Many Taiwanese users, and I believe many Hong Kong users too, will wonder how iPad could be a useful device at all. Apple states on its tech specs that the initially supported languages include English, French, German, Japanese, Dutch, Italian, Spanish, Simplified Chinese and Russian. There is no Traditional Chinese.
Language support is a topic on its own in modern operating system. It is mainly about three things:
Text display and localization (font, layout support, system and application messages)
Natural language processing (text correction, text-to-speech, voice recognition, information discovery, indexing)
When an Apple device supports a language, it’s mostly full support of 1. and 2., and varying degree of 3. On Mac OS X 10.6, for example, Traditional Chinese is supported with fonts, localized menus and messages. OS X also has a few Traditional Chinese input methods. It doesn’t have spelling checker for the language, but it has some natural language processing capability on which the input methods and system-wide indexing (Spotlight) rely.
It’s curious why iPad does not come with Traditional Chinese support in its current version while it does Simplified Chinese. Perhaps it’s not yet fully localized, or perhaps the input methods are not ready yet. Curious, then—since Simplified Chinese is there, and Apple can just use the same code base that is used in iPhone and iPod touch. But on the other hand, iPhone did not come with appropriate Traditional and Simplified Chinese until OS 2.0, and a lot of other languages now available on iPhone (Korean and Hebrew, to name just two) are not supported on iPad yet.
Psychologically, however, this gives iPad a bad impression to Traditional Chinese users. Many American companies made this terrible mistake assuming that Traditional Chinese can make do with Simplified Chinese interface. It’s true that most of us can read the latter without any problem, just like an American English user can read the original British edition of Harry Potter with hardly any difficulty. But being able to read is different from wanting to read. Simplified Chinese has a vastly different vocabulary, especially in computer terms. Font and rendering preferences differ, too.
I sincerely hope Apple does not think that way, although Apple’s track record is not stellar. It messed up its new Traditional Chinese font, Hei TC, that practically renders the font unusable in design and publishing. iPhone is also shipped with that defective font. You can bet that many of us do not feel confident in Apple’s way of doing things, let alone its perceived priority, when it comes to Traditional Chinese.
Even if the next version of iPad’s OS supports Traditional Chinese (and hopefully with the Hei TC problem repaired), a bigger question looms: Which input methods will it include? The Traditional Chinese landscape is not like the Simplified Chinese one, where Pinyin is taught in school and covers 95% of the user base. In Taiwan, pupils are taught the Bopomofo phonetic system, a kana-like set of symbols that represent Mandarin Chinese sounds. About 80% of users in Taiwan use that. There are a number of other input methods. In Hong Kong, where (so I’ve been told) only college students take elective Cantonese romanization classes, component-based input methods like Cangjei is more popular.
The iPhone has three Traditional Chinese input methods: handwriting, Pinyin and Bopomofo. All three are finely implemented but not outstanding to say the least. The problem being this: You don’t write much on the iPhone. You tweet, send SMS, keep some notes, compose one-liner emails, and that’s pretty much all. I can live with slow input methods on the iPhone. But imagine using that with Pages on iPad? It’s going to be like having your fingers chopped.
The biggest problem, though, is that there will be no alternative to iPad’s built-in input methods even if comes with what iPhone has now. If you aren’t satisfied with Mac’s input methods, you can install a third party package. On iPhone OS there is no such thing. Yes, there are jailbreak input methods, and I know many people who jailbreak solely for the damned input method, but jailbreak is never mainstream and not a reliable way to solve a problem that Apple should tackle.
When I say “a problem that Apple should tackle”, I don’t mean that Apple should try to solve everything. It excels in the fields it knows the best, but it has weakness in others. As much as Apple does not make every app on the App Store, it should let other developers solve the problem for the platform.
I can see why Apple might not love to open up input method on iPhone. I’ve actually had discussions with many people in this field since late 2007. Text input component can exercise a lot of control over the system, especially if it lives in your app’s address space. There are inherent security problems too.
But, once again, Mac OS X has shown us that the problem can be overcome, and even iPhone OS has a solution for providing a limited kind of third-party service. Since Mac OS X 10.5, the new input method architecture is based on Objective-C Distributed Objects (DO), which is a high-level mechanism of inter-process communication. Input method modules are no longer attached as a loaded plug-in and can only receive what the OS allows them to know. Suppose there is a kind of “Settings” bundle that functions like that. Mac OS X also employs some watchdog mechanism (which is vastly improved in 10.6) to ensure that those special processes do not block or crash the system, and a similar kind of watchdog can be put into iPhone OS to make sure that input method modules behave.
The question is whether Apple is willing to do it, or does it really think, like many multinationals, that Traditional Chinese users can make do.
I currently run my own company (as a corporation as defined in Taiwan business code) and have experience of having started a partnership (later helped its conversion to corporation). All small businesses (or “micro” businesses if using EU definition).
One thing the experiences have taught me is to understand the significance of operating as a business entity—that you can work with people as an organization to provide your services to the market. There are several importance differences between providing your service as an entity and doing that as an individual, even for many small businesses the content of the work is much the same. That you’re able to handle risk better as a corporation (which is also one of the reasons why corporation is one of the most important, yet often undervalued, organizational innovations in modern history) and that you’re able to provide service to both other entities and other people and issue invoices (compared to individual work-for-hire contracts, one at a time) are just two advantages among many.
Setting up and running your own company, however, always involve more work than just working as an individual. And that’s how I come to appreciate the friendly business environment in Taiwan. It still has lots that need improving, but the quality of a system is always a comparative matter. Red tapes exist everywhere. In that aspect, there is hardly any hidden cost in setting up a company—for example there’s no favor you need to ask for. I’ve learned that it’s not so in many other places in the world.
Many of us live in countries where business freedom is taken as granted, just like other types of freedom we enjoy. In fact it’s not necessarily a given. Interventionist measures, protectionist regulations, restrictive zoning laws, complicated tax codes, obscure accounting rules and filing requirements, lack of flexible payment gateway, insufficient intellectual property protection—they can stifle many aspiring companies. It’s not a cliché to say that we want to cherish what we have and work to make it better. We also need to understand the importance of a good supporting structure and be wary of changes that could discourage entrepreneurism.
In Taiwan’s case, I think the supporting structure had been strong and helpful when ours was more a manufacturing- and export-oriented economy. Doing service on an international scale and providing service beyond Mandarin Chinese-speaking markets are not our strongest thing—yet. I met a Swiss company owner at WWDC ‘09, and he could use a representative office in San Francisco set up by some Swiss government organ that aims to help specifically software startups get literally a foothold in the very competitive Bay Area. As I don’t believe in big government that does everything (mostly badly), a TAITRA equivalent in service sector might help business owners in Taiwan better. And that is one thing that I as a software company owner care about.
I don’t watch TV, and I don’t have access to services like Hulu, but it’s hard not to run into tweets on the latest Leno-Conan spat. With the help of Taiwan’s NMA News (動新聞), anyone can understand what has been going on in well under 2 minutes. (Via @kerim)
Yes, the same animated news that featured Tiger Woods.
Disruptive journalism technology at work. In some unexpected ways…
For all the advantages of Mac OS X over Windows, there is one thing that my IT friends always sneer at: That OS X has the inferior remote desktop protocol.
Apple makes remote management quite easy. Its Apple Remote Desktop, although expensive, has great user interface and is indispensable if you manage over a bunch of machines. Since OS X 10.5, it has become even easier and cheaper if you just need a quick access to another box—the screen sharing feature is built-in gratis.
This being said, OS X’s core performance of remote desktop management itself has a bad reputation. If you compare it to that of Windows, OS X’s remote desktop is slow. If you use Microsoft’s own client software to talk to a remote machine, it uses its proprietary protocol and only requires essential drawing commands. This is why the remote desktop is stripped of any theme when drawn in the client side (and many people like it). OS X, on the other hand, builds on the open VNC protocol and transmits lots of raw pixel information. That OS X has far more visual effects and its UI is more sophisticated is of no help. It seems that when OS X’s window system was designed, remote desktop was never taken as a high priority, even if the window system is also client-server based, like most modern desktop GUI systems are.
Another interesting consequence when all those design decisions got lumped together is that you can never have more than one person using the desktop remotely. That is, if someone is logged into the local desktop console, it is impossible for you to login remotely and have your own desktop served to you. Programmatically it has also limited the usefulness of OS X as a graphics server, as window-backed drawing processes must either be under an actively logged-in user (and that has to be in the local desktop console) or root, which is highly undesirable if your graphics program powers some web service. My experiences show that Quartz, when being used with higher-level abstraction layers like Cocoa, is far superior to any graphics package out there (e.g. ImageMagick), especially with its supported range of formats (especially PDF) and excellent graphics model, but that graphics server permission thing limits its usefulness.
On the other hand, the decline of the desktop—if we can term the trend as such—is evident in the way that not much has been changed in the basic design of modern window-based GUI, and remote desktop is among the laggards. And when laptops are so cheap (but bandwidth still a problem), it’s easier to build a service architecture around client-rendered web pages and server-provided web services. Serving desktops remotely is out.
Tumblr’s reblogging design says a lot of its peculiarity.
I find it fun reading my Tumblr dashboard everyday, in an age where blog seems done and RSS in decline—established media and writers make it a norm, whereas the rest of us have moved on to “microblogging”. I find truly insightful writings and excellent blogs on Tumblr. It has a light mix of social network, like the elements of following, but only a very light mix. It’s largely anonymous, like a masquerade ball: Many of its default themes don’t even show your own intro, and that is if you bother to fill in that info at all.
Then there’s this no-comment thingy. Not that you can’t do it, as there are definitely ways to extend, but it’s just not there as a built-in. Instead, reblogging is made insanely easy. Almost too easy that it could have become a retweet soup that deluged much of the Twitter scene, but apparently it doesn’t.
I’d say that reblogging to Tumblr is what comments have been to “traditional” blogging. By allowing comments, each blog entry becomes a forum and a point of attention on its own. Tumblr, on the other hand, makes easy adding your own words when reblogging. And this is important: Because it’s you, the one who reblogs, that is the center of the game, and you get a voice of your own. At the same time the original author also knows your existence. It is unlike the entry-comment design, in which comment is relegated to a secondary, albeit supportive, role. I’ve followed a number of interesting tumblogs by following the reblogging history of an interesting post.
A judgment of plagiarism requires that the copying […] induce reliance by [the intended readers]. By this I mean that the reader does something because he thinks the plagiarizing work original that he would not have done had he known the truth. […] If [the reader is] a teacher he gives a bad student a good grade, to the prejudice of other students in the class (if the students are graded on a curve), thinking the student’s paper original.
I’m citing this passage because I believe many of us, in our early student days, had this experience: you wrote something that the rest of the class copied, but you got a mediocre grade, if not the worst.
Many years ago I helped someone write her Fortran 77 homework (which was taught in her department of some engineering school, using g77), and many of her classmates copied her homework, some by copying the built executable, some by copying the printout of the source code, which carried errors. Surprisingly all those copiers got better grades than she did (some 90s, some 80s, out of 100, and she only got 70 something).
It was a very comical experience, not least because I was the ghostwriter. I tend to think I never became proficient in Fortran (or decided not to be) because of that episode.
It was the first time any country had created an abstract logo to brand itself but today everybody thinks that all you need to do to rebrand a country is make a logo.
I think that’s out of date. Logos are old fashioned. These days, you can communicate almost instantly with the world. And the internet means that message has got to be verbal. How many logos do you ever notice when you’re surfing online? If I were in charge of making a new image for Spain now, or a new nation such as Montenegro for that matter, I would not be thinking about a logo. All you need is to be clear about what you’ve got to say. In this age of mass communication in English all you need are a few carefully chosen words.
His comment is definitely worth reading as he also talks about how Spain Tourism’s logo was created by none other than Joan Miró.
