On Thu, Jul 3, 2008 at 1:34 AM, Gerard Meijssen gerard.meijssen@gmail.com wrote:
Hoi, As I have to spell it out for you:
- Wolof: 3,612,560 people
- Swahili: 772,642 first language 30,000,000 second language users
- Xhosa: 7,214,118 people
- Zulu: 7,214,118 first language 15,700,000 second language users
Dejavu status:
zu Zulu 100% (52/52) 100% (52/52) 100% (52/52) sw Swahili 100% (52/52) 100% (52/52) 100% (52/52) xh Xhosa 100% (52/52) 100% (52/52) 100% (52/52) wo Wolof 100% (66/66) 100% (66/66) 100% (66/66)
(http://dejavu.svn.sourceforge.net/viewvc/*checkout*/dejavu/tags/version_2_24... unfortunately the status hasn't been updated in a little while)
For those who are unware the Dejavu fonts are a family of high quality truetype fonts (http://dejavu.sourceforge.net/wiki/index.php/Main_Page) with a fairly wide coverage of unicode.
Dejavu does not yet cover all of unicode (the most notable big gaps are the Chinese, Japanese, and Korean character sets). The primary reason Dejavu does not yet cover everything is because quality is the highest priority (well below being free). A comprehensive font which is ugly or otherwise has poorly matched characters (incompatible metrics and such) is not something that many people will want to use if they have any choice. Most modern systems can substitute characters from other fonts if the selected font has holes in it, so Dejavu allows that to happen until it has quality replacements that fit nicely with the rest of the characters.
Dejavu's focus on quality means that it's suitable for everyday use by everyone with a supported language, and not just by extreme-polyglots who are perhaps more tolerant of ugly print. Dejavu has been the default font set on the current GNU/Linux distributions for some time now.
I'm sure the Dejavu folks would welcome contributions from Wikimedians interested in improving the coverage and quality of the font.
It may also make sense at some point to declare it to be the 'standard font' for the projects, and prefer it via CSS for improved consistency as well as support for the panoply of bizarre characters that even English Wikipedia utilizes. (at least for users who are willing to download it and install it... at well over a megabyte for the complete character set, sending it dynamically is pretty much out of the question! ;) )
(Preferring something like the currently non-free code2000 font would also provide better coverage of unicode... but on most projects doing so would get you shot: that font is not visually appealing in the slightest)
It is people who speak languages. It is people we aim to provide information to.
As to the Mozilla Foundation; the point is that our aim is to provide information to all people. It is essential that the infrastructure is there to achieve it. Fonts are essential in this game. The Mozilla foundation is to provide functionality to people that are already on the web. In what we aim to do, we do not say that people have to be online in order to be part of our potential public.
Indeed. Fonts are important. But they are important to a great many people outside of Wikimedia. Other projects like Fedora Linux, and OLPC depend on good fonts even more than we do, so the bulk of the work has already been done by others and is already freely available without any action on our part.
What gaps remain must be filled, but that is work that should be coordinated through the real heavy lifters in this space, and not Wikimedia which does not have a past history of font production or distribution.
Cheers
Hoi, I like Dejavu. Dejavu at this moment does not cover what we need. We do need Chinese, Japanese and Korean; to be fair we need more then UTF-8 is able to offer. What I intended to do is indicate ways of spending money. The question is how to spend money and how to get the most bang for our buck.
Creating content in for instance the four languages that I singled out, would be relatively cheap. We can generate a lot of content that will apeal to the people for whom it is their mother tongue or for whom it is a major language.
Creating new fonts or making existing fonts Open Source is a worthwhile goal. We can work on the extension of Dejavu, it is not unlikely that we can Open Source existing fonts more cheaply. In the end, the real difference is made when an end user is provided with information in a better way.
In the original thread the argument was about spending money. We should when it furthers our goal. Fonts is in my opinion one. In the end, when we spend money we want to have a result. When we are to invest in Dejavu, the work done has to be welcomed by the people behind Dejavu.
In my opinion, the first order of busines re fonts is getting something out that is available and free. Esthetics are secondary. When the WMF provides the funding, it will be heavy lifters doing the work not our volunteers.
Thanks, GerardM
On Thu, Jul 3, 2008 at 7:58 AM, Gregory Maxwell gmaxwell@gmail.com wrote:
On Thu, Jul 3, 2008 at 1:34 AM, Gerard Meijssen gerard.meijssen@gmail.com wrote:
Hoi, As I have to spell it out for you:
- Wolof: 3,612,560 people
- Swahili: 772,642 first language 30,000,000 second language users
- Xhosa: 7,214,118 people
- Zulu: 7,214,118 first language 15,700,000 second language users
Dejavu status:
zu Zulu 100% (52/52) 100% (52/52) 100% (52/52) sw Swahili 100% (52/52) 100% (52/52) 100% (52/52) xh Xhosa 100% (52/52) 100% (52/52) 100% (52/52) wo Wolof 100% (66/66) 100% (66/66) 100% (66/66)
( http://dejavu.svn.sourceforge.net/viewvc/*checkout*/dejavu/tags/version_2_24... unfortunately the status hasn't been updated in a little while)
For those who are unware the Dejavu fonts are a family of high quality truetype fonts (http://dejavu.sourceforge.net/wiki/index.php/Main_Page) with a fairly wide coverage of unicode.
Dejavu does not yet cover all of unicode (the most notable big gaps are the Chinese, Japanese, and Korean character sets). The primary reason Dejavu does not yet cover everything is because quality is the highest priority (well below being free). A comprehensive font which is ugly or otherwise has poorly matched characters (incompatible metrics and such) is not something that many people will want to use if they have any choice. Most modern systems can substitute characters from other fonts if the selected font has holes in it, so Dejavu allows that to happen until it has quality replacements that fit nicely with the rest of the characters.
Dejavu's focus on quality means that it's suitable for everyday use by everyone with a supported language, and not just by extreme-polyglots who are perhaps more tolerant of ugly print. Dejavu has been the default font set on the current GNU/Linux distributions for some time now.
I'm sure the Dejavu folks would welcome contributions from Wikimedians interested in improving the coverage and quality of the font.
It may also make sense at some point to declare it to be the 'standard font' for the projects, and prefer it via CSS for improved consistency as well as support for the panoply of bizarre characters that even English Wikipedia utilizes. (at least for users who are willing to download it and install it... at well over a megabyte for the complete character set, sending it dynamically is pretty much out of the question! ;) )
(Preferring something like the currently non-free code2000 font would also provide better coverage of unicode... but on most projects doing so would get you shot: that font is not visually appealing in the slightest)
It is people who speak languages. It is people we aim to provide
information
to.
As to the Mozilla Foundation; the point is that our aim is to provide information to all people. It is essential that the infrastructure is
there
to achieve it. Fonts are essential in this game. The Mozilla foundation
is
to provide functionality to people that are already on the web. In what
we
aim to do, we do not say that people have to be online in order to be
part
of our potential public.
Indeed. Fonts are important. But they are important to a great many people outside of Wikimedia. Other projects like Fedora Linux, and OLPC depend on good fonts even more than we do, so the bulk of the work has already been done by others and is already freely available without any action on our part.
What gaps remain must be filled, but that is work that should be coordinated through the real heavy lifters in this space, and not Wikimedia which does not have a past history of font production or distribution.
Cheers
foundation-l mailing list foundation-l@lists.wikimedia.org Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/foundation-l
On Thu, Jul 3, 2008 at 2:27 AM, Gerard Meijssen gerard.meijssen@gmail.com wrote:
Hoi, I like Dejavu. Dejavu at this moment does not cover what we need. We do need Chinese, Japanese and Korean; to be fair we need
Today Dejavu is intended to be paired with other fonts to complete the CJK coverage. This quite well works on systems with reasonable font handling. If you want a single download you could always package them togeather.
But sure, it's incomplete. I did acknowledge that.
[snip]
more then UTF-8 is able to offer.
Could you expand on that somewhat? Why are you saying UTF-8 specifically?
Hoi, The MediaWiki software supports all scripts that are in UTF-8. This means that scripts that are not supported in UTF-8 and by inference the languages that use these scripts are not supported in MediaWiki. We have a request for one such language; American Sign Language. The people behind SignWriting, the script that allows ASL to be written, are developing an extension that will allow for ASL to be used in MediaWiki. It would be optimal when all scripts find their way in UTF-8.
NB SignWriting is only one out of a multitude of scripts that are not supported by UTF-8.
The idea of packaging a set of fonts and promoting this to our users is something I would find REALLY nice :) Thanks, GerardM
On Thu, Jul 3, 2008 at 8:32 AM, Gregory Maxwell gmaxwell@gmail.com wrote:
On Thu, Jul 3, 2008 at 2:27 AM, Gerard Meijssen gerard.meijssen@gmail.com wrote:
Hoi, I like Dejavu. Dejavu at this moment does not cover what we need. We do
need
Chinese, Japanese and Korean; to be fair we need
Today Dejavu is intended to be paired with other fonts to complete the CJK coverage. This quite well works on systems with reasonable font handling. If you want a single download you could always package them togeather.
But sure, it's incomplete. I did acknowledge that.
[snip]
more then UTF-8 is able to offer.
Could you expand on that somewhat? Why are you saying UTF-8 specifically?
foundation-l mailing list foundation-l@lists.wikimedia.org Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/foundation-l
On Thu, Jul 3, 2008 at 4:09 AM, Gerard Meijssen gerard.meijssen@gmail.com wrote:
Hoi, The MediaWiki software supports all scripts that are in UTF-8. This means that scripts that are not supported in UTF-8 and by inference the languages
Ah. Okay. Please say unicode and not UTF-8 in this context.
Because of the wealth of broken software out there there are still people who believe that UTF-8 can only cover the subset covered by UCS16, rather than all of unicode. Your specific mention of UTF-8 let me wondering if you were suffering from that particular misconception.
Of course, mediawiki has been used in the past with other encodings than unicode. ... But supporting these scripts with alternative encodings would be a return to the bad old days. Better to get them included in unicode. ... a matter well outside the scope of this list. :)
wikimedia-l@lists.wikimedia.org