Date: Wed, 19 Sep 2001 17:40:36 +0200 From: Terje Bless <firstname.lastname@example.org> To: "gregory j. rosmaita" <email@example.com> cc: firstname.lastname@example.org Message-ID: <email@example.com> Subject: Re: QA comments on new W3C validator format [ Apologies for the late reply. ] On 14.09.01 at 16:19, gregory j. rosmaita <firstname.lastname@example.org> wrote: >while i want to thank you for your expeditious response, i must take issue >with the argument that the infrastructure must first be in place before >accessibility issues are addressed - this is what has become known in the >various disabled communities as "trickle-down" accessibility, and the >results of such a retrospective approach to accessibility concern results >in persons with disabilities not having access to the facility at a >critical point in its development, which is not only patently unfair and >unnecessarily exclusionary, but which runs counter to the efforts of the >W3C as manifested in the Web Accessibility Initiative (to which i am an >invited expert) I'm going to quote you in full below because that message bears repeating, but I'll respond directly to this paragraph which appears to be the crux of the matter. I fear I have expressed myself poorly in my previous reply. I'll attempt to remedy that here; but, as usual, I will probably end up compensating for lack of clarity with verbosity. Buyer Beware! :-) Also, as is my standard disclaimer, I speak for no one but myself. In particular, I do not speak for the W3C and I have no authority over the Validator. I'm just some guy with an interest in the web that happens to send patches to Gerald every once in a while (and write long and mostly incoherent messages to this list at ungodly ours of the day). :-) Now onto the meat of the matter... My intent was not to suggest that Accessability was a low-priority issue. Quite to the contrary, it's a very high priority issue. As I've mentioned elsewhere recently, I'd like the Validator to be a shining example of accessible web design if at all possible. I've been told there is no such thing, and I doubt we could achieve it in any case, but I would still very much like for that to happen. However, this cannot happen with the Validator as it stands today. It cannot happen for the same reason that the Validator cannot be localized to different languages, or installed locally, or customized to suit different installations, or have new features added within a reasonable timeframe. The reason is, to put it bluntly, that the current code base sucks. Badly! It used to be quite clean and perfectly suited to do what it was intended for, but it's been added to and tinkered with for quite some time, while the web has evolved at "Internet Time" with new technologies and formats, and has now reached a point where it is approaching the state of being unmaintainable. Before we try to solve these problems we need to rip it apart and rebuild it more or less from scratch. Several good ideas for how to do that has been presented, but the gist of it is modularity and abstraction. In particular reference to Accessability and your three suggestions, it's now taken me several hours to make three small changes to what is conceptually a small web page (and I haven't finished yet). Why? Because the HTML is hard coded in the source code of the validator, and scattered about in little fragments that interact in unpredictable ways. Every such little change is more likely then not to require similar hoop jumping. No matter whether the issue is Accessability, Internationalization, or Maintainability. Asking the expertise to review it in the state it is now would be an insult and a waste of everyone's time. Those issues that can actually be fixed will take far too long and be nasty hacks that will break the next time we change the code in that area. What I meant to convey in my previous message was that we need to refactor the code and build a solid foundation before we can deal with these kinds of issues properly; not that other issues were more important! Meanwhile we'll do our damndest to fix those issues we _can_ fix as they are reported. And Accessability problems will tend to have a _higher_ priority then others, not less. I hope that better explains my intentions regarding Accessability. I obviously do not speak for the W3C -- or, for that matter, anyone but myself -- but insofar as I have any influence on development priorities for the Validator, those are my stated goals. That having been said, there are probably several issues we could address short of the mentioned full rewrite. It will also be much easier to build Accessability into the Validator during the rewrite if we are aware of those issue beforehand. So regardless of whether we will be able to address a particular issue, we'd still like to hear about it! If we can fix it immediately we will, and if not we'll add it to the to-do list so it will be addressed when the opportunity presents itself. In particular, I suspect from your earlier comments that we have just added another barrier to accessability with our navigation menu -- as evidenced on the front page of <URL:http://validator.w3.org/>, or on every page of the test server on <URL:http://validator.w3.org:8001/> -- located at the beginning of the page. I would very much appreciate any comments you have on that issue, and any suggestions you have on how to improve it (whether specific to accesability or of a general nature). >and, while i encourage you to seek feedback from WAI-IG, i am at a loss as >to why WAI-PF was not/has not been consulted or asked to participate or at >least monitor the reformat. Anyone who wishes to comment on the validator are more then welcome to do so and we'll endeavour to meet all concerns as time allows. They (or rather the WAI-IG, as I'm not familiar with the WAI-PF) have not been invited to monitor the reformat because 1) we aren't in a position to do very much with their input as mentioned above and 2) it didn't occur to me that the reformat involved anything of substance that would be of interest to them. What issues fall within the remaining intersection can be dealt with by responding to user reports as they are discovered. If someone with experience in the area of Accessability -- whether that "someone" is a group, acting as a representative of a group, or only on behalf of him or her self -- volunteers to be a "beta tester" for Accessability in the Validator, there is little that would make me happier. We normally run new versions in development on test servers so interested parties can have a peek at what's in store and hopefully catch any mistakes we've made before it goes public. Accessability issues are bugs like any other bug and testing for them is essential to finding them so they can be fixed. If anyone (this is CC to the list) wants to volunteer for this, just send me a message and I'll give you the information you need. There is no need to committ to some kind of long-term obligation; just watching this list, and maybe email@example.com, will keep you up to date if you just want to make a few comments occasionally. If you have any questions feel free to send them to the list or to me directly (the list has better chance of a timely reply, BTW). If anyone wants to be Grand Poobah of Accesability for the Validator, I can arrange to let you know whenever a change that might affect Accessability is about to oocur and run a testing cycle with you in addition to the usual regression tests we run. The development takes place on this list or on #Validator on OpenProjects IRC (the W3C also maintains internal channels for this sort of thing); both are public and anyone that wants to contribute are more then welcome. Bug reports and feature requests can be sent to the list without subscribing. Not all things are fixed immediately, but there are many highly talented people on the list that answer questions and I endevour to make sure every message gets a reply even if it's just to acknowledge receipt of the message. Hopefully I've managed to convey more accurately what I meant. If I didn't, please do let me know. Next time I'll see if I do any better with brevity then I did with verbosity. :-) On 14.09.01 at 16:19, gregory j. rosmaita <firstname.lastname@example.org> wrote: >aloha, terje! > >while i want to thank you for your expeditious response, i must take issue >with the argument that the infrastructure must first be in place before >accessibility issues are addressed - this is what has become known in the >various disabled communities as "trickle-down" accessibility, and the >results of such a retrospective approach to accessibility concern results >in persons with disabilities not having access to the facility at a >critical point in its development, which is not only patently unfair and >unnecessarily exclusionary, but which runs counter to the efforts of the >W3C as manifested in the Web Accessibility Initiative (to which i am an >invited expert) > >accessibility isn't something to be addressed once all of the bugs and >kinks have been worked out - that is not only a bad implementation >decision, it is paternalism of the most pernicious type - pernicious >because it is unconscious. that last word, unconscious, is the key - as >long as persons with accessibility needs are kept out of the developmental >loop through the inaccessibility of the resource being developed, you are >not only denying yourselves of the expertise of those of us with highly >developed technological competencies, but you are aping the developmental >insensitivity which efforts such as the WAI exist to obviate. > >accessibility can't be bolted on afterwards - it must be considered >throughout the redesign/update process; anything else is not only >insufficient, it is an insult. it is not sufficient to log accessibility >issues - they must be dealt with as they are identified, for unless >accessibility is a built-in consideration of the design/update/maintenance >process at EVERY stage of that process, true accessibility may prove >extremely difficult to retrospectively achieve. accessibility cannot be >an afterthought or "back-burner" issue - it must be addressed and >considered at every step of the developmental/maintenance cycle. > >and, while i encourage you to seek feedback from WAI-IG, i am at a loss as >to why WAI-PF was not/has not been consulted or asked to participate or at >least monitor the reformat. > >terje, i do not state the case so forcefully as a personal rebuke to you >or anyone else who labors to provide one of the most useful resources on >the web (valid markup, after all, is the first step towards accessible >markup), and i do appreciate your putting the 3 issues i brought to your >attention on the "immediate to-do list" - it is, rather, my intent to make >as clear as possible to those who work on the validator the importance of >considering accessibility at EVERY step of the process. > >gregory. (please read disclaimer below) > >ps: while i am the IG-member-at large to the WAI Co-Ordination group, this >emessage represents my PERSONAL opinion ONLY and does not represent a >consensus of the WAI-CG - to put it another way, i am not speaking on >behalf of the WAI-CG, but as an individual materially affected by this >implementation decision.