Talk:Sun and moon letters

As noted in Talk, I've tried to make this a little more approachable for the non-expert reader, who may not be familiar with some terms in phonology or with Arabic letters.Herbivore 03:36, 3 January 2007 (UTC)[reply]

Merger

I noticed someone placed a merge tag on the article. I think it might be an acceptable thing to do, but the issue here is what will the merged article's name be. There is no common distinctive name for the letters. Any suggestions? - Anas ^talk? 22:02, 21 March 2007 (UTC)[reply]

The most logical I think would either be Sun and Moon letters or Moon and Sun letters. The Storm Surfer 15:38, 26 March 2007 (UTC)[reply]

I just merged them to Sun and moon letters. Hopefully I've managed not to screw it up ;) The Storm Surfer 21:14, 2 April 2007 (UTC)[reply]

Order of moon letters

In my browser (Firefox/Iceweasel 2.0.0.3 on Debian etch) the list of moon letters doesn't match the transcription. I see the letters in the order gim, ha, kha, `ain, ghen, fa, qaf, kaf, mim, he, alif, ba, <nothing>. On the edit page the order of letters differs from this, and I ran into problems with left/right direction when editing, so somebody with more knowledge about editing Arabic text should look into this. 89.14.40.212 07:43, 17 May 2007 (UTC)[reply]

ch

In Maltese, "ċ" /ʧ/ (which doesn't descend from an Arabic sound) is also treated as a sun letter, presumably because "t" is also one and "ċ" is kind of like "tx" /tʃ/. But "ġ" /ʤ/ is a moon letter, presumably because it derives (in native words) from jeem ج, which is a moon letter. -- pne (talk) 20:24, 20 May 2009 (UTC)[reply]

Graphic - Readability

The graphic's gray text should be blackened and probably enlarged for readability. It conveys no information to anyone who can infer that the color coding only applies to the Arabic text, which should be everyone. LokiClock (talk) 05:48, 17 December 2009 (UTC)[reply]

I don't think so. Just click at the graphic and you will see an enlarged and excellent readable image. --Obersachse (talk) 06:28, 17 December 2009 (UTC)[reply]

The image was made to supplant the article, not to stand alone, and such images should be readable at the size they're meant to be displayed at within the article whenever possible. It's still gray, not even dark gray, which makes it difficult to read at any size- it's a matter of contrast. Unless you have a good reason, you should never reduce the contrast between text and its background. LokiClock (talk) 12:42, 17 December 2009 (UTC)[reply]

Okay, I've now assembled evidence to prove my point about the color. W3C's WCAG documents state certain contrast thresholds for visual usability. One can read them here. Using this applet to take measurements, I compared the text color of the image, #808080, with white, #FFFFFF, and the background color of image boxes, #F9F9F9, in the "Vector" skin. According to the minimum requirements (1.4.3) for usability, normal text should have at least a 4.5:1 contrast ratio with its background, and the restrictions are loosened for large text to 3:1. Even against pure white, the contrast is as low as 3.95, which is closing in on the minimum for large text (for which the applet reports the contrasts "sort of" compliant), well below acceptable for normal sized text, and simply illegible for text at the size with which it is displayed in the article. Against the image box color, the contrast level becomes 3.75, which the applet judges to be flat out incompliant. Again, these are minimum standards. That is, "if the colors pass these, it's possible that someone can read them." For comfortable reading, the enhanced requirements (1.4.6) prescribe a contrast of at least 7.1:1, 4.5:1 for large text. "The user can just click the image to enlarge it" is not a valid argument against adjusting the image to follow the guidelines. The document declares that, for the purposes of compliance with the guidelines, "font size is the size when the content is delivered." Again, the image was made for an article, so the delivery point is the article. This article, while focused on the effects of anti-aliasing, shows great differences in reader preference from 10 to 12 pt type. Other opinions, research, etc. regarding type usability can be easy found through google. LokiClock (talk) 16:09, 6 January 2010 (UTC)[reply]

Would someone who speaks German please inform the author of this discussion? LokiClock (talk) 13:55, 18 February 2010 (UTC)[reply]

I thought it was You? --Obersachse (talk) 16:12, 18 February 2010 (UTC)[reply]

I mean, inform the author of the image of this discussion's existence. LokiClock (talk) 16:20, 18 February 2010 (UTC)[reply]

I did it. --Obersachse (talk) 17:34, 18 February 2010 (UTC)[reply]

Thank you muchly. LokiClock (talk) 18:39, 18 February 2010 (UTC)[reply]

He said "What can I contribute to this discussion? I see no problem with the readability. Otherwise I would have increased the fonts or colored them darker. I agree, if someone wants to do this, even though I can not see the sense." --Obersachse (talk) 20:07, 18 February 2010 (UTC)[reply]

Well, if he doesn't do the editing himself, it's good to have his consent for replacing the graphic. Note, in summary: 1. The requirements of the W3C usability guidelines, which the image fails to pass, and that even though the type may be large enough at full size, at display size it is in fact too small to even properly sample (note aliasing issues), much less read. 2. The failure of #808080 on #F9F9F9 or #FFFFFF to pass this automated color contrast test. These shortcomings demonstrate that, even for people without vision problems, the images are unpleasant to read. LokiClock (talk) 20:48, 18 February 2010 (UTC)[reply]

As the image is an SVG, using the size of the full view is a vacuous argument for its adequacy of usability; You can make it any size you want if you save it and open it in a SVG viewing program. Again, the W3C guidelines prescribe that the usability of type should be judged by its size at content delivery (i.e., as it displays within the article), not on a possible arbitrary scale that the user could view the image at. It also takes up much more space than is actually needed to represent the content by drastically contrasting the scales of the information. LokiClock (talk) 20:48, 18 February 2010 (UTC)[reply]