b53 is up in /beta -- beta tester pls note this version introduced lot of deep memory allocation changes, be very sensible about errors and report soon.
this is especially for windows users
- i'm getting rule $9$9 as invalid while $9$8 (and the likes in the same sequence) don't show any message
- in a3 mode while using -1 ?l?d the input mask screen shows, as example, while running 6 characters ?1?1?1?1?1?1 [12]. Wasn't expected to be [6] instead ? Funny, now can't reproduce this and it's back to normal !
Thx,
problem 2 (-a 3, mask length) will be fixed w/ next beta! (mask length when using hash modes that internally use unicode etc, were kind of always "doubled")
I have no clue how I can reproduce problem #1. It seems to work here.
Please let us know what hash type you use and what exactly is in your rule file. Maybe give the whole command line string + (hexdump / content etc) of rule file.
Thx again
The problem #1 has be reproduced.
The test: pick a few md5 hashes, use best64 or shorter rule file and we don't get any error message.
Now edit your rule file and append to the end of that rule file any simple rule, like $1 $1 (double $1 to
append 11). Now start running we get the message "Skipping invalid or unsupported rule in file best64.
rule in line xx $1 $1".
- if we add that line at any position (1st line or 2nd line, whatever) but not adding an extra line, no error message
- adding one more rule line, this last one appears as an invalid or unsupported rule
- however we have a the rule display with the correct total of rules in the rule file
I tried to reproduce #1 but failed.
I expect this to be a cr/lf error somehow. Therefore it's important to have a hexdump from a rulefile that creates that error.
Also, did you use b57?
Please provide a hexdump from that position you added. HxD is a fine hex editor for windows. Screenshot is fine.
Is catalyst 13.9 still the recommended driver version for b57?
No, b57 includes workarounds to all known bugs in catalyst drivers starting from 13.4
You are free to use any driver you want as long as it is >= 13.4
(11-26-2013, 11:53 AM)atom Wrote: [ -> ]I tried to reproduce #1 but failed.
I expect this to be a cr/lf error somehow. Therefore it's important to have a hexdump from a rulefile that creates that error.
Also, did you use b57?
Please provide a hexdump from that position you added. HxD is a fine hex editor for windows. Screenshot is fine.
The message: Cannot convert rule for use on GPU in file e:\tools... in line 350: s_@
The rule file screenshot:
Excellent, I'm now able to reproduce it.
Will be fixed soon.
ok, fixed with beta60 - up in /beta