29: チェイントーク処理不能、OnAiTalk/OnAITalk、OnSurfaceRestoreなど

ばぐとら研究所統合ToDo管理システムへようこそ!バグ報告に限らず様々な要望を扱うシステムです。
まずは同じような内容がないか [検索] した後、無ければ [新規レポート] で作ってみましょう。
Windows 8以降、高速スタートアップが実装された関係で、SSP以外の部分に起因する不具合発生が出てきました。
何か挙動がおかしいかな、と思ったら、一度再起動(シャットダウン->起動ではなく)してみてから再度確かめてみてください。

最新状況

名前 スイート
概要 チェイントーク処理不能、OnAiTalk/OnAITalk、OnSurfaceRestoreなど
状態 [?]意見求む
作成日時 2013-04-18 19:50:56
最終更新日時 2014-12-28 21:08:19

履歴

1 | 2013-04-18 19:50:56 | [−]新規
AccVeronica
I'm afraid I can only use English to write the report, really 
sorry for not having good Japanese skill.

version:
SSP: 2.2.99
YAYA: 5-45-1

As tama logged, in yaya_shiori3.dic ,error happened in 
"ChainTalk" Event, "error E0092 : ()、もしくは[]に
続く演算子が見つかりません".

At the same time, If RandomTalk is going to show the talk right 
after the one with a CHAIN.IDName, It will print 
"CHAIN.IDName,(talk content)", and keep printing it in 
following times of random talk.

I tried both of the old and new version of your yaya_shiori3, 
but I am not sure is this bug and others is affected by changing 
OnAiTalk into the new SSP event OnAITalk.

I have to say the ghost I'm debugging is one with multiple modes 
and I set two of the modes leave out the kero side(for 
example,"\0\s[0]\1\s[-1]\p[2]\s[20]" in 
OnSurfaceRestore). However, the no.10 of surface appears most of 
the time when OnAITalk starts under these modes, talks just 
happen after that.

And the event "OnSurfaceRestore". It should happen 30 
seconds after the talk is printed, but only in this ghost it 
delays to more than 90 seconds. I could not find out why it only 
happen on this ghost(checked the related functions in 
yaya_shiori3.dic), it keeps happening no matter what I do.

2 | 2013-06-14 06:51:31 | [?]意見求む
C.Ponapalt
Hello AccVeronica, Sorry for *very very* late response, because
I missed update notification of this BTS.

> As tama logged, in yaya_shiori3.dic ,error happened in 
> "ChainTalk" Event, "error E0092 : ()、もしくは[]に続く演算子が見
> つかりません".
> 
> At the same time, If RandomTalk is going to show the talk right 
> after the one with a CHAIN.IDName, It will print 
> "CHAIN.IDName,(talk content)", and keep printing it in following 
> times of random talk.

I think this trouble will be solved using latest 
"konnoyayame"
system dictionary.
http://ms.shillest.net/ghost/konnoyayame.nar
Please use this archive and find "ghost/master/system" 
directory.

> I have to say the ghost I'm debugging is one with multiple modes 
> and I set two of the modes leave out the kero side(for 
> example,"\0\s[0]\1\s[-1]\p[2]\s[20]" in OnSurfaceRestore). 
> However, the no.10 of surface appears most of the time when 
> OnAITalk starts under these modes, talks just happen after 
> that.
> 
> And the event "OnSurfaceRestore". It should happen 30 seconds 
> after the talk is printed, but only in this ghost it delays to 
> more than 90 seconds. I could not find out why it only happen on 
> this ghost(checked the related functions in yaya_shiori3.dic), 
> it keeps happening no matter what I do.

I can't figure out why these problems happened, but I guess
"Option -> Old Specification Support" will be 
activated and
"Option -> Raise AI Talk From SHIORI" will be 
deactivated,
because these symptoms caused by this settings in many cases.
Please check options described above.

3 | 2013-06-14 13:30:14 | [?]意見求む
AccVeronica
For the event of restoring surface, I was trying to find a 
solution, but maybe not to use the event would be better.

And chain talks... I just tried the solution, seems the error in 
talks cannot be resolved yet. Actually the system dictionaries 
are the newest ones, and I was not changing any part of it.
Is it relevant with the charset of ghost or platform? I always 
use UTF-8, and seems even SSP the platform would have some bugs 
when it is not using Japanese as system lang(I never saw anybody 
in Japan complained to you).

And with more time to check, now I wonder if it has some 
problems after using array for the randomtalks, as I want to 
know how many talks I've written.
It is like:

RandomTalk
{parallel Talks}

Talks:array
{
(contents)
}

I found that every time when the bug happened, it printed the 
sentence in beginning of the event first (actually it would not 
stop reading until it found a chain ID), then the error occurred 
and it began to repeat printing "CHAIN.IDname,(talk 
content)", and content is the sentence after the last one 
with a chain ID. It can be concluded that when reading through 
the part and selecting the first sentence as the result of 
random operation, it could not stop anywhere the script reads 
"\e", until failure in coping with the first chain ID. 
Then it would just raw print the last part of the contents, if 
it can be seen as divided by chain IDs.

Strange that I use array and parallel in more than one 
ghost(some also have modes or talk event groups), with all your 
newest system dictionaries, but only this one is in trouble with 
this bug. I haven't found why they operated differently.

Thus, I guess the bug is not because anything wrong with the 
ChainTalk part, but somewhere else. Still I hope you can find 
out why the bug happens like that, and tell me what can do to 
get rid of it.

4 | 2013-06-14 19:49:57 | [?]意見求む
C.Ponapalt
Hmm... I want to investigate your ghost to solve these errors.
Can I get whole ghost data, especially script files under 
ghost/master/ ?

5 | 2013-06-14 22:22:19 | [?]意見求む
AccVeronica
> Hmm... I want to investigate your ghost to solve these errors.
> Can I get whole ghost data, especially script files under 
> ghost/master/ ?

Absolutely. If you'd like to look in this case, please 
download... actually I'm not quite clear with the option of file 
type below and even caused some error. So I leave the link here 
(apologize for that).
http://altitude8d9.ucoz.com/1forBTS.zip

This package is attached with minimized shell and balloon for 
convenience(1.74 mb).
It might be difficult to read because of highly messed up 
scripts, and some debugging notes, sorry!

6 | 2014-12-28 21:08:19 | [?]意見求む
スイート
> This package is attached with minimized shell and balloon for 
> convenience(1.74 mb).
> It might be difficult to read because of highly messed up 
> scripts, and some debugging notes, sorry!
うーん、特定のシェル&バルーンで起こってるのかなあ?

[リプライをつける]
Bug Tracking System 影舞 0.8.8
Powered by Ruby 1.8.7