Re: [WSG] Semantic Form - Person's Title
Richard Czeiger wrote: Can I get a consensus that this is actually the right way to do it? It feels right, but I'd like the opinion of my venerable peers :o) Looks right to me. Note that in a large site with a lot of form pages (online banking, for example) putting meaningful IDs for all the inputs is wasted effort. They just need to be unique and to correlate with the appropriate label. For speed, I use sequentially numbered IDs for radios or checkboxes; r1, r2, r3. I find it makes the process much faster on autopilot, as it were. Otherwise you spend at least half your time thinking up names that don't suck... If the other inputs already have meaningful ids (Dreamweaver can insert then automatically when you name them, for instance) I'd use the default ones, otherwise I'd use something like t1, t2, t3 and so on for those too. Cheers Ian -- _ zStudio - Web development and accessibility http://zStudio.co.uk Snippetz.net - Online code library File, manage and re-use your code snippets links http://snippetz.net ** The discussion list for http://webstandardsgroup.org/ See http://webstandardsgroup.org/mail/guidelines.cfm for some hints on posting to the list getting help **
Re: [WSG] Absolute Positioning-A Naive Question (Maybe)
Felix Miata wrote: The author here is the lead layout developer in the Mozilla project: http://dbaron.org/log/2005-12#e20051228a Well, I don't read anything in the author's cited reference for CSS floats - http://www.w3.org/TR/CSS21/visuren.html#floats - to imply that floats were intended solely for floating small bits of content in a paragraph. The source quoted specifically refers to block-level boxes as well as line boxes. Floating DIVs that create new positioning contexts is perfectly within the intent of CSS as I read it. The reliance on extremely obscure details the author refers to is justified by nothing more than a post by himself over what happens to the flow around floats when floats have negative margin-box dimensions. I am sure this _is_ tricky, and the intended behaviour in CSS may well be unclear. But this would apply to someone applying negative margin to floats in a small local setting just as much as in the major layout containers that we also float today. I don't think the Baron reference is sufficient evidence for the assertion that using floats for layout is an abuse of them. On the contrary, I have seen several references in the last few years that stated floats *were* the preferred layout method by the W3C CSS working group. Who's right? Without detailed research into the mailing list archives of the working group, there's not much point arguing about things like this. The point is, floats observe the letter of web standards as we have them today, and provide a robust method of achieving layouts that don't work as well or are harder to achieve using other positioning methods. Cheers Ian -- _ zStudio - Web development and accessibility http://zStudio.co.uk Snippetz.net - Online code library File, manage and re-use your code snippets links http://snippetz.net ** The discussion list for http://webstandardsgroup.org/ See http://webstandardsgroup.org/mail/guidelines.cfm for some hints on posting to the list getting help **
Re: [WSG] Screen reader recommendations???
Michael Yeaney wrote: Are there any recommendations for screen readers to test with I'd like to at least 'preview' what our site(s) sound like to such a user. I don't know what country you're in, but in the UK, HAL from Dolphin is very popular. It's very similar to JAWS but a lot, lot cheaper. http://www.dolphinuk.co.uk/ JAWS and Window-Eyes are the world leaders, though WE is probably the third most-used screen reader in the UK, behind HAL and JAWS. IBM Home Page Reader is actually the easiest for doing testing in, but it isn't a true screen reader and there are sometimes important differences in the way things are read. Very few people use this in real life compared to the others. JAWS is by far the most common one - the IE of the screen reader world HTH Cheers Ian -- _ zStudio - Web development and accessibility http://zStudio.co.uk Snippetz.net - Online code library File, manage and re-use your code snippets links http://snippetz.net ** The discussion list for http://webstandardsgroup.org/ See http://webstandardsgroup.org/mail/guidelines.cfm for some hints on posting to the list getting help **
Re: [WSG] Do you still support 4.0 browsers?
[EMAIL PROTECTED] wrote: But in all seriousness, if you were setting up a website for a client who has never been on the web before (no server logs to analyse) and is marketing their gates/fencing business, would you try and support 4.0 browsers? Has the time come to just have a disclaimer on the site stating support for 5.0 browsers or above? In my opinion, the word support is not a useful one to use here, because it means different things to different people. To a standard-based developer, it would probably mean can access content and functionality. To a graphic designer or marketing manager, it is more likely to be interpreted as show pixel-perfect design You have different layers to a web site: * content * functionality (meaning server side, usually) * design and layout * interaction enhancements (AJAX-driven UI and all that malarkey) I would personally interpret support as meaning can access content and functionality. Version 4.x browsers would not be expected to display any design/layout or interaction enhancements by most web developers, I suspect. Whether this view is shared by most marketing managers could be another matter :) Version 5.x browsers would be expected to display design/layout and as much interaction enhancements as they can, though the scripting should be unobtrusive so that if something isn't going to work, there should be no evidence that the user is missing out on something. IE 6 and better browsers should get the whole works, IMO. Having said all that, it is a relatively easy matter to supply some basic or even relatively sophisticated visual formatting to the version 4.0 browsers, though the main CSS should be hidden from them as the faults in these older browsers may make the site unusable if they can see the proper CSS or even crash the browser. These days, if the need arose, I would write a style sheet for Win IE 4, another for NN4, and deliver them using conditional comments for IE and using JavaScript for NN4 (test on document.layers). NN4 cannot display CSS if JS is off, so it all hangs together quite well. HTH Cheers Ian -- _ zStudio - Web development and accessibility http://zStudio.co.uk Snippetz.net - Online code library File, manage and re-use your code snippets links http://snippetz.net ** The discussion list for http://webstandardsgroup.org/ See http://webstandardsgroup.org/mail/guidelines.cfm for some hints on posting to the list getting help **
Re: [WSG] Linking to top of page
Curby wrote: What is the recommended way for linking back to the top of the page? I can't link to the id of my H1 because of my CSS. The name attribute of the A tag is deprecated/removed[1]. And while some people might say use the scrollbar or press Home I'm wondering if anyone has experimented with these: Why not use h1 id=foo? (You say because of your CSS, but what does this mean?) Then link to mypage.htm#foo - all legit and above board Seems to work for most people :) Cheers Ian -- _ zStudio - Web development and accessibility http://zStudio.co.uk Snippetz.net - Online code library File, manage and re-use your code snippets links http://snippetz.net ** The discussion list for http://webstandardsgroup.org/ See http://webstandardsgroup.org/mail/guidelines.cfm for some hints on posting to the list getting help **
Re: [WSG] Linking to top of page
Lachlan Hunt wrote: It depends where the H1 is positioned. If it's not at the top, the page won't be scrolled to the top. If the H1 isn't at the top of the content, then I'd say there's a pretty good case for saying that the H1 is in the wrong place. The top of the content is where it should go, not the top of all the nav and other guff that may be on the page as well, in my opinion. Cheers Ian -- _ zStudio - Web development and accessibility http://zStudio.co.uk Snippetz.net - Online code library File, manage and re-use your code snippets links http://snippetz.net ** The discussion list for http://webstandardsgroup.org/ See http://webstandardsgroup.org/mail/guidelines.cfm for some hints on posting to the list getting help **
Re: [WSG] Linking to top of page
Curby wrote: Lachlan is right, I use image replacement to reposition and stick an image in the upper-lefthand corner of the viewport instead of the normal H1 text. It stays visible in the corner for compatible browsers. Example: http://curby.net/doc/layout/jello-new.html Sorry to disagree but looking at your code, you should be using #content as the back to top anchor destination. Your use of a decorative graphic element on the left edge is irrelevant to the issue. Hope this helps Cheers ian -- _ zStudio - Web development and accessibility http://zStudio.co.uk Snippetz.net - Online code library File, manage and re-use your code snippets links http://snippetz.net ** The discussion list for http://webstandardsgroup.org/ See http://webstandardsgroup.org/mail/guidelines.cfm for some hints on posting to the list getting help **
Re: [WSG] Should logo not link to the homepage?
Justin Owens wrote: * the img alt text read 'logo' but the link went to home Your alt attribute should be modified to correctly represent the actual image. IMHO, 'logo' is not descriptive enough to be used as alternative text for a linked image. In my opinion, it is important for linked images to have alt text that explains the function of the link - the alt text should be more or less the same as the link text ought to be if the link was on text instead of an image. The alt should not be a literal description of the image in this case. So, the logo should say something like MSNBC home page so that it includes functional branding but clearly conveys the destination of the link. To consider two extreme examples if the MSNBC logo was linked to the home page, alt text of MSNBC would be the least helpful, and alt text of Home would be the most correct. I think the best course of action is a compromise incorporating both ideas like that above. -- _ zStudio - Web development and accessibility http://zStudio.co.uk Snippetz.net - Online code library File, manage and re-use your code snippets links http://snippetz.net ** The discussion list for http://webstandardsgroup.org/ See http://webstandardsgroup.org/mail/guidelines.cfm for some hints on posting to the list getting help **
Re: [WSG] Should logo not link to the homepage?
Patrick Lauke wrote: How about a dual approach of using ALT that describes the image (MSNBC) and a title on the link to provide additional advisory information of where the link is going to ('MSNBC home page'), i.e. Great minds and all that? If you reread the previous bit of my post you'll see: 'So, the logo should say something like MSNBC home page' We seem to be in agreement :) Sorry my post was a bit disordered -- _ zStudio - Web development and accessibility http://zStudio.co.uk Snippetz.net - Online code library File, manage and re-use your code snippets links http://snippetz.net ** The discussion list for http://webstandardsgroup.org/ See http://webstandardsgroup.org/mail/guidelines.cfm for some hints on posting to the list getting help **
Re: [WSG] Google pages and the XHTML doctype
Gunlaug Sørtun wrote: Neither Google, nor most of those who might use that or similar services, care all that much about XHTML - or any other standard. Now, which Task Force over at http://www.webstandards.org/ should take on the task of fixing this Google service? Should keep any Task Force busy for a long - long - time. Outcome? ...? Doesn't hurt trying though. I think this is a little harsh. I have just done a quick test on the service, and I see no font tags. What I do see is a thoughtful, earnest attempt to deliver content managed publishing for ordinary people in a way that tries to comply with current web standards. No font tags, no tables or spacer gifs; it even has nice touches like an empty script element to avoid FOUC since they are going for an all-out CSS layout. The OP could just as well have published a positive note that Google have introduced a really nice Web 2.0 app in its early stages. Yes it has some validation issues that will be easy for them to iron out, but they are giving web standards a serious go. Kudos to Google, on this occasion. I think they have made a good start on the service. They tell you it's a work in progress when you sign up, so you can expect it to look a bit rough. Cheers Ian -- _ zStudio - Web development and accessibility http://zStudio.co.uk Snippetz.net - Online code library File, manage and re-use your code snippets links http://snippetz.net ** The discussion list for http://webstandardsgroup.org/ See http://webstandardsgroup.org/mail/guidelines.cfm for some hints on posting to the list getting help **
Re: [WSG] Breadcrumb as Section Heading H1
Kevin Futter wrote: Anyway, for the benefit of others interested in this thread/topic, the upshot from the above link seems to be that the pipe character (|) is the best compromise currently available as a screen reader-friendly element separator. I profoundly disagree with that. The vertical bar is the worst thing you can use in a breadcrumb trail, regardless of how it reads in a screen reader. This is conventionally used on the web as a separator for sibling links, and it really doesn't work as a breadcrumb separator for sighted users. I don't think a list is appropriate markup for breadcrumbs and prefer the conventional You are in: home products foo Screen reader users in our user testing on sites with this construct had no problems with it reading home greater than products greater than foo, and reported that it was a very useful aid. You should not overlook the vital importance of the visual direction implied by the character. This is a crucial means of showing that it is a hierarchical relationship. Not all accessibility is about screen reader users after all. The conventional breadcrumb with is a working compromise which has proven to be a good tool in usable sites. Cheers Ian -- _ zStudio - Web development and accessibility http://zStudio.co.uk Snippetz.net - Online code library File, manage and re-use your code snippets links http://snippetz.net ** The discussion list for http://webstandardsgroup.org/ See http://webstandardsgroup.org/mail/guidelines.cfm for some hints on posting to the list getting help **
Re: [WSG] Breadcrumb as Section Heading H1
Ian Anderson wrote: http://www.standards-schmandards.com/?2004/11/06/6-the-sound-of-the-accessibl e-title-tag-separator Anyway, for the benefit of others interested in this thread/topic, the upshot from the above link seems to be that the pipe character (|) is the best compromise currently available as a screen reader-friendly element separator. The vertical bar is the worst thing you can use in a breadcrumb trail Just realised that the article quoted by Patrick is in fact related to separators in the document title. In that case, I agree that the vertical bar is an appropriate separator. For example, quite often it is desirable to put the site name and the specific document title in the title element, and the order may be either way round: Adobe Systems | Buying software online Here, I don't think directional separators have value, unlike in a breadcrumb trail. It's also common to omit intermediate levels of the hierarchy in a page title. Cheers Ian -- _ zStudio - Web development and accessibility http://zStudio.co.uk Snippetz.net - Online code library File, manage and re-use your code snippets links http://snippetz.net ** The discussion list for http://webstandardsgroup.org/ See http://webstandardsgroup.org/mail/guidelines.cfm for some hints on posting to the list getting help **
Re: [WSG] Quick Site Check - CSS Problem?
David Nicol wrote: I would appreciate it very much if you could look at this site: http://www.visitshetland.com/ ... This person was using IE6, on a brand new laptop. Site looks fine on Windows XP SP2, IE6 Could it be that one of the CSS files failed to load properly for this user - perhaps the main one - and what is left is what you get from the normal unstyled rendering plus a couple of things in the IEMac or IE5 style sheets? Some network caching or proxy server in their Internet access may conceivably have made the problem persist beyond a simple refresh, I suppose. You could try disabling your main CSS and see if the view resembles that user's? HTH Cheers Ian -- _ zStudio - Web development and accessibility http://zStudio.co.uk Snippetz.net - Online code library File, manage and re-use your code snippets links http://snippetz.net ** The discussion list for http://webstandardsgroup.org/ See http://webstandardsgroup.org/mail/guidelines.cfm for some hints on posting to the list getting help **
Re: [WSG] TARGET in 4.01 Strict
Lachlan Hunt wrote: Many users hate popup windows. There are no valid use-cases or reasons for opening a popup window, don't do it. I disagree with this statement. In my opinion, there are several very good use cases. The primary one is help windows, where instructions can be compared with the items in the main window. A variant of this is TCs in forms, where there are serious consequences for the user if they move off the page while in the middle of completing a form - they may lose some or all data. I think generalisations like users hate popup windows are perhaps a little unhelpful. I am sure a lot of people dislike popup adverts, but to dismiss all use of popups on this basis is clearly to throw the baby out with the bath water. That is absolutely no better than using the target attribute. In fact, for accessibility reasons, it is worse. This is not true, in my experience. Popup windows with JavaScript are more usable than with target, because they are usually sized smaller than the main window (so not causing confusion), and caused users with disabilities no problems at all in any of the user testing we've done where popups were encountered in tasks. Also, because the button UI is usually turned off, there are fewer issues with users starting to surf in the new window and losing their window history, which is a problem using target _blank. In fact, screen reader users may well perform tasks faster in popup windows because there is a lot less clutter around the content. Provided they are used appropriately, and the user is informed of their use in situations where they are not normally found, in my view there is no huge problem with them. Cheers Ian -- _ zStudio - Web development and accessibility http://zStudio.co.uk Snippetz.net - Online code library File, manage and re-use your code snippets links http://snippetz.net ** The discussion list for http://webstandardsgroup.org/ See http://webstandardsgroup.org/mail/guidelines.cfm for some hints on posting to the list getting help **
Re: [WSG] accessible drill-down into a nested list
Paul Novitski wrote: Tell me if this would be a better scenario: When you select a menu item, the page reloads with a set of breadcrumbs that spells out the history of selected menu items, such as: I think you are correct to be concerned about the issue, but this may not be the optimal solution. If you consider the requirements of a screen reader user, their task is one of wading through immense amounts of irrelevant stuff trying to find the thing of interest at that moment. What's of most interest at the moment you're describing is hearing the *new* options that are now available. The breadcrumb (hierarchical location) is also hugely appreciated as a means of keeping track of where they are in the site. From user testing we've done with screen reader users, the most important thing is that the page title and main heading on the page are descriptive. The page title is read first as the page loads, and then the behaviour depends on the screen reader in use. JAWS will typically skip over stuff that has been seen before, and try to jump to the first new content on the page. If the page title and heading don't change, this can be very destructive for these users, as they start trying to backtrack or reload the page to see what has gone wrong. In the scenario you describe, is the page more or less identical except that a new submenu has appeared in the navigation area? I think this would be very harmful UI for screen reader users. The chances of them locating the submenu are remote, and the chances of them realising that it represents hierarchically subordinate options are too. The best solution for these users may be to create menu pages that contain the submenu links as primary content. Ensuring that navigation links come after content links in the source order may also be very beneficial, as these downward or onward links are much more likely to be what the user is looking for. Somewhere else on the page, perhaps last in the markup, would be the full menu including all menu items at each selected level. A jump to navigation link early on the page could get you there quickly. I think this would be immensely bad design for screen reader users. This is a site map. What you may be missing is that too many links are the bane of a screen reader user's life. They rely on using links as a kind of binary tree to navigate the site - the last thing they benefit from is hearing links again that they have already discarded as not of interest. They go back much more than sighted users in order to find a link they heard before. The other interesting thing is that screen reader users build a mental map of a site that is nothing like the real architecture, based on the links they hear. If every link is on every page, all pages sound the same to them, because about half of a user's time on each page is spent listing the links. When the links on each page are mostly unique, screen reader users perform better in tasks. So, have a site map linked off each page, but don't include extra links on every page - these are bad for screen reader users, not helpful, in my opinion Hope this helps Cheers ian -- _ zStudio - Web development and accessibility http://zStudio.co.uk Snippetz.net - Online code library File, manage and re-use your code snippets links http://snippetz.net ** The discussion list for http://webstandardsgroup.org/ See http://webstandardsgroup.org/mail/guidelines.cfm for some hints on posting to the list getting help **
Re: [WSG] Target sued over non-accessible site
Absalom Media wrote: Amount of Javascript disabled based on various client profiles I've got: My site: Less than 0.1% Commercial music site: Less than 0.5% Commercial / education health care site: Less than 0.7% What methodology are you using to identify humans as opposed to search engines and other robots? I am used to seeing figures of around 5% in global stats for non-JavaScript users, and assumed this is grossly distorted by something. Very interested to find out more... Cheers Ian -- _ zStudio - Web development and accessibility http://zStudio.co.uk Snippetz.net - Online code library File, manage and re-use your code snippets links http://snippetz.net ** The discussion list for http://webstandardsgroup.org/ See http://webstandardsgroup.org/mail/guidelines.cfm for some hints on posting to the list getting help **
Re: [WSG] IE7 Compatibility Team
Thierry Koblentz wrote: height:1% or height:0 or whatever height you set gives layout to an element, which is not the case with display:inline. Making sure an element hasLayout is a big tool in the box when it comes to fix IE bugs. Just FYI, over on CSS-D there was a thread today about a web chat with the MS IE7 developers, in which hasLayout was mentioned and I thought this specific point was quite important: IE7 will respect height: 1%, which if not filtered away from it could break many layouts. They are retaining hasLayout as an internal property and recommend using zoom: 100% for inducing hasLayout as a replacement technique for height: 1%. More here http://sltclan.com/images/cj/ie7.html Cheers Ian -- _ zStudio - Web development and accessibility http://zStudio.co.uk Snippetz.net - Online code library File, manage and re-use your code snippets links http://snippetz.net ** The discussion list for http://webstandardsgroup.org/ See http://webstandardsgroup.org/mail/guidelines.cfm for some hints on posting to the list getting help **
Re: [WSG] DIV Target
Samuel Richardson wrote: If you want to load entire web pages embedded into the current page you will have to use the iframe, if just want to change simple text/html within a div then you will have to use the innerhtml property (or use this method that came up on delicious this morning: http://slayeroffice.com/articles/innerHTML_alternatives/) I agree that this would be better done using DOM scripting as a more standards-based approach, like it says in that article. I'm new to DOM scripting myself, but found this book by Jeremy Keith to be excellent; DOM Scripting http://www.domscripting.com/book/ It's suitable for those with very little scripting experience as well as those with intermediate or better JavaScript skills - I skipped the first 60 pages and then found the content and pacing to be exactly what I needed to understand DOM-based alternatives to things like innerHTML and old-fashioned image swapping techniques. I think it would be a great addition to the shelf for those that want a good introduction to the area, considering degradable code and accessibility at every stage; the book is well written and very easy to follow. What it's not is a recipe book of flashy effects. Going back to the original request, my concern would be about how accessible the solution would be without JavaScript; you'd need to ensure that the content was either supplemental to the main purpose of the page, or that the page degraded well so that the content was still available without JavaScript. Hope this helps Cheers Ian -- _ zStudio - Web development and accessibility http://zStudio.co.uk Snippetz.net - Online code library File, manage and re-use your code snippets links http://snippetz.net ** The discussion list for http://webstandardsgroup.org/ See http://webstandardsgroup.org/mail/guidelines.cfm for some hints on posting to the list getting help **
Re: [WSG] Most semantic XHTML markup possible - your thoughts
[EMAIL PROTECTED] wrote: But this definition list code I feel is not the most semantic way to solve the problem (not using the dl as it was intended). However, I also feel that what I have currently is also not perfect (and if it is, is em better than strong??). Hi Nathan, why don't you think a definition list is appropriate? Seems spot on, to me... Definition lists vary only slightly from other types of lists in that list items consist of two parts: a term and a description...Another application of DL, for example, is for marking up dialogues, with each DT naming a speaker, and each DD containing his or her words.[1] div id=customer-details h3Welcome Frankh3 dl dtCustomer Identification Number (CID):/dt dd100 223 578/dd dtLast accessed/dt ddFeb. 12, 2006 at 9:47pm/dd /dl /div There is a clear relationship between the label and the content, so in my opinion this content would suit either a definition list or a table. A table would be overkill here, but still theoretically appropriate because the number for CID would have no independent meaning without the associated label. I don't think there is any semantic or practical difference between em and strong, personally. I would be guided by how appropriate the traditional visual rendering of these is to the content. e.g. the name of a sea-going vessel is traditionally italicised, so I would use em in that case. (Not that it comes up a lot.) Hope this helps Cheers Ian PS - Hello all on WSG - this is my first post :). Looks like a great list. [1] http://www.w3.org/TR/html4/struct/lists.html#h-10.3 -- _ zStudio - Web development and accessibility http://zStudio.co.uk Snippetz.net - Online code library File, manage and re-use your code snippets links http://snippetz.net ** The discussion list for http://webstandardsgroup.org/ See http://webstandardsgroup.org/mail/guidelines.cfm for some hints on posting to the list getting help **
Re: [WSG] Most semantic XHTML markup possible - your thoughts
[EMAIL PROTECTED] wrote: However, in my case, I just felt that 100 223 578 does not define Customer Identification Number (CID). ...So am I wrong in my thinking about definitions lists?? Can a random bunch of digits such as 100 223 578 really be a definition that means Customer Identification Number?? If someone can qualify these questions, then I guess I will be convinced that this is the most semantic way to solve my problem. Well, my understanding is that DL was not intended to be used solely as strictly as your interpretation suggests. Again, what the HTML 4.01 spec says is: Definition lists vary only slightly from other types of lists in that list items consist of two parts: a term and a description...Another application of DL, for example, is for marking up dialogues, with each DT naming a speaker, and each DD containing his or her words. If you look at the latter example W3 give, this does not fit the strict interpretation of the DD being an *actual* definition of the term either. Many people use DL for marking up list items and labels, where again there is a semantic association but not a classical definition as such. I feel this is perfectly appropriate. Hope this helps Cheers Ian -- _ zStudio - Web development and accessibility http://zStudio.co.uk Snippetz.net - Online code library File, manage and re-use your code snippets links http://snippetz.net ** The discussion list for http://webstandardsgroup.org/ See http://webstandardsgroup.org/mail/guidelines.cfm for some hints on posting to the list getting help **