Hi All
Can I clear any doubt on Burmese Localization Situation.
*anyone can use any myanmar Unicode 5.1 - compliant font to look at it.*
For Burmese, It's different from English situation. Microsoft Windows XP and older operating system are widely used by Burmese People. Microsoft Windows Vista natively support Burmese Rendering. Ubuntu 8.04 natively support Burmese Rendering. Main Problem is lack of rendering support by Microsoft Windows older version. We have to patch/update Uniscribe Rendering Engine in older Operating System. That's technical issues.
*If there are Unicode 5.1 fonts, perhaps incomplete, perhaps complete, then fine. People can use those.*
Absolutely, There are no problem concern with fonts. There are 3 Unicode Fonts for Burmese. At First, my.wikipedia.org is not suited for other ethnics characters and Buddhist teaching concern. That's fine. Why we need to wait anymore. But for the Zawgyi fans and old operating system users might not be seen proper Burmese characters because encoding standards are different from Zawgyi and Unicode 4.1 What ever we need to think long run. I strongly recommended to store Unicode 5.1.
*What would the problem be with this? Would it render the site unusable for people who are using only Zawgyi?*
We should support any Wikipedian who are willing to write with any font/encoding. But we can convert those input texts with different encoding by internal converter. It's meaning to say Zawgyi text can be converted into Unicode 5.1. I suppose to do accessibility problem is key issues for every language. Michael figured out that multiple platform would be fine. I would like to propose again that multiple font encoding must support in my.wikipedia.org
Be clear enough. I do not support storing non-Unicode 5.1 Standards. But we must support converted HTML in Zawgyi, MyaZedi and such dated Fonts.
Well, Here is solution to put converter in wiki. http://mmbotconvert.vndv.com/ It's off-line converter for Zawgyi<>Unicode 5.1. It can be engaged with Wiki Textbox. Let me explain 1 scenario, A Wikipedian will go Myanmar CyberCafe. Cybercafe may not allow to install Unicode 5.1 Fonts. But In their machine, They installed Zawgyi Font. So, Our proposed solution is to easily access with installed fonts.
How can we solve user accessibility issues? Only Javascirpt file, It should be attached in MediaWiki. We have accomplished converter development in javascript. We are going to test unit testing in Myanmar Orthography List. So, you may clear that why we are taking care on non-Unicode Fonts. We had solution to convert those thing.
So, There are no problem in Fonts. Only problem in accessibility issues.
Thanks for your time and effort on Myanmar Wikipedia.
Ngwe Tun
2008/7/25 Ngwe Tun ngwestar@gmail.com:
Hi All
Can I clear any doubt on Burmese Localization Situation.
*anyone can use any myanmar Unicode 5.1 - compliant font to look at it.*
For Burmese, It's different from English situation. Microsoft Windows XP and older operating system are widely used by Burmese People. Microsoft Windows Vista natively support Burmese Rendering. Ubuntu 8.04 natively support Burmese Rendering. Main Problem is lack of rendering support by Microsoft Windows older version. We have to patch/update Uniscribe Rendering Engine in older Operating System. That's technical issues.
My current understanding is that usp10.dll is protected by Microsoft's EULA which prohibits copying usp10.dll onto a computer. Some developers can license usp10.dll but only for local use with their application, the license prohibts them from replacing the system version of usp10.dll
There are also issues with font fallback mechanisms. As Michael pointed out, you'd have disasterous results if the fallback selected a pre-Unicode 5.1 font implementation.
In theory a user could select a default font for a script in their web browser's preferences/options. A few barriers exist. Firefox has no UI mechanism to select to specify a default Myanmar script font. Selecting a default font in the Windows version of Internet Explorer is inhibited due to a long standing bug in mlang.dll, IE will not recognise a Myanmar Script font. The only remedy is to edit the Windows registry. A task that the average user may not be able to perform.
The only real solution is to handle font fall back through CSS, identifying the most suitable and most commonly available of the Unicode 5.1 fonts and specifying each within a CSS rule.
Andrew
At 07:45 -0400 2008-07-25, Jimmy Wales wrote:
Andrew Cunningham wrote:
The only real solution is to handle font fall back through CSS, identifying the most suitable and most commonly available of the Unicode 5.1 fonts and specifying each within a CSS rule.
Who can make that happen?
Htoo Myint Naung and I had proposed to do so when I wrote to you at the end of May. You may be a little behind me in reading mails on this list though.. earlier this morning I wrote this:
I've suggested getting my.wiki sorted out so that it is user friendly in terms of getting people fonts and input methods (like the Kannada and/or Telugu sites do). I think we should investigate compliant fonts and their capabilities, put them in the CSS ranked in terms of universality, clean up what needs cleaning up, and then work on the conversion of the other material.
Yes, I am caught up now. It sounds like you and Andrew agree.
okisan, do you agree?
Michael Everson wrote:
At 07:45 -0400 2008-07-25, Jimmy Wales wrote:
Andrew Cunningham wrote:
The only real solution is to handle font fall back through CSS, identifying the most suitable and most commonly available of the Unicode 5.1 fonts and specifying each within a CSS rule.
Who can make that happen?
Htoo Myint Naung and I had proposed to do so when I wrote to you at the end of May. You may be a little behind me in reading mails on this list though.. earlier this morning I wrote this:
I've suggested getting my.wiki sorted out so that it is user friendly in terms of getting people fonts and input methods (like the Kannada and/or Telugu sites do). I think we should investigate compliant fonts and their capabilities, put them in the CSS ranked in terms of universality, clean up what needs cleaning up, and then work on the conversion of the other material.
On Fri, Jul 25, 2008 at 9:04 PM, Jimmy Wales jwales@wikia.com wrote:
Yes, I am caught up now. It sounds like you and Andrew agree.
okisan, do you agree?
If he REALLY means it, it has already been done and we don't have to go this much of conversation.
Okisan
At 10:17 +0630 2008-07-25, Ngwe Tun wrote:
For Burmese, It's different from English situation. Microsoft Windows XP and older operating system are widely used by Burmese People.
Yes, and they have to migrate to Unicode 5.1. Otherwise we will simply have continued encoding chaos on my.wiki. I don't see any other option.
*If there are Unicode 5.1 fonts, perhaps incomplete, perhaps complete, then fine. People can use those.*
Absolutely, There are no problem concern with fonts. There are 3 Unicode Fonts for Burmese. At First, http://my.wikipedia.orgmy.wikipedia.org is not suited for other ethnics characters and Buddhist teaching concern. That's fine.
Excuse me, but I object here. The Myamnar block in Unicode contains many characters. Some are used in Sanskrit in Burmese. Some are used in minority languages. "Buddhist teaching concerns" are legitimate articles for the Myanmar Wikipedia, and users may wish to cite Sanskrit words there. So you cannot say that the Sanskrit characters in Unicode are irrelevant or should be ignored. They may be part of data. Some Unicode 5.1 fonts support this behaviour, some do not. I have proposed to Wikimedia that we investigate what level of support each font has so that users can choose which ones they want. This is POSITIVE comptetition. A font (let's say Padauk or Parabaik) which doesn't correctly render a Sanskrit word might be improved. Or it might have fallback rendering. But at least we would know.
Similarly, the Myanmar scripts minority characters may find their way into texts on the Myanmar wikipedia. Look at http://en.wikipedia.org/wiki/Thorsteinn_P%C3%A1lsson and http://en.wikipedia.org/wiki/Ilham_Aliyev and http://en.wikipedia.org/wiki/Lech_Walesa. Each of these men uses letters in his name which are unrecognized by English speakers. Yet those letters are used in the English wikipedia -- sometimes for the article title, sometimes just in the text. So when Ngwe Tun says that "http://my.wikipedia.orgmy.wikipedia.org is not suited for other ethnics characters" I cannot agree with him. Font developers should be encouraged to support the whole block of Myanmar characters, at least at some level. If not, then we can state so, and users can choose the fonts they wish.
Why we need to wait anymore. But for the Zawgyi fans and old operating system users might not be seen proper Burmese characters because encoding standards are different from Zawgyi and Unicode 4.1 What ever we need to think long run. I strongly recommended to store Unicode 5.1.
On this Ngew Tun and I are agreed. :-)
*What would the problem be with this? Would it render the site unusable for people who are using only Zawgyi?*
We should support any Wikipedian who are willing to write with any font/encoding. But we can convert those input texts with different encoding by internal converter.
That's just going to be
It's meaning to say Zawgyi text can be converted into Unicode 5.1. I suppose to do accessibility problem is key issues for every language. Michael figured out that multiple platform would be fine. I would like to propose again that multiple font encoding must support in http://my.wikipedia.orgmy.wikipedia.org
I don't think this is feasible. You're going to have to have a suite of browser plug-ins (platform specific) which will be smart enough to detect what encoding people are using and convert losslessly back and forth between the encodings. And there are several encodings and pseudo-encodings out there. And then there is the question of when *this* gets phased out in the future. Ten years from now when the last Windows XP computer dies? It just delays the migration. Also what happens when somebody selects text in the browser window and pastes it into an e-mail? You'll have encoding incompatibility there too, and your plug-in will not help you there.
Be clear enough. I do not support storing non-Unicode 5.1 Standards.
Good.
But we must support converted HTML in Zawgyi, MyaZedi and such dated Fonts.
I can't see how this can be supported on the Wiki side.
Well, Here is solution to put converter in wiki. http://mmbotconvert.vndv.com/http://mmbotconvert.vndv.com/ It's off-line converter for Zawgyi<>Unicode 5.1. It can be engaged with Wiki Textbox. Let me explain 1 scenario, A Wikipedian will go Myanmar CyberCafe. Cybercafe may not allow to install Unicode 5.1 Fonts.
The customer should complain.
But In their machine, They installed Zawgyi Font.
They shouldn't. I don't think we should encourage the perpetuation of non-Unicode fonts.
Ngwe Tun wrote:
How can we solve user accessibility issues? Only Javascirpt file, It should be attached in MediaWiki. We have accomplished converter development in javascript. We are going to test unit testing in Myanmar Orthography List. So, you may clear that why we are taking care on non-Unicode Fonts. We had solution to convert those thing.
This sounds simple enough. The javascript already exists, right? And it works? And it will allow people with older browsers to use Wikipedia today, while also providing a sensible path forward for the future?
That sounds wonderful! How can we make that happen on my.wikipedia.org?
--Jimbo
At 07:44 -0400 2008-07-25, Jimmy Wales wrote:
This sounds simple enough. The javascript already exists, right? And it works? And it will allow people with older browsers to use Wikipedia today, while also providing a sensible path forward for the future?
That sounds wonderful! How can we make that happen on my.wikipedia.org?
If you are going to allow "any" UTF-8 font to be used on my.wiki then I do not see how I can be of service to you. What I can do is help to ensure that we know what the capabilities of Unicode-5.1 compliant fonts are, and to ensure that compliant fonts are listed in a reasoned order in the CSS. What is being suggested here sounds as though it would simply perpetuate and encourage the use of non-standard fonts. I don't believe that is the right thing to do, and it will just cause more work later.
Michael Everson wrote:
At 07:44 -0400 2008-07-25, Jimmy Wales wrote:
This sounds simple enough. The javascript already exists, right? And it works? And it will allow people with older browsers to use Wikipedia today, while also providing a sensible path forward for the future?
That sounds wonderful! How can we make that happen on my.wikipedia.org?
If you are going to allow "any" UTF-8 font to be used on my.wiki then I do not see how I can be of service to you. What I can do is help to ensure that we know what the capabilities of Unicode-5.1 compliant fonts are, and to ensure that compliant fonts are listed in a reasoned order in the CSS. What is being suggested here sounds as though it would simply perpetuate and encourage the use of non-standard fonts. I don't believe that is the right thing to do, and it will just cause more work later.
Well, let me be clearer on what I am suggesting.
If there are people who are simply unable to upgrade or able but don't know how, I think we should find a way to accomodate them while at the same time encouraging them to upgrade. Everything should be stored in a Unicode-5.1 compliant way, and then input method used for those on older systems should strictly enforce this.
This is a question of effectiveness. Is it more effective to simply switch to a system on Wikipedia that many people are unable to use at all, or is it more effective to get those people involved in writing content that will be valuable for the future while teaching and educating them (and turning them into sales people for upgrading!)?
I will say this much, though: allowing people to use an input method which "breaks" anything is absolutely something I would oppose, because that would create more work later. So I think we agree on that part.
The part where we may not agree is in the area of: what do we tell people on the site who do not have a system which can do things in the modern way?
1. Go away, you have an old computer - not welcoming 2. Go to this other site and work in the old way, make friends with a different community, do work that probably replicates work done on the real site, etc. - the worst possibility, because we will never unite the communities in the end, trust me, I know about this 3. Welcome! Here is how you can input using your old system, but really you should upgrade. Please upgrade. Here is how. Please upgrade. You can keep doing this for now but it is old-fashioned. Please upgrade.
--Jimbo
At 08:10 -0400 2008-07-25, Jimmy Wales wrote:
If there are people who are simply unable to upgrade or able but don't know how, I think we should find a way to accomodate them while at the same time encouraging them to upgrade.
I see. For how long?
Everything should be stored in a Unicode-5.1 compliant way, and then input method used for those on older systems should strictly enforce this.
I understand. Certainly the same problem has been encountered in India. But on the Tamil Wikipedia, for instance, where there are also many legacy fonts and those are not being supported, as far as I can see.