see, despite the fact that U of Calgary purchased long-term extended warranty support on this machine, it only covers hardware. So, when a piece of Dell-supplied software doesn't work, I'm asked to pay $129 for a support subscription, on a machine that's only several months old. And, if what I've read is true, if I managed to read Dell support they'd just read me a script telling me to restart the computer and update the software anyway.
What's going on with customer service these days? In the meantime, I've uninstalled all the junky software Dell sent, and I'm doing updates manually with my service tag and their website.
Saturday, February 23, 2008
There is no "back"; there is no way to see what happened. A box appeared and disappeared before I saw it, and now the install is cancelled and must be restarted.
Argh! As I'm writing this, it just happened again! I was running the install and it went to completion. It popped up some sort of confirmation box and as I was typing, again
I think if an application has to steal keyboard focus, it might as well steal visual focus too. As much as I dislike the disruptive full screen blankout of the User Account Control system in Vista, at least it gets your attention. The problem with that one is that you can't say "go away, I'll decide later". Something visually unavoidable (even on 2 monitors, which is where I have most trouble with this) but with the option to minimize it for later would be great. Dialogs that appeared on the currently active monitor would also be better. Additionally, when a dialogue box steals keyboard focus, the OS should be smart enough to recognize whether you were vigorously typing before the automatic focus shift occurred, and ignore keyboard events until a short pause in typing occurs, indicating the user probably has stopped typing and has realized their application lost keyboard focus. This would help avoid inadvertent
Or, perhaps even better, don't steal focus for events like "the install finished". Flash the orange indicator in the taskbar/dock and wait for the user to be ready to acknowledge it.
In a fit of energy, I decided to stop procrastinating. I was inspired by some listening and reading I did... while procrastinating. Tim Pychyl, a Carleton University Professor, has a great website on procrastination -- the research focus of his group at Carleton. He also has a great podcast called "iProcrastinate". Anyway, my favourite episode of the podcast basically consists of him chastising the listener for "cyberslacking". Basically, this is the self-delusion that because you are sitting at work, at your computer, you are "working". We all know, deep down, that isn't true. He rails against Facebook, email, IM, online news, etc. and says these are forms of entertainment, not forms of useful work activity (no matter how much we might try to convince ourselves that using Facebook is for research into social networks).
This really resonated for me. I often go home feeling like I've missed an entire day, and that is because I've been "cyberslacking" for hours, rather than getting up and leaving my computer if I've made a conscious decision not to work, or otherwise intentionally shutting off distractions and getting to it. As hard is it might be, I think he's right, I need to turn off IM and email notifiers for several hour blocks, and force the Facebook use and the newspaper to my at home hours. That way, maybe I'll actually get some at home hours. This isn't rocket science, it's something I've been battling for a while, but having a Dr. who researches procrastination somehow just validates the problem for me.
My favourite quote: "Cyberslacking is making you a mouse potato". OK, well maybe "desk potato" would be more congruent with the original phrase, but whatever, it's a nice saying.
So, as of today, no more mouse potato. At least until I forget about this blog post. Enough cyberslacking with my blog.
Tuesday, February 12, 2008
[Sidebar: I took that screen shot with the Vista "Snipping Tool". You draw a red box to define the capture area, but the inclusion of the red box in the screenshot was unexpected!]
So, when you want to enter something like a path (mine is 434 characters), you have to scroll this tiny box. And, yes, you have to scroll it by positioning the cursor in the box and holding down left or right (or using a keyboard shortcut like home/end). Once you set it, you have the same problem if you want to read it -- the form for viewing environment variables is fixed at a small width that appears to be designed for 640x480 monitors.
Some of the unpolished aspects like this, the buggy UAC system and slow printing, along with the the persistent awfulness of the calculator and notepad applications make me wonder why Microsoft didn't finish Vista before publishing it.
Friday, February 8, 2008
There is currently a referendum and election underway at the University of Calgary Student's Union. Upon looking into some policy, it seems they union here has an even clearer procedure and policy about how to approve referendum questions. In a nutshell, 5% of the student body has to sign a petition supporting the referendum. The petition must contain the wording of the full question to be put to students, and the final question is then approved by the Student Legislative Council through a process of 2 readings. Referendums about fees have to meet three additional criteria:
i) information on the purpose of the fee;
ii) the university sessions for which the fee applies; and
iii) the level of the assessment of the fee in all faculties,
programs, and sessions in which the fee level is different.
Also, "If the SLC decides the wording of a referendum or plebiscite
included in a petition is unclear or potentially in violation of the
Constitution, bylaws or Procedures of The Students’ Union, the SLC may, by SLC resolution, refer the question to the Review Board for clarification, grammatical correction of the wording or form, or a ruling on the validity of the question" (SLC Election By-law).
So, then I have to ask what the SLC was thinking when they didn't send this question for review and grammar correction:
"Do you support an increase of $1.00 for each full-time and part-time students, in all faculties, per session (fall/winter/spring/summer) for the continued growth, maintenance, and operations of NUTV, a non-profit society."
Yes, it says "for each full-time and part-time students" and ends with a period. The session information is separated by a clause from the dollar amount, muddying the readability. NUTV is not defined other than as a "non-profit society." Some question!
Compare it to another question from the same referendum:
"The Gauntlet is the independent student newspaper of the University of Calgary. Currently, Gauntlet fees are $3.50 per session for full-time students and $1.75 per session for part-time students. Do you support an increase in the Gauntlet fee of $1.00 per session for full-time students and $1.00 per session for part-time students?"
I guess I should expect better writing from a newspaper than a tv station. This question provides context (current fees) where the other one doesn't. This provides more of an ability for voters to judge the value they are getting for the current fee. In order for student governments to be well-respected organizations, I think they need to take issues like referendum validity seriously. The SU of U of C seems to have not followed it's own by-laws in not correcting the NUTV question. They should also add a fourth clause to referendum questions regarding fees: (iv) current levels of the fee, if any.
Well, I'm not an undergraduate, or a U of C student, so it's really none of my business. I guess I'm a political junkie and I miss being involved in student governance. Or, I just need to get a life.
Wednesday, January 30, 2008
This all seemed really convenient because yesterday I downloaded a new Java package that has documentation shipped in DocBook format, an XML Schema format that can (supposedly) be compiled into a PDF, HTML, or other sorts of files. You would think that there would be something like "docbook2pdf" to compile it, but looking over the DocBook Wikipedia pages, and Googling all about DocBook, such a thing doesn't seem to exist. That would be too easy (if you think LaTeX-like document creation is easy).
Instead, I followed the ironically named "really quick guide to DocBook for Windows", which required me to install various cryptically named packages like "fop" and "saxon" (+ 3 more). Then I downloaded the "easy DocBook scripts for Windows" and edited the files to point to the install locations of the aforementioned packages. What a pain in the ass. Java has Javadoc for a reason! Compiling help documentation should not be so difficult.
In the end, I thought I was all done. I even registered "compile DocBook" as a right-click item for xml files. Then, I click it and get about 4 pages of error messages (sounds like LaTeX).
DocBook was started in 1991 and development continues today. I wish they would develop a simple front end for
Computers, you've got a LONG way to go. Makes me excited, but also dismayed, to work in HCI.