SmartVersion | Tools + Scripts

if you want to create svf you need two files first the main file [any extention] then file to be extracted from svf file using main file. i hope you can understand it clearly now .

means you need both en & fr langauge iso's to create a fr language based svf file so that svf file will extract fr langauge iso using main en language iso file.

&

one must have en language iso file with svf file plus svf extractor to extract fr language iso file using both en language iso plus differential svf file.

 

want to create a Windows FR ISO file, the latest version, from the svf file

Where do I find the .svf file?

 

sorry i thought you want to create a differential svf file from two files.
now i got it clearly that you want to extract fr langauge iso file using a svf file you are having with you.
for that you need main language file too which is used to create that differential svf file.

 

Yes
Where do I find svf _fr

 

. = full stop

[Stay Happy n Healthy]

 

You are in the wrong thread! Take a look at:
Thread: Windows 10 SVF Repository, e.g, here: click me

 

Can you create a copy of Windows this way?

Wind 10 consumer 20H2 19024-928-apr to
Wind 10 (1809_17763) apr 2021

Svf:

(1809_17763) apr 2021

 

As I said before, you are in the wrong thread! This is about the software / tool SmartVersion and its options and not intended for such requests.

 

Did you try compressratio 522 ?

 

You can recompress without recalculating delta


smv m original.svf recompress522.svf -compressratio 522 -recompress

And you need only the original.svf file, not the first and second version (existing delta is recompressed)

Another idea, if you have small memory amount : first, make delta with a high -nbhashbits (so all memory will be used for delta) and -compressratio 160 or 171.

Then recompress with 522

 

Hi @gvollant
Just tried it now....this is just too cool.
I didn't time it but to loose additional 39 MB is impressive

@LostED
-compressratio 522

 

I'll release smartversion source in github , probably before july

 

compressratio 522 is probably the best (there is error in the 'usage')

-nbhashbits 27 take more memory and is faster than -nbhashbits 24. Result must be same

 

All these years, I am using 7z (LZMA2, dictionary size 1536, 24 GB of RAM is enough) to compress these "big" uncompressed (160) SVFs. Use more resource, yes, but it is always the best. Hope for some implementation of LZMA2.

Will try new zstd soon. As I can see, nice solution for computers with less RAM.

There are, at least for me, four bugs. I hope I find the time to explain. One of them is serious, IMO, because SHA1 of output file does not match. Have to say, I checked last year, I do not know now, is bug(s) removed or not.

 

Buf that can corrupt result must be fixed... Can you reproduce it ?

Zstd have far fastest extraction than Lzma2
I can imagine create a mode with LZMA and very large dictionnary

Note ; compressratio 522 need Smartversion 3.82 to extract , several GB of memory and probably 64 bits binary

 

Yes, x64 would be mandatory

This was seen using zstd (-compressratio 192) on x86 iirc.
@GezoeSloog found the best setting in this context with lzham (-compressratio 259) and use this nice solution in his scripts:

Code:

if %Processor_Architecture% equ x86 set SmvRatio=259

Kind regards

 

Unlike deflate/gzip/zlib, lzma2/xz and of course zstd, lzham projet is not updated.
So I suggest not compress using lzham

 

You need to login to view this posts content.

 

Thank you a lot.
Pehaps we can reproduce with smaller SVF file (with same logic of file renaming inside each version)