81Dojo Kifu: [kinsoko(JPN, 2-Dan) vs wacpre(JPN, 5-kyu)], Opposition, Black’s 3rd-file Rook.

I did not know how to attack and did nothing in this game.
My bad points are

• 40: I should drop pawn to 93 not pushing pawn to 74.
  
• 54: I should drop bishop to 32 not to 63.
  

This was a first game in 2012 and I want to continue playing shogi this year, even if I was not a good player. I will try to study opening strategy again (and again).

81Dojo Kifu: [wacpre(JPN, 5-kyu) vs fu4sacrifice(THA, 1-Dan)], Free-style.

Here is my bad points.

• At 35; pushing pawn to 24 (P-24) would be better than moving rook to 58 (R-58).
  
• At 63; dropping pawn to 22 (P*22) might be good, not pushing pawn on 6th-file (P-65).
  
• At 67; dropping rook to 77 (R*77) could be better than moving gold to 66 (G-66).
  

I could not deal with white’s central rook attack. As play going on, black (my side) was getting worse and worse. I can not use pawn well in this game.

Playing shogi constantly is a good way to become stronger. (I know, I know…) I would like to play one more game in this Holiday season.

I finally get a tendency; it’s “Opposition Seeker”!

Too lots of “Free Style”? I think I have to study how to attack and deal with it.

I found that you can paste any diagrams from 81Dojo easily: push “Share Kifu” button on the right side and check “Diagram (on cloud service by fantakeshi)”. You can choose a way to share it. From Cloud Shogi Diagram Generator, you can create diagram manually.

81Dojo Kifu: [luxtorpeda(POL, 9-kyu) vs wacpre(JPN, 6-kyu)], Free-style.

Here is my bad points. Diagram shows the better move (a highlight piece on the board).

• 24 move; Maybe I should abandon my silver and take black’s knight by promoted pawn (Tx89) than taking pawn by silver (Sx34).
  
• 34 move; Taking gold by dragon (Dx46) might better than taking pawn (Px24).
  
• 42 move; Checking by bishop to 15 (B-15) is better than dragon runs (D-46), but I had no confidence to mate from here.
  

Is it easy to see a game review with diagrams or without it? With them it easy to think, but it becomes longer note than without them and it might hard to read. umm…

I think from now on I could play fewer games than ever, but I go on play Shogi and write such kind of short reviews to become more stronger.

参考にしたのは、以下のサイト様：
気の向くままに ～備忘録@wiki～ – Ubuntu Linux 11.10
transitive.info – プログラミング用フォント Ricty を使う

やり方はプログラミング用フォント Rictyにすべて書かれているので、以下はメモ書き程度のものだと。

まずは、必要なものをダウンロード。
現時点での最新版のRicty（Ricty 3.1.2）と、ダウンロード : M+とIPAの合成フォントからMigMix 1Mをダウンロード。
まずは、MigMix 1Mをインストール。最後のfc-listでちゃんとinstallされたかを確認する。

$ unzip migmix-1m-20111002.zip
$ mkdir MigMix
$ cp -v migmix-*/*.ttf MigMix/
$ sudo cp -rfv MigMix /usr/share/fonts/truetype/
$ sudo fc-cache -fv
$ fc-list | grep MigMix

次に、FontFrogeとInconsolataをinstallし、その後Rictyをinstallする。

$ sudo aptitude install fontforge ttf-inconsolata
$ tar zxf Ricty-3.1.2.tar.gz
$ cd Ricty
$ sudo cp -rfv /usr/share/fonts/truetype/ttf-inconsolata/Inconsolata.otf .
$ sudo cp -rfv /usr/share/fonts/truetype/Migu/migu-1m-*.ttf .
$ sh ricty_generator.sh Inconsolata.otf migu-1m-regular.ttf migu-1m-bold.ttf （時間がかかる）
$ mkdir Ricty
$ mkdir RictyDiscord
$ cp -v Ricty-* Ricty
$ cp -v RictyDiscord* RictyDiscord
$ sudo cp -rfv Ricty RictyDiscord /usr/share/fonts/truetype/
$ sudo fc-cache -v -f
$ fc-list | grep Ricty

最後に、.mlterm/aafontに以下を記述。

ISO8859_1 = Ricty-iso10646-1:100;
ISO8859_1_BOLD = Ricty-iso10646-1:100;
ISO10646_UCS4_1 = Ricty-iso10646-1:100;
ISO10646_UCS4_1_BOLD = Ricty-iso10646-1:100;
JISX0201_ROMAN = Ricty-iso10646-1:100;
JISX0201_ROMAN_BOLD = Ricty-iso10646-1:100;
JISX0201_KATA = Ricty-iso10646-1:100;
JISX0201_KATA_BOLD = Ricty-iso10646-1:100;
JISX0208_1983 = Ricty-iso10646-1:100;
JISX0208_1983_BOLD = Ricty-iso10646-1:100;

これで随分と見やすくなった。よしよし。

With #Ubuntu, you can do everything you can do with other operating systems. But faster. And for free! ubuntu.com/download/ubunt…—
  (@wacpre) October 19, 2011

まずは、上記のURLからwubi.exeを取ってくる。install時のsnapshotはとり忘れたが、上のページに書いてあるとおりにすれば問題無し（Ubuntuに16GBを割り当てた）。ダウンロードに約１０分、インストールに10分弱。その後再起動したら、WindowsのMBRが立ち上がり、OSの選択画面となる。実に簡単である。

さっそくログインすると・・・まったく画面が写らなかった。
そこで、Ubuntu 2Dを選択してログイン。今度はきちんと見れた。Gnomeが進化したか、Ubuntuが進化したのか分からないが、使い勝手はよさそう。
が、いかんせん古いPCなので、WindowManagerは別なものに変更することにした（fluxboxとawesomeを入れたけど、メインはawesome）。
ちなみに最初の状態では、2.8GB程度しか使っていなかった様子。つまり4GB程度割り当てていたら十分かもしれない。

その後、Ubuntu 10.04 LTSやTag/Ubuntuなどを見返しながら、初期設定をば（仮想コンソールの削除など）。

apt-getしかなかったので、まずはaptitudeをinstall。その後はaptitudeを使用して、以下のソフトをinstall。

$ sudo apt-get install aptitude
$ sudo aptitude install awesome awesome-extra fluxbox vim gnuplot mlterm rox-filer g++ ptex-bin vpnc nautilus-dropbox

ちなみにDropboxは以前とは異なり、$ dropbox startで起動するようになっていた。

Firefoxを立ち上げて、Google chromeをダウンロードして、インストール・・・んが、ここで以下のエラーが出たので、これらもインストール。

dpkg: 依存関係の問題により google-chrome-stable の設定ができません:
google-chrome-stable は以下に依存 (depends) します: libnspr4-0d (>= 4.7.3-0ubuntu1~) …しかし:
パッケージ libnspr4-0d はまだインストールされていません。
google-chrome-stable は以下に依存 (depends) します: libnss3-1d (>= 3.12.3) …しかし:
パッケージ libnss3-1d はまだインストールされていません。
google-chrome-stable は以下に依存 (depends) します: libxss1 …しかし:
パッケージ libxss1 はまだインストールされていません。
google-chrome-stable は以下に依存 (depends) します: libcurl3 …しかし:
パッケージ libcurl3 はまだインストールされていません。

$ sudo aptitude install libspr4-0d libnss3-1d libxss1 libcurl3

後は、不要なサービスを停止。今回は以下のサービスを停止した。

bluetooth
dns-clean
grub-common
pppd-dns
pulseaudio
saned
sppech-dispatcher

ひとまずこのぐらいで。それにしてもねぇ・・・

WindowsXPじゃ耐えきれなくなって、Ubuntu投入。wubiなるものを使うと実に簡単にdual bootができた。便利になったものだ。　dual bootなんてものは、こわごわやって、失敗するとWindowsから再インストール・・・ なんてものだったけどね。今は違うのね。—
  (@wacpre) October 19, 2011

Just for the information. Cloud Shogi Diagram Generator has an English version. You can create any diagram you want at here.

81Dojo Kifu: [wacpre(JPN, 6-kyu) vs yukitakahashi(FRA, 3-kyu)], Opposition, White’s 4th-file Rook.

Here is my bad points.

• 27 move; Pushing pawn is rushing. I should move silver (S-46) and rook (R-38).
• 37 move; I should run my silver to 55, 57 or 37, not moving rook. I just lost my silver by nothing.
• 59 move; Dropping rook to 69 (R*69) would be better than dropping bishop to 37 (B*37).
• 71 move; Moving rook to 46 (R-46+) is better than moving to 61 (R-61+).

I could not attack well. What should I do? I have no idea…

I could win a 47th game.

81Dojo Kifu: [wacpre(JPN, 5-kyu) vs yakyeze(ESP, 11-kyu)], Free-style.

At 15th move, I should have drop pawn (P*46), not pushing pawn at 2nd file (P-25).

I lost the game with 81KID.

81Dojo Kifu: [81KID(NLD, 5-kyu) vs wacpre(JPN, 5-kyu)], Free-style.

Here is my bad points.

At
• 26th move: Silver to 95 (S-95) is better than promoting pawn (P-97+).
• 36th move: I should have drop pawn to 84 (P*84), not running my rook (R-92).
• 70th move: Dropping silver to 97 (S*97) is better than dropping knight to 86 (N*86)
• 74th move: I should have taken white’s dragon (Kx52), not running my king (K-33).

I think, learning Jouseki and solving Tume problems is not enough to become stronger. Playing Shogi is also important.
In these days, I can not play Shogi every week. But, I will continue to play shogi.

In Side Pawn Picker #2, you can see how it goes if black does not move his gold but push his pawn on 2nd file (P-24: ▲２四歩).

You can see how the games goes from a basic position of Side Pawn Picker (below diagram) at Side Pawn Picker #3.

Black threatening bishop exchange in next move, and white has three options to prevent it: B-33, N-33 and Bx88+(: △８八角成).

If white made bishop exchange (Bx88+: △８八角成) and

1. takes side pawn by his rook (Rx76: △７六飛), it goes to Double Side Pawn Picker.


From this diagram, black has three moves P*77, N-77 and S-77(: ▲７七銀 main line). It’s a very complicated but very interesting.

2. drops his pawn to 27 (P*27: △２七歩), black will take it by silver (Sx27: ▲２七銀). Now, white drop his bishop to 4e (B*4e: △４五角).(Side Pawn Picker, B*4e Strategy #1).



After several moves (including rook exchange) from the above diagram, black drops his lance to 36 (L*36: ▲３六香).

White has two options from here: Bx36 (: △３六角) (Side Pawn Picker, B*4e Strategy #2 (Bx3f variation)) and S*66 (: △６六銀).

When white drops silver (S*66: △６六銀), black will take knight by lance (Lx33+: ▲３三香成). White will reply this by moving silver to 67 (Sx67+: △６七銀成) (Side Pawn Picker, B*4e Strategy #3 (S*6f variation)) or moving bishop to 67 (Bx67+: △６七角成) (Side Pawn Picker, B*4e Strategy #4 (S*6f variation)).

3. drops his pawn to 38 (P*38: △３八歩), it goes to Side Pawn Picker, P*3h Strategy #1.

I thank to クラウド将棋局面図ジェネレーター, where you can create diagrams as shown in this press.

まずはDataStoreなるものに置けるかどうかのテストから。
参考サイト様：Python Datastore API – Google App Engine – Google Code

#!/usr/bin/env python
# -*- coding:utf-8 -*-

#OAuth
consumer_key = ‘???’
consumer_secret = ‘???’
oauth_token = ‘???’
oauth_token_secret = ‘???’

import tweepy
#Create OAuth handler
auth = tweepy.OAuthHandler(consumer_key, consumer_secret)
#Set access token to OAuth handler
auth.set_access_token(oauth_token, oauth_token_secret)
#Create API
api = tweepy.API(auth_handler=auth)

from google.appengine.ext import db
class Status(db.Model):
id = db.IntegerProperty()

reply_id = api.mentions()[2].id
statusid = Status(id=reply_id)
statusid.put()

重要なのは最後の６行。
前半の三行で、Statusというclass(?)を定義してやって、
後半の三行で、そこにidを入れる。
reply_id～については、Road to make a twitter bot 06を参考に。
（mentionで[2]となっているのは、test用のBOTへのmentionが少ないので・・・）

ちなみにGoogle App EngineでDatastore Viewerを見て、idが入っていればOK。

ちなみに元のtweetは

@wacpre_bot01 もぐらのテスト　１—
  (@wacpre) July 09, 2011

で、後は入れたデータを取り出してやればOKかと思ったが・・・
これがなかなかできないので断念。

そこで、以下の形にしてみた。
参考サイト様：勉強も兼ねて、お見合いbotを作ってみた。 – アジャイルSEを目指すブログ
Growth Curve: Google App Engineで動かすTwitter Botをつくってみた

#!/usr/bin/env python
# -*- coding:utf-8 -*-

#OAuth
consumer_key = ‘???’
consumer_secret = ‘???’
oauth_token = ‘???’
oauth_token_secret = ‘???’

import tweepy
#Create OAuth handler
auth = tweepy.OAuthHandler(consumer_key, consumer_secret)
#Set access token to OAuth handler
auth.set_access_token(oauth_token, oauth_token_secret)
#Create API
api = tweepy.API(auth_handler=auth)

from google.appengine.ext import db
class Status(db.Model):
id = db.IntegerProperty()

#get mentions
import re
num = 2
while num > -1:
reply_id = api.mentions()[num].id
id_check = Status.all().filter(‘id =’, reply_id).count()
if id_check == 0:
mentions = api.mentions()[num].text
word = re.search(u’もぐらのテスト’, mentions)
if word:
user = api.mentions()[num].author.screen_name
date = datetime.datetime.today()
post = ‘@’ + user + u’ もぐもぐら’ + ‘ at ‘ + str(date.hour) + ‘:’ + str(date.minute) + ‘ on ‘ + str(date.month) + ‘/’ + str(date.day)
api.update_status(post, reply_id)
num -= 1
statusid = Status(id=reply_id)
statusid.put()
else:
num -= 1
else:
num -= 1

（これまたnumの最初の定義でnum=2となっているのは、test用のBOTへのmentionが少ないので・・・）

id_check = Status.all().filter(‘id =’, reply_id).count()の部分で、reply_idがDatastoreに含まれているかどうかを判別。
（filter()やcount()に関しては、Query クラス – Google App Engine – Google Codeを参照）
もし含まれていないようであれば、id_checkは0となる（はずな）ので、tweetされた文章にある文字列（ここではもぐらのテスト）が含まれているかの判定へと移る。

実際にテストしてみた結果が以下のとおり。

@wacpre_bot01 もぐらのテスト ２—
  (@wacpre) August 20, 2011

上記のtweetに対してちゃんと以下のようにreplyしてくれている。もちろんDatastoreに入っているidを持つmentionには反応していない。

@wacpre もぐもぐら at 1:35 on 8/20—
  (@wacpre_bot01) August 20, 2011

その後しばらく動かしていても、同じtweetに反応することはなかった。

Datastore Viewerを見るとちゃんとidが追加されている。

長かった道のり（Road to make a twitter bot）だけど、これにてひとまず完成！

（また何かやりたいことが思いついたら、やってみるけどしばらくはいいかな～）

81Dojo Kifu: [81KID(NLD, 7-kyu) vs wacpre(JPN, 5-kyu)], Opposition, Black’s Opposing Rook.

Here is my bad points.

• At 30 move: Moving silver (S-65) is too bad. My silver was taken by nothing. King move (say, K-42) would be better.
• At 46 move: Pushing pawn at 5th-file (P-56) is better than dropping bishop to 79 (B*79).

043
At the next game with 81-BABY, I can do nothing and lost.

81Dojo Kifu: [wacpre(JPN, 5-kyu) vs 81BABY(NLD, 8-kyu)].

My bad points is 17th move. I should drop silver to 49 not 58.

044
I played one more games with 81-BABY and I could win.

81Dojo Kifu: [81BABY(NLD, 8-kyu) vs wacpre(JPN, 5-kyu)], Free-style.

At 30 move, dropping knight to 67 (N*67) is better than promoting rook (R-87+).

045
And I had a one more game with 81-KID and lost again.

81Dojo Kifu: [81KID(NLD, 6-kyu) vs wacpre(JPN, 5-kyu)], Opposition, Black’s 3rd-file Rook.

Here is my bad points.

• At 24th move: Moving silver (S-71) is too bad. Dropping pawn to 83 is better (P-83).
• At 30th move: is better than dropping bishop to 65 (B*65).
• At 64th move: Moving gold (G-32) is not good. Moving king might be better (K-51).
• At 68th move: Moving gold from 32 to 42 is better than 51 to 42 (G-42).
• At 70th move: Taking horse by gold (Gx33) is better than taking promoted silver by silver (Sx32).

046
One more! I had a game with 81-KID.

81Dojo Kifu: [wacpre(JPN, 5-kyu) vs 81KID(NLD, 6-kyu)], Opposition, White’s Central Rook.

At 67th move, dropping gold to 62 (G*62) would be better than dropping silver to 79 (S*71).

Finally I could win.

In these games, I did not make castle at all. It’s a very bad point. Not rushing attack, but making a castle is a good way to win, I think.

Additionally, I have to attack better way, which I always have no ideas.

I am still weak at playing shogi. To become a better shogi player, I have to spend more time to learn and play shogi!