> I'd like to propose a project about Input Method for
> OpenSolaris.
>
> The scope of this project would be:
>
> 1. To support IIIMF (which is currently the defualt
> IM framework on Solaris) on OpenSolaris, including
> building/packaging, bug fixing, and adding new
> features.
>
> 2. To support more input method frameworks on
> OpenSolaris, such as SCIM and UIM, both are popular
> IM frameworks on various Linux distributions. And to
> monitor new IM technologies on *IX systems, such as
> IM-Bus (which would be the next generation unified IM
> framework).
>
> 3. To host intelligent input method implementations,
> such as SunPinyin (developed by Sun Asian G11N
> Center, shipped since Solaris 10, and would be
> opensource'd to OS.o), which is a SLM (Statistical
> Language Model) based IME.
>
> This project could provide more choices to customers,
> and attract more developers, especially in Asian
> areas. Thus help the adoption of OpenSolaris in
> Asia.
>
> This project will be endorsed by Internalization
> Community, the leaders initially would be Fuyuki
> Hasegawa, Naoyuki Ishimura, and myself (all of us are
> Sun employees) . We will invite more IM developers
> around the world, as the lead engineers.
>
> Any suggestions and comments are welcome!

Hi Yung (& Ienup),

As we discussed in private mails, I have assigned a high school intern to try 
to port UIM-pinyin-big5 to Solaris. However, after an in-depth research into 
the Chinese-specific localization infrastructure, he convinced us to at least 
temporarily switch back to SuSE. (This kid has been accepted to the accelerated 
BA/MD program at Northwestern. No doubt that his uniquely qualified 
Solaris-related experience that was prominently discussed in his CV helped 
quite a bit; this is a deeply sad conclusion for him--no less for us.)

Unlike other features which we can always manage to improvise (& we have), the 
"input method" is something, like the keyboard itself, that we have to use 
everyday, & there is no substitute. I am also surprised to discover how the 
Chinese-related input method mechanisms that are implemented in Linux have 
improved so much within such a short period of time. As far as SuSE is 
concerned, of course the fact that the president of Novell-Taiwan himself also 
uses SuSE (on his notebook) plays an undeniable role. I remember complaining to 
him, after playing with the final release candidate of SuSE 10.1, why 
UIM-pinyin-big5 was dropped. Lo and behold, it was put back in the official 
release (after the beta test period has long closed).

The fact that Solaris Express sticks with IIIMF--to the exclusion of all 
others--is also a key reason why we decided not to pursue any possible patch-up 
solutions. As I mentioned to you in my private mail, it is my impression that 
IIIMF, at least the Chinese character input portion thereof, is practically six 
feet deep if not with Sun's support. While IIIMF used to be the only show in 
town, the field of Chinese input is now booming with more polished methods such 
as SCIM, OXIM, GCIN, etc. From our own experience, we have found that SCIM is 
definitely more stable and much more responsive than the IIIMF implemented in 
Solaris Express. Even Red Hat, once the other biggest supporter of IIIMF, has 
switched to SCIM.

With regard to SunPinYin, I feel that the NewPinYin is doing at least as a good 
job if not better. There is probably no reason to re-invent the wheel. Since 
the Chinese input method in Solaris is so far behind from the industry leaders 
(believe me, it pains me greatly to say these words), the most important order 
of business at the present time, if I can be excused for making such an 
irrespectful statement, is to duplicate what is already available. Don't try to 
do anything exotic. Once Solaris is brought back to the same leveled footing, 
the on-screen keypad that Solaris already has will make Solaris stand head and 
shoulders above everything else. And let's not forget about the Solaris kernel 
(yes, the kernel), and the fact that, thanx to Sun's xwindows team, a Solaris 
user can install the nVidia driver without even missing a breath).  Etc., etc.

So, we will be taking a hiatus from Solaris for a while. I definitely hope this 
will not last long. In revamping the Chinese input method(s) for Solaris, I 
would suggest keeping one potential "customer" in mind: Dr. David Yen. Dr. Yen 
and I went to school together at National Taiwan University more than 35 years 
ago. He has, and will have more, a circle of very influential friends in China 
and Taiwan who are key players in the fields of computers. communications and 
other consumer electronics. If you can develop a Solaris 
desktop/notebook/whathaveyou system that will make him antsy to show it to his 
friends (like Steve Jobs demoed his iPod/iPhone), then you know you have got 
the job done.

I have confidence that the current Chinese localization downside will be timely 
solved. A couple of days ago, Sun's Chairman was talking about merging the odf 
and (China's own) uof file formats. While this endeavor may take a while to 
precipitate, one area that may have an immediate impact, at least as far as 
Solaris desktops are concerned, is the inclusion of the (China's) AVS codec. 
Codecs are always a big pain for Linux users. Once the Chinese locale problems 
are solved, inclusion of the AVS codec in Solaris will very likely generate 
huge interests.
 
 
This message posted from opensolaris.org
_______________________________________________
opensolaris-discuss mailing list
[email protected]

Reply via email to