- From: Ivan Enderlin <w3c@hoa-project.net>
- Date: Wed, 14 May 2008 20:10:02 +0200
- To: Justin James <j_james@mindspring.com>
- Cc: 'public-html' <public-html@w3.org>
Justin James a écrit : > Ivan - > > I am intrigued, but I think that it would be important for you to enunciate > the practical and theoretical differentiators that you envision between a > standard text input and the proposed search type. Outside of marking it at > the code level so that tools can recognize its purpose, what would it do? > For example, you mention applying a special style. People can already do > that with CSS. Also, it is my (probably over stated) opinion (which the > direction of HTML seems to agree with) that style and function should be > wholly separated, so I don't think that allowing the browser to apply a > special style to it, in and of itself, is a good argument in favor of it. > That being said, part of me really likes the idea, but I am having a hard > time finding concrete reasons to like it. :) > Hey Justin, I am agree with you about the style point (the third point — see iii. —) : a style sheet could happily do the trick. But the first and second point are essentially the goal of my suggestion. When I try to make an accessible website, I often add an anchor in the navigation bar : #search_field. The search form is a special form because when you are familiar with a website, you use the search engine regularly. If we add this semantic attribute, web browsers could add a specific shortcut — for example — to the search field. It's like the nav, article, section ... tag. If we have more special points in the document, it will be easier to move into the document. But as the accessibility is not restricted to handicapped persons, here are the positive aspects for many users. I search a keyword on a website, and unfortunally, I don't find what I look for. I exchange website and retry a new search : I will appreciate that the search field suggests me my previous keyword(s). We can do that nowadays if the search fields have the same value of the attribute name (if the UA allows this feature), but it's not simple to do. And if the UA does not propose this feature, developpers will use AJAX to remember to keyword in cookie or something like that, but it's for only for one website. If we have a search type for input tag, UA will automatically remember the entries and suggests you the next time. We can see farther if the UA suggests keywords you entered in the OS search field (e.g. Spotlight for Mac OS X). I think it's more ergonomic and it will ameliorate accessibility, semantic etc. The risk is to produce a new type for each form group (login, subscribe etc.). But the search type is different because of the multi-document search for example. I do not know if I was very comprehensible ... Best regards :). > -----Original Message----- > > > Hey folks, > > There was previously a short discussion about the input type="search" > attribute value of Safari. I think, it's not a bad thing and should have > more importance. > So, here what I think : > i. it allows to the cache operating system to interfere ; > ii. it allows a better accessibility ; in fact, the search field is > a special field (the most used/needed ?) ; > iii. allows a better comprehension by the lambda user because the > operating system should apply a special style (like Safari does). > > I wonder if it should be a good idea to add this feature in the HTML 5 > spec, no ? > > Best regards. > > -- Ivan Enderlin Developper of Hoa Framework http://hoa-project.net/
Received on Wednesday, 14 May 2008 18:10:48 UTC