The limits of free advice

(NOW WITH UPDATES)    Long before he was chair of the TTC, I knew Adam Giambrone was 21st-century enough to surf the Web. I know because I sat some distance behind him in City Hall committee rooms and watched him dick around on IE6 during dull moments. We definitely need to get him on Firefox or Opera, but this at least shows he acts like every other 29-year-old in Canada. It may make him cool in some generic way, but it does not mean we should do TTC work for free at his behest.

It was revealed over the weekend that, upon his ascent to the TTC throne, Giambrone called in head Spacer Matt Blackett for a two-hour meeting. I doubt I would disagree significantly with any of the suggestions Blackett might hypothetically have, although some of them must surely involve financial deals between Spacers and the TTC. (It goes without saying that the TTC will eventually sell the Spacer subway buttons or other fan merchandise.)

But as soon as that meeting was publicized, there was Giambrone in the Star (2007.01.01, p. B1) calling for the addition of a trip planner to the TTC site (also in December). Then it was revealed that Giambrone and Blackett agreed to combine the forces of several blogs, including Spacing’s, to poll readers about what they wanted from a redesigned Web site.

While this online consultation is needless duplication (and I say that as someone who, in November, used oldschool roadblock methods to promote my micropatronage drive), there are bigger problems:

• It empowers people to say the same thing over and over again. Yes, we know we need a trip planner. Even the head of the TTC knows that.
• It tackles only half the problem – the latter half. Everyone’s talking about features. But you can’t add features until you have a decent code base (standards-compliant HTML). And how do you make a trip planner accessible? Has anyone other than me done any thinking about that?
• It will tempt TTC mandarins to add new features to their existing platform. Do you really think the TTC wouldn’t take the cheap and expedient way out on any issue that does not involve rolling stock? We already know they’re looking at “the ‘quick fix’ issues on the TTC Web site” (ACAT minutes [PDF], 2006.03.30).
• It misleads an organization that already doesn’t know how to put a Web site together that this new plan (surveying the public via the same blogs they’ve always ignored) is the right way to do it. Public-opinion polls, which is all these things are, can be useful to driftnet for some people’s wants and needs. But if that’s all you go by, then you end up with today’s site with a few features grafted on. This faux-consultation does not take the place of professional needs analysis, prototyping, and testing with browsers, with adaptive technology, and with users, including users with disabilities. (And that doesn’t mean just blind people.)

In short, it’s nice that everyone thinks they’re contributing, and maybe that’s what everyone really is doing, but what the TTC needs is a cadre of standardistas, a generous development timetable, lots of time and money to test everything, and a resistance to bad ideas.

Let me put this another way. Giambrone may surf the Web, but he’s a Windows IE6 user, making him almost the worst possible client. The entire TTC is Windows plus IE6. These people, by definition, don’t know the first thing about real Web sites and will conclude that if it works in their browser the whole job is done. We are not making a Web site just for you, or me, or Giambrone, or TTC staff. It has to work for everyone who wants to use it, with only rare and unavoidable exceptions. And there is one way to do that: Use (nearly-)valid, semantic HTML, comply with accessibility guidelines, and test everything.

The TTC’s level of ignorance would lead to a transformation of the site from its current state of tables for layout, unusable drop-down menus, crawling marquees, and impossible graphic design and usability to a site that looks marginally better but is still a complete disaster under the hood. Or it might just go all-Flash, like Viva.

In short, all this ostensible consultation may lull the TTC into committing a Failed Redesign. They’ve already been sued over inaccessibility and lost (PDF). I don’t think they’re competent enough even to get to the following point of understanding – that anything less than the absolute state of the art will expose them to another human-rights complaint that they will invariably lose. Fixing the Web site might easily cost $100,000; fighting a losing human-rights complaint will cost that much on top of whatever they spend on a Failed Redesign.

Incidentally, at most I’d be able to vet and improve any new TTC site, but one person couldn’t do the whole job alone. Since there are no standards-compliant shops in Toronto, none at all (Wishingline is a sole proprietorship), the whole job would probably go out of town, which further implies that standards-compliance and accessibility vetting might also go out of town. Blackett might license his subway buttons to the TTC, but it is far from a sure thing that I’d see a cent of work from a TTC redesign. Whether I get hired or not has no bearing on anything I’ve said here. There is no other way to do it.

Updates

Just as I figured

• (2007.01.06)    There’s Adam Giambrone in the Globe today calling this exercise what it is: “almost like free consulting.”

  “Almost like”? They have a billion-dollar budget and we don’t. Why are we consulting for free?

  It is also revealed that the owners of the four blogs canvassing public opinion are to meet with Giambrone. I have asked to be invited to the meeting.

• In a Giambrone panegyric, the Star notes that Giambrone (now) has a Mac. Since all extant Mac browsers are reasonably or highly standards-compliant, any new and improved site will be really improved on his machine.

OK, this is getting ridiculous

• (2007.01.08)    Two more stories in the paper about the TTC redesign and how those plucky bloggers are just tryin’ to help:
  1. A piece in the Star that had little new to say, except for Robert Ouellette’s telling admission that “[i]t’s free consulting.” Yes, but that’s the problem.
  2. A better article by my esteemed colleague Marc Weisblott in the Globe (not onliné) that at least mentions “[a] need for consistent design and accessibility standards” and has a nice quote from Thickslab: “They don’t want to listen to anyone outside because they’re convinced they know what they’re doing. They think that they’re running the system well, and that what they need is more money, and there’s nothing they can do without more money.” (Admittedly, a proper redevelopment of the site will cost lots of money.)
• Matt Blackett engaged in his usual bullshit by refusing to respond to my request to be included in the meeting with Giambrone. A jittery-sounding Ouellette hemmed and hawed and more or less declared that he was in this for the good of the proletariat while I am out to make a buck. While the former might possibly be true, the latter isn’t.
• None of these blog editors has Web-development knowledge, and it is a non-starter for them to intensify their status as gatekeepers by packaging information that is already available online. I assume TTC staff, even if they are shackled to Windows and IE6, can read actual blog comments; nobody needs them edited and printed out.

Enough is enough

I have other things to do, but I’m getting tired of this nonsense. Since I believe in transparency and openness, unlike the Spacers with their top-secret projects and closed-door meetings, here is the letter I sent to TTC acting general manager Gary Webster. Of course the TTC needs a trip planner and an API, but this is what people should be telling them.

I am writing with free, noncommercial advice and a warning about the now-much-discussed redesign of the TTC Web site.

Background

I work in Web standards and Web accessibility. I wrote the book Building Accessible Websites (New Riders, 2003). I run the Webstandards.TO social club for standards-compliant Web developers. I lecture around the world and I do consulting for clients, the latter of which is not the purpose of this message.

Everyone knows the site is broken…

Nobody outside TTC headquarters believes the current TTC Web site is even vaguely functional or appropriate in any sense whatsoever.

…but a new site could be just as broken

However, the TTC has so little Web and computer knowledge – everyone in the organization uses Windows and Internet Explorer, almost the worst possible environment – that the TTC is at enormous risk to replace its current dysfunctional site with one that isn’t as ugly but is just as broken.

Web standards

Perhaps unbeknownst to the TTC, since 2000 Web developers have come to understand the benefits and importance of Web standards, that is, coding Web sites correctly, with valid HTML (which gives the document structure), valid CSS (giving it appearance), and correct scripting (like JavaScript, giving it behaviour). My book is one of about a dozen on the topic. Web standards aren’t anything new in other cities, but are pretty much ignored in Toronto.

Standards-compliant development is the sole correct method to develop Web sites. Among many other things, Web standards means no use of tables for layout (there are six on the current homepage), no self-running sounds or multimedia (like the cute but intrusive subway chimes), no overlong drop-down menus that activate by themselves, no opening of new windows without user foreknowledge (seven on the current page), and no images without text equivalents (several).

It is additionally understood that Web addresses should be compact and predictable –

• ttc.ca/31
• or ttc.ca/Greenwood
• or ttc.ca/greenwood

and not

• ttc.ca/multilingual/multilang.htm (actual example)
• or ttc.ca/portal/!ut/p/.cmd/cs/.ce/ 7_0_A/.s/7_0_24P/_s.7_0_A/ 7_0_24P/_l/en?docid=EC002001 (Ontario government example).

Web accessibility

Complying with Web standards also ensures at least minimal accessibility for people with disabilities. A separate set of standards, the Web Content Accessibility Guidelines, can and must be intelligently followed to ensure that as many people with disabilities as possible can use a site.

Web standards and Web accessibility are inextricable; you can’t have the latter without the former. And without Web accessibility, an organization stands to be accused, probably correctly, of providing unequal treatment for or discriminating against people with disabilities. The TTC will be aware of what can happen when such accusations reach the tribunal stage.

Testing

It is by now well established that even a standards-compliant site must be tested, including testing with people with various disabilities (not just blind people).

Multilingual content

The TTC is unaware that information in languages other than English does not have to be provided as a picture of text. Arabic, Chinese, Farsi, Greek, Punjabi, Russian, Tamil, Korean, Turkish, Ukrainian, and Vietnamese can and must be presented as real text, not pictures as is currently seen on the site. [Author’s note: That’s particularly atrocious given that two of those languages use Latin script.]

PDF

I serve on the PDF/Universal Access Committee, working on a definitive specification for accessible PDF. PDFs are overused by the Commission and have almost no defensible role on a public Web site. The few PDFs that may be permissible must be tagged and follow other accessibility requirements. No existing online TTC PDFs meet those specifications.

Local development expertise

Because I run the club, I know for a fact that there are no standards-compliant Web-development shops in Toronto. None. There are individual developers, but no companies. A site as large as the TTC’s cannot be redeveloped by an individual.

I have not seen the RFP that allegedly went out last year for a Web redesign, but I have no reason to think it called for compliance with Web standards and accessibility guidelines, nor that it required extensive user testing.

Expensive usually means worse

A large Web site requires a content-management system. The more expensive a CMS is, the worse its standards compliance is, without exception. (Vendors often promise standards compliance that they can’t deliver. Those promises are easily disproved by looking at the vendors’ own sites.) It will benefit a government body to accept that low-cost or zero-cost software usually has higher standards compliance built in.

Free advice

I suppose it’s commendable in some ways that the young chair of the TTC has solicited comments from the public about the TTC Web site via four Weblogs. While this is useful as a survey of some members of the public, in actual fact it is quite unfair for a billion-dollar corporation to canvass for public opinion at no cost. It is unethical to allow the editors of these blogs to act as gatekeepers of such public opinion.

What I think will happen

• The TTC, with its poor to nonexistent Web knowledge, will not know how to evaluate firms that submitted bids.
• The TTC will not be willing to accept that its original RFP, as I speculate, did not ask for the right things and should be cancelled and replaced with an updated RFP.
• The TTC will accept a bid from a local company that will turn in an untested site that uses tables for layout, incorrect code, pictures of non-English text, and poor accessibility. PDFs will be commonly used.
• The site will work just fine in Internet Explorer, the only browser it was tested in.
• Because it works well in its own staff computers, the TTC will ask the Commission to approve its public use. If anyone from the public should point out failings in standards compliance and accessibility, the response will be that the money has already been spent and the new site is better than nothing.

At that point in this scenario, some questions will come to mind:

• Why did the TTC ignore even the free advice it canvassed?
• Why didn’t it re-tender the contract with requirements for Web standards, accessibility, and testing?
• If and when a human-rights complaint is filed, what will be the defence? Shouldn’t money for lawyers have gone into fixing the Web site properly in the first place?

A question I can ask irrespective of any scenarios: Will the TTC respond so poorly to being told that their entire exercise needs revision that it will charge right ahead with its existing plans?

Not a commercial solicitation

I reiterate that this is not a commercial solicitation. I didn’t bid on the earlier RFP, I’m not angling for Web-related business now, and, as an individual developer, I would not be viable as a single bidder for a revised RFP. Since the Chair has already canvassed for free advice, it is incumbent on Commission management to treat this message as what it is – free and informed advice.