No, there is nothing to do.
It's a limitation of the reimport technique with content that is not located inside the original archive (memory_files are like temporary files).

it works perfectly as described here:
http://aluigi.altervista.org/misc/simra ... wfiles.txt

try this script:
http://aluigi.org/papers/bms/others/skydrift.bms
don't worry if you receive an error with some of the big archives

the tables are encrypted/obfuscated so I don't see an easy way to extract the content (which is not obfuscated)

I fixed the behaviour with Insane2 but I pass on Star Conflict

set name unicode name remember to use the latest version of QuickBMS that I have released yesterday, it has a better support for unicode strings note that the handling of utf16 endianess depepends by the current endian status (I tell you in case you get strange chars), the one of that string is lit...

I guess I have found a problem that may be related to this, I'm checking it

Do you have an example command to replicate the problem? Here I have tried using -D with absolute and relative folders, from the input folder, from the output folder, using same input and output folder and so on but in my minimal tests I did there was nothing wrong. I used just a simple script like ...

try placing echo y in pipe before quickbms:

cool idea the negative filter, I have just added it to quickbms 0.5.21a: -f "!{}.txt" will extract all the files except the txt ones -f "{}.mp*;!{}.mp3" all the mp* files (like mpg or mpeg) except the mp3 files why I use {} instead of *? because windows is so stupid that even if ...

I don't know what programming language you are using but with quickbms this is not complex: math ZSIZE = 0x1234 # placeholder math SIZE = 0x5678 # placeholder log MEMORY_FILE 0 0 put ZSIZE long MEMORY_FILE put 0 long MEMORY_FILE put SIZE long MEMORY_FILE append log MEMORY_FILE 0 ZSIZE append get SIZ...

@chrrox
do you have the same problem if you use {} instead of * in your filter?

parsing strings with quickbms is a pain just because the language wasn't intended for it. for LINEBREAK you need to use the binary type because \n is a C escape char handled only by the bynary type. signed_short is handled in read mode like get VAR signed_int, and yes it's my fault to forget "s...

cws2fws (the following is the original homepage):
http://iancoog.altervista.org/C/cws2fws.rar

or my script:
http://aluigi.altervista.org/papers/bms ... ws2fws.bms

I have created a folder containing 2 files named asdf.dat and 00000000.dat and then I launched "quickbms script.bms that_folder" without having problems.

do you have a way to replicate that problem?