DiffStat(pRepo, rc15, beta16);

あー，3/28に間に合わせる予定がー． :(((
SoftMACの方は終わっているのに，HardMACがまだまだ低品質．
で，今日までの分．
[ZOMG, 3/28, t3h date expired. :(((
SoftMAC is WORKSFORME, HardMAC is still b0rked.
Then, here's diffstat of rc15 to beta16.]

 b/vhdl/AGC_CA.vhd   |  137 +++++++++++++++++++++++++
 b/vhdl/AGC_FA.vhd   |  279 +++++++++++++++++++++++++++++++++++++++++++++++++++
 b/vhdl/SYNC_MA.vhd  |   94 +++++++++++++++++
 bench/DUMB_RF.vhd   |   16 ++
 vhdl/AAO_UGT.vhd    |   13 +-
 vhdl/AGC_NG.vhd     |   57 +++++++++-
 vhdl/AUGT.vhd       |   11 +-
 vhdl/FPGA2.vhd      |   37 ++++--
 vhdl/NCDLY.vhd      |   70 ++++++++----
 vhdl/PHY.vhd        |    9 +
 vhdl/PWRC.vhd       |  177 ++++++++++++++++++++++++--------
 vhdl/RSSIC.vhd      |  283 +++++++++++++++++++++++++---------------------------
 vhdl/RSSI_AGC.vhd   |    5 
 vhdl/RX_BBP.vhd     |   63 ++++++-----
 vhdl/RX_IOBlock.vhd |   13 +-
 vhdl/TX_BBP.vhd     |   21 ++-
 vhdl/UGT.vhd        |   76 +++++++------
 vhdl/UND_BLOCK.vhd  |   35 +++---
 vhdl/UNI_BLOCK.vhd  |   73 ++++++-------
 vhdl/UNP_BLOCK.vhd  |   74 +++++++------
 20 files changed, 1135 insertions(+), 408 deletions(-)

土日でどれ位ヤレるのかが勝負．:P
[Sat/Sun. Am I ready to ride'on kamikaze, this spring storm. :P]

DiffStat(pRepo, rc14, rc15);

ホンマモンのhot codeって，やっぱdiffだと思うのですよ，ハイ．
つー訳で，現在release candidateしまくりな例のrepositoryで約一週間分
に相当する1.0_rc15 - 1.0_rc14．
[Yup, t3h real hot code means the diff, I thought. So, here is a diffstat
from the repository between 1.0_rc14 and 1.0_rc15, it's about 1 week or so.]

 a/vhdl/AGC.vhd             |  203 -----------------
 a/vhdl/PIDCONT.vhd         |  265 -----------------------
 a/vhdl/RSSICONV.vhd        |   78 ------
 b/bench/BENCH_AUGT.vhd     |  106 +++++++++
 b/bench/BENCH_UNPID.vhd    |  231 ++++++++++++++++++++
 b/vhdl/AAO_UGT.vhd         |  159 +++++++++++++
 b/vhdl/AGCSTM.vhd          |  211 ++++++++++++++++++
 b/vhdl/AGC_NG.vhd          |  465 ++++++++++++++++++++++++++++++++++++++++
 b/vhdl/AUGT.vhd            |  195 +++++++++++++++++
 b/vhdl/UND_BLOCK.vhd       |  153 +++++++++++++
 b/vhdl/UNI_BLOCK.vhd       |  156 +++++++++++++
 b/vhdl/UNP_BLOCK.vhd       |   97 ++++++++
 bench/BENCH_MODU_DMODU.vhd |   67 +++--
 vhdl/CCAD.vhd              |   91 ++++++-
 vhdl/FPGA1.vhd             |   28 +-
 vhdl/FPGA2.vhd             |  400 ++++++++++++++++++-----------------
 vhdl/GPS3_U301.vhd         |   61 +++--
 vhdl/GPS3_U801.vhd         |   62 ++---
 vhdl/LOCKD.vhd             |   49 ++--
 vhdl/NCDLY.vhd             |   59 ++---
 vhdl/PWRC.vhd              |  318 ++++++++-------------------
 vhdl/RSSIC.vhd             |  265 +++++++++++++----------
 vhdl/RSSI_AGC.vhd          |  512 ++++++++++++++++++++++++---------------------
 vhdl/RX_BBP.vhd            |  247 +++++++++++----------
 vhdl/UGT.vhd               |  119 ++++------
 25 files changed, 2967 insertions(+), 1630 deletions(-)

んー，やっぱ思ったより少ない．
うへー，仕事してねぇー． :P
[Hmm, it's also kinda little, I expected. Well, b/c I was not so working
hard? :P]

HotCodeStat(&me);

今弄っているブツのrepositoryのSLOCがどれくらいあるのか晒してみる．
以前のナニからパクって来たモノや他人のブツを元に変更したモノは含めない．
「hot codeのSLOCがどれくらいあるのかが分かるのでは？」と言う目論見． :P
[I'll investigate t3h stuff in current repo's SLOC,
exclude previous project's ones and based on others' code.
I think it will show me my hot code SLOC capability, maybe. :P]

 bench/BENCH_AUGT.vhd       |  106 ++
 bench/BENCH_DSMP2.vhd      |  119 +++
 bench/BENCH_DSMP8.vhd      |  122 +++
 bench/BENCH_FAKE_ADDA.vhd  |  204 +++++
 bench/BENCH_MODU_DMODU.vhd |  283 +++++++
 bench/BENCH_ROUND16.vhd    |   89 ++
 bench/BENCH_UNPID.vhd      |  225 +++++
 bench/BENCH_USMP2.vhd      |  121 +++
 bench/BENCH_USMP8.vhd      |  124 +++
 bench/DUMB_RF.vhd          |  151 +++
 bench/FAKE_ADDA.vhd        |  297 +++++++
 bench/STA.vhd              |  785 ++++++++++++++++++++
 bench/STA2STA.vhd          |  264 ++++++
 do/Bench.do                |   30 
 do/Compile.do              |  424 ++++++++++
 stim/AP_STUB.asm           | 1731 ++++++++++++++++++++++++++++++++++++++++++++
 stim/STA_STUB.asm          | 1735 +++++++++++++++++++++++++++++++++++++++++++++
 stim/include/mk_dchirp.h   |   23 
 stim/src/mk_dchirp.c       |   67 +
 stim/stimgen_DCHIRP.sh     |  155 ++++
 stim/stimgen_TXBB.sh       |  198 +++++
 stim/update_mif.sh         |    3 
 vhdl/AAO_UGT.vhd           |  150 +++
 vhdl/AGCSTM.vhd            |  187 ++++
 vhdl/AUGT.vhd              |  187 ++++
 vhdl/CCAD.vhd              |   63 +
 vhdl/DMODU.vhd             |  375 +++++++++
 vhdl/DPRAM_VINIT.vhd       |   88 ++
 vhdl/DSMP.vhd              |  143 +++
 vhdl/FP1_CLK.vhd           |  110 ++
 vhdl/FP2_CLK.vhd           |   94 ++
 vhdl/FPGA1.vhd             | 1018 ++++++++++++++++++++++++++
 vhdl/FPGA2.vhd             | 1264 ++++++++++++++++++++++++++++++++
 vhdl/GPS3_U301.vhd         |  616 +++++++++++++++
 vhdl/GPS3_U801.vhd         |  518 +++++++++++++
 vhdl/MODU.vhd              |  576 ++++++++++++++
 vhdl/MONSEL1.vhd           |  113 ++
 vhdl/MONSEL2.vhd           |  419 ++++++++++
 vhdl/NCDLY.vhd             |   75 +
 vhdl/PLSEXP.vhd            |  115 ++
 vhdl/PWRC.vhd              |  214 +++++
 vhdl/ROUND.vhd             |   85 ++
 vhdl/RSSIC.vhd             |  266 ++++++
 vhdl/RSSI_AGC.vhd          |  406 ++++++++++
 vhdl/RX_BBP.vhd            |  395 ++++++++++
 vhdl/TX_ADJ.vhd            |  269 ++++++
 vhdl/TX_BBP.vhd            |  334 ++++++++
 vhdl/UGT.vhd               |  127 +++
 vhdl/UND_BLOCK.vhd         |  143 +++
 vhdl/UNI_BLOCK.vhd         |  150 +++
 vhdl/UNP_BLOCK.vhd         |   94 ++
 vhdl/USMP.vhd              |  149 +++
 52 files changed, 15999 insertions(+)

実際の所，この物理SLOCの半分以上はコメントやらお決まり文句のカタマリ．
つーことで，思ったより少ない．
単純計算で論理SLOCで10k以下なので，まだまだ修行あるのみ．
と言うか，テストコード(*.asm)がデカ過ぎ． :P
[Hmm, half of physical SLOC are comments or generic stuff for the syntax.
So, I'd have to say it's kinda nuts, I have to do code more.
Well, t3h test code (*.asm) sucks, too bloat. :P]

GentooJP.NomiKai++; /* OSC2007 Spring */

今日はOSC2007 Tokyo/Spring最終日（二日目）でした．
以下，忘れないうちにメモメモ．
[Today is OSC2007 Tokyo/Spring finale.
These note for my stupid limbo.]

• matsuuさんが持ってきたPS3と某S社のテレビを展示．
  中身のGentooより「SPUとかはどうなのよ？」的な展開が繰り広げられた． :P
  [matsuu displays PS3 w/ S0ny TV. Many comers were talking about
  its SPU functionalities and its software development status,
  instead of installed one. :p]


• 持参したPowerBookを放置プレイ．
  genkernelでこれにはまったり，emerge systemとか．
  [Leave PowerBook alone. genkernel fails due to this bug, but work around.
  Then, emerge system or somthing.]


• trombikさんのcfengine/puppetの発表．
  金曜日にやっていたヤツを10分(?)に短縮．
  [trombik preach down to us WRT neat sysadmin stuffs.
  But it's really rapid, b/c he has only 10 min or so.]


• idaniさんのGentooバッヂを売った/買った．
  在庫 - 10．
  [idani sells cool0rz Gentoo badge, I got new one.
  Stok - 10, ATM.]


• オラ本を買った．
  bin hackとmake．
  [Got O'REILLY books.
  bin hack and make.]


• SUSUBOARDを見てくるついでに，未踏とかどうだったのかを教えてもらう．
  ついでにFPGA技術を買った． :)
  [Admired SUSUBOARD, hearing WRT MiTou thingy.
  Of cource, I got awesome brochure is so-called "FPGA tech". :)]


• Rob Pike氏ネタでPlan9な人と話す．
  #ifdefはダメダメ.
  [9ish guy show me Rob Pike autographed ones.
  "#ifdef considered harmful."]


• 耐久性が二ヶ月に向上したUdaの生演奏．
  例のとんかつ屋が店仕舞いをした事を知る． :(((
  [Live Uda, this new one improved its durable, about ~2mon.
  BTW, t3h TonKatsu restaurant has been closed． :(((]


• lightning talkを拝聴．
  OSS開発者は更なる愛とイタリア料理とチョコレートを必要としています． :DDD
  [Lighting talk, a finale of this event.
  So, now we know OSS devs need more love/italian/chocolates. :DDD]


• 喫茶店でコーヒー．
  MAC層以下な事とか特許とか．
  [Gotta coffee.
  Something about lower layer than MAC and patents thingy.]


• 懇親会参加，なんか色々話した．
  データがデカ過ぎるのでファイルシステムとかDBとかの使い方とか，
  Google Desktopは慣れるとアレとか，
  購入するPCのメモリは多くしてくれとか，
  サーバのステータスのメールに関する教育的指導とか，
  CJK herd増員の計画とか．
  [NomiKai, misc ones are stuffed, feel like ones are missing, though.
  Large/Huge data, file system selection, use of DB dilemma.
  Google Desktop is now really cool, if you like it.
  So, plz give more RAM if you buy new PC.
  A inquisition about flood status mail from servers.
  CJK herd needs more really something.]

早いトコ，今の仕事を終わらせてムニャヘニャするべ． :P
['Kay, I'll do something related Gentoo ASAP when I've done this project.]

修正: typoとか．
[FIX: typos or so.]

(pRF->PWR > RefPWR) ? pRF->ATT++ : pRF->ATT--;

"でむぱのぱわー"を計測，ちょっと調整が必要だったが何とかなった．
残りの問題は，受信側PHY下層のダイナミックレンジ拡張とアレなSME
の仕様をまとめる事．
ナンダカンダで次の納期まで二週間位しかなったりするのは罠ですか，
そうですか． :P
[Measure power of the radio wave. Well, it still
needs to tune a bit, but that's OK ATM.
Then, two major subjects are, 1) increasing dynamic
range of lower layer of reception PHY, 2) mark down
some spec of t3h SME functionalities.
There is only 2 weeks as rest of workable hours?
Blah... :P]

で，今週末はOSC．
予想通り，土曜日のみの参加になりそうな気配．
[BTW, OSC will take place at this weekend.
WTH, I can join that bunch only at the Sat.]

TuneParam(&pRF->ATT);

客先の電波暗室にてRFでナニが通る事を確認．
明日も"でむぱのぱわー"を計測するので早朝から出張也．
巧く行けばこれで一段落するのでOSCには参加出来る見込み．
[In the shield room, RF in/out connected. Tomorrow, we'll
measure the power of radio wave, so I have to get up earlier
than today. If thongs go well, I can enjoy OSC.]

色々騙し騙しなトコが有るので，マジ納入までに何とかしなければ．
先ずは，inner loopとouter loopで五倍近く遅れ時間が違うので，
その辺りからかな？
で，離散時間最適レギュレータでもやらかすか？ :DDD
[Yes, some really somethings are, though. First of all, I'd have
to purge dumb difference delay (x5!) of between inner loop and
outer loop. Oh, wait.. I won't miss t3h "optimal regulator in
discrete time domain". :DDD]

ONCT.karma++; EUC.karma--;

ONCTの某恩師の方が酔っぱらって電話してきた．
今弄っているFPGA内のPID制御器があり得ない出来だった事とか，
あんまりここに書けないモニョモニョな事とか話してたら午前五時． :P
[A former teacher of ONCT call me who is drunk.
We talked about, e.g. a really rock0rz PID controller stuff and
several really somthings. It was already 5:00 AM. :P]

取り敢えず，これを教材に使うとか言い出すのは勘弁して下さい． :DDD
[Oh, plz give it a break. So, don't use this crap as a resume of control
theory class. :DDD]

TuneParam(&pPHY->RSSI_AGC);

IFでナニが通る事を確認．
次はRFでなのだが，小ロットで手に入るRFなチップの出来が低品質な為，
RSSI_AGCと呼んでいるブロックへの入出力に色々補正をしなければなら
ない事が発覚．
[Connected peer IF in/out.
Next is RF in/out tuning, but t3h suck0rz RF chip set is really
retarded. It's mandatory to some adjustment/correction.]

具体的にどの辺りが低品質かと言うと，制御入力が出力に筒抜けになる為，
目標値との偏差が大きい時の粗調段階でシステムが発振する． :(((
これにはテキトーな一次補正で対処．
[E.g. t3h RF chip set through its control input to plant output.
So, it makes hunting when error of targeted value is kinda mighty
in coarse adjustment phase. :(((
We needs to use one-loop correction for it.]

問題はここからで，復調部分にかなり高次のFIRフィルタを噛ませている
関係上，かなり遅れ時間が有る．しかもOFDM特有の時間領域でのトゲが
かなりウザい． :(((
FPGAの容量もギリギリなので二進重み付け平均で対処予定．
[The real subject is.
Part of OFDM demodulation has a high-order FIR filter, it has big
latencies. And then, generally, OFDM symbol in time domain has short
cycle peak. It's really annoying. :(((
I'd have to fix by using binary weighted moving average b/c the
FPGA's logic stuff's capacity limit.]

あとは適当にモニタ信号に色々突っ込んで順を追ってパラメータを調整して，
RFの個体差が誤差の範囲に入る事を祈るだけ． :P
[Well then, to pull some signals related this gain controller and
step by step paramter tuning is all I have to do.
Oh wait, I won't forget to pray t3h RF modules have no considerable
individual variability, though. :P]