code logs -> 2008 -> Wed, 04 Jun 2008< code.20080603.log - code.20080605.log >
--- Log opened Wed Jun 04 00:00:42 2008
00:20
< Vornotron>
WHy not unleash your Vast Army Of Robots?
00:20
< Vornotron>
This is the very reason I showed you struct.
00:21
<@Shoukanjuu>
I don't have robots
00:21
<@Shoukanjuu>
They are not my style
00:21
<@Shoukanjuu>
Instead, have 255% growth rates
00:21 * Vornotron beats Shoukanjuu with Python.
00:21
<@Shoukanjuu>
xx
00:27 MyCatVerbs [~mycatverb@Nightstar-13709.lurkingfox.co.uk] has quit [Client exited]
00:31
<@Shoukanjuu>
You will pay for this, Vornotron >:O
00:31 * Shoukanjuu sets Vornotron to Jeigan status
00:32
<@McMartin>
Shou, in case you missed this point
00:32
<@McMartin>
Vorn is suggesting you use an Army of Robots to do the mass edits.
00:32
<@McMartin>
Instead of hand-stepping through 108 characters on your own.
00:32
<@Shoukanjuu>
That'd be fine and good...if I knew the addresses that needed editing
00:33 MyCatVerbs [~mycatverb@Nightstar-13709.lurkingfox.co.uk] has joined #code
00:33 mode/#code [+o MyCatVerbs] by ChanServ
00:34
<@Shoukanjuu>
Or even if the data didn't have seemingly random values in the middle of what I'm trying to find and correct
00:35
<@McMartin>
Whoo, struct dumps with aligned data values.
00:35
<@Shoukanjuu>
I would instead be able to do it in a matter of minutes, instead of hours
00:35
<@McMartin>
Alternately, relying on a specific allocator and saving pointers
00:35
<@Shoukanjuu>
I'm using a hex editor and and noting and editing hex values
00:36
<@Shoukanjuu>
I can only see these things in hexadecimal D:
00:36
<@McMartin>
I'm guessing what they were doing.
00:36
<@McMartin>
Pointers only make sense in hex.
00:36
<@Shoukanjuu>
Of course
00:37
<@Shoukanjuu>
It's not as simple as FE7, though
00:37
<@Shoukanjuu>
Or any of the other <9 fire emblems
00:37
<@ToxicFrog>
Is this a rom hack you're working on?
00:37
<@Shoukanjuu>
It's not really a rom hack
00:37
<@ToxicFrog>
If so, have you checked out the RHDN forums and #romhacking.net on Darkmyst?
00:38
<@ToxicFrog>
Because they often have tasty info.
00:38
<@ToxicFrog>
Aah.
00:38
<@Shoukanjuu>
The idea was to go into the data of FE10 and pick out the growths for each character therein
00:38 Derakon[AFK] is now known as Derakon
00:38
<@Shoukanjuu>
This would be easy. It should be easy.
00:38
<@Shoukanjuu>
It isn't as easy as I planned it to be
00:38
<@Shoukanjuu>
The growths that are accepted as the norm are wrong in some places
00:38
<@McMartin>
That's roughly speaking a ROM hack, just on a modern executable
00:38
<@McMartin>
The basic concept of Binary Patch still applies.
00:39
<@Shoukanjuu>
Instead of 8bits for 8 stats, there are 10 bits for 8 stats, 2 of which have no correlation with what I'm doing
00:39
<@Shoukanjuu>
So I can't just search for a byte...the problem with searching with too little bits is that there are often multiple of the same thing, not all have anything to do with what I want
00:42
<@Shoukanjuu>
Of course, I cvould be wrong in my terminology because I'm an idiot
00:44
< Vornotron>
So your problem is you don't know how to find the things automatically?
00:44
<@Shoukanjuu>
I can find the things using a "find and replacE"
00:44
< Vornotron>
then you should be able to find them automatically.
00:44
<@Shoukanjuu>
But since there is no way to know where the random values are located
00:44
<@Shoukanjuu>
I can't search for the entire string
00:44
<@Shoukanjuu>
28 0F 50 0E 28 23 50 14 0A 52 is a prime example
00:45
<@Shoukanjuu>
I meant bytes earlier, again, I'm an idiot
00:46
<@Shoukanjuu>
That'd be in decimal, 40 15 80 14 40 35 80 20 10 84
00:47
< Vornotron>
okay, so?
00:47
<@Shoukanjuu>
The 14 and the 10 are what I mean. They aren't in the same place in other strings
00:47
<@Shoukanjuu>
They aren't even always 0E and 0A
00:48
< Vornotron>
I'm really not sure what you're getting at
00:48
<@Shoukanjuu>
So I search for 2 values. The way I found this one was with 28 0F...
00:49
<@Shoukanjuu>
I can't do it all automaatically because these random values are in what I'm searching for, and if I search for too few bytes, I have many entries to sift through...Too many, and I won't get anything I can use
00:50
<@Shoukanjuu>
So I'm doing it by trial and error. Also: It keeps me busy
00:52
<@Shoukanjuu>
This is further compounded by the values I am searching for not e xisting, because the "expert" who found them was DOING IT WRONG.
00:52
<@Shoukanjuu>
It led me to believe that people just played the game a few times and took averages, making them the growths and thinking that nobody would know the difference
00:52
<@Shoukanjuu>
Because it has happened before
00:54
<@Shoukanjuu>
It wouldn't be so tedious if there wasn't a need to do this with a certain amount of...attention
00:54
<@Shoukanjuu>
I've made headway, though, which is good
01:04
<@Shoukanjuu>
The main thing I was trying to do was get the addresses for the characters, to help in rom hacking...>.>;
01:22 Serah [~Z@87.72.35.ns-26506] has joined #Code
01:22 mode/#code [+o Serah] by ChanServ
01:44 Thaqui [~Thaqui@Nightstar-12992.jetstream.xtra.co.nz] has joined #code
01:44 mode/#code [+o Thaqui] by ChanServ
02:50
<@Shoukanjuu>
>.>?
02:51
< Vornotron>
>.>?
02:51
<@Shoukanjuu>
Just checking if I'm stil lalive
02:52
< Vornotron>
no.
06:19 Derakon is now known as Derakon[AFK]
06:19
<@Shoukanjuu>
Subdomain...not working correctly...why
06:58 UndeadAnno [AnnoDomini@Nightstar-29192.neoplus.adsl.tpnet.pl] has joined #Code
06:58 mode/#code [+o UndeadAnno] by ChanServ
07:28 Kazriko [~kaz@Nightstar-26352.gdj-co.client.bresnan.net] has quit [Client exited]
07:32 Kazriko [~kaz@72.174.26.ns-12779] has joined #code
07:32 mode/#code [+o Kazriko] by ChanServ
10:31 You're now known as TheWatcher[afk]
10:51 Vornotron is now known as Vornicus-Latens
11:02 Thaqui [~Thaqui@Nightstar-12992.jetstream.xtra.co.nz] has left #code [Leaving]
11:02 Chalcedon [~Chalcy@Nightstar-488.ue.woosh.co.nz] has quit [Quit: Leaving]
11:46 You're now known as TheWatcher
13:20 UndeadAnno [AnnoDomini@Nightstar-29192.neoplus.adsl.tpnet.pl] has quit [Ping Timeout]
13:27 UndeadAnno [AnnoDomini@Nightstar-29546.neoplus.adsl.tpnet.pl] has joined #Code
13:27 mode/#code [+o UndeadAnno] by ChanServ
15:27 Wingless102983 [~Wingless1@Nightstar-10648.hsd1.mn.comcast.net] has joined #code
16:33
<@ToxicFrog>
Ok, the underlying issue is that my <Name> pattern simply doesn't work inside the larger grammar, for some reason.
16:33
<@ToxicFrog>
It never matches on stuff that it should.
16:35
<@ToxicFrog>
...
16:35
<@ToxicFrog>
This may be because I wrote AtomicName as (!%reserved !%WS !%eol .+) instead of (!%reserved !%WS !%eol .)+
16:37
<@McMartin>
That sounds ominous, yes
16:44
<@ToxicFrog>
The latter matches everything up to the next whitespace, reserved character, or end-of-line marker, which is the desired behaviour.
16:44
<@ToxicFrog>
The latter matches everything ever provided it doesn't start with one of those.
16:44
<@ToxicFrog>
Er, the former.
16:45
<@McMartin>
I kind of figured it wasn't "Anything but reserved, then anything but whitespace, then anything but EOL, then anything"
16:47
<@ToxicFrog>
The ! are actually non-consuming assertions, so it's "and", not "then"
16:48
<@ToxicFrog>
And now that this is fixed, QuotedName is working as well. Excellent.
16:51
<@ToxicFrog>
Now I just need the trees for the 11 kinds of definition, each with its own distinct subgrammar.
16:51 * ToxicFrog weeps
17:01
< NSGuest-6780>
:(
--- Log closed Wed Jun 04 17:11:33 2008
--- Log opened Wed Jun 04 17:11:38 2008
17:11 TheWatcher [~chris@Nightstar-29731.dsl.in-addr.zen.co.uk] has joined #code
17:11 Irssi: #code: Total of 22 nicks [12 ops, 0 halfops, 0 voices, 10 normal]
17:11 mode/#code [+o TheWatcher] by ChanServ
17:12 Irssi: Join to #code was synced in 59 secs
17:16 You're now known as TheWatcher[afk]
17:45
<@ToxicFrog>
I have to say, PEGs are totally sweet and I should have been using them a year ago.
17:45
<@ToxicFrog>
Maybe I should do a RIT post on them.
17:45
<@ToxicFrog>
Or cs_lectures.
17:45
<@ToxicFrog>
Actually-
17:45 * ToxicFrog prods Vornicus-Latens
17:46
<@ToxicFrog>
Has cs_lectures done anything on regexes, CFGs, formal grammars, parsing, etc?
17:57
< NSGuest-6780>
?
17:57
< NSGuest-6780>
cs_lectures?
17:57 NSGuest-6780 is now known as Doctor_Nick
17:58
<@ToxicFrog>
community.livejournal.com/cs_lectures/
17:58
<@ToxicFrog>
run primarily by Vorn, and...not updated since 2006, looks like.
17:58
<@ToxicFrog>
when did that happen?
17:58 Doctor_Nick is now known as NSGuest-6809
17:59
< NSGuest-6809>
ga
18:00
< NSGuest-6809>
hmm
18:03 UndeadAnno is now known as Pete
18:12
<@ToxicFrog>
Hmm.
18:13
< NSGuest-6809>
HMMM.
18:13
<@ToxicFrog>
Can I create a PEG that matches bash-style here-docs?
18:13
< jerith>
PEG?
18:13
<@ToxicFrog>
Parsing Expression Grammar.
18:14
<@ToxicFrog>
http://en.wikipedia.org/wiki/Parsing_expression_grammar
18:14
<@ToxicFrog>
It's what you get when you crossbreed CFGs and regexes.
18:15
<@ToxicFrog>
(a regex or CFG that matches here-docs would also work, as either can be expressed as PEGs)
18:16
< jerith>
You can do it with multiline regexen.
18:16
< jerith>
For the simple case, at least.
18:18
< jerith>
r'<< (\S+)\n.*?\n\1\n'
18:18
< jerith>
Assuming . matches newlines as well.
18:18
< jerith>
There may also be more whitespace allowed.
18:30
<@ToxicFrog>
Assume no backrefs.
18:31
<@ToxicFrog>
Which are not actually part of formal regular expressions, they're just a popular and convenient extension.
18:55
<@ToxicFrog>
woot
18:55
<@ToxicFrog>
DefCondition
18:55
<@ToxicFrog>
<- 'define' %WS 'condition'->"def:condition" %WS {<Name>} %eol
18:55
<@ToxicFrog>
(!<EOF> <Condition>)* -> {}
18:55
<@ToxicFrog>
<EOF>
19:38 Wingless102983 [~Wingless1@Nightstar-10648.hsd1.mn.comcast.net] has quit [Client exited]
19:39 GeekSoldier [~Rob@Nightstar-7685.dip.t-dialin.net] has joined #code
19:39 mode/#code [+o GeekSoldier] by ChanServ
19:44 You're now known as TheWatcher
20:28 Vornicus-Latens is now known as Vornicus
21:20 Pete is now known as UndeadAnno
22:12 UndeadAnno [AnnoDomini@Nightstar-29546.neoplus.adsl.tpnet.pl] has quit [Quit: "General! Strike me down, then have every last man, woman and child put to the sword. There's no point going on anymore, not now that it's all gone..." "At once, sire."]
22:19 Mojo1978 [~Mojo1978@Nightstar-17832.hsi.ish.de] has joined #code
22:48 C_tiger [~c_wyz@Nightstar-16806.nycmny.east.verizon.net] has quit [Ping Timeout]
22:48 C_tiger [~c_wyz@Nightstar-16806.nycmny.east.verizon.net] has joined #code
22:48 mode/#code [+o C_tiger] by ChanServ
--- Log closed Thu Jun 05 00:00:35 2008
code logs -> 2008 -> Wed, 04 Jun 2008< code.20080603.log - code.20080605.log >