I have a fun time reading various predictions for 2007, e.g. Wired, Wireless, Funny, Google and Information Week.

Some sample includes:

– Internet Traffic Doubles to 5,000 petabits per day by the end of 2007. And 80 percent of it is peer-to-peer file sharing, mostly Skype video and BitTorrent.

– Google GDrive will finally be launched

– Navigation becomes rather more important on mobiles. Mobile search doesn’t.

Some of the funny ones includes

– RFID and Web Services by Information Week [JS: hahahahahahahaha….thats so 2004]

– Nokia releases the mxx2115 cell phone incorporating an MP3 player, video games, digital camera, digital camcorder, text messaging, calendar, GPS, calculator, satellite radio, e-mail, downloadable video clips, web browser, voice recorder and digital thermometer before realizing that it neglected to include the ability to make phone calls. [Hint: Nokia 770]

I am tempted to add my own set of predictions but nay, I leave it for another day.

Happy New Year

As we start on a new year, lets look at the 2006 IPv4 Address Use Report

The current (jan 1st, 2007) figure for 2005 is 175.52 million addresses. Together with adjustments for earlier years, this brings the total addresses available to almost exactly 1.3 billion, down from 1468.61 million a year ago. This is out of 3706.65 million usable IPv4 addresses, so 2407.11 million addresses are currently given out to either end-users or Internet Service Providers.

Lets also put a stop the myth that “MIT has more IP addresses of whole of China”, something that is no longer true for a couple of years. Yet sadly, some still chant it, as recently as a couple of weeks ago by a NUS lecturer that I immediately put a stop to.

China is the 4th largest IP holder now after US, JP and EU.

We are slowly recovering from the disaster submarine fiber cut due to the earthquake in Taiwan. It is definitely one of the worst I remember…the last time a fiber cut affect us that badly is in 1994.

First, lets clarify a few misconceptions: submarine cables are not under the seabed. It is just lay across the sea. It is only when it crosses major shipping lines and fishing zones (mostly near the coast) that it is lay (5 to 10m below) under the seabed. Therefore, it is not unusual for submarine cable to be cut, mostly due to deepsea fishing.

So when one submarine cable is severed, the operator generally have “restoration”, either on their own (self-healing cables) or using their competitors cables. This minimized the disruption. What is unusual is that this earthquake in Taiwan sever all the major cables (see above for the fibers near Taiwan).

As far as reports goes, SeaMeWe-3, ACPN2, C2C and EAC (ANC) (link) are severed. In other words, all the major submarine cable at the same time. Under other circumstance, when SeaMeWe-3 cuts, we can fall back to APCN2 or C2C or EAC. But when all of them are cut at the sametime, there is really nothing to fall back upon.
Read the rest of this entry »

Saw this on slashdot today: Chinese student arrested after making 150 million yen selling items for online RPG

Wang Yue Si, 23, came to Japan on a student visa in April 2004. He started selling items such as weapons and currency for online games through an Internet auction site in April this year, without obtaining the appropriate residency status…

Police suspect that Wang has sold a total of 150 million yen in virtual items and sent more than 100 million yen to China. (Mainichi)

The title focus on the student being deported after making US$1.3M. I think the real story is the fact that someone can actually made US$1.3M buying and selling virtual weapon and currency in online games.

Incidently, Blizzard step up its effort to ban farmers and sellers in the world of warcraft recently. A friend also informed me that Chinese game-farms are also seeing lower yield from their investment; an average account gets banned in a month as compared to 3-6 months before. This means they have to recreate an account and re-level a new character to level 60, wasting 10-12 game time.

The result is that the price of the in-game currency jumped nearly four fold, from US$50-60 per 1,000 gold to over US$200 per 1,000 gold now. The virtual black market is exactly like real black market, increase in policing only increase the black market price but never kill it.

I related this to Joi in the game over the weekend and he replied “Blizzard should just “legalized” it”. I agree.

It is a strange morning to wake up and to read the following two news:
– Lack Of telecom infrastructure drives WiMAX adoption in AP; and
– WiMAX Will Not Dominate in the Short-Term, Analysts Say.

Arent analysts fun? One say “Yes yes, WiMAX will grow in AP” and the other say “No no, WiMAX wont happened until 2010”. If you are confused, fear not, I will add to the confusion and say “Both are right!” :-)

There are several business case for Mobile WiMAX. Broadly speaking, (1) ADSL/Cable replacement and (2) personal mobile broadband.

When most people think of broadband in Asia Pacific, they generally look at Japan, Korea & Singapore and then forgot that for the rest of the developing countries in Asia Pacific, most if not all are struggling with broadband pentration. Therefore, as an ADSL/Cable replacement or competitor, Mobile WiMAX is a pretty good alternative.

But when you look at the more advance countries where ADSL is cheap and in abundance, there is a trend that users are using internet where-ever they are, on the train, on bus etc. Japan definitely leads the thrend here and in this personal mobile broadband game, the short time is likely to be win by the mobile alternatives like HSPA, Ev-Do etc.

However, most of the mobile operator are still voice centric – most of the 3G operators are on R5 which is still a voice/data hybrid network, leaning towards voice more. Therefore, as personal mobile broadband subscriptions grows, it is not certain the existing mobile technology is going to support a large number of subscribers since it was never planned as such. So while the short term personal mobile broadband looks bad for Mobile WiMAX, the long term is a different story*.

So I guess it depends which market you are after and in which country. This is why I am not so gungho on Mobile WiMAX in countries like Japan, Korea, Singapore but spending most of my time in Malaysia and Indonesia instead.

*My personal take however is that mobile technology is likely going to evolve and embrace OFDMA. It is going to be more WiMAX-like which will compete with the pure Mobile WiMAX we see today.

A couple of news caught my attention today. With the impeding Microsoft IE 7 support of IDN, the industry has renewed its interest in IDN. Already, CircleID is buzzing with articles on IDNs in the last couple of weeks and various commenters noted that IDN registrations is expecting to go up. Look no further, it is going up. My visit to CNNIC last month confirmed they have increased the Chinese Domain Names registration by a significant portion to 300k registrations. At a prenium price (20 USD a name), Chinese Domain Names is generating as much revenue for CNNIC as the English domain names (a lot more but a lot cheaper too).

But what catch my attention is this wonder article written by Geoff Hutson on Internationalization of Internet. It is definitely a great summary covering various events surrounding IDNs.

Just two note:

1. On DNAME, I wasnt a big fan of using DNAME. I think ICANN is misguided that the DNS infrastructure cannot support too many TLDs and that using DNAME adds a level of complexity than it needs to be.

A label is a label, does not matter if it is on the 2nd level or the top level. If we dont use DNAME on 2nd level, I dont see why we need to use DNAME on the top level.

But I wont stand in the way for ICANN to experiment with DNAME IDN TLDs. Any little step forward is a step forward.

2. The article didnt mention RFC 4690 : Review and Recommendations for Internationalized Domain Names (IDNs). It covers a great deal of the technical complexity of IDNs, what we know works and some of the potential technical pitfalls we are concerned with in the existing system.

I wrote to Patrik and John a few weeks ago and here is what I said:

I agreed with issues raised in the doc, many of them are already
well-known before the original set of IDN RFCs was publish.

However, I think it would make sense for the system to implement a bit
longer to see whether the problem mentioned is real or just a
theortical possibility. It might also hurt the IDN progress if IETF
undertakes a review at this moment in time, as more likely then not,
implementors will then wait for the completion of the review before
further implementation gets done.

Other the minor disagreement with the timing, I applaud the work done
on RFC 4960 :-)

Patrik replied to say they have taken the timing into consideration so I left it as that.

A few months ago, a friend (who requested to rename nameless) forward me a website called M2Z. It was already covered by GigaOM, and Tom Evslin.

It was a application for 20Mhz of spectrum on the 2.1Ghz (2155-2175Mhz), nation-wide to roll out a free high speed broadband wireless access in United State (see their application). Before you write this off as another wannabe, this is brain child of Milo Medin, founder of @Home in 1995, and John Muleta, the former head of FCC’s Wireless Telecommunication Bureau. The backers include the top names on Sandhill Road, Kleiner Perkins, Caufield & Byers, Charles River Ventures and Redpoint Ventures.

Their goal is to provide 512kbps wireless broadband to 33% of the US population by 3rd year, 66% in the 5th year and 95% by the 10th year. Looks like they are giving ClearWire the competition they need.

But an email conversation with the unnamed friend recently remind me about this proposal as well as some of the questions I had.

1. Why 2.1Ghz? I can understand that 2.3Ghz and 2.5Ghz is practically gone in United State but still, 2.1Ghz is reserved for 3G expansion for many countries. I think they are going for a uphill battle.

2. 20Mhz, is probably just sufficient for a decent wireless broadband roll out. Some would even argue we need 40Mhz to be safe, using 10Mhz channels on n=3 configuration with 1 channel on reserved.

3. I suppose 512kbps is probably the reasonable speed they can do on 5Mhz channel (given they are asking for only 20Mhz). But could you possibility imaging we still using 512kbps 10 year from now? But then again, having broadband is better than no broadband at all for many parts of US so maybe they have a case.

Anyway, good luck to M2Z. I hope they got what they asked for (or better, more than what they ask :-)

ps: I am willing to bet they have a higher chance after Bush leave office. :-)

Two days ago, the Indonesian regulator issued an interesting whitepaper: Penataan Spektrum Frekuensi Radio Layanan Akses Pita Lebar Berbasis Nirkabel (Broadband Wireless Access/BWA). You can download the PDF here but it is all in Bahasa Indonesia.

The whitepaper clarify a lot of confusion over the spectrum plan in Indonesia (from 1.5Ghz to 10.5Ghz) as well as the next step for different blocks as well as the timeline. More important, it also describe the procedure for the pending 2.3Ghz allocation in Feb 2007 (See Section 5, Seleksi Pita Frekuensi BWA 2.3 GHz). Particularly of interest is this paragraph

Kriteria penyelenggara yang diperbolehkan mengikuti seleksi, dibagi atas 3 (tiga) kategori, yaitu KATEGORI A, KATEGORI B dan KATEGORI C,
yaitu:

Kategori A, penyelenggara jaringan telekomunikasi yang memiliki
infrastruktur telekomunikasi serta telah memiliki penomoran pelanggan dan fasilitas interkoneksi, yaitu …

Salah satu prasyarat utama untuk pemenang lelang pada kategori A adalah dikenakan kewajiban untuk memberikan fasilitas esensial jaringan seperti tower maupun fasilitas esensial lainnya kepada pemenang seleksi kategori B dan kategori C.

I am not well versed in Bahasa Indonesia but basically this is what it says. 2.3Ghz will be allocate in 3 category. Category A is basically refers to the incumbent operators. (Category B is other existing licensee and Category are new entrants). And get this, the winner of Category A (ie incumbent) is mandated to share towers, facilities etc with Category B and C winners!

Not only it even out the allocation so that it is not dominated by the incumbent, it basically comes with infrastructure sharing :-)

A few days ago, IDA called for bids to operate the centralized number portability database for Singapore.

The Infocomm Development Authority of Singapore (IDA) issued the Request for Proposal on the Centralised Number Portability Database Solution (RFP) to invite interested ICT companies to submit proposals for the provision and operation of the Centralised Number Portability Database in Singapore. The RFP arose from IDA’s 2 August 2006 announcement to implement True Number Portability in Singapore from fourth quarter of 2007.

Wasn’t Singapore one of the first country in Asia to have number portability? Well, as I noted before: “What we have in Singapore now is “Call Forwarding” and we are trying to move to “Onward Routing” or “All Call Query”. Both will give us true number portability (ie, the Caller ID will match your number)”

Incidently, a friend called me up a few weeks ago as he was assigned by his company to look at this thing called “number port that government is calling for bid”.

Friend: Hey, you know anything about number portability?

Me: Erm, yea. Kind of…what can I do to help?

Friend: So hur, who is most likely to win?

Me: Well, Neustar has put a lot of effort in this area for a while so they are in a good position whereas I am not aware of any Telecordia activities.

Friend: So should my company partner Neustar?

Me: Sure why not? If you can convience them to partner, good for you!

Friend: But but..why would they partner with us?

Me: Thats for you to figure out right?

Friend: Shouldnt government make it such that no foreign companies can bid for it?

Me: Huh? Why?

Friend: So they have a reason to partner with us.

Me: … (WTF!)

My dear friend, Singapore government no longer works in that way…not anymore in the Telecom sector.

I am a fan of Gmail as well as the Gmail mobile. I use it extensively to check my mails on my mobile phone. The mobile web portal Gmail has is clean and fast and I like it alot. I also use the POP feature of Gmail to retrieve the email onto my mobile phone Messaging. While not fantastic, it is okay for an offline system.

But nothing beats what I found yesterday, a Gmail mobile application by Google. Just point your mobile web browswer to http://gmail.com/app and it will auto-detect your phone and ask you to install the relevant application onto the phone as below.

And it is really really cool! It has all the shortkeys you expect. Press “8” twice to mark the email as spam, “9” to archive it, “4” to reply, “5” to reply all and “1” to search etc. But I have a 3 wish lists

1. The application is not tolerant to unstable mobile connections, which is pretty common until the mobile data quality is improved.

2. All editing are done in the native edit pad provided by Symbian on a horrible black on blue interface. I am sure they can do better

3. Please please do it for Google Calendar soon!