W3C Gets a Makeover

[joebert]

October 18th, 2009, 1:34 am

http://www.w3.org/

I already looked, the most I can find is a missing character encoding warning on an obscure page.

[digitalMedia]

October 18th, 2009, 3:18 am

Oh thank goodness! Much more aesthetically pleasing.

[mk27]

October 18th, 2009, 6:19 am

Yeah, nice. And thank god they didn't go with fixed width. I have this nightmare because a few months ago the guy who maintains "perldoc.org" revamped the site and it went from one of the nicest doc sites around to an absolute piece of crap (that was not just my opinion, either, the thread at perlmonks.org was nasty, and the guy was there. When I said I felt lucky to have mirrored the old site so I could just use that instead, people started to pm me for it).

When I sent him a letter explaining fixed width was for tabloids and adolescents, not professional documentation, he wrote back explaining "a London Times study had proven 8-10 words per line on average is ideal". For who? Their semi-literate readers? Or their pseudo-literate advertisers? Out to lunch.

And if you want 8-10 words per line, shrink your browser window...do the rest of us have to be penalized? Anyway, whew, thank god nothing that stupid happened over at W3C.

[joebert]

October 18th, 2009, 6:33 am

I'm a fan of short lines and short paragraphs, but 8-10 words-per-line is just ridiculously short.

Quote:

Yeah, nice. And thank god they didn't go with fixed width. I have this nightmare because a few months ago the guy who maintains "perldoc.org" revamped the site and it went from one of the nicest doc sites around to an absolute piece of crap (that was not just my opinion, either, the thread at perlmonks.org was nasty, and the guy was there. When I said I felt lucky to have mirrored the old site so I could just use that instead, people started to pm me for it).

When I sent him a letter explaining fixed width was for tabloids and adolescents, not professional documentation, he wrote back explaining "a London Times study had proven 8-10 words per line on average is ideal". For who? Their semi-literate readers? Out to lunch.

And if you want 8-10 words per line, shrink your browser window...do the rest of us have to be penalized? Anyway, whew, thank god nothing that stupid happened over at W3C.

According to common readability formulas, it would take a 6th grade reading level to understand that post on average.

[mk27]

October 19th, 2009, 8:24 am

joebert wrote:

I'm a fan of short lines and short paragraphs, but 8-10 words-per-line is just ridiculously short.

Hey, there is nothing wrong with that. I honestly find 100-140+ characters per line the most convenient to read (it is less movement for your eyes, but whatever), which is like minimum 20-25 words, and I *universally* find fixed width pages with a "newspaper size" column really, really irritating. It is not a newspaper, it is a widescreen monitor. I'll leave a site for that unless I absolutely have to get the info (like I said, thank god I have a copy of the older perldoc, since perl5 won't be changing now perl6 is out).

Anyway, this is another thing which IMO should be strictly the user's preference -- you should not force theses stupid little columns down other people's throat because YOU or some loser consultant for the London Times thinks it's cute.

[digitalMedia]

October 19th, 2009, 9:11 am

You have an interesting perspective mk27. I find it to be in complete opposition to my own preferences and experience, but very interesting, nonetheless.

For me, when the text is the width of my wide screen, I tend to read the first 10 to 15 words and move on to the next line without reading the rest. It's just more comfortable for me. In fact, I find it quite exhausting to read across the full width of my screen.

I really don't think the vast majority of web developers and printers use columns because of a desire to be cute or because of a loser consultant. I think it has more to do with 500+ years of experience in putting words on a page.

This is just my own opinion, however.

I think there are other major contributing factors in readability like negative space, good hierarchical visual organization, type face, good writing practices, etc.

It would be interesting to read a quality study on comprehension and comfort for consumers of different media types.

[digitalMedia]

October 19th, 2009, 10:30 am

Well, I just got done reading several studies and opinion pieces regarding line length and there seems to be no consensus or solid conclusions to be had. In fact, much of what I read seems to contradict itself on several levels, or be contradicted elsewhere. I was floored with the vast number of facets that can be considered in these matters.

Some consistent results seem to be comprehension of screen media was dramatically less than that of printed media due to the consumers tendency to scan, rather than read.

I found some studies to be interesting because they reported statistical results that were at odds with subjective results(i.e. readers performed better with what they claimed to be less comfortable or desirable). Generally, I think most methodologies of the studies could be questioned and/or refuted.

So, I guess we're all left to our own preferences and opinions.

[mk27]

October 19th, 2009, 10:46 am

digitalMedia wrote:

For me, when the text is the width of my wide screen, I tend to read the first 10 to 15 words and move on to the next line without reading the rest. It's just more comfortable for me. In fact, I find it quite exhausting to read across the full width of my screen.

Yes, but you can NARROW THE BROWSER WINDOW IF YOU HAVE A PROBLEM. Like I said, it should be about providing the user with CHOICES, not DICTATING to them how they must do everything.

Quote:

I really don't think the vast majority of web developers and printers use columns because of a desire to be cute or because of a loser consultant. I think it has more to do with 500+ years of experience in putting words on a page.

Well, in fact I have a bookshelf right here (and a web browser). I think if you open a book you will notice there is usually about an 80+ character line. Not 40, which is most fixed width column pages. In any case, it is not in columns (except the bible maybe, or a magazine) SO, truth be told, the VAST MAJORITY of printers (and web developers, AFAICT) do not use columns, they fill the page. The ones that do are simply a significant minority.

Even if this were not the case, NASA did not fly to the moon by relying on doing things the same old way. You can sit on earth and just stare at it if that's your dictum.

But let freedom reign. (I just wish it were freedom for the user, is all...)

Quote:

I think there are other major contributing factors in readability like negative space, good hierarchical visual organization, type face, good writing practices, etc.

We can agree on that! Some fixed width pages are much better than others*. But look at the new perldoc:

http://perldoc.perl.org/

It's freaking terrible. Not only is it fixed width, but like half of that is stupid columns down the side with clutter-stuff that could have tucked away in a menu, etc. It looks like JUNK MAIL -- something I find on the foyer floor and throw in the trash.

I know this is a style some people like, I just find it remarkably ludicrous. Thank God W3C dodged a bullet here. There are columns down the side, but on my screen the center is still about 1000px, so I get a good 15+ words per line even with a 22pt font. Because I have a widescreen monitor on purpose, just like the other choices I make. When I'm allowed to make them for myself. If I seem a bit rabid, honestly, if someone asked me my top three pet peeves they's be:

1) War
2) Hunger
3) fixed width web pages

* and vis font face/size (I know, we've had this debate before) that is something the user can set according to their own preferences, according to what they feel is best for them. What on earth could be better than that? (Nothing, since it can be set to override )

[digitalMedia]

October 19th, 2009, 11:03 am

Quote:

Well, in fact I have a bookshelf right here (and a web browser). I think if you open a book you will notice there is usually about an 80+ character line. Not 40, which is most fixed width column pages. In any case, it is not in columns (except the bible maybe, or a magazine) SO, truth be told, the VAST MAJORITY of printers (and web developers, AFAICT) do not use columns, they fill the page. The ones that do are simply a significant minority.

Actually, I would say 40 characters per line is way too little for my own comfort unless I'm reading brief summaries that lead to full articles elsewhere. Otherwise, I prefer 80 to 110, I think. I don't feel comfortable going beyond that, though.

I would argue, also, that most periodic literature does use multi-column layout. Looking at several magazines I have here, I can't find one that uses single columns.

I also have an O'Reilly book on ActionScript 3 here which uses two columns (1 for the primary content and 1 for asides). The primary content column, where there are no asides, is about 3/5ths of the page width - about 80 characters per line in a 10pt Times.

I think there's also something to be said for familiarity. People who grew up with print as their primary medium are probably more likely to claim a preference to that as a similar structure where people who grew up with screen media are probably far more flexible.

Also, in reference to the Bible, I have both two column per page Bibles and one column per page Bibles. While I'd prefer to read Paul's epistles in a one column per page format (because of his propensity for run on sentences), Psalms are far more easily read in two column layouts.

Let's design our own study!!!

//Oh, on the perldoc thing, I find the color scheme more offensive than anything.

[digitalMedia]

October 19th, 2009, 12:07 pm

mk27 wrote:

SO, truth be told, the VAST MAJORITY of printers (and web developers, AFAICT) do not use columns, they fill the page. The ones that do are simply a significant minority.

Sorry man, this line keeps running through my head, and I just don't think it bears scrutiny.

Consider this page, if you will, http://www.mozilla.com/en-US/firefox/features/

...fixed width, multi-column. At no point is there a single column. Yet the page is perfectly readable. The only contention I have here is that the layout suits the information. There isn't a universal standard being applied here. The design is based on necessity.

Consider http://www.opera.com/

Again, fixed width, multi column. The six paragraphs on this page would look silly if they were to stretch across the full width of the screen/page.

http://www.ubuntu.com/ the same

The vast majority of web pages are fixed width and do have multiple columns - as far as I can tell. Again, I think it is, and should be, driven by content and not universal application.

http://www.opensource.org/ This is a great hybrid, btw.

//While it may be true that most fiction is printed on single column pages, instructional books, books heavy on graphics and insets and books where pages are wider than 6 to 7 inches don't. If they are wider than that and single column, there is a significant amount of negative space surrounding the column.

[mk27]

October 19th, 2009, 12:50 pm

digitalMedia wrote:

Actually, I would say 40 characters per line is way too little for my own comfort unless I'm reading brief summaries that lead to full articles elsewhere. Otherwise, I prefer 80 to 110, I think. I don't feel comfortable going beyond that, though.

Okay, but it is generally accepted (as best I can remember from doing my BA, in English and the Humanities) that a average (English) word is five characters. So before the days of word processors which do word counts, you could just count the pages if you had to write 2000 words. That means 80-110 should be 15-20 words per line, including a space between words. That I have no problem with, altho I do really, for myself, find that minimal -- I'd rather read a file with 140 char lines than one with 65, by a long shot. But, as I will shortly demonstrate, most fixed width pages actually have 50 characters per line or less (and you can go check that yourself).

Quote:

I would argue, also, that most periodic literature does use multi-column layout. Looking at several magazines I have here, I can't find one that uses single columns.

I agree. I actually don't like magazines much -- nothing in particular about them, I just almost never come across one that I find interesting enough to read for more than 5 minutes. I rarely even finish an article (and I rarely start one, either). But I totally enjoy books and always have. I think most of that 500 year history is about books. The "developments" that occurred with printing in the 20th century I am sure have to do with making room for ADVERTISING. I'm pretty neutral about advertisizing, I don't quite consider it the work of the devil or anything, but I think most people, given a choice, would prefer to avoid it. You tolerate it in magazines and web sites because you know that it helps to pay for it, and maybe that is even a good thing.

But designers who have been immersed in this superficial and semi-literate world for long enough I think seem occasionally to have this "zombie" response where they just automatically must include columns of clutter even if it is not paid ads. Now, I have ADHD, but even I find this silly and condescending. Or maybe because I have ADHD??

Quote:

I also have an O'Reilly book on ActionScript 3 here which uses two columns (1 for the primary content and 1 for asides). The primary content column, where there are no asides, is about 3/5ths of the page width - about 80 characters per line in a 10pt Times.

I actually have Actionscript Cookbook and Essential Actionscript 2.0 here by O'Reilly (both from the library; I haven't really read them much yet since I don't have as much time as I was hoping for flash). A normal O'Reilly page is 80 characters right across.

Now, I know part of my personal pet peeve here is because of the 22px font. I know I said 22pt before, but I make it a habit to not set font sizes (beyond <h>) so I don't pay attention to them and after doing a little digging realized "22" in FF refers to pixels. Which is quite a bit smaller than 22pts. So going with a 1000px width, we should get 50 characters per line. That's without sidebars that are 250 px wide each. Even with a 10px font (which is going to be smaller than average) a 500-600px column is still only 50-60 characters. That is 25-30% less than 80. At 22px, it is just stupid.

So this is my point: LEAVE THE USER THE FREEDOM TO CHOOSE. I know everybody is enamoured with their own creativity, but I really think an obsession with font sizes and page width -- whereby you must set these exactly in order to get your page "just right" -- is a sure sign this creativity has been coerced in the wrong direction.

I do not do much professional web work, but I know that clients are an issue here. I have had someone I was doing work for point me to sites he liked and they were all fixed width, and when I showed him my idea he said, literally, "Well, it does not look like a magazine like these other sites; there should be space down the sides". Pushing my point a bit, I wrote him an explanation about not constraining the user interface (which is not a magazine) and about distributing whitespace evenly, rather than putting it down the side and squishing all the content into the middle, and (personal opinion) that the reason fixed width sites exist is because they are easier to code. Then I said if he really insisted, I would do that, but I thought it was a bad idea. The next day he agreed. Of course, not everyone is like that and I am sure clients are a sad factor here.

Quote:

Let's design our own study!!!

It probably would be a good poll. However, it's not my intent to win a popularity contest vis. how many words you like in your line, because that implies that this is an issue which everyone should vote on and let the majority decide for everyone else. Even if I were (maybe I am!) in the majority on this, I would not want to make it a rule for everyone else. Which, like I've been trying to emphasize, the user (often can) make a choice buying a monitor, and certainly they have a choice manipulating the browser window and setting the preferences. That is the nature of the beast. It just seems backward and ignorant to me to decide to pretend these things are not real and that you must present an 8 1/2" x 11" flyer, because of this dufus at the London Times or where-ever.

Quote:

Oh, on the perldoc thing, I find the color scheme more offensive than anything.

Yeah, there were some things said about that at perlmonks too I almost felt sorry for the guy until I exchanged a series of emails with him (kudos, at least he was interested in my criticism) and it became clear he was sticking to his guns, etc, in which case you get what you deserve if people repeatedly say look, honestly you have made a mistake and screwed up here. Get over it.

This is the old perldoc; FF is 1680px wide with a minimum 22px font. Works great! Looks great! Now browsing over to the new one, the word devolution springs to mind. I understand if the guy wants to do comic books instead -- nothing wrong with that -- but really (as much as I might sometimes like it if it were) perl documentation is not a comic book, or Newsweek. It is technical documentation...

Attachments:

[digitalMedia]

October 19th, 2009, 12:55 pm

Well, I'm not sure we've achieved any conclusion. But, I do think this is a fascinating converstion. Thanks!

[mk27]

October 19th, 2009, 1:12 pm

Hmmm.

digitalMedia wrote:

Sorry man, this line keeps running through my head, and I just don't think it bears scrutiny.

Consider this page, if you will,

It might even be "most"* but I still don't think that's a justification. "Most" people will help to destroy the planet and deny it like angels, etc. Either it will stay that way forever, or else things will change as people start to recognize there could be a better idea. That's people, I guess. Kind of like, oh, I don't know -- haircuts from the eighties.

WRT the sites, the firefox one is loathsome -- it looks sloppy, for starters. Those columns have like four words per line. They are readable, sure, but I'd still rather see that information in a single column with 10-20 words on a line.

Also, that stuff is just blurbs. The Opera site is much much better IMO. Both that and he Ubuntu site, if you actually click to read an article, the article is presented as a single column replacing two or three of the others. A front/splash page is it's own thing -- in this case, these are all essentially product advertisements. Clearly a goal with advertising is to throw as much crap as you can in someone's face using a minimum of SPACE. Altho that should not be so true on the web, I think the mentality remains.

I do not have respect for something just because it was done by Coca-Cola, BTW.

The opensource.org one is quite nice. Now that is my idea of a well done fixed width page. For starters, it is 1200px -- but I am still positive it would look better as variable width, so that the page always maintains the same relationship to the window as it does when viewed at or slightly over it's width, ie, with little blank down the sides. This can be a problem if the window is like 600px wide, but down to that point everything will work out fine. And no one has a 600px wide display.

I will set sidebars fixed width, since you don't want unnecessary text wrapping there, etc. But if you have two 250px sidebars, you will still have 300px left in the middle on an 800px display (which I don't think people have anything that small, either). So it is more like 500px minumum, and if the content is text (that's what we're talking about), it's fine. It looks better, it works better, it is more "dynamic". IMO. It may be slightly trickier for the developer but I think not so much...

* certainly it is not most sites that I actually spend time at, which are usually forums, (and fixed width forums REALLY suck) or documentation, which most of that is not fixed width either. Eg, W3C

[joebert]

October 19th, 2009, 8:03 pm

What we need is a simple way to determine programatically whether an index page is fixed or fluid width.

Here's the most visited sites according to Alexa and what format they're in.

Google: Fixed width, center aligned (SERPS, left aligned);
Facebook: Fixed width, center aligned;
Yahoo: Fixed width, center aligned;
YouTube: Fixed width, center aligned;
Bing: Fixed width, center aligned (SERPS, left aligned);
Wikipedia: Fluid width;
Blogger: Variable, multiple sites;
MSN: Fixed width, center aligned;
Baidu: Fixed width, center aligned (SERPS fixed width, left aligned);
Myspace: Fixed width, center aligned;
Twitter: Fixed width, center aligned;
Microsoft: Fixed width, center aligned;
QQ: Fixed width, center aligned;
Wordpress: Main site fixed width, center aligned, variable otherwise, multiple sites;
Rapidshare: Fixed width, center aligned;

I'm seeing an overwhelming majority of fixed width sites being on top of the popularity chart here.

[digitalMedia]

October 20th, 2009, 4:28 am

If I was going to design a study, my main considerations would be comprehension and willingness to consume information. Speed would be interesting, but it wouldn't be important to me where comprehension and willingness were high.

I would want to look at line length, line height (leading), proper punctuation and grammar, font faces, alignments, interstitial formatting (bold, italic, color), overall formatting (fore and background color), overall page size, negative space surrounding a body of text.

I would also want to look at the effects of columns in different ways; When a body of text occurs across columns, when a body of text is juxtaposed to columns of related and non-related content.

I would want test subjects in controlled and non-controlled environments, looking at screen media and paper media. I would want to evaluate performance and subjective response.

I'm not really interested in a popularity contest, but in some conclusive results - again related to comprehension and willingness to consume information.

I wonder if we could get a grant.