User talk:Kymara: Difference between revisions

insert into temp_milled_flour

SELECT source, substring_index(param2,';',1) AS count, id AS gameEvents_id, timedate FROM gameEvents_2009_08_17

WHERE event ='quest' AND param1 = 'jenny_mill_flour' AND param2 <> 'done' ;

insert into temp_milled_flour

SELECT source, substring_index(param2,';',1) AS count, id AS gameEvents_id, timedate FROM gameEvents_2009_02_19

WHERE event ='quest' AND param1 = 'jenny_mill_flour' AND param2 <> 'done' ;

insert into temp_milled_flour

SELECT source, substring_index(param2,';',1) AS count, id AS gameEvents_id, timedate FROM gameEvents_2008_08_21

WHERE event ='quest' AND param1 = 'jenny_mill_flour' AND param2 <> 'done' ;

-- yes I know about unions I just like to see row counts.

alter table temp_milled_flour

add index Index_tempflour_source(source);

alter table temp_milled_flour

add column last boolean default 0;

update temp_milled_flour t join

(select source, max(timedate) as maxdate from temp_milled_flour group by source) t2 on t.source = t2.source

set last = 1

where maxdate = timedate;

-- oh yes i remember now why i thought this was going to get hard - i can't use id like i planned.

select * from temp_milled_flour where last = 1 and timedate < '2009_08_17';

-- if they didn't pick it up then delete it - as when they pick it up later they will get the points

-- oh and they might pick it up between now and the next release too :/

-- anyway it's only worth to check when it's the last mill they did

/* delete FROM stendhal.temp_milled_flour where last = 1 and (select id from gameEvents where event ='quest' and param1 =

'jenny_mill_flour' and param2 = 'done' and temp_milled_flour.source=gameEvents.source and gameEvents.id > gameEvents_id limit 1) is null; */

Okay, looks like I should have done this first:

Please install msysgit 1.6.1 or above before install tortoisegit http://code.google.com/p/msysgit

== For building achievements to list at [[Stendhal Achievements]] ==

They still need a bit of categorisation for the page though:

<source lang = "sql">

select distinct

concat('{{',

concat_ws('|',

'Achievement',

concat('title=',title),

concat('description=',description),

concat('difficulty=',(case base_score when 10 then 'easy' when 50 then 'medium' when 1000 then 'hard' else 'todo' end))),

'}}')

from testserver.achievement

order by category, identifier;

</source>

==Pending achievements==

<source lang = "sql">

insert into pending_achievement(charname, achievement_id, param, cnt)

select source, 17, "", count(*) from temp_elf_princess group by source;

stendhal> insert into pending_achievement(charname, achievement_id, param, cnt) select name, 26, param1, 1 from looted_all_items where timedate<'2010-11-24' and name <>'' and param1 IN ('chaos armor', 'chaos boots', 'chaos cloak', 'chaos helmet', 'chaos legs', 'chaos shield');

Query OK, 20004 rows affected (0.46 sec)

Records: 20004 Duplicates: 0 Warnings: 0

stendhal> insert into pending_achievement(charname, achievement_id, param, cnt) select name, 34, param1, 1 from looted_all_items where timedate<'2010-11-24' and name <>'' and param1 IN ('shadow armor', 'shadow boots', 'shadow cloak', 'shadow helmet', 'shadow legs', 'shadow shield');

Query OK, 13100 rows affected (0.20 sec)

Records: 13100 Duplicates: 0 Warnings: 0

stendhal> insert into pending_achievement(charname, achievement_id, param, cnt) select name, 40, param1, 1 from looted_all_items where timedate<'2010-11-24' and name <>'' and param1 IN ('black armor', 'black boots', 'black cloak', 'black helmet', 'black legs', 'black shield');

Query OK, 1632 rows affected (0.17 sec)

Records: 1632 Duplicates: 0 Warnings: 0

stendhal> insert into pending_achievement(charname, achievement_id, param, cnt) select name, 61, param1, 1 from looted_all_items where timedate<'2010-11-24' and name <>'' and param1 IN ('golden armor', 'golden boots', 'golden cloak', 'golden helmet', 'golden legs', 'golden shield');

Query OK, 9562 rows affected (0.26 sec)

Records: 9562 Duplicates: 0 Warnings: 0

-- after 0.92 if we want a kill enemy army achievement, can do

-- highest number so far is about 40 and there's nothing in older gameEvents tables

insert into pending_achievement(charname, achievement_id, param, cnt)

SELECT source, XX, "", count(*) FROM gameEvents WHERE event='quest' AND param1= 'kill_enemy_army' AND param2 like 'done%' group by source;

-- 0.94

insert into pending_achievement(charname, achievement_id, param, cnt)

select name, 95, param1, 1 from looted_red_items where timedate<'2010-11-24' and name <>'';

</source>

== turn overflows table ==

<source lang="bash">

grep "Turn " log/server.log* > turn.txt

sed 's/^[^:]*://' < turn.txt | sed 's/ WARN.*: //' > turn2.txt

</source>

<source lang = "sql">

CREATE TABLE overflow (

odate DATE,

otime TIME,

o1 INTEGER,

o2 INTEGER,

o3 INTEGER,

o4 INTEGER,

o5 INTEGER,

o6 INTEGER,

o7 INTEGER,

o8 INTEGER,

o9 INTEGER,

o10 INTEGER,

o11 INTEGER,

o12 INTEGER);

LOAD DATA LOCAL INFILE '/tmp/turn2.txt'

INTO TABLE overflow FIELDS TERMINATED BY ' ' LINES TERMINATED BY '\n';

alter table overflow add column timedate datetime first;

update overflow set timedate = concat_ws(' ',odate, otime);

alter table overflow drop column odate;

alter table overflow drop column otime;

rename table overflow to overflow_sum

create table overflow (timedate timestamp, lck integer, snxt integer, visit integer, end integer, trans integer, percept integer, save integer, wnxt integer, begin integer, ulck integer, stats integer, kickdb integer, sum integer);

insert into overflow select timedate, o1, o2-o1, o3-o2, o4-o3, o5-o4, o6-o5, o7-o6, o8-o7, o9-o8, o10-o9,o11-o10,o12-o11, o12 from overflow_sum order by timedate;

</source>

== Translation ==

<pre>

2010

\#arianne.01-04.log:23:42

\#arianne.08-07.log:12:35

2011

\#arianne.01-04.log:23:42

\#arianne.02-04.log:10:15

</pre>

== Maze Challenge ==

<source lang = "sql">

-- score board for Maze Challenge, to run on 15th October

select

charname,

points,

@old := (select points from halloffame_archive_recent r1 where r1.charname = r2.charname and fametype = 'M' and day < '2011-10-01' order by day desc limit 1) old,

points - @old as difference

from halloffame_archive_recent r2

where fametype = 'M' and day = date(now())

having difference >0

order by difference desc;

-- experimental procedure to get max maze score for a given player (i.e. the ones in that list) in a certain time period

DROP FUNCTION IF EXISTS getMaxMazeScore //

CREATE FUNCTION getMaxMazeScore(

name VARCHAR(32),

enddate DATE,

ndays SMALLINT

) RETURNS INTEGER

COMMENT 'Get max maze score for player in certain time period'

BEGIN

DECLARE maxscore INT DEFAULT 0;

SET maxscore = 0;

WHILE ndays >= 0 DO

select points,

@old := (select points from halloffame_archive_recent r1 where r1.charname = name and fametype = 'M' and day < date_sub(enddate, interval ndays day) order by day desc limit 1) old,

points - @old as difference

from halloffame_archive_recent r2

where fametype = 'M'

and day = date_sub(enddate, interval ndays day)

and r2.charname = name INTO @p, @o, @d;

IF @d > maxscore THEN

SET maxscore = @d;

END IF;

SET ndays = ndays-1;

END WHILE;

RETURN maxscore;

END //

-- now put it all together in one super amazing query that runs in about 35 seconds!! wow super amazing hey!!

-- (it would be way slower if the function call was not done 'outside' the scores table as the scores table uses a 'having' so is only small after it's completed. incase you wondered why I did that)

select charname, difference as totalscore, getMaxMazeScore(charname, date(now()), 14) as maxscore

from

(select

charname,

points,

@old := (select points from halloffame_archive_recent r1 where r1.charname = r2.charname and fametype = 'M' and day < '2011-10-01' order by day desc limit 1) old,

points - @old as difference

from halloffame_archive_recent r2 where fametype = 'M' and day = date(now())

having difference >0

order by difference desc)

scores;

-- output is name, total score, max score, for all participants

</source>

== making a video ==

<pre>

sudo apt-get install gtk-recordmydesktop

</pre>

Run it. Select area to record, press Record.

Stop recording by pressing the square on the top right of your menu bar where the clock is normally.

Encoding takes a while.

Saves as out.ogv in your home folder by default.

Sound etc was okay on my netbook though it seemed to pick up the mic and you can hear typing.

Couldn't get sound to capture at all on my desktop.

Youtube didn't like the .ogv file. Said it was too long, it was 12 mins, 15 is max? maybe the problem is the format.

Convert to AVI:

<pre>

sudo apt-get install mencoder

</pre>

Use command as per [http://info123.org/tutorials/54-how-to-convert-ogv-file-format-to-avi-or-any-other-format]

<pre>

mencoder -idx out.ogv -ovc lavc -oac mp3lame -o intro_stendhal_201201.avi

</pre>

Weirdly the avi was 36 minutes but the original was 12, nothing happens after 12 minutes, weird? Trim to only keep the first 00:12:12

<pre>

mencoder -endpos 00:12:12 -ovc copy -oac copy intro_stendhal_201201.avi -o intro_stendhal_201201_1.avi

</pre>

following advice at [http://ubuntuliving.blogspot.com/2008/03/splitting-avi-file-into-smaller-parts.html]

Then google accepted it

== Benfords Law fun ==

For a blog post or something, is interesting stats :)

<pre>

-- xp of characters

mysql> select left(xp,1), count(*) from character_stats where xp > 0 group by left(xp,1);

+------------+----------+

| left(xp,1) | count(*) |

+------------+----------+

| 1 | 5555 |

| 2 | 3122 |

| 3 | 2380 |

| 4 | 2344 |

| 5 | 2854 |

| 6 | 996 |

| 7 | 788 |

| 8 | 667 |

| 9 | 633 |

+------------+----------+

-- age of characters

-- max is 2940058

mysql> select left(age,1), count(*) from character_stats where age > 0 group by left(age,1);

+-------------+----------+

| left(age,1) | count(*) |

+-------------+----------+

| 1 | 8810 |

| 2 | 4962 |

| 3 | 3456 |

| 4 | 2417 |

| 5 | 2854 |

| 6 | 1545 |

| 7 | 1402 |

| 8 | 1250 |

| 9 | 971 |

+-------------+----------+

9 rows in set (0.02 sec)

-- number of kills per day

-- 2160 is max

mysql> select left(cnt,1), count(*) from kills group by left(cnt,1);

+-------------+----------+

| left(cnt,1) | count(*) |

+-------------+----------+

| 1 | 527945 |

| 2 | 272803 |

| 3 | 164413 |

| 4 | 116935 |

| 5 | 81453 |

| 6 | 64920 |

| 7 | 49086 |

| 8 | 40612 |

| 9 | 32843 |

+-------------+----------+

9 rows in set (0.61 sec)

-- size of stacks of items added to other items

mysql> select left(param2,1), count(*) from itemlog where event = 'merge in' group by left(param2,1);

+----------------+----------+

| left(param2,1) | count(*) |

+----------------+----------+

| 1 | 10636326 |

| 2 | 3470611 |

| 3 | 2667505 |

| 4 | 1920918 |

| 5 | 1397535 |

| 6 | 972027 |

| 7 | 757029 |

| 8 | 683354 |

| 9 | 667485 |

+----------------+----------+

9 rows in set (5 min 38.02 sec)

</pre>

== paperchase check ==

<source lang="sql">

select source, param2 from gameEvents

where event = 'quest'

and param1 = 'paper_chase_2013'

and param2 like 'done%'

order by round(substring_index(substring_index(param2,';',2),';',-1));

</source>

== sokoban check ==

<source lang = "sql">

select `char`, `round`, score as timetaken, (1000000*round - score) as points from

(SELECT

source AS `char`,

least(max(round(substring_index(substring_index(param2,';',2),';',-1))),59) AS `round`,

max(round(substring_index(substring_index(param2,';',3),';',-1))) AS `score`

FROM gameEvents

WHERE event = 'quest'

AND param1 = 'sokoban'

AND param2 LIKE 'done%'

GROUP BY source ) t

order by points desc

limit 10;

</source>