Internet Explorer and Windows might not be optimized for this Blog

by Michael S. Kaplan, published on 2008/12/22 03:01 -05:00, original URI: http://blogs.msdn.com/b/michkap/archive/2008/12/22/9246308.aspx


After reading It isn't really RED versus GREEN, Cheong asked:

I've searched for a while in charmap.exe and yet to find a font that'll display these characters in WinXP. That brings me to a question: Is there a font that ships with Windows (not necessarily WinXP or Vista... just any version. Maybe even a future version. :) ) that'll display most non-user-defined range of Unicode characters? Or at least that'll display most of the "sponsor characters" in your blog?

Sorry if this question has already been answered, because I've been yet to come up with effective keywords for this question.

For good measure, after a less than 24 hour wait he asked over in the Suggestion Box:

I'm refering to article titled "It isn't really RED versus GREEN"

I've searched for a while in charmap.exe and yet to find a font that'll display these sponsor characters in WinXP. That brings me to a question: Is there a font that ships with Windows (not necessarily WinXP or Vista... just any version. Maybe even a future version. :) ) that'll display most non-user-defined range of Unicode characters? Or at least that'll display most of the "sponsor characters" in your blog?

Sorry if this question has already been answered, because I've been yet to come up with effective keywords for this question.

Now I am not making fun of him, it was actually the Suggestion Box comment that reminded me about the question.

So what was done here was the right thing to do, at least insofar as right and wrong can be determined by whether a blog is inspired.

Don't look now, we are heading into an über-metablogong topic -- Blog morality! Where right and wrong is only defined in terms of the content of a Blog!

I'll revisit that topic again some day, since although it has the risk of playing out really dumb, I think I'll probably enjoy it. Though for now I'll go back to Cheong's actual question.... :-)

There is no font that ships with any version of Windows currently available that will handle any possible character that an ornery person like me sets up sponsorship details with.

On the whole, I have found that due to limitations like the one I described in The importance of Tagalog to Burmese, aka "Of course I'd lie to you, I'm a font!" that for many of the ranges outside the ones listed in the table in that blog, Firefox seems to do a better job finding fonts that will work than Internet Explorer does.Not so much that I'd yet claim such blogs on this Blog are "optimized for Firefox" because I still have an ECMA Script bug to track down. But maybe one day....

In he meantime, I find that for most of Plane 1 if you have a font that supports the character on your machine that Firefox seems to find it more often than IE does -- even IE8.

IE seems to be optimized for the fonts supporting languages that Windows claims support for, which one again suggests that a browser less interested in its idea to its OS has the opportunity to do better here. But that is just a theory.

My IE team advice would be to either take over MLang so you can update it, or dump it like a bad habit and find a way to support features that require forward looking thinking, like the aforementioned Tagalog/Burmese/Mongolian bug.

But to get back to Cheong's question.... :-)

I also do not know of future plans for such a font. Though with each version coverage gets better it is harder to imagine coverage of very obscure historic scripts with limited user communities a being a priority.

So I'd guess that, given the way I often to go to the more obscure parts of Unicode, you could not claim that Windows is optimized for this Blog, either.

Of course you have to realize that this argument is quite a specious tail-wagging-the-dog type argument, and not only because claiming that the reason one reads this blog is for the sponsoring character is like the person who claims to be reading Playboy for the articles (with the exception of Microsoft employees and I believe the July 1994 issue!).


This blog brought to you by(U+1700, aka TAGALOG LETTER A)


parkrrrr on 22 Dec 2008 9:23 AM:

Everson Mono Unicode and Code 2000, both available on the web as shareware, support all of the symbols who sponsored your Red vs. Green post, but charmap.exe on Windows XP does not show those characters.

Oddly enough, Firefox also does not show them, but it also doesn't show the "1700" in a box that it shows for the sponsoring character for this post.  I think Firefox on my system is selecting Bitstream Cyberbit, which claims to support 246 code points (only 176 are defined!) in the "Miscellaneous Symbols" block, but in fact maps most of those code points to some blank glyph.

None of the fonts I have installed on this system - including those three - claims any support for anything in the Tagalog block.  (I'm using Babelpad version 1.10.1 to determine which fonts support which blocks; it claims to support Unicode 5.0.0.)

John Cowan on 22 Dec 2008 1:48 PM:

Bitstream Cyberbit is an ancient font, it lies like a rug, it sucks like a pump, dump it!

parkrrrr on 22 Dec 2008 3:35 PM:

Having heeded John Cowan's advice, I can say with certainty that Firefox was indeed trying to use Bitstream Cyberbit, and those symbols now show up for me on the Red vs. Green blog.  It's now using Code 2000 (also an ancient font, honestly) for those code points.  

Also, having done further tests, I've determined that charmap on Vista x64 does display the "Red vs. Green" code points when you select a font that has them, but only if you select "Group by: All."  If you select "Group by: Unicode Subrange" and pick the so-called "Symbols & Dingbats" subrange, it display code points on both sides of those, but it does not show those.  This is a step up from Windows XP, but still not perfect.  Other much, much newer versions of Windows apparently continue to have the so-so Vista behavior.

Michael S. Kaplan on 22 Dec 2008 4:38 PM:

Vista x64 is just Vista compiled for the x64 platform, so I am not sure what you meant by that last bit....

Jan Kučera on 23 Dec 2008 9:57 AM:

Honestly, I would be happy enough if the IE8 was supporting the combining diacritical marks!

Andrew West on 23 Dec 2008 2:03 PM:

I think that parkrrrr was comparing Vista x64 with XP not with Vista.

At any rate, as I am sure I have mentioned before, charmap's Unicode Subrange groupings are still constrained to Unicode 2.0 (released in 1996), and so not only does charmap only show Unicode 2.0 subranges, but it will only show characters in a particular "subrange" that were in Unicode 2.0 ... so no characters encoded in the last twelve years will be shown in Unicode Subranges mode! I generally don't go in for Microsoft-bashing, but honestly, charmap has to be a contender for the suckiest MS application ever.

parkrrrr on 23 Dec 2008 4:39 PM:

Feel free to moderate this out of existence, as befits it, but I meant that win7 still has that bug.

parkrrrr on 23 Dec 2008 4:49 PM:

To further summarize, since I'm not sure which "last bit" you refer to:

Windows XP doesn't display those codepoints at all in charmap.exe, no matter whether you select "Group by: All" or "Group by: Unicode Subrange."  

Windows Vista (including x64, because that's the exact subspecies I was running when I did the test yesterday, and I wanted to be specific) displays those codepoints when you group by all, but not when you group by Unicode subrange.  

I haven't tested it on Server 2003 or 2008.

The XP behavior was clearly a bug, but it seems like the Vista behavior is, too.  It might be nice if it were fixed, because grouping by all is painful in a font like Code 2000.  (Also, as long as I'm wishing, it might be nice if the subranges offered by charmap.exe were actually the official Unicode names for the blocks.)

Michael S. Kaplan on 23 Dec 2008 7:10 PM:

I pointed out several of the CharMap issues in

http://blogs.msdn.com/michkap/archive/2008/02/27/7897776.aspx

previously, including the grouping one. Character coverage gets better, but thus far grouping does not since no one seems to know how to update the groupings....


go to newer or older post, or back to index or month or day