Re: I18N, fonts and mobile (was Re: After Access – Challenges Facing Mobile-OnlyInternet Users in the Developing World)

On 5 Aug 2010, at 08:55, Stephane Boyera wrote:

> (me changing the topic to be closer to what we are discussing)
>
> Let me first start with a big thank to Richard for taking the time  
> to write this blog entry. I had the plan to do such a thing, but was  
> lacking the expertise. So this is very helpful.
>
> Now some  comments on the point made by David
>
>> * Opera Mini uses what is installed on the phone (I think it has  
>> its own
>> generic serif/sans-serif too as a fall back but I'm not sure if  
>> that is
>> still the case). I'd expect phone in India to be shipped with a font
>> that can render the local language (or is that assumption too much to
>> ask?).
>
> I think this is a very good question. I don't know if it is too much  
> to ask, but this is surely something this group could recommend to  
> handset manufacturers and operators.
> This has an impact on web browsing as well as on other text-based  
> technologies (SMS,  USSD).
>
>> The Opera Mini UI is localised into many of the languages used in
>> India, so I expect it has to work at some level.
>
> Is there any data collected and publicly available about about many  
> users are using the localized UI ?

Not that I know of. I could ask internally but it is probably  
confidential. I’ll try to find out. We do talk about how in certain  
countries such as India, we feel users are getting their first taste  
of the web via mobile browsing, rather than the desktop. The desktop  
is mostly English in countries like India where there is a large  
population (at least in the "educated" segment of the population) that  
speak English. I expect that to not be as much the case on mobile (at  
least for non-smart phones), but I don't have any data to present  
right now to prove that hypothesis.

For web sites, we do have aggregated data of which sites users of  
Opera Mini (per country) visited. The full data is not public (for  
obvious reasons) but we publish the top 10 for our top countries each  
month (India is one of those, and many of the others are developing  
nations such as Indonesia, China, South Africa, Vietnam and Nigeria),  
and we do a regional focus each month. Last month was Africa:

http://www.opera.com/smw/2010/06/

Indonesia is not too interesting to this debate (although it is our  
number one country now, above Russia) as they use a latin script.  
Russia and Ukraine both use Cyrillic and all their top sites are  
available in cyrillic. I know from first hand experience that phones  
in Russia support cyrillic fonts (and many Russians do not speak  
English), so the majority of our CIS-based users are likely using the  
Russian localised version (or Ukrainian, etc.) In India most of the  
top sites are global. Only one is local (but ironically from Pakistan,  
not India). Another "zedge.net" seems global, but at least looking at  
the current users on the home page, they are all from developing  
markets such as Philippines, Egypt and India. At least some of those  
global sites are available in Hindi, etc (google, orkut, wikipedia,  
etc), but we don't have data on if they are using the localised  
version of those sites. The long tail would tell us more, as it won't  
be almost exclusively your typical big sites like orkut, facebook,  
google, wikipedia etc.

The top sites for China are in Chinese, and I'm pretty certain the  
situation there is the same as Russia where phones (and browsers)  
support simplified Chinese. South Africa and Nigeria don't tell us  
much as English is common in those countries, and the local languages  
are mostly latin-based these days (I think). Opera Mini is localised  
into Hausa, Igbo and Yoruba though. With English being the official  
language of Nigeria, I expect content availability to be a bigger  
issue. Vietnam has a number of local sites in the top ten (such as http://dantri.com.vn/) 
  which default (at least) to Vietnamese. So for a number of countries  
at least, there doesn't seem to be a problem across the full chain of  
sites, phones and browsers of supporting non-latin (or extended latin)  
characters. India is different to countries like Vietnam and Russia,  
as there are more people that speak English (and more official  
languages) so there may be less incentives to supporting those non- 
English languages by sites or phone makers, but with the size of the  
Indian market I'd expect there would be a market for it to happen if  
it doesn't already.

Looking further back to our report on the Middle East: http://www.opera.com/smw/2010/01/ 
  there are a number of sites in the top 10 sites of the top 10 middle- 
east countries that are either in Arabic, Persian or Hebrew, so I  
expect it is also not an issue in those markets either. But again,  
they are countries where the local language is dominant over languages  
such as English.


> For instance i'm wondering if the current major issue is on the  
> phone side and lack of fonts, or on the content side, and lack of  
> content available and/or correctly written in one or the other  
> indian languages .

This one I can't answer for India yet, but isn't a problem across the  
board for other countries like Russia, China, Taiwan, Vietnam, etc.

>> Opera Mini has multiple millions of
>> users in India (and across the Indian sub-continent) so that gives me
>> further suspicion it works out of the box there.
>
> "works out of the box" ? that's all the question imho. Is there  
> users that do not speak english, or are the multiple millions of  
> users the english speaker ? Here again, if there are public data on  
> the number of sites accesses in one of the indian languages this  
> would be helpful.

I'll try to get hold of this if possible.

>
>> As mentioned later in this thread, CSS3 Web Fonts (not HTML5) can  
>> also
>> be used, as you can link to one of the freely available fonts (which
>> give you a licence to use Web Fonts). This works in Opera Mobile, but
>> not Opera Mini (which is far bigger in India as it works on lower end
>> devices). One reason for that is Web Fonts are quite expensive in  
>> terms
>> of resources, bandwidth and storage requirements. Having to download
>> hundreds of kilobytes (not to mention megabytes for many CJK Asian
>> fonts) isn't very kind on the user when they are on a slow network  
>> and
>> paying by the kb. Font subsetting can help of course. The WOFF format
>> handles subsetting well I believe, but is not widely supported on  
>> mobile
>> browsers, or even desktop browsers yet.
>
> Those are also very interesting information. how web fonts fits with  
> mobile are surely another topic of interest and discussion. It might  
> be worth developing a note on all these issues ? the blog post from  
> Richard would be a great first draft we could enrich with further  
> question identifying issues, challenges, resources, recommendation  
> etc.
>
> Is there interest within the group to proceed with such a work ?
>
> Steph
>>
>> [1] http://gs.statcounter.com/#mobile_browser-IN- 
>> monthly-200907-201008
>> [2] http://www.opera.com/developer/tools/
>>
>> On 6 Jul 2010, at 16:18, Shwetank Dixit wrote:
>>
>>> Hi Prashant,
>>>
>>> Let me investigate the issue...and I'll get back to you. Many mobile
>>> browsers, including Opera Mini/Mobile, support unicode and thus,
>>> regional fonts...but this is dependent on font support in the phone
>>> which many times limited.
>>>
>>> I'll get back to you on how to get it working with your sites, and
>>> could be something of use to add to this group as well.
>>>
>>> Cheers,
>>>
>>> On Tue, 06 Jul 2010 19:37:06 +0530, Prabhas Pokharel
>>> <prabhas@mobileactive.org> wrote:
>>>
>>>> Shwetank,
>>>> I got super excited when I heard that, but I just tried browsing
>>>> those sites
>>>> on Opera Mini 5 beta on my Android phone (Nexus One, running Froyo,
>>>> bought
>>>> in the US), and the unicode fonts at aajtak.com (or a few other  
>>>> sites I
>>>> tried) don't display properly. If it will after configuration, that
>>>> could be
>>>> huge!
>>>>
>>>> Let me know,
>>>> Prabhas
>>>>
>>>>
>>>> On Tue, Jul 6, 2010 at 6:16 AM, Shwetank Dixit  
>>>> <shwetankd@opera.com>
>>>> wrote:
>>>>
>>>>> This has at least *partly* have to do with the way developers
>>>>> include fonts
>>>>> in web pages. If the page uses Unicode, then at least in Opera  
>>>>> Mini and
>>>>> Mobile (and most probably, most other major mobile browsers),  
>>>>> there
>>>>> shouldn't be a problem (for example, aajtak.com uses unicode to  
>>>>> display
>>>>> hindi fonts, works fine)
>>>>>
>>>>> However, many people use .EOT fonts to display regional indian  
>>>>> language
>>>>> text in web pages. The problem with that is that .EOT fonts only
>>>>> work with
>>>>> Internet Explorer, and thus can't be displayed properly even by  
>>>>> other
>>>>> desktop browsers, let alone mobile ones.
>>>>>
>>>>> Prashant: I don't know whether the local indian language fonts on
>>>>> mobisitegalore use unicode or not. If it does, and it still  
>>>>> doesn't
>>>>> display
>>>>> properly on Opera Mini/mobile, then please let me know off-list,  
>>>>> and
>>>>> I'll
>>>>> investigate the issue further.
>>>>>
>>>>>
>>>>> On Tue, 06 Jul 2010 15:20:27 +0530, Stephane Boyera  
>>>>> <boyera@w3.org>
>>>>> wrote:
>>>>>
>>>>> Hi Prashanth,
>>>>>>
>>>>>> This is a very good point. I believe that there is not much
>>>>>> attention from
>>>>>> the I18N community on the support of languages in mobile browser.
>>>>>> This is
>>>>>> something that i've on my agenda. As a first step i discussed
>>>>>> recently with
>>>>>> a colleague and we will develop a doc around the key elements to
>>>>>> deliver
>>>>>> content in a specific languages.
>>>>>>
>>>>>> Steph
>>>>>>
>>>>>> Le 06/07/2010 04:42, Prashanth a écrit :
>>>>>>
>>>>>>> Dear Stephane,
>>>>>>>
>>>>>>> As far as India is concerned the biggest issue is rendering of  
>>>>>>> local
>>>>>>> language fonts in the mobile browsers. Note that I am not even
>>>>>>> pointing
>>>>>>> to user having to key in text in local languages, but very
>>>>>>> unfortunately
>>>>>>> the Indian fonts do not even get displayed in the mobile  
>>>>>>> browser.
>>>>>>> Today we have over 80,000 mobile websites created using our tool
>>>>>>> www.mobisitegalore.com and over 50% come from India, but there  
>>>>>>> is
>>>>>>> no way
>>>>>>> that they can created a mobile website in Hindi or Tamil  
>>>>>>> because the
>>>>>>> fonts would just not render in the mobile browser.
>>>>>>> This is an important missing link as most people in India  
>>>>>>> cannot read
>>>>>>> English.
>>>>>>>
>>>>>>> Warm regards
>>>>>>> S.Prashanth
>>>>>>> CEO
>>>>>>> www.akmin.com
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> On 05-Jul-2010 5:17 PM, Stephane Boyera wrote:
>>>>>>>
>>>>>>>> Dear All,
>>>>>>>> i want to share with you a very good paper from S. Gitau, J.  
>>>>>>>> Donner
>>>>>>>> and G. Marsden.
>>>>>>>> The paper is attached
>>>>>>>> References are:
>>>>>>>> Gitau, Shikoh, Marsden, Gary, & Donner, Jonathan. (2010). After
>>>>>>>> access
>>>>>>>> - Challenges facing mobile-only internet users in the  
>>>>>>>> developing
>>>>>>>> world. Proceedings of the 28th international conference on  
>>>>>>>> human
>>>>>>>> factors in computing systems (CHI 2010) (pp. 2603-2606). New
>>>>>>>> York: ACM.
>>>>>>>>
>>>>>>>> One of the first papers i see focusing on the barriers of using
>>>>>>>> mobile
>>>>>>>> web in developing countries by people without PC access and
>>>>>>>> experience.
>>>>>>>> Some good suggestions for operators and other players.
>>>>>>>>
>>>>>>>> Steph
>>>>>>>>
>>>>>>>
>>>>>>
>>>>>
>>>>> --
>>>>> Shwetank Dixit
>>>>> Web Evangelist,
>>>>> Site Compatibility / Developer Relations / Consumer Products Group
>>>>> Member - W3C Mobile Web for Social Development (MW4D) Group
>>>>> Member - Web Standards Project (WaSP) - International Liaison  
>>>>> Group
>>>>> Opera Software - www.opera.com
>>>>>
>>>>> Using Opera's revolutionary e-mail client: http://www.opera.com/mail/
>>>>>
>>>>>
>>>>>
>>>>
>>>>
>>>
>>>
>>> --
>>> Shwetank Dixit
>>> Web Evangelist,
>>> Site Compatibility / Developer Relations / Consumer Products Group
>>> Member - W3C Mobile Web for Social Development (MW4D) Group
>>> Member - Web Standards Project (WaSP) - International Liaison Group
>>> Opera Software - www.opera.com
>>>
>>> Using Opera's revolutionary e-mail client: http://www.opera.com/ 
>>> mail/
>>>
>>
>> David Storey
>>
>> Chief Web Opener / Product Manager, Opera Dragonfly
>> W3C WG: Mobile Web Best Practices / SVG Interest Group
>>
>> Opera Software ASA, Oslo, Norway
>> Mobile: +47 94 22 02 32 / E-Mail/XMPP: dstorey@opera.com / Twitter:  
>> dstorey
>>
>>
>
> -- 
> Stephane Boyera		stephane@w3.org
> W3C				+33 (0) 5 61 86 13 08
> BP 93				fax: +33 (0) 4 92 38 78 22
> F-06902 Sophia Antipolis Cedex,		
> France
>

David Storey

Chief Web Opener / Product Manager, Opera Dragonfly
W3C WG:  Mobile Web Best Practices / SVG Interest Group

Opera Software ASA, Oslo, Norway
Mobile: +47 94 22 02 32 / E-Mail/XMPP: dstorey@opera.com / Twitter:  
dstorey

Received on Thursday, 5 August 2010 13:47:15 UTC