Comments on: @Font-Face Worsens In Opera 10.5 Beta, But Honestly, It’s OK

By: James John Malcolm

James John Malcolm — Mon, 01 Mar 2010 10:58:19 +0000

Thanks – I’ve got more time in a few weeks, so that’s fine. Synthesized text is the biggy for me as it represents how well fallback will work. (I’m thinking about font-weight:number; it’s a natural extension, but probably of limited use.) What other @font-face support would you like to have tested?

Mail address should be in your cms, attached to the comment.

By: Richard Fink

Richard Fink — Sat, 27 Feb 2010 23:51:29 +0000

James,
Thanks. I’ll catch up with your work ASAP. Might be a few weeks though, realistically.
However, I’ll try to get you any IE screenshots I can, sooner.
BTW – do you have any free time to spare for additional testing outside the sphere of synthesized text? Do you intend to expand your investigations of @font-face support?
Help is always needed. And input on related typographical stuff.
Let me know.
(Don’t think I have an email address for you, either.)

Rich

By: James John Malcolm

James John Malcolm — Thu, 25 Feb 2010 22:35:31 +0000

Hi Richard,
Good to hear about IE synthesizing fonts properly. IE not being included in my results is primarily due to my lack of time and the limited availability of IE on my machine (convert to eot, boot into windows, run IE tester, etc.).

The methodology is simple. The test case is made up of 3x the same text. Nr. 1 has a font file assigned to each member in the family (myfont-bold.ttf to font-weight:bold), Nr. 2 has a font file assigned to the regular member of the family and Nr. 3 has simply Georgia (&Serif) defined because it’s so different from Fontin Sans. The fallback font for Nr. 1 and 2 is also Georgia to make clear when fallback is happening.

The optimal results should show nr. 1, 2 and 3 all matching each other in terms of text styles (what’s bold italic in one should be bold italic in the other). Nr. 1 and 2 should match each other quite closely, although small differences are allowed as the browsers have to guesstimate all the other weights and styles.

The resulting support table is just a summary of all the screenshots. If you have a screenshot of IEs rendering, it’d be very welcome.

By: Richard Fink

Richard Fink — Thu, 25 Feb 2010 15:15:56 +0000

Hi James,
I came upon your post and test pages awhile ago with a promise to myself to double-back and analyze because I was having a little trouble following the methodology. Probably my fault, not yours. I will take a close look though, and repro using EOTs and check your results. I can tell you that with a four member font family, IE synthesizes Bold, Italic, and Bold Italic, exactly as you would expect – from the regular/normal font of the family, not from a fallback font.
I also know from applying @font-face to this blog that it will synthesize “Small Caps”. (For example, the heading “Recent Posts” uses font variant and IE handles it OK. Chrome, for example doesn’t.)
I’d be interested to know why your tests didn’t include IE – what were the barriers to testing? Was it file conversion?

By: James John Malcolm

James John Malcolm — Thu, 25 Feb 2010 10:05:28 +0000

I’ve run into the same problems with Opera 10.x as you have (funnily enough while creating a design for an opera-sponsored event). In my testing Opera 10.5b (on mac) has improved slightly over 10.5a (and certainly over 10.10). I’m miffed they’ve added ligature support before proper family support.

I’d love to hear your comments on my browser support table: http://james.gameover.com/index.php/2010/css3-font-face-browser-support-table/
I’d be especially interested in extending it with IE’s results.

By: dan bloom

dan bloom — Tue, 16 Feb 2010 01:11:05 +0000

click on my name to see LINK

By: dan bloom

dan bloom — Tue, 16 Feb 2010 01:10:40 +0000

Richard

Hear ye, hear ye, Juneau Empire prints my oped today on snailpapers, your reax and POV, pro or con?

Danny