Coffeearc

Hi Yann,

thank you for the nice words! It feels good to see that my archiver is put to good use.

Andreas

Hello Andreas

Regarding the bug you found in the decoder of LZP2 version bundled with CoffeeArc on my website :
This is modified version of LZP2,
which changed quite a few things in the File Manager source in order to improve performance & some issues.
But unfortunately it created others.

In fact, modifying the file manager seems a bad idea now, because it is quite complex.
LZP2 compression code itself is very simple, just a single week of work, starting from "i don't know C language".
Then, I've been using 6pack because, well, as a programmer beginner, i needed some help on I/O handling,
so i used an existing code.

Now, things have changed (a bit), and i should be able to build my own commandline file manager,
with probably better success & control than by trying to modify someone else's code.

This will take a while, but should be worth it.
So expect some improvement after this week-end.
I will keep you informed, and of course test the new version when it's ready with your latest CoffeeArc release.

Regards

Hi Yann,

that's calming, at first I wasn't sure if it was a bug in my code.
I think it's a good idea to rely on existing libraries - I also used one for I/O which saved me a lot of work and increased stability. But I also know that you can easily get in trouble though, if something's not fitting a 100% and you have to make some changes.
All in all, LZP2 seems like a really fast and yet efficient compressor, especially for only a week of coding.
So, good luck with your work- and let me know of any results!

Andreas

Hi Andreas

I believe to have corrected the issue you mentionned with the new I/O manager.
I've tested it with CoffeeArc v0.55b, and it seems to work correctly.

You can find the new LZP2 profile for CoffeeArc here :
http://img20.xooimage.com/files/0/7/8/profiles-e22ad4.zip

This new i/o manager is able to read/write from/to HDD in parallel with compression itself.
This means a faster overall experience for the user.

This is also a first step towards multi-threaded compression.

Regards

Hi Yann,

I briefly tested your new version and had no problems creating and extracting an archive. I used a large number of files and enabled checksum verification in the profile which succeeded, so I think that it works now!

Parallelization is definitely a nice feature, I plan to realize it in one of the next releases in Coffeearc too, by running two or more instances at once. Though I think in case of such fast compression tools like yours this could perhaps do more harm than good since the HDD is probably the bottleneck... but who knows - we'll see!

Thanks for the news and best regards,

Andreas