questions on Rules used to reject plains

[aprizm]

Hey crackerz,
I was checking out the rules and I have a question.
Some of the rules reject candidates...
here is my question:

are the rules to reject plains acting on the source or the results of the rule?

lets say we have a rule like this : <10 $1$2$3
and a word : password

would this be rejected because :
source: password (8 characters)
result: password123 (11 characters)


also is there a difference between those rules :
<10 $1$2$3
$1$2$3 <10

the thing is I want to reject candidate that are less than 8 char AFTER all the transformation.

btw im sorry my native language is not english.

[aprizm]

wow I just spent last night working on these rules to finally end up realizing theyre not working with GPU AAAAAAAARGH.

suggestion: Bring this feature to ocl

another question: does the skipping (rejecting) really help the performance or its negligible?

[atom]

thats not completely true, they are just not working for fast hashes like md5. for slow hashes like wpa they work, as slow hashes use the hashcat cpu rule engine.

skipping candidates based on rule rejection would make cracking slower because it adds another branch to it while it doesnt hurt to check a hash if you can do multimillions a second