yes in inno setup i have error with your cls but not in cmd
but profager's cls works great

Same error? Will see

bpkv0.2.6 is not compressing .bk2 files from GRID 2
Its not showing percentage, also the output is 0 byte size, any idea whats d issue here?
i tried with .bik files and it proved successful

I have a question about DLZ.

srep before or after DLZ processing?

before...

try dlz:cc2222:cm43:cu1:em1:d64m

Thanks.

Unfortunatelly there are no infos about switches

For what are those switches?

there is in russian using google translate to english :
General options:
-slp - includes an attempt to allocate large pages of memory to increase the speed of packaging.
Options that affect any data:
-d - specifies the dictionary for the package. Example: -d128m. The default is -d64m;
-fb - sets the maximum match size for fast parsing in the parser. Similar to the corresponding option in lzma, but with the maximum possible value of 337. Default is -fb128;
-mc - specifies the maximum number of search cycles. Similar to the corresponding option in lzma. The default is -mc1024.
-ep - includes contextual modeling for the entire range of coincidence distances. In my tests for dds gives a permanent win. The default is -ep1.
Options that affect all data except dds:
-lc - sets the size of the context of the previous character to encode literals. It affects only normal data (all except dds and dds raw). Similar to the corresponding option in lzma. The default is -lc8.
-lp - sets the size of the context of the current position for encoding literals. It affects only normal data (all except dds and dds raw). Similar to the corresponding option in lzma. The pb parameter is always set to -lp. The default is -lp4.
Options that affect only dxt dds:
-dc - sets the size of the context of the previous character to encode a literal that was not accepted by the parser as a "dxt literal". It only affects the encoding of dxt dds. Possible values: 0..8. The default is -dc8.
-em - Enable or disable the encoding of special literals immediately after the repetition. It only affects the encoding of dds. Possible values: 0..1. The default is -em1.
-cu - Specifies the type of update of the literals statistics. It only affects the encoding of dxt dds. The possible values ​​are: 0..3. Which compression mode packs more efficiently depends on the data. The unpacking time increases with increasing value -cu (At -cu0, the unpacking time is approximately equal to the time lzma, at -cu3 - 2 times longer). The default is -cu0.
Options that affect only raw dds:
-pl - sets the model for the raw image. The possible values ​​are: 0..3. 0 - disables detection of dds raw. 2 - the data is preprocessed, which consists in the usual subtraction of the current pixel value and the previous one, 3 is also preprocessing, but the value calculated by the MED predictor is subtracted from the value of the current pixel. By default -pl1;
-pc - Specifies the size of the context for encoding the literal. It only affects raw dds coding. The possible values ​​are: 0..8. The default is -pc8.

CREDIT TO EDISON007...

dlz:d256m:mc1024: pc0: pl1:dc0:em0

optimal settings

Hi all !

Is it possible to use BPK with 4x4 ?
If yes ,how ?
I really dont understant the 4x4 parameters. 4x4:b128mb....
What is b128mb and depends on what ?
TIA....
:)

You can't.

The size of the block of packed data.

is it possible with cls of srep provided by profrager to avoid creating temp files on disk.
i guess no one seems interesting in posting reply.

I already used it on few games(not wise yet only ogg), it work fine and save up to 10%. I say dont worry and use it everywhere you can.

I also tried on bk2 before and it worked fine. Try command tools instead of cls.


Btw regarding dlz and lolz, I recommend using them only on .dds files. Thats where they are specialized. Even in .dds files gain was questionable vs speed, saving ~1gb(or ~10%) with added ~2-3h+ of compression. Meh, internal FA's LZMA with lc8 is often more rational option. We simply hit the wall with LZMA where any more % gain add logarithmic scale of time wasted(and power).

how can we use oggre?