[css-d] Alternative font sizing
Folks, let me propose a scenario to you and get your ideas on how useful/useless you would find it. As you know CSS 2 allows absolute and relative font sizing. Of course relative refers to the font size of the parent element, but I've often found myself pining for the ability to use a ratio relative to the size of the parent element itself rather than the parent's font table. This is primarily because while I love the concept of liquid designs, such layouts often fail in terms of usability when long-ish text blocks run longer than the print-standard of 50-70 characters per line. A fixed width design is significantly weaked by high-res displays, which makes a forced standard line length too often too small. An em/ex-based design width is OK, but requires the user to adjust the type size manually. If one accepts this as a legitimate problem, it seems to me the most obvious solution would be to provide a method of basing the current em space not on the parent element's em space, but on a percentage of the parent element's width. If dynamic, this would change the font size based on the width of the element particular to each user, but would still allow for the user to override the page's display with their own +/- adjustments. My colleague and I have been playing with this concept, and implementation is possible and pretty straight-forward with a little Javascript, but I wonder if such stuff would be of interest to anyone else? Jared Stein Director of Learning Media Development Distance Education, Utah Valley State College, MS 149 http://www.uvsc.edu/disted/ phone: (801) 863 8929 office: Learning Center 221d __ css-discuss [EMAIL PROTECTED] http://www.css-discuss.org/mailman/listinfo/css-d IE7b2 testing hub -- http://css-discuss.incutio.com/?page=IE7 List wiki/FAQ -- http://css-discuss.incutio.com/ Supported by evolt.org -- http://www.evolt.org/help_support_evolt/
Re: [css-d] Alternative font sizing
Jared Stein wrote: Folks, let me propose a scenario to you and get your ideas on how useful/useless you would find it. As you know CSS 2 allows absolute and relative font sizing. Of course relative refers to the font size of the parent element, but I've often found myself pining for the ability to use a ratio relative to the size of the parent element itself rather than the parent's font table. This is primarily because while I love the concept of liquid designs, such layouts often fail in terms of usability when long-ish text blocks run longer than the print-standard of 50-70 characters per line. A fixed width design is significantly weaked by high-res displays, which makes a forced standard line length too often too small. An em/ex-based design width is OK, but requires the user to adjust the type size manually. If one accepts this as a legitimate problem, I don't, but I won't get into that right now - see below. it seems to me the most obvious solution would be to provide a method of basing the current em space not on the parent element's em space, but on a percentage of the parent element's width. If dynamic, this would change the font size based on the width of the element particular to each user, but would still allow for the user to override the page's display with their own +/- adjustments. My colleague and I have been playing with this concept, and implementation is possible and pretty straight-forward with a little Javascript, but I wonder if such stuff would be of interest to anyone else? Personally, I would *not* like it like that. It would mean that when I visit a site and find the line-length too short or too long, I'd widen or narrow my window, only to have zero result, plus seeing my font-size get very large or small. I'd then have to change my font-size in the browser to what I like, then most likely still want to adjust the window width some more, meaning having to adjust the font-size again, and so on. Next I'd surf to a different site, which has a different page setup, and different widths of boxes in it, and I'd have to start over again. Also - have you tried how that system works on a page with a fixed width sidebar and a flexible width content div? AIUI, narrowing the window would make the font-size of the content div smaller, while the font-size of the sidebar stayed intact, meaning that when I adjust my font-size in the browser, the content div will display my preferred font-size, while the sidebar content went too large. Nah, really, can't see any benefit at all of this idea... -- Els http://locusmeus.com/ http://locusoptimus.com/ __ css-discuss [EMAIL PROTECTED] http://www.css-discuss.org/mailman/listinfo/css-d IE7b2 testing hub -- http://css-discuss.incutio.com/?page=IE7 List wiki/FAQ -- http://css-discuss.incutio.com/ Supported by evolt.org -- http://www.evolt.org/help_support_evolt/
Re: [css-d] Alternative font sizing
Folks, let me propose a scenario to you and get your ideas on how useful/useless you would find it. I browse with a 3200x1200 twin monitor setup. If I ever expand the browser window to anywhere near full screen size on one of the monitors, it'd be because there's content clipped at a smaller size and I want to see it all: the whole of a table, a large image and so on. I can't imagine that anyone with any sense regularly uses a full- screen browser window at more than 1280; the whole point (for as many people as I've ever talked to about it) of a large screen for browsing purposes is to have multiple apps or windows visible at the same time, not to have a single window open at full screen size. If the content of the window - or the text within the site - were to zoom with the size of the browser window, it would either be infuriating (browsing with an effective body text size of 16px - 18px - 20px), or impractical and going against text size settings I've chosen manually, as an end user. I want to retain control over the size of text on my screen, not have a remote content manager or designer decide for me and have the text enlarge on me without choosing a larger size myself. You've stated... such layouts often fail in terms of usability when long-ish text blocks run longer than the print-standard of 50-70 characters per line If that's the case, I'd suggest that a flexible layout is less suited to the site or content in question than an alternative. Liquid layouts, with flexible main content columns, can be difficult to read and follow once they exceed, say, 500px (or 70 - 80 chars). I and many people who have tested layouts for us - online and off - find fixed-column layouts with respectably proportional columns much easier to recognize and read, than those with average amounts of text and flexible or full-width main content columns. HTH Regards Mark Howells www.permanenttourist.ch __ css-discuss [EMAIL PROTECTED] http://www.css-discuss.org/mailman/listinfo/css-d IE7b2 testing hub -- http://css-discuss.incutio.com/?page=IE7 List wiki/FAQ -- http://css-discuss.incutio.com/ Supported by evolt.org -- http://www.evolt.org/help_support_evolt/
Re: [css-d] Alternative font sizing
Jared Stein wrote: Folks, let me propose a scenario to you and get your ideas on how useful/useless you would find it. ... My colleague and I have been playing with this concept, and implementation is possible and pretty straight-forward with a little Javascript, but I wonder if such stuff would be of interest to anyone else? I don't see the problem since we have more alternatives than the ones you presented. a: Fluid combined with min/max... http://www.gunlaug.no/contents/molly_1_01.html ...works in most cases. b: Elastic... http://www.456bereastreet.com/archive/200504/fixed_or_fluid_width_elastic/ ...combines regular em-based width with max-width fixed to browser-window. However, if you are creating a working solution anyway, then I wouldn't mind seeing it live so I can evaluate it. One can never have too many options. regards Georg -- http://www.gunlaug.no __ css-discuss [EMAIL PROTECTED] http://www.css-discuss.org/mailman/listinfo/css-d IE7b2 testing hub -- http://css-discuss.incutio.com/?page=IE7 List wiki/FAQ -- http://css-discuss.incutio.com/ Supported by evolt.org -- http://www.evolt.org/help_support_evolt/
Re: [css-d] Alternative font sizing
Good points. One problem I still see, however, is that if you use a fixed width for text at say, 500px, then on your high res setup the text is impossibly small, and enlarging the text manually creates too-large text in too small an area. Not necessarily. Using a commonly used example, 10px Geneva/Verdana is easily legible to people with reasonable vision when using 1600px on a suitable monitor. Take a look at http://www.jungfrau-zeitung.ch/ artikel/?cq_*5a5b87e9=ivxPU=87988m0d where the main content column is about 450px (I think!). That's easily legible for me on a 20 monitor at 1600x1200; if I want to be daft and set 1600 x 1200 on a 17 monitor - which is a good way to get eye strain anyway :-) - then I'll increase the text size in the browser. The layout here is flexible enough and planned to accommodate zooming up to about 250 - 300% before the layout starts to break, which is, I would imagine, plenty for people who have poorer eyesight. The point is, if I as a user have a non-standard setup, then I should be able to choose to increase the text size myself if I want to, rather than have a larger text forced on me. Regards Mark Howells www.permanenttourist.ch __ css-discuss [EMAIL PROTECTED] http://www.css-discuss.org/mailman/listinfo/css-d IE7b2 testing hub -- http://css-discuss.incutio.com/?page=IE7 List wiki/FAQ -- http://css-discuss.incutio.com/ Supported by evolt.org -- http://www.evolt.org/help_support_evolt/
Re: [css-d] Alternative font sizing
Hmm, maybe it's just me getting old and wanting my text bigger! Still one lingering thought: as user displays increasingly become higher resolution, there certainly will be a point--and I imagine not too far down the road--where what is currently readable by most people with the standard layout and font size will not be easily legible to a broad base with 'default' high-res displays. I think your 16x12 example is a good one, but really what would be wrong with setting it thus on a 17 monitor if the page display was more suitable to high-res? I'd question the logic of that. You have to have really good eyesight to be able to use any standard OS at 1600 x 1200 on a 17 display. The point is that as a user, you should be able to increase the text size if YOU want to, not as the designer forces you to. My point wasn't about cramming a big resolution into a small window space, more that most people I know who use 16x12 are working with big monitors and normal (c. 1024 - 1200) browser window widths. If you're using a small monitor, cranking it up to 1600x1200 and then having to use a full-width browser window whilst zooming everything to be able to read it, why not just use 1280 or 1024? If displays do increase in resolution, and remain at 17 or 20, then that'll mean that the text will be pretty illegible and 72 dpi web images pretty tiny if the resolution goes above 1920! Getting back onto the CSS tracks (here on the CSS list!), you'll avoid all kinds of pointless, expensive and time-consuming development by defining text sizes in such a way as to allow for client-side resizing, rather than a nifty technical solution which I, for one, would find annoying as a user. I may be Superman and be able to read 10px Geneva on a 15 monitor at 1600x1200, but if not, then I'll increase the text size if I need to. I'd respectfully say that you're digging yourself into a hole which you really don't need! Regards Mark Howells www.permanenttourist.ch __ css-discuss [EMAIL PROTECTED] http://www.css-discuss.org/mailman/listinfo/css-d IE7b2 testing hub -- http://css-discuss.incutio.com/?page=IE7 List wiki/FAQ -- http://css-discuss.incutio.com/ Supported by evolt.org -- http://www.evolt.org/help_support_evolt/
Re: [css-d] Alternative font sizing
Yes, I see your point that this is the case for most people with standard OSs and browsers and resolutions, but wouldn't we like to move beyond 72dpi eventually? My own point of view (and this is where my 'logic' comes in) is that higher-res is always better if the software supports it. For example I have a little 14 laptop screen set to 1400x1050 (or higher depending on the project); I personally set Win to show large fonts so everything's readable, but it seems OSs of the future would better detect the res/screen and adjust the base font size automatically--even if it's just to set a baseline for the default or 'normal' setting. Oh, and the javascript for this took all of 20 minutes to write and test, so it's really not a big investment of time, and I don't see it as a hole. Currently the js has font-size dynamically increase/decrease as the window is resized, but it could calculate the window size on load and set the font-size once, with no font-size +/- as the window resizes--really this is no different than designers setting a base font-size for any element off the bat, except that in this case the base font-size is not some arbitrary 12pt/16px/0.9em or whatever, it's calculated based on the user's initial window size. Anyway, a few have commented that this is a bit off-topic, so I'll invite any further points/counterpoints offlist! Jared Stein Director of Learning Media Development Distance Education, Utah Valley State College, MS 149 http://www.uvsc.edu/disted/ phone: (801) 863 8929 office: Learning Center 221d Mark Howells [EMAIL PROTECTED] 3/7/06 10:38 AM Hmm, maybe it's just me getting old and wanting my text bigger! Still one lingering thought: as user displays increasingly become higher resolution, there certainly will be a point--and I imagine not too far down the road--where what is currently readable by most people with the standard layout and font size will not be easily legible to a broad base with 'default' high-res displays. I think your 16x12 example is a good one, but really what would be wrong with setting it thus on a 17 monitor if the page display was more suitable to high-res? I'd question the logic of that. You have to have really good eyesight to be able to use any standard OS at 1600 x 1200 on a 17 display. The point is that as a user, you should be able to increase the text size if YOU want to, not as the designer forces you to. My point wasn't about cramming a big resolution into a small window space, more that most people I know who use 16x12 are working with big monitors and normal (c. 1024 - 1200) browser window widths. If you're using a small monitor, cranking it up to 1600x1200 and then having to use a full-width browser window whilst zooming everything to be able to read it, why not just use 1280 or 1024? If displays do increase in resolution, and remain at 17 or 20, then that'll mean that the text will be pretty illegible and 72 dpi web images pretty tiny if the resolution goes above 1920! Getting back onto the CSS tracks (here on the CSS list!), you'll avoid all kinds of pointless, expensive and time-consuming development by defining text sizes in such a way as to allow for client-side resizing, rather than a nifty technical solution which I, for one, would find annoying as a user. I may be Superman and be able to read 10px Geneva on a 15 monitor at 1600x1200, but if not, then I'll increase the text size if I need to. I'd respectfully say that you're digging yourself into a hole which you really don't need! Regards Mark Howells www.permanenttourist.ch __ css-discuss [EMAIL PROTECTED] http://www.css-discuss.org/mailman/listinfo/css-d IE7b2 testing hub -- http://css-discuss.incutio.com/?page=IE7 List wiki/FAQ -- http://css-discuss.incutio.com/ Supported by evolt.org -- http://www.evolt.org/help_support_evolt/
Re: [css-d] Alternative font sizing
On 06/03/07 11:49 Mark Howells apparently typed: I can't imagine that anyone with any sense regularly uses a full- screen browser window at more than 1280; the whole point (for as many people as I've ever talked to about it) of a large screen for browsing purposes is to have multiple apps or windows visible at the same time, not to have a single window open at full screen size. This user's 24/7 browser window is 1160x938, on a 1280 wide screen (and more than 1280 wide on his 1792 wide screen). I reset it that way with Web Developer anytime I find it necessary to temporarily change it to anything else. That's 90.6% of my screen width, and 100% of my available screen height. Not everyone uses a large screen in order to have many non-overlapping windows open simultaneously. Some people, notably people over-40 and low-vision users generally, need things bigger just to see, and often do it precisely as an alternative to printed matter that can't be had with print big enough to read without straining, or at all. Fullscreen isn't necessarily a bad idea even on a large display with high resolution: http://mrmazda.no-ip.com/SS/accessibility.jpg This doesn't seem to be on-topic though. -- Blessed are they whose ways are blameless, who walk according to the law of the Lord.Psalm 119:11 NIV Team OS/2 ** Reg. Linux User #211409 Felix Miata *** http://mrmazda.no-ip.com/auth/auth __ css-discuss [EMAIL PROTECTED] http://www.css-discuss.org/mailman/listinfo/css-d IE7b2 testing hub -- http://css-discuss.incutio.com/?page=IE7 List wiki/FAQ -- http://css-discuss.incutio.com/ Supported by evolt.org -- http://www.evolt.org/help_support_evolt/
Re: [css-d] Alternative font sizing
On 06/03/07 12:13 Mark Howells apparently typed: Using a commonly used example, 10px Geneva/Verdana is easily legible to people with reasonable vision when using 1600px Just how do you define reasonable vision? I have no problem comfortably reading 16px on 17 at 1024x768, 20px on 19 at 1280x960, 22px on 19 at 1400x1050 or 28px on 19 at 1792x1344, as they're all near exactly the same physical size, 12pt (nominal), which happens to be the size most web users prefer [1]. OTOH, 10px on an the last of those isn't big enough for me to distinguish from a straight line, and much too small for comfort on the first two. on a suitable monitor. Take a look at http://www.jungfrau-zeitung.ch/ artikel/?cq_*5a5b87e9=ivxPU=87988m0d where the main content column is about 450px (I think!). That's easily legible for me on a 20 monitor at 1600x1200 For me to use that page as I find it I estimate I would need at least a 26 display for 1400x1050 resolution. Here's 1400x1050 on the lowly 19 that suits my budget: http://mrmazda.no-ip.com/SS/markho1.jpg http://mrmazda.no-ip.com/tmp/showcasec.html (needs JS off to prevent redirect) This doesn't seem to be on-topic though. [1] http://psychology.wichita.edu/optimalweb/text.htm -- Blessed are they whose ways are blameless, who walk according to the law of the Lord.Psalm 119:11 NIV Team OS/2 ** Reg. Linux User #211409 Felix Miata *** http://mrmazda.no-ip.com/auth/auth __ css-discuss [EMAIL PROTECTED] http://www.css-discuss.org/mailman/listinfo/css-d IE7b2 testing hub -- http://css-discuss.incutio.com/?page=IE7 List wiki/FAQ -- http://css-discuss.incutio.com/ Supported by evolt.org -- http://www.evolt.org/help_support_evolt/
Re: [css-d] Alternative font sizing
On 06/03/07 13:00 Jared Stein apparently typed: it seems OSs of the future would better detect the res/screen and adjust the base font size automatically--even if it's just to set a baseline for the default or 'normal' setting. Many Linux systems have already been doing that for quite some time, probably around three years. It simply requires a competent (unbroken) version of X, a proper DDC display, and good video drivers, the latter of which is unfortunately somewhat uncommon if the video hardware is new. X can use the display's DDC output to set DPI accurately, and with it all font sizes that are set in points are accurately sized. Today that means only somewhat short of everything on the X desktop. The main exceptions are Gecko browser defaults (set in px), and web pages that explicitly size things in other than such absolute units as pt, mm, cm, and in. -- Blessed are they whose ways are blameless, who walk according to the law of the Lord.Psalm 119:11 NIV Team OS/2 ** Reg. Linux User #211409 Felix Miata *** http://mrmazda.no-ip.com/auth/auth __ css-discuss [EMAIL PROTECTED] http://www.css-discuss.org/mailman/listinfo/css-d IE7b2 testing hub -- http://css-discuss.incutio.com/?page=IE7 List wiki/FAQ -- http://css-discuss.incutio.com/ Supported by evolt.org -- http://www.evolt.org/help_support_evolt/
