CHI-WEB Archives

ACM SIGCHI WWW Human Factors (Open Discussion)

CHI-WEB@LISTSERV.ACM.ORG

Options: Use Forum View

Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
Joe Clark <[log in to unmask]>
Reply To:
Joe Clark <[log in to unmask]>
Date:
Thu, 15 Dec 2005 14:18:24 -0500
Content-Type:
text/plain
Parts/Attachments:
text/plain (145 lines)
>From:    Jake McMurchie <[log in to unmask]>
>
>>Date:    Wed, 7 Dec 2005 16:57:13 -0500
>>From:    Joe Clark <[log in to unmask]>
>>
>>Please do not specify Arial. Specify Helvetica if you must, 
>>particularly for display type.
>
>So what do you recommend given that the majority of computer users, 
>especially on the web, are stuck with a handful of ubiquitous 
>windows fonts?

Use those. Verdana/Georgia/Trebuchet were at least designed and 
tested for onscreen use. You've got tons of other options, too.


>From:    Christopher Fahey <[log in to unmask]>
>
>"Helvetica versus Arial" discussions are usually pointless and 
>religious in nature.

Only when carried out by the ill-informed. The design differences 
between Helvetica and Arial have been documented by experts. It isn't 
a matter of preference, which is another way of dismissing the design 
features of typefaces as mere decoration.

>HTML/CSS permits us to specify multiple fonts. If you prefer 
>Helvetica, go right ahead and use it and specify Arial as the 
>*second*, fallback font. This is HTML 101, folks.

It's CSS 101, actually.

>There is no big taboo about using fonts that most people don't have 
>because you can always specify a universal backup font.

Which, in this case, would be font-family: [list of fonts], sans-serif;.

>One typography-geek nuance to be aware of, however, is that for font 
>sizes below around 14 pixels or so, Arial has 'pixel fonts' designed 
>specifically to make it look good on screen at small sizes, 
>rendering them without anti-aliasing.

Helvetica also has bitmaps, which is what "'pixel fonts'" are 
actually called in this context. The difference is that Arial's are 
newer and are handmade. I still remember the original LaserWriter 
fonts and how appalling the Helvetica bitmaps were; *standard* 
shipments of Helvetica do not have hugely-improved bitmaps even 20 
years later. However, if you shell out the money and buy Helvetica 
Neue or suchlike, the bitmaps are considerably improved.

In addition, small-type rendering onscreen is seriously affected by 
antialiasing. (The rest of Christopher's discussion of that issue, 
not excerpted here, is incorrect.) Some people turn it off altogether 
below a threshold. Then you get into CRT vs. LCD distinctions, 
foreground and background colours, and the influence of visual 
impairment. While it is difficult to make categorical statements 
about typefaces' performance characteristics, it is much less 
difficult to do so when the viewing conditions are specified. In 
those cases you can actually *test* font performance.


>From:    Jim Griesemer <[log in to unmask]>
>
>Helvetica vs. Arial -- This is the second time I have heard a 
>preference of Graphic Designers referred to as a "religious 
>argument" (the first time was from an Engineer). This strikes me, 
>not only as disrespectful to religion, but also disrespectful to 
>Graphic Designers and Graphic Design as a discipline. I submit, if 
>Graphics Designers state that a font is preferable, there are good 
>reasons for that preference, especially if it is acknowledged to be 
>from the vast majority.

Indeed. It's called actual knowledge backed up by experience. If 
critics are tempted to dismiss that knowledge because it is mere 
"opinion," they should be consistent and apply that standard to 
usability testing, too, where in some cases *all* the data is 
"opinion."

>If the Usability *and* Engineering Communities want to encourage 
>Graphic Designers to join in with the ethos of Usability (and I 
>believe we should join), contributions from Graphic Designers must 
>be received with respect for their years of experience in the 
>discipl[in]e, and not be dismissed as pointless. Graphic Designers 
>address the vagueries of continuity, visual connotation, and general 
>"look and feel" -- all important issues to business, and I might 
>add, the users. These may be hard to quantify,

Actually, they increasingly are not. Not only is it possible to 
demonstrate to a crowd of doubters that typefaces have performance 
characteristics (I've done it), it is possible to test those 
characteristics. And graphic designers must be open to such testing. 
It isn't *all* "opinion."



>From:    Christopher Fahey <[log in to unmask]>
>
>I called it a "religious argument" because neither side will ever be 
>able to 'prove' the other wrong

Yes, we can, given a specification of the viewing conditions. Then 
you additionally reach past the quantitative and simply *trust our 
word for it* because we're experts. Isn't that the same thing you 
tell your clients about usability? Why else are they asking for your 
opinion? You know more than they do. And the same applies here.

>It in no way demeans the field of graphic design to point out that 
>many design debates rest, ultimately, on unquantifiable judgement 
>calls by educated practicioners

Um, no. To continue with the previous theme, not only is it possible 
to test performance of individual fonts, it is also possible to test 
the effects of entire *layouts*. Ask Kevin Larson of Microsoft what 
he's been up to in that regard.

>Many graphic designers prefer Arial, by the way.

When you say "many," do they all use Windows, do they have 
graphic-design degrees, are they (in Ontario) registered graphic 
designers, and in any event could they all fit into a London black 
cab?


>From:    Jim Griesemer <[log in to unmask]>
>
>Of course, as you pointed out, this is all strictly unquantifiable.

He may have pointed that out, but he was incorrect. "This" not only 
is quantifiable in many ways but the number of *ways* in which "this" 
is quantifiable inreases each year.


-- 

     Joe Clark | [log in to unmask]
     Accessibility <http://joeclark.org/access/>
     Expect criticism if you top-post

    --------------------------------------------------------------
           Tip of the Day: Postings must be in plain text
     CHI-WEB: www.sigchi.org/web POSTINGS: mailto:[log in to unmask]
              MODERATORS: mailto:[log in to unmask]
       SUBSCRIPTION CHANGES & FAQ:  www.sigchi.org/web/faq.html
    --------------------------------------------------------------

ATOM RSS1 RSS2