Blind Confidential (Blog) Thursday, September 13, 2007 Google's web interfaces with screen readers By Chris Hofstader Since arriving in Massachusetts, I've had access to a wireless network but I do not have my father-in-law's password for what was his Comcast account and the huge ISP/Cable/Phone Company requires authentication to use its outgoing server and it doesn't permit anonymous rerouting to other SMTP servers. And, no, I cannot ask my father-in-law for his password as he died in March and hasn't spoken to me since. Ordinarily, I write Blind Confidential in MS Word and use the "Send" item under the File menu to email the post to blogger.com. Once the Blogger for Word button bar stopped working, I found that the email approach, described to me first by Jeff Bishop (link to his Desert Skies blog above) in a phone conversation a few months ago, suited my needs better than anything else I could find. I understand that Word 2007 has some kind of interface for people who write blogs that sends information directly to the host but I only have Office 2003 on the laptop I brought with me up north and cannot comment on how well it works with a screen reader. So, in order to stay in touch, post blog entries and communicate with people on my various research projects, I created a gmail account and also tried to use the blogger online interface. On this laptop, I have two Windows screen readers installed namely, JAWS and System Access so I cannot comment on HAL, Window-Eyes or any of the others. When one goes into the gmail page with JAWS, they quickly learn that they cannot use the site unless they click on the link that reads, "If you are using a screen reader, click here for basic html," or something very similar. With System Access one can start using the site and the dynamic content is updated properly, tables are recognized as such and the links that JAWS reports as plain text actually work. Even in the "basic html" interface, JAWS exhibits some peculiar problems. In the multi-line edit fields, when one tries to type a capital letter that is also one of the quick keys in the JAWS virtual buffer support, the leading screen reader announces that "This feature is only available when using the virtual buffer on the Internet." Oddly, this only happens the first time that any of the capital letters are typed in such an edit field per session. Thus, capital F, O, I, B and the other quick keys cause a temporary error when typing. This isn't an enormous problem unless, like me, you type very quickly and, when you review your message later, you find that some words are missing their initial letter. Still, this is more of an annoyance than anything else and I'd assume that it would be an easy bug to fix. Because I'm running Visual Studio a lot, I tend to also run JAWS as the scripts that Jamal Mazrui and the guys on the blind programming mailing list have written as a team, are so good that VS .Net works better with JAWS than any other screen reader/IDE combination out there that I have tried. [If you are interested in these scripts or any of Jamal's cool and highly accessible programs, go to his web site: http://www.empowermentzone.com or one of the other sites that provide ways to download this software.] I don't always feel like jumping from one screen reader to another just to read mail or send a quick response to someone so I have grown kind of accustomed to using JAWS with gmail although I would prefer the System Access level of support. The blogger interface also works better with System Access than with JAWS but it is not as smooth as the SA gmail support. Yesterday, as many of you noticed, my post "The RIM, RAM, SAM Scam" contained a bunch of garbage and two copies of the text I copied from MS Word and pasted into the blogger edit field. I don't know how or why this happened but, somehow, the text I copied from Word got combined with text in the JAWS virtual buffer and when I pasted it into the edit field, it looked pretty crappy. I did the blog post right as my wife and I were running out the door to visit an old friend in Jamaica Plain so I didn't review the post and, given my luck with web interfaces lately, it, of course, came out miserably. Generally, though, the screen readers I tried (more so in the JAWS case than SA) need to improve a bit before I would say that the gmail or blogger interfaces are truly usable. SA, as I state above, does an excellent job with gmail and performs adequately in blogger. JAWS requires that one use the blind guy ghetto "basic html" interface for gmail and works dreadfully in the blogger pages. I'm told that JAWS 9 is supposed to do revolutionary things on the Internet so I hope that when 9.0 is released, it does at least as well as System Access on pages built with AJAX that have a lot of dynamic content. Mike Calvo wrote an interesting post on the "Who's to Blame" topic on the Serotek blog yesterday (http://www.serotek.com/blog). I recommend that BC readers check it out as I think he provides a more comprehensive discussion of the issue than any of the other blogs I've read recently. I still think that ATIA, the industry association for access technology companies, should try to coordinate an effort to develop a document that web developers can use to better understand what AT users will see, hear or feel when on their web sites. The precise design of user experience should probably remain in the hands of the AT companies as features like Quick Keys and others are issues on which these companies compete and I, for one, want the screen reader vendors to continue to try to innovate in order to beat each other at the cash register. At the same time, though, I feel strongly that web developers should have a easy set of reference materials on which they can set expectations for how their pages will work with AT. Mike Davies, the actual author of the blog post I accidentally attributed to someone else last week, said in a comment he posted that he would not like to have different expectations for behavior in different screen readers and that he would also not like putting a "best if read with screen reader X" statement on a web page as this would be bad for standards and guidelines and would likely muddy the waters of web accessibility. I believe this sort of thing is inevitable whether the web sites state that they work better with one screen reader or another or leave such a statement off and let the users guess which AT might work best on which sites. I feel strongly that the AT companies should try to adhere to the user agent guidelines as closely as possible; sadly, though, I think that the leading screen reader vendors will do whatever best suits their business model rather than what best suits their users and rely on companies like google to provide a blind guy ghetto "basic html" alternative to all of the cool new dynamic content that people who do not depend on AT can enjoy. Afterward As the easiest thing I could find to fix the "RIM, RAM, SAM Scam" article was to delete it and repost the entire thing, I also deleted the comments posted before I put the corrected version up. Will Pearson and Chairman Mal had sent in interesting comments and, if they read this, I hope they will repost their comments as I found them entertaining but I don't think they were online long enough for many others to see them -- End posted by BlindChristian at 11:06 AM http://blindconfidential.blogspot.com/2007/09/googles-web-interfaces-with-screen.html BlindNews Mailing List Subscribe: BlindNews-Request@xxxxxxxxxxxxx with "subscribe" as subject Unsubscribe: BlindNews-Request@xxxxxxxxxxxxx with "unsubscribe" as subject Moderator: BlindNews-Moderators@xxxxxxxxxxxxx Archive: http://GeoffAndWen.com/blind RSS: http://GeoffAndWen.com/BlindNewsRSS.asp More information about RSS feeds will be published shortly.