from pythonmagick.image import Image
from pythonmagick.color import BLUE
i = Image('foo.jpg')
i.format = 'PNG'
i.rotate(45, BLUE)
i.save('flip.png')

The above reads in a JPEG image, reformats to PNG and rotates by 45 degrees with a background colour of blue. The image is then saved to a new file.

This may look a lot like PIL code, so why not use it instead of ImageMagick? ImageMagick covers a much wider range of image formats than PIL, and supports better algorithms for image resizing.

Check out the bindings by installing them with easy_install:

easy_install http://svn2.assembla.com/svn/pythonmagickwand/trunk/

On a side note, I’m quite impressed with Assembla, the site which I’m using for my SVN and Trac repo. It’s about as good as something I would deploy myself, but with much less hassle. Honestly, these days there’s less and less reason for me to spend hours setting up stuff myself, and that’s a great thing.

If you’ve caught the Nana fever and the only solution is more Nana, be sure to also check out her DiverCity podcasts on BlogTO.

What I need is a command-line tool to do some basic PDF preflighting (eg. non-CMYK elements, non-embedded fonts, images below 300dpi, characters which may cause difficulties in printing, etc.). Is that so much to ask? When the PDF spec is 1300 pages long, it is. Still, don’t think I won’t try and take a stab at it.

The TTC chairman confirmed my forecast of 8 months for a new website, giving an estimate of 6 to 8 months. That figure does not include the transit planner, which will be phased in for 2008. The planner has been budgeted for, although Giambrone said that the TTC has not yet decided whether it will be created in-house or will be out-sourced. He did state, however, that they were not looking to Google Transit due to fears that their service might require integration costs and ongoing fees in the near future. I hate to say this, but it seems like a wise move to forgo Google Transit for now. There have been complaints that Google Transit for Portland, whose system was included in Google Transit’s launch, provides sub-optimal trips which cost riders more than the ones provided by TriMet’s own planner.

I recently discovered that the TTC considered a trip planner in 2004 but shelved it due to integration issues and budgetary constraints. When I asked what was different now, Adam Giambrone cited increased outside pressure and an acknowledgement by TTC management that a planner was necessary. Before it can be created, however, its IT department must amalgamate transit data into one database, confirming my suspicion that it currently exists across a number of systems.

Speaking of Portland’s TriMet system, Bibiana McHugh, TriMet’s GIS Manager, is a huge fan of using open source for public transit. McHugh has proposed that TriMet and other public agencies in the Portland metropolitan region ‘organize [their] data and make it universally accessible and useful.’ This relates to both what I’ve talked about with open-sourcing the TTC website, and with the Transit Camp’s open TTC session. I hope that the TTC is paying attention and taking notes.

Update: Bibiana McHugh has left a comment which addresses some issues mentioned above.

The great thing about the Transit Camp was its openness. When the call for sessions came up, I encouraged my girlfriend, who wasn’t entirely certain how she could contribute, to draw on her strengths and introduce a PR-related topic. She did, but unfortunately her idea of TTC-endorsed parties inside stations along the subway line was superseded by the Newmindspace subway party presentation which was unannounced at the time, so only two people showed up.

I attended a session on opening up the TTC data, which centred around creating an API for third-party developers. That was one of my pie-in-the sky suggestions for a new TTC website, so I thought it would be interesting. It was, but much of the discussion involved creating a trip-planner based on data available on the website. Offshoots of that grand scheme, such as creating an interface for determining the time until a bus appears at a specific stop (an idea I’ve thought of but haven’t touched), were discussed as well.

One of the problems with bringing in people from the community is that those who don’t have a background in computer science often don’t realise how complicated certain programming tasks are. Transit route planning involves calculating the shortest path with minimum delay between two nodes and is NP-complete, meaning that no polynomial-time solution exists and that, even with intelligent search algorithms, massive amounts of processing power is required for an imperfect but reasonably fast solution. Even Google is unable to create a perfect solution to this problem, as anyone who has made frequent use of Google Maps to get directions can attest, and they can only return a timely result because they have vast amounts of computing power available. Factor in bus schedules as a third-dimension, and the problem becomes quite cumbersome. The TTC, with its more than 350 routes and over 12000 stops, is the third-largest public transit system in North America. It’s commendable that individuals want to attack this problem, but automated trip planning and scheduling is best left to those with doctorates and massive amounts of computing power.

In the afternoon I attended Joe Clark’s eye-opening “Why TTC Signage is Fucked” presentation. As my girlfriend will attest, I am often bothered and frustrated by inappropriate signage and directions, so this talk affirmed a lot of my own thoughts and made me realise just how bad and widespread the problem is. Sadly, it seems as though the solution to bad signage, apart from massive lobbying, is a lawsuit brought about by death or injury, at least until the TTC becomes more enlightened.

Next up were the design slams. I joined one of the slams on redesigning the TTC’s homepage. While not a task that an hour or so can solve, it was nevertheless an interesting exercise. My main interest was multilingual support. Language drop-downs and language selection pages are not the answer as far as I’m concerned. With which language do you present the drop-down or link? I’m also unconvinced of the reliability of using the Accept-Language header sent by the browser. I feel that a properly-encoded display of the top languages covering upwards of 95% of Toronto’s population and its visitors is a better solution, but one that is also fraught with problems.

As I mentioned before the camp, my intent was to connect with people, and connect I did. I managed to meet and speak with Adam Giambrone, Joey deVilla, Jay Goldman, Joe Clark, Bob Brent and David Pritchard, among others. Having not been to this kind of gathering since the release of my transit map, it was a little strange to be instantly recognised by name. I was introduced as both “the guy who did the Google Toronto Transit Map” and “the guy who remixed the website suggestion spreadsheet“, although I think the most flattering comment came when the team who created a concept for a new bus shelter described a map on their embedded LCD as “a Crazed Monkey type map.”

All in all, the Toronto Transit Camp was a success for me and no doubt a success for the community as a whole. Did it live up to the hype? Only time will tell whether it will have any lasting impact. Representatives from the TTC were present, so hopefully the ideas presented there will stick in their minds and be recalled when it comes to policy change.

A proper TTC website will probably cost upwards of $1 million and will probably take at least 8 months to be done right. Any company involved in its creation will need web designers and developers who are not only familiar with creating high-traffic, intuitive websites with multilingual support, but should have the power to wean the TTC off its dependence on PDFs for bulletins and notices. That company will also need to employ user-interface and accessibility experts, translators and lawyers well-versed in the legal ramifications of accessibility. Each of these roles will be important for providing full translations of the TTC website for the city’s half-dozen or so most popular languages as well as ensuring that the website can be accessed by everyone, young or old, hearing or sight impaired. In short, everyone from a sight-impaired person accessing the web over dial-up and a Chinese-speaking senior should have no problems using the site. The accessibility lawyer will be needed to ensure that all guidelines (such as those of the Web Accessibility Initiative) are met to cover the TTC’s ass in case of liability. These experts will not come cheap.

In addition to the above roles, the TTC website will require one or more individuals who will be tasked with the responsibility of signing off on the finished product. These people will be responsible for seeing that the project meets the highest guidelines and must have the political backing so that their say is final and that the project isn’t complete until they say it is.

No company in Toronto can offer all these services. In fact, probably only a handful of companies world-wide have the personnel, experience and technical know-how to create a website suited to avoid lawsuits similar to that of NFB v. Target. In addition to being expensive, these companies are probably backlogged with work.

Toronto has some intelligent people who can provide the skills and knowledge, but they are few and far between and already have jobs. What this city needs is a sort of Justice League of web experts who can occasionally group together for the common good. Single-day initiatives like the Toronto TransitCamp are not the answer.

Probably the best solution is to open-source the TTC website, but make it more cathedral than bazaar. Invite people knowledgeable about accessible, multilingual websites to submit, maybe even as a cross-discipline project for senior and grad students within the University of Toronto. Have one or two people tasked with deciding which changes are integrated to the final product. Gather requirements, design, develop, test and repeat until done. The website will take longer (maybe a year or possibly two) but it will be cheaper and will be more likely to meet the public’s requirements.