My use of two displays is generally as follows.

Display #1 – Actual work being done.

Display #2 – Information about models / classes / etc… that I don’t get automatically from my text editor (Sublime Text 2)

The iPad through Air Display sucks for #1; the lag is real and irritating.  For #2, however, it shines.  I have it open to the appropriate page right now, and it probably helped me shave about 15-30 minutes of research off my daily routine.

Considering how rarely I want to change the content on screen #2, I wonder if using an e-ink screen could be plausible in the near future…

If users navigated to the newly updated PHPMyFAQ server by name, they layout was broken – but going to it by the IP address would have no such effect.  There was no weird javascript issue that differed between the two.

It turned out that the server name was listed as an Intranet site, and was being displayed in a different display mode than if one would navigate to it by the IP address alone.  Disabling intranet auto-detection resolved it, but I am not really sure WHY it happened.

However, the landscape for studying Kanji has changed almost completely from what it was when I was a student. Changed for the better.

During my student days, I studied Japanese by creating / listening to vocabulary tapes and by writing thousands of flash cards. I would shuffle them and test myself with them for hours in the library. I eventually wrote down every example in my textbook (“Yokoso”). While proving my determination, it was not particularly practical. Cards got stuck together, there were mistakes while writing them down, and it was not easy to keep track of which words or Kanji I was bad at.

There was one “hack” to learn Japanese; Heisig’s Remembering the Kanji [1]. Unfortunately, it was incomplete with more than half of the book requiring you to create your own mnemonics for the most complex Kanji in the book. Eventually this became impossible and eventually, sometime between making email penpals and moving to Japan, I stopped hand-writing Kanji.

I tried several ways to get back in the education groove, but I rarely found useful help. There is some Nintendo DS software that I tried (２５０万人の漢検), but it is worthless, using barbaric rote memorization and not maintaining any kind of useful user history.

So what is left? Quite a bit actually!

First, I got a private instructor, courtesy of the company I am working at. An hour of guaranteed tests and written homework analysis is priceless (or 4000 yen an hour). I also have 3 text books that I received (Nihongo So-Matome and a reading book).

Two, Anki and other SRS[2] systems are quite practical and helpful for reinforcement. They take care of the problem of repeating problematic words, and allow you to automate your flow. I personally am using StickyStudy [3] and adding a list of hiragana readings / kanji with a little script I put up on Github [4].

Three, there is Kanjidamage, a logical culmination and perfection of Remembering the Kanji. While I generally don’t link to sites (because it makes me feel like a big shill), I really recommend their mnemonics and overall method. Whereas I would spend hours puzzling over Heisig’s overly-serious mnemonics, the Kanji Damage ones are simple and absurd enough to be memorable. Moreover, it is not limited by being a book – it is wherever I have an internet connection.

My process goes like this.
-> Create a word list from the Kanji in the kanji book.
-> Run it through my python script and add it to Stickystudy
-> Go through the Kanji in browse mode, while writing down the Kanjidamage mnemonics for the ones I don’t remember.
-> Do the test mode in StickyStudy, while writing the Kanji on paper. If I make a mistake, review the mnemonics from Kanjidamage and go on.
-> Exam every week on Thursday.

I don’t know if this process will be successful. However, it feels a lot more effective than my previous forays, probably because it is doable ANYWHERE I can sit down.

[1] http://en.wikipedia.org/wiki/Remembering_the_Kanji
[2] http://en.wikipedia.org/wiki/Spaced_repetition
[3] http://www.stickystudy.com/
[4] https://gist.github.com/2730102

The ratings on the Japanese app store [1] are in the 3 star range. This is incredibly odd, because expensive apps (Mahjong Police was 10$ or 1200 yen) rarely have bad ratings. The commentators mainly complain that the game is impossible to beat, and that the AI acts counterproductively. For example, other detectives who are ostensibly on your side tend to do insane things to reduce your points, or find ways to attack you directly in the game (when they should logically not be doing so).

I needed help, and uncle Google was always so helpful when I had problems with old games. I found great information for everything from Star Trek 25th anniversary to Taito’s Groove Coaster. Obviously don’t expect Mahjong Police to have useful results in the English language; however, it was shocking that I found no useful results on the Japanese-language Google search. I found the official site, some useless affiliate sites that offer no useful information [2], and links to older versions of the game [3]. What’s worse is that the iTunes page did not show up, and the negative reviews are effectively censored. Instead, this somewhat-flawed game is portrayed as an electronic equivalent of rose-scented fair-trade gold that was mined by Starbucks baristas.

This is a symptom of a deeper problem; the “trusted web”.

I define the “trusted web” as being much like the “trusted media”; it is a site that an average Japanese user would not feel uncomfortable visiting (where comfort is proportional to positive portrayals in the “trusted media”). The value of this is not lost on Japanese entrepreneurs. “Trusted” bloggers shamelessly accept bribes[4] to write about products and the backlinks they generate are not indicators of truth, nor of interest, just the size of the bribe. The “trusted web” is really a set of treacherous chains that prevent its users from accessing genuine content (for any meaningful definition of genuine).

I have no doubt that there are interesting discussions about this (and other topics). I fear, however, that these are hidden in ghettos like 2ch and parsing the signal from the noise may not be that easy. Maybe the idea of crawling 4chan’s predecessor makes people a bit twitchy, but god damnit, I want to know how to play this game well, not how to buy it a second time.

Extracting truthful information or discussions from the Japanese web probably means finding indicators other than links. Until someone does a better job of crawling the real Japanese social web, the Japanese web will be the tool of marketers and not a serious source of information.

Please break their chains on your results Goog, they make you look bad.

(Which is to say, they keep me angry as I keep losing this game over and over again.)

1. http://itunes.apple.com/jp/app/mahjong-police/id293982887
2. http://www.affiliatejyuku.com/diary/archives/242
3. This might have been helpful but I did not find anything useful for older versions of the game either, sadly. This review of the GBA version was all I could find, but did not provide much in the way of help.
4. Without pointing any fingers, there are a few startups that I happen to know who used this to their advantage. You may choose to disbelieve me and ignore this blog.

A final note.

For what it is worth, I do think that Mahjong Police is a decent game. At least the story mode. (I can win the other two modes without much effort).

My guess is that the story mode is programmed to train you to recognize different achievable hands. Instead of going for a pure random approach, it gives you X paths to victory and then punishes you for not recognizing them. If that’s the case, it is brilliant, and playing it will no doubt help me while playing with friends and Yumiko’s father. (I can’t understand what else it could be becuase, joking aside, I’m not _that_ bad at Mahjong).

However, with my mother on the phone, angrily demanding my 8 gigs of wedding pictures, I realized that twiddling my thumbs wasn’t going to help much.

Since I sort my photos by directory, it seemed that taking advantage of that order would be the most effective way to go about uploading my pictures. Thus Flickr CLI was born.

(Yeah, I know there is another Flickr CLI on Lifehacker, but it hasn’t been updated in years and doesn’t work with Flickr’s new absurd login methodology).

You can find it here:

Flickr CLI in Python

Download the source code (not the mac installer, it has the same problem) and run the following:

./configure --disable-perl
make
make install


My python bindings ended up in /usr/local/lib/graphviz/python.

Japanese comments / text all mangled while trying to get past a text injection vulnerability for which there is no update available.

I am pretty upset about it right now, mostly because I had to spend an hour and a half fixing this. Not sure what I will do, but migrating to blogger would sound almost worth it (if Google hadn’t been offing its services one at a time).

Not sure what I’m going to do. I did manage to save the text for the most obviously important Japanese language article I wrote, and I do have backups here and there. I’m an idiot I guess.

(I tried installing through Brew but it didn’t work.  Might want to try that if your install fails)