25 replies [Last post]
matte
matte's picture
Offline
Enthusiast
Last seen: 5 years 49 weeks ago
Joined: 2004-03-18
Posts: 142
Points: 0

In my never ending search for copy and pastable text, I want to use a table to hold some images in each of many records rendered on a page (because tables are ignored when you paste to plain text). But I would hate to have screen readers always reading off these tables which are merely for layout, and have no content value.

Is there some trick that can hide a table from a reader?

Tyssen
Tyssen's picture
Offline
Moderator
Brisbane
Last seen: 1 year 2 weeks ago
Brisbane
Timezone: GMT+10
Joined: 2004-05-01
Posts: 8201
Points: 1386

The only way would be to use

The only way would be to use some sort of javascript UA-sniffing to detect a screenreader and then direct them to a different page (or serve different content on the same page - either way, you're going to have to create two lots of content).

Is that sort of method reliable? No; sniffing screenreaders is even less reliable than sniffing for normal browsers.

Is it something that would be worth doing? Not for the amount of work involved and the fact that your desired outcome is far from guaranteed.

How to get help
Post a link. If you can't post a link, jsFiddle it.
My blog | My older articles | CSS Reference

matte
matte's picture
Offline
Enthusiast
Last seen: 5 years 49 weeks ago
Joined: 2004-03-18
Posts: 142
Points: 0

Thanks tyssen. I have now

Thanks tyssen.

I have now posted on several accessibility forums, and no one seems to want to answer me. Is it considered acceptable to use small tables for widgets, such as button panels, etc/, at all? And if it is, are there tricks for minimizing the redundancy of readings for repetive layout elements?

There are now many popular Javascript components and libraries (EXTjs, YUI, Moca, Prototype, Jqery, Dojo,etc), and they can all use tables for toolbars, lists, menus, etc. With so many of these in use, it mystifies me that i cannot find more opinions about how to improve their accessibility. Those with no keyboard support are not even worth consideration, but even those that have it, only have spotty support.

It's very frustrating that organizations which supposedly exist to promote accessibility, do so little to actually demostrate best practices. There is one project, called FLUID, that is presumably developing open source widgets with accessible considerations, but i can't find a working demo on their site. Seems to be mostly an initiative in progress with a lot of ambition, but not too much to show for it, thus far.

Ed Seedhouse
Ed Seedhouse's picture
Offline
Guru
Victoria British Columbia
Last seen: 12 weeks 19 hours ago
Victoria British Columbia
Timezone: GMT-7
Joined: 2005-12-14
Posts: 3570
Points: 675

You use tables for tabular

You use tables for tabular content, such as multi column output from a database, where each column has a natural meaning that is the same for all rows. For such content a table is the only proper markup, and it's what they were designed for.

For anything else tables are simply inappropriate, especially if you are trying to code for cross browser compatability and especially when coding for different kinds of displays.

Read up on "semantic" html coding. Accesability is largely but not wholly an outcome of good semantic coding practices. At least semantic code makes accesability ever so much easier to achieve.

Basically, if you are writing your html with any particular kind of output device or "look" in mind you are probably not coding it correctly. Proper html ideally has nothing to do with presentation. CSS is for presentation. The two concerns are separate and should be dealt with separately. Semantic coding practices along with CSS allow that.

Ed Seedhouse

Posting Guidelines

Watch out! I am carrying irony, sarcasm and satire, and know how to use them.

matte
matte's picture
Offline
Enthusiast
Last seen: 5 years 49 weeks ago
Joined: 2004-03-18
Posts: 142
Points: 0

So what you are saying is..

...that all table based components, toolbars, menus,etc.,---which have become very popular with web2.0/ajax developers -- are basically no-nos in all cases? Many of them have keyboard access. Do I assume they are simply deciding that making them somewhat accessible is acceptable given the benefits they provide to regular users?

All over the web, people claim to be sensitve to the issues, at least in theory. But nobody actually seem to know many of the details of them. Even the so called experts don't want to be specific very often. If tables were always bad, why then would such accessibility experts say to use "table summarary attributes" to describe such layout tables. Such uses are not tabular data, and their use is obviously widepread enough to warrant this recommendation. Is this just out of recognition that people will use them anyway?

I wonder how many desigenrs flocking to javascript widget libraries using tables know what trouble they might be creating for themselves, should accessibility mandates become important to their products.

Tyssen
Tyssen's picture
Offline
Moderator
Brisbane
Last seen: 1 year 2 weeks ago
Brisbane
Timezone: GMT+10
Joined: 2004-05-01
Posts: 8201
Points: 1386

matte wrote:In my never

matte wrote:
In my never ending search for copy and pastable text

I thought you were using tables because they let you copy and paste text the way you want to (I'm only going on your say-so on this cos I don't really know), not because the authors of javascript libraries have marked up their examples that way. Just because the examples of a particular framework are marked up a particular way doesn't mean it's the best way to do it.

matte wrote:
Do I assume they are simply deciding that making them somewhat accessible is acceptable given the benefits they provide to regular users?

I wouldn't assume anything. They might be using tables cos they've always used tables and cos they didn't think about doing it any other way. The fact that it's done in tables doesn't necessarily mean it's the best way or the only way to do it. But without seeing specific examples of what you're talking about, it's hard to comment further.

How to get help
Post a link. If you can't post a link, jsFiddle it.
My blog | My older articles | CSS Reference

Ed Seedhouse
Ed Seedhouse's picture
Offline
Guru
Victoria British Columbia
Last seen: 12 weeks 19 hours ago
Victoria British Columbia
Timezone: GMT-7
Joined: 2005-12-14
Posts: 3570
Points: 675

matte wrote:...that all

matte wrote:
...that all table based components, toolbars, menus,etc.,---which have become very popular with web2.0/ajax developers -- are basically no-nos in all cases?

They are only no-nos if you are intending to write valid, standards based, accessible websites using html and CSS in the way they are intended to be used by the folks who created the web.

No one can stop you from using all these components and toolboxes just because they happen to be crap. But since this whole site is dedicated to standard based design probably you shouldn't expect a helpful answer from us because many of us gone beyond all that stuff and are happy to have done so, and trying to forget it as fast as we can to quell the nightmares and flashbacks. Smile

Popular doesn't mean good.

Ed Seedhouse

Posting Guidelines

Watch out! I am carrying irony, sarcasm and satire, and know how to use them.

Tyssen
Tyssen's picture
Offline
Moderator
Brisbane
Last seen: 1 year 2 weeks ago
Brisbane
Timezone: GMT+10
Joined: 2004-05-01
Posts: 8201
Points: 1386

Ed Seedhouse wrote:But since

Ed Seedhouse wrote:
But since this whole site is dedicated to standard based design probably you shouldn't expect a helpful answer from us because many of us gone beyond all that stuff and are happy to have done so

On the contrary, people should expect helpful answers no matter what the nature of their enquiry with the possible exception being those who ask in rude or disrespectful ways. And of those 'many of us', most also don't hold to the notion that we're somehow better than the person asking just because we develop web sites in a certain way. :rolleyes:

How to get help
Post a link. If you can't post a link, jsFiddle it.
My blog | My older articles | CSS Reference

Ed Seedhouse
Ed Seedhouse's picture
Offline
Guru
Victoria British Columbia
Last seen: 12 weeks 19 hours ago
Victoria British Columbia
Timezone: GMT-7
Joined: 2005-12-14
Posts: 3570
Points: 675

Tyssen wrote: On the

Tyssen wrote:

On the contrary, people should expect helpful answers no matter what the nature of their enquiry with the possible exception being those who ask in rude or disrespectful ways. And of those 'many of us', most also don't hold to the notion that we're somehow better than the person asking just because we develop web sites in a certain way. :rolleyes:

I don't think I'm "better" than anyone who doesn't code to standards and I don't think most of the other memberd do either. "Different" != "better".

But it has been my observation that those who don't find it harder to get help with their old code practices than they might in other less dedicated forums. We tend to try to convert them to the standards approach and that's sometimes hard for new members to understand.

There are certainly some of us, like you for instance, who do go out of their way to help with legacy coders more than others. I'm not one of them myself because I'm lazy, old, and often tired and grouchy. Did I mention ugly? And I haven't had enough time for drinking lately.

As my sig line says, of course, I could be wrong.

Ed Seedhouse

Posting Guidelines

Watch out! I am carrying irony, sarcasm and satire, and know how to use them.

matte
matte's picture
Offline
Enthusiast
Last seen: 5 years 49 weeks ago
Joined: 2004-03-18
Posts: 142
Points: 0

This certainly has gotten

This certainly has gotten off topic, but since it has, let me say that I am simply trying too understand how I can make modern UI components that are as accessible and compliant as they might be -- if at all practical and possible. If this were not important, I doubt W3 would have an extensive road map for the future of such toys.

The issue that I, and many seem to be facing, is how to get from here to there until they do. My application uses a LOT of dynamic data and requires a great deal of interaction that just isn't going to work in a device independent context. And I'm not going to cripple its utility for 95% of the users out there, simple because the standards have not evolved which can bring similar tools to the remaining 5% with disabilities or other challenges. Naturally, I would like it to work, and be at least adequate, if not optimal, and without having to code a completely separate interface for a large and expensive project.

Just the word "standards" in this context raises the hair on my neck sometimes. In my view, most of these "standards" were not really standards to begin with. They became de-facto standards with a lot of industry push and pull having little to do with what was needed, or right, and with a lot of manipulative funny business from MSFT and other industry stakeholders.

It seems to me that most of the "practices" we have now, have been workarounds, hacks, and brute force compromises that became official standards, often as a result of messy, circuitous and very politicized process. W3 has long acknowledged that many things came about more by popular demand and industry fiat, than by concise planning, or precise architecture.

Most web UI component concepts were primitive in the mid 90s, relative to what PC-based computing could achieve. I think it's perfectly realistic and sensible for the the web2.0 developers to want to modernize the browser and make it do what we've taken for granted on PCs for 20 years, or so.

And since only about 10% of all websites even make a token gesture toward accessibility anyway, I think it's a worthy goal to at least be concerned with making widgets and other modern UI components as accessible as possible, whether a current "standard" or not. It seems almost self-evident (to me, at least), that the real culprit is not the components we create, but that the device manufacturers and reader programmers haven't caught up. And the only way to push them to do that is to keep innovating forcefully, forcing them to play catch-up even faster.

And hell, HTML itself ignored standards that already existed, and that's been half the problem with it. But that's the nature of innovation, eh? The game isn't always played fairly or elegantly, but the ball still lurches on down the field.

Thanks for the discussion, and tolerance for we newbs Smile

I still think it would be great if there were a separate forum here, just for accessibility issues. I am sure many of the mods have a lot to offer on this topic.

Chris..S
Chris..S's picture
Offline
Moderator
Last seen: 1 year 49 weeks ago
Timezone: GMT+1
Joined: 2005-02-22
Posts: 6078
Points: 173

Back on topic sort of.

Back on topic sort of.

If you hunt around this forum, you'll find a number of posts by Lorraine on aspects of accessibility. As I recall, one of the points regarding the use of tables was if tables are used for layout then they must be able to be read naturally, ie. left to right, top to bottom, table by table.

I would think that your best method of handling screen readers is to offer them the same page content formatted in a different manner. What is the likelihood of screen reader users attempting to use the mouse to copy/paste? How likely are they to find your icons useful?

The accessibility guidelines don't require that one page is accessible to all. Providing a links to the same content in differently formats is ok.

matte
matte's picture
Offline
Enthusiast
Last seen: 5 years 49 weeks ago
Joined: 2004-03-18
Posts: 142
Points: 0

We will have to have some

We will have to have some alternatives for pop up dialogs and other things anyway. My goal is to minimize the number of places we need these alternate interfaces. I would like to deploy this generic "item" structure in many applications, and it would be really annoying to have to code alternative pages for every one of them.

I have asked a few accessibility specialists on other sites, and they have yet to say OH, DON"T DO THAT! They seem more concerned that I simply have ALT attributes on the icons. But that's a no brainer. Still, I am still not convinced it's ok to use a layout table for my icons (and some images in other contexts), but no one wants to give me a firm thumbs up or down on this. Really frustrating. Perhaps there isn't a firm answer, but hell, I'd just like someone to say THAT authoritatively too!

Just for the hell of it, here's an example I emailed to someone an hour ago:

-------------------------
...so the issue is not how to use ALT attributes, but merely whether the use of a "formatting table" was acceptable -- AT ALL -- in this context. We need to position our icons in a very specific way, and only a small table will give us just what we want.

It seems to us that such a layout would create extraneous audio clutter for a reader, for each and every record in a list. We are unsure if this would be considered a clear violation of accessibility standards or guidelines. If not, then we're good to go. We just need someone in the know to give us a clear thumbs up or down about it (we're insecure that way Smile.

Here is roughly how each item record in given list would appear schematically. Obviously, the icons would be inside A tags with appropriate ALT attributes, etc.

ICON1 ICON2

Text of paragraph here. Would this an acceptable use of a table, without violating accessibility standards or principles?

matte
matte's picture
Offline
Enthusiast
Last seen: 5 years 49 weeks ago
Joined: 2004-03-18
Posts: 142
Points: 0

Well, a definitive answer.. at last

I spoke to Jim Thatcher, who is a rich source of knowledge on accessibility. I'm sure many are aware of him, but he was sure a find to me. He replied that the use of a table in this instance is perfectly acceptable. Thanks to all here for their input.

For reference, here's some background on him:

Thatcher's... work culminated in the development of one of the first screen readers for DOS in 1984-85, called IBM Screen Reader. Jim Thatcher served as Vice-Chair of the Electronic and Information Technology Access Advisory Committee (EITAAC) which was empanelled by the Access Board to propose standards for Section 508; he chaired the subcommittee on software standards. Later he wrote the course on Web Accessibility for Section 508 for ITTATC, the Information Technology Technical Assistance and Training Center, which was funded in support of Section 508. The course is available here, http://jimthatcher.com/webcourse1.htm.

matte
matte's picture
Offline
Enthusiast
Last seen: 5 years 49 weeks ago
Joined: 2004-03-18
Posts: 142
Points: 0

More on using tables for record layouts

I've gotten a few requests to clarify what I mean by using tables for paragraphs, records and other mini-layouts. So let me post an example here. I'd be interested to hear any and all opinions, objections, rants, flames, or other comments about why such structures could be seen as a bad thing -- if they are.

Background: Many sites use tables for item layout. Even Google uses them for their listings. We cannot see a reason why we should not use them in this way, but have never been clear on whether it's really considered accessible or not. Yes, we could do alternative pages that remove them for reading devices, but that would not be practical for all content. So we need a default structure that is "good enough."

If it's perfectly kosher to use tables this way, it would make our application dramatically easier to program, style, and customize. (and copy and paste, too Smile. Below is a schematic example of what we want to do, presuming it's not ruled a complete accessibility or standards faux pas.

Icon Buttons

Each and every paragraph of our documents, blogs, forum posts, articles, comments, and other text objects would be placed in such a generic container before receiving additional CSS styling.

Etc.

matte
matte's picture
Offline
Enthusiast
Last seen: 5 years 49 weeks ago
Joined: 2004-03-18
Posts: 142
Points: 0

And this response from Jim Thatcher

(Reprinted With his permission)
"The purists will never say tables are OK – but for accessibility they are! Commercial sites are trying to go tableless, for example IBM.com and Priceline.com – in fact jimthatcher.com. But the truth is that for accessibility, the tables almost always do not hurt. Your layout suggested below is just fine. It would not be a problem. Don’t forget the alt-text on those icons."

Ok, so at least one accessibility guyt won't complain about paragraphs formatted with tables. And Google uses them not for query result listings. Is there ANY reason they could not be used for any type of content in our CMS-like system? They are less brittle and buggy for many things we want to do. So what red flags remain? Who is there left to offend? What traps might we be setting for ourselves. Your crystal balls welcomed Smile

Tyssen
Tyssen's picture
Offline
Moderator
Brisbane
Last seen: 1 year 2 weeks ago
Brisbane
Timezone: GMT+10
Joined: 2004-05-01
Posts: 8201
Points: 1386

OK, so from an accessibility

OK, so from an accessibility point of view your example is OK, but I have to say that there doesn't appear to be anything in your example that couldn't be done without tables and it seems to me a 'no brainer' that you wouldn't use tables in such a case.

How to get help
Post a link. If you can't post a link, jsFiddle it.
My blog | My older articles | CSS Reference

Ed Seedhouse
Ed Seedhouse's picture
Offline
Guru
Victoria British Columbia
Last seen: 12 weeks 19 hours ago
Victoria British Columbia
Timezone: GMT-7
Joined: 2005-12-14
Posts: 3570
Points: 675

matte wrote: Ok, so at least

matte wrote:

Ok, so at least one accessibility guy won't complain about paragraphs formatted with tables. And Google uses them not for query result listings. Is there ANY reason they could not be used for any type of content in our CMS-like system? They are less brittle and buggy for many things we want to do.

No they are not. Table code is convoluted and hard to understand compared to straightforward semantic code. Tables are fine for their intended use, but will cause many problems when they are abused, as you seem to want to do.

Quote:
So what red flags remain? Who is there left to offend? What traps might we be setting for ourselves. Your crystal balls welcomed Smile

Who cares who you offend? There are much better reasons for avoiding tables except for tabular data than who you may or may not offend. It's much harder to maintain than proper coding practices for one thing. I know from personal experience, having to deal with a legacy site that abuses tables. It is a nightmare.

Try googling on "tables for layout" for a start. The first link returned is a good place to begin reading, if a bit simplistic. The majority of links returned on the first page of such a google search is anti table (which ought to tell you something), but there are dissenting viewpoints as well.

So read then all and make up your own mind.

Ed Seedhouse

Posting Guidelines

Watch out! I am carrying irony, sarcasm and satire, and know how to use them.

matte
matte's picture
Offline
Enthusiast
Last seen: 5 years 49 weeks ago
Joined: 2004-03-18
Posts: 142
Points: 0

Well..

Ty, that's the problem. There is a reason for it. Several, actually. The example above is greatly simplified. There are many issues andnuances discussed in several overlapping threads pertaining to my outlines, numbering, and icon issues over several weeks.

The first, is that these "items" must be used in a hiearchical, collapsing outline structure, while preserving the icon/button integrity, outdented/indented (generated) numbering, and many other issues, while still TRYING -- repeat TRYING to maintain a reasonably straightforward copy and paste result when pasted as plain text.

After weeks of painstaking experiements, and the invaluable help of you, and this great communnity, I have reduced the issues to one where the best middle-ground solution, that can still be copy and pasted (almost perfectly in all browsers), requires the darn table as a container. Obviously, I've expended no small effort to avoid it, but just can't seem to get there. I am doing ONE LAST implemtation of this solution now. After that, I am going forward unless someone has a compelling reason why I should not. This should not be a seen as a "kick me" sign on my back. I am not looking for more punishment. I just don't want to start the final mile of an elaborate template rendering engine that uses a core record structure I might regret. Color me paranoid Smile

matte
matte's picture
Offline
Enthusiast
Last seen: 5 years 49 weeks ago
Joined: 2004-03-18
Posts: 142
Points: 0

Arrrh.. kicked myself

Just went over my notes, and remembered the fatal flaw with a tabled container (I was sure I had one).

There would be no fluidity around floats. A floated box or image could only be above, below, or adjacent to a text paragraph. No wrapping around the float from even one paragraph, and certainly not two. Thus, there would be no simulating the wrap behavior of a P tag, list item, or div.

Argggggggggggggh.

Oh well, at least the tabled icon structure helps. The downside is only that a copy and paste produces TWO blank lines between every paragraph. But other than that, everything is rather nice.

matte
matte's picture
Offline
Enthusiast
Last seen: 5 years 49 weeks ago
Joined: 2004-03-18
Posts: 142
Points: 0

All good points, Ed. I

All good points, Ed. I understand the layout objection pretty well, and my use was more atomic than a page layout. Our pages are rather fluid. This was simply concerned with list items, posts, teasers, comments, and other items rendered dynamically within the main layout. The live data.

Given that I have already found a good reason to not do it after all, is the "icon" container described above at least acceptable, if less than ideal? The paragraph would be a P tag, and only the icon buttons within each would be contained by a small left- floating table. Here again, was my schematic illustration from earlier:

ICON1 ICON2

Text of paragraph here. Would this an acceptable use of a table, without violating accessibility standards or principles?

Ed Seedhouse
Ed Seedhouse's picture
Offline
Guru
Victoria British Columbia
Last seen: 12 weeks 19 hours ago
Victoria British Columbia
Timezone: GMT-7
Joined: 2005-12-14
Posts: 3570
Points: 675

matte wrote:All good points,

matte wrote:
All good points, Ed. I understand the layout objection pretty well, and my use was more atomic than a page layout. Our pages are rather fluid. This was simply concerned with list items, posts, teasers, comments, and other items rendered dynamically within the main layout. The live data.

Trouble is, the display of an item you insert will be affected materially by the rest of the page unless you have assured that all the other content is valid html. Invalid html can mess up things far away from the errors.

Quote:
Given that I have already found a good reason to not do it after all, is the "icon" container described above at least acceptable, if less than ideal? The paragraph would be a P tag, and only the icon buttons within each would be contained by a small left- floating table. Here again, was my schematic illustration from earlier:

ICON1 ICON2

Text of paragraph here. Would this an acceptable use of a table, without violating accessibility standards or principles?

"Acceptable?" I'm not a CSS or html God you know, just a guy with a particular opinion.
I'm not in the habit of checking every page I go to for good design practices, either. Why would I have any authority over what is "acceptable"?

What I think is that if you are inserting stuff dynamically into a web page you should insert as little as possible, preferably text only. Failing that if you are iterating over some database output for example, output the semantically appropriate element along with it. Often that may well be a table, but not for presentation purposes, just because a table is appropriate for marking up tabular data.

I also think it's probably a big mistake to be thinking about presentation so early in your design process. Get your system outputting a valid and semantic html file in a strict doctype first, then you can decide how to style it. You might even style it in many different ways for many different display devices. I believe it will save you much work and complication and end up with a better result if you do it that way. But that's my opinion, not the word of God or anything.

Ed Seedhouse

Posting Guidelines

Watch out! I am carrying irony, sarcasm and satire, and know how to use them.

matte
matte's picture
Offline
Enthusiast
Last seen: 5 years 49 weeks ago
Joined: 2004-03-18
Posts: 142
Points: 0

Ed, as I've partially

Ed, as I've partially explained in other threads, we are not really building a site, as much as CMS tool, of a particular sort. It will render many kinds of data, hierarchically, as outlines in some contexts, and simple linear or nested lists of article threads in others. The data is all bound into collections with a unique new kind of tagging technololgy. So I am focused on finding a flexible and generic container that can be deployed in as many contexts as possible, but only as a default skeletal envelope. From that model, more detailed and concise styling is added using the internals of the system.

I wish we could simplify the mission further, but that's sort of the whole point of the tool. To help others do things that are generally hard.

I appreciate your comments and feedback greatly. YOu are an authority to me. By "acceptable, " I mean that I am concerned about what standards cops and critics might say at some future date about big mistakes we may have made. With luck, our product will be used by people doing a lot of government work, and 508 accessibility can and will be a hurdle we/they must face. We just cannot focus on any of that just yet. We are trying to avoid the preventable errors. Clearly, keeping things as simple and fluid as possible will serve us well, and discussions like this are helping us do that.

Ed Seedhouse
Ed Seedhouse's picture
Offline
Guru
Victoria British Columbia
Last seen: 12 weeks 19 hours ago
Victoria British Columbia
Timezone: GMT-7
Joined: 2005-12-14
Posts: 3570
Points: 675

Well there are, what? Maybe

Well there are, what? Maybe some dozens of reasonably successful CMS's out there already. All of them seem to do more or less what you are setting out to do. "Why reinvent the wheel" is the question that might be asked. But then if someone's paying you money to develop a new one, why not?

But the final code that is created by any CMS surely should meet the same standards as any good standards based web site. Semantic html, separation of content and presentation, CSS for style and position, html used only for marking up meaning and structure, not presentation, and always valid html and a strict doctype. But I do not rule the html universe, obviously.

I am not even an authority to myself, so if you make me one for yourself you are making a big mistake, in my opinion.

Ed Seedhouse

Posting Guidelines

Watch out! I am carrying irony, sarcasm and satire, and know how to use them.

matte
matte's picture
Offline
Enthusiast
Last seen: 5 years 49 weeks ago
Joined: 2004-03-18
Posts: 142
Points: 0

I suppose "CMS of a

I suppose "CMS of a particular sort" was merely a way to give you a frame of reference. We're not simply trying to solve this problem for "a site," but rather, for a tool which renders content for a site (many, actually). Perhaps I should have said "an application with some CMS-like features?" If we're reinventing a wheel, the wheel needed some work Smile. HTML was never designed to be a robust, graphical layout and presentation markup language so nearly everything done today is the result of some hack, force, tweak, or technique often completely unheard of until someone did it.. first, or better.

Anyway.. obviously I am concerned with each and every point you raise, or we wouldn't be having this discussion. At each stage, my entire goal has been to find ways to solve a problem in the most compliant way possible. I really felt I had been clear about this, but I apologize, if I've not beeen. But perhaps you could go back and solve the problem the other mods and I had worked on for about a week: the one where a dynamically generated number could be rendered exactly as OL list would do it, IF such an element could do decimal based, outline numbering across platforms (which it cannot). If you can do that with one of those existing wheels, I can start all over with a wagon that you feel uses better standards. I'd be happy to try.

Ed Seedhouse
Ed Seedhouse's picture
Offline
Guru
Victoria British Columbia
Last seen: 12 weeks 19 hours ago
Victoria British Columbia
Timezone: GMT-7
Joined: 2005-12-14
Posts: 3570
Points: 675

matte wrote:HTML was never

matte wrote:
HTML was never designed to be a robust, graphical layout and presentation markup language

So what you are really saying here, it seems to me, is that html is not the language you want. But html is the very basis of the world wide web. There are some very good reasons why it is like that - the web was never intended to do what you want it to do. The web is not paper! Well actually it is sometimes paper, but it is also much more than that. If you design as if it were paper you are just going to make it into a mess.

Quote:
so nearly everything done today is the result of some hack, force, tweak, or technique often completely unheard of until someone did it.. first, or better.

I think I'll bow out of this now, with a reference to two extremely important and valuable web sites that say it all much better than I can. The first is "Web Matters", and the second is "Truth & Consequences of web site design". If you are interested by what I've said here, these both say it much better than I have.

As to your original problem, I invite you to consider the possibility that the reason it is not easy to do is because you are asking something of the medium it was never designed to do, and that this very fact might be a good reason not to do it.

Anyway, that's my piece, so take care.

Ed Seedhouse

Posting Guidelines

Watch out! I am carrying irony, sarcasm and satire, and know how to use them.

matte
matte's picture
Offline
Enthusiast
Last seen: 5 years 49 weeks ago
Joined: 2004-03-18
Posts: 142
Points: 0

You say the web is not "paper"...

..but in fact,if you read some of the early writings of Lee, Goldfarb, Nelson and many others at the time, that is precisely what it was meant to be. An electronic form of communication that could mimic most of what existed in the print world, and then transcend it over time. HTML was supposed to be the beginning of a simpler form of SGML that any scientist or other publisher could learn to use without knowing the far more robust SGML syntax. But it was not conceived as some ideal, closed-end solution in itself. It was just a beginning and not an end.

Many of these things I complain about, and am trying to work around, have been requested for in HTML and CSS working groups by people far more experienced than I am. They include more content generation, layout control, tagging, variables, and countless other things left out of HTML and CSS because they were never designed by people concerned with these issues at the time. Tim B. Lee envisioned HTML to be a structural container for many data and media types to evolve within it. He never figured that browser turf wars would muddy the water as they have.

I suppose it comes down to being content with what is, rather than what should or could be. When I read your remarks, it's almost as if you're suggesting that the people screaming for SGML/XML, laTex, or PDF-like presentation techniques on a web page--as well as the people who pushed for the innovations and standards we now enjoy--are just whiners. It's easy to tell me that I shouldn't want to do decimal numbering, because HTML kinda sucks sometimes (but it's "good enough for government work). But it's not so easy to solve my problem. Flash, Flex and now Silverlight get attention precisely because they address some of these legendary failings, albeit with some top-heavy overhead.

I am hardly the first to suggest that HTML is an inadequate language for what the web has become today, and just because it's not in industry's interests to fix it fast, doesn't excuse those inadequacies, nor obviate the need to work around them at times. I don't see it as unreasonable to expect tools to work well. I would rather rail against the shortcomings at times, and push for ways to improve them someday. CSS experts and hackers provide novel ways of working around many of those inadequacies. And if they didn't, sites like CSScreator wouldn't be so popular.

While it's fun to have, I really didn't want to waste your time, or mine, on this discussion (but I needed coffee anyway). I just wanted to create a versatile, generic container that could support a style of numbering that can be seen in countless printed format examples. Formats that were invented to communicate ideas and information, and have endured for a few centuries now. Just because browsers were written in a rush so scientists could annotate their papers, and porn producers could format their banner ads, and that the HTML LIST element is hostile to marker and box-model tweaks, is hardly a strong rationale for accepting or defending the status quo. I shouldn't have to abandon my goal of a clean dynamic decimal numbering scheme for paragraphs that any magazine or book publisher could output in LaTex in about 10 seconds. Have you ever seen Englebart's HyperDocs, or OPML solutions like RadioLand's Active Renderer? All of these things push HTML to the outer limits at times, often by breaking or invalidating it. Should their creators just fold up their tents because they had to push that envelope to get a desired effect or novel utility? Most of Web2.0 ambitions are UI tools and techniques that work around the limitations of HTML in order to provide a richer user experience. I think that should be the goal, and not rote acceptance of what exists in the here and now. In short, like many people, I just want to make HTML do what I want, and not force my wants to conform to what it does.

As always, a pleasure.

"The vaunted "information superhighway" would hardly be worth traveling if the landscape were dominated by industrial parks, office buildings, and shopping malls. Thanks to the Text Encoding Initiative, there will be museums, libraries, theaters, and universities as well." -- Charles Goldfarb