Thanks Chaals,
I knew you were involved in SVG accessibility from the early days, but August 2000? My goodness! A brief reread informs me that another is needed. James Deering recently e-mailed me a font for these five glyphs. He says it took only “a couple of hours which was a nice break from what I am working on “. He used Fontographer and Font Studio. Take a look at http://cs.sru.edu/~ddailey/svg/HTML5logo5.svg I would claim that the new thing is preferable to the old thing (http://www.w3.org/html/logo/ ) in about 627 different ways (and 627 is not even a prime number!). In fairness to James, though, I suspect he would appreciate me reminding that we have no idea quite what strange material one may have embedded in that WOFF font – another reason to prefer SVG fonts! Most importantly, the HTML and the 5, in the revision, are text rather than polygons. This means that current search engines (rather than the wondrous search engines of the future that will parse pictures as fluently as they parse text) can understand what the 8 polygons actually are (one each for “H”, “T”, “M” and “L” and four for the segments of the “avenue” of the 5 (it is really a path walking around the avenue than an avenue in typical SVG). It is about 1/20th of the file size (just remembering the original version posted by the W3C before it was “cleansed”). And even without <title> and <desc> it is accessible to screen readers and search engines. Here are a couple of concerns though, about my proposed “improvement”: 1. SVG fonts would be better since then a screen reader would have access to the geometry and could tell the viewer that the H is, in fact, a rectangle with two symmetric rectangles removed from the axis bisecting its horizontal midline That make sense to some people, since it conveys the physics of how the H would tumble if it were to tumble. 2. The “5” character is , in fact, a five, rather than four polygons stitched roughly together at the edge where we expect the left and right halves of the calligraphy to meet. (In the original HTML5 logo, they did not actually align because of rounding errors and/or the artist’s hand tremor!) 3. The left half of the “5” and the right half of the “5” are colored differently, not because they are separate objects filled with different colors, but because the shading of the 5 changes at midpoint. 4. The browsers that understand WOFF do a “pretty good job” of agreement about interpreting HTML5logo5.svg. This fact can seen by looking at http://cs.sru.edu/~ddailey/svg/HTML5logo4.svg . In that file, the W3C given polygons are overlaid with the HTML5 woff font. The latter has been displayed with a green stroke so we can compare the degree of agreement across browsers. You’ll notice that to get “perfect” alignment, I had to rely on the <text> “kerning” attribute that not all browsers support. Hence the HTML chars migrate a wee bit rightward in some cases. 5. When it comes to logos (and flags), corporations (and nations) have been known to get fussy at times, perhaps reflecting the personalities of their communications officers! 6. The issue is even more complicated, since, in reality, don’t we wish the “5” to inherit its shading from the differential colors of the shield behind it? And don’t we really wish the “5” to inherit the geometry of the “badge/shield” that frames it? 7. Browsers don’t agree about how to fill polygons and fonts with gradients. Even though the geometry agrees between the top and lower shapes in http://cs.sru.edu/~ddailey/svg/HTML5logo6a.svg , most browsers choose to find the midpoint of the gradient differently between the left and right halves of the “5” glyph depending on whether it is drawn as a polygon or a WOFF font. I think that disagreement about where to differentiate midpoints of a gradient is probably attributable to browser error, since it is doubtful that the spec designers desired to leave that aesthetic choice to the implementers. I’ll be fussing on and on about these issues more at SVG Open: http://www.svgopen.org/2011/papers/43-Geometric__Accessibility/index.html Thanks to James Deering, Chaals, and others for support as I formulate some of my still unclear thinking about these issues. Cheers David From: [email protected] [mailto:[email protected]] On Behalf Of Charles McCathieNevile Sent: Sunday, October 09, 2011 7:04 PM To: [email protected] Subject: Re: [svg-developers] another WOFF question On Thu, 06 Oct 2011 15:33:55 +0200, David Dailey <[email protected] <mailto:ddailey%40zoominternet.net> > wrote: > Suppose, for purposes of accessibility, we wished to replace the paths in > the HTML5 logo [1], by actual characters in a font.... > It is quite easy for an author to create an alphabet of exactly five SVG > glyphs for use in the logo, satisfying all of the assistive needs > outlined above, by issuing five statements like: > > <glyph unicode="H"><path d=" > M108.382,0h23.077v22.8h21.11V0h23.078v69.044H152.57v-23.12h-21.11v23.12h-23. > 077V0z " /></glyph> > > (this uses the actual path of the "H" in the HTML5 logo) > > This takes 90 seconds of work using SVG fonts. (Plus the minute and a > half per glyph to draw the letters in either emacs, Illustrator or > Inkscape). As another example, I wrote a font to produce the W3C logo,in the work to create the original and now sadly out-dated SVG accessibility note http://www.w3.org/TR/svg-access I think it took about 10 minutes to make the three letters in a text editor, plus about 10 to print them onto graph paper and get the coordinates. (Plus about a day to get an answer from Jon Ferraiolo about where exactly the flip from zero-down to zero-up on the y-axis actually takes place). > So, the question is this: How easy or painful is it to create a typeface > for just 5 glyphs in WOFF? A few minutes looking at the Web suggests that doing it by hand would be misery, but using a tool should be reasonably straightforward. I found http://onlinefontconverter.com/ which claims it will convert my font from one format to another ... Which would appear to allow for > Does one have to go through the relatively elaborate process of > designing an entire font, or can one create five simple glyphs more or > less like above? As I undertand, you can create just the glyphs you plan to use. > Once one has drawn the shapes, is it a 100 hour process or a 90 second > process?...I honestly have no idea what it would be like to cobble > together a font of 5 simple glyphs, but my suspicion is that it is a > couple of orders of magnitude worse than 90 seconds, hence rather > clobbering the probability that authors would do it. Seems to be less than "orders of magnitude more difficult", but a bit harder. Indeed, the easiest way I can see is to write an SVG font, and then convert it. Cheers Chaals -- Charles 'chaals' McCathieNevile Opera Software, Standards Group je parle français -- hablo español -- jeg kan litt norsk http://my.opera.com/chaals Try Opera: http://www.opera.com [Non-text portions of this message have been removed] ------------------------------------ ----- To unsubscribe send a message to: [email protected] -or- visit http://groups.yahoo.com/group/svg-developers and click "edit my membership" ----Yahoo! Groups Links <*> To visit your group on the web, go to: http://groups.yahoo.com/group/svg-developers/ <*> Your email settings: Individual Email | Traditional <*> To change settings online go to: http://groups.yahoo.com/group/svg-developers/join (Yahoo! ID required) <*> To change settings via email: [email protected] [email protected] <*> To unsubscribe from this group, send an email to: [email protected] <*> Your use of Yahoo! Groups is subject to: http://docs.yahoo.com/info/terms/
