Can you include .3MF to the list of re-compressible formats?

[spwolf]

i hate editing as you wont get email notification with edited parts :).

Eugene is our compression engineer, working on actual format (same guy did patchbeam engine, rar speed improvements, wz jpeg decompression, etc).

[davidsplash]

i was just pointing out features in existing formats that would be nice to have included in pa. why right a new format unless you want a format that includes everything.

I apologise its very difficult using android and your forumn nearly impossible

[mwb1100]

@spwolf:

same guy did patchbeam engine

You guys did patchbeam in-house?

Nice. For some reason I had assumed it was technology bought/licensed from a 3rd party.

[spwolf]

@mwb1100:

You guys did patchbeam in-house?

Nice. For some reason I had assumed it was technology bought/licensed from a 3rd party.

all ours…

[spwolf]

@TBGBe:

Not sure what you mean with these existing formats (conversion?)-
zip
cab
setup.exe/data1.cab
tar/tar.gz/tgz/tar.bz2
rar

re-compression… if you compress them as they are, you will gain nothing… if you extract and compress them, between 15 and 50% gain in compression is possible.

[Mameluke]

I would most likely want to see the following in this order if possible please :)

1. Internal and External (RAR style) recovery record (with choice of percentage just like RAR).
2. JPEG compression (got many pictures that are quite large)
3. Strong Encryption with ability to encrypt file names (Like Winrar and 7z)

Thank you :D

Mame

[davidsplash]

would love to see a format that actually can compress media files better than we have at present. At the moment most media formats are donme by several formats i would like one format to do them all e.g wave mp3 and jepg format compression all in one archive format at present i have to use several formats in order to get the best comopression for each media format like sound or images

[davidsplash]

a format that preserves data like who made it and what version of the archive format made it. usefull for people who need this info

[By-Tor]

I’d still like to see wz jpeg compression for zipx.

[spwolf]

@By-Tor:

I’d still like to see wz jpeg compression for zipx.

it is still eventual possibility… however, while we researched jpeg compression last year, our first version was slightly better compressing than WZ Jpeg and 3x faster on dual core machine… and that could be done even better with 20% better compression than wz and up to 3x faster speed on dual core machines :).

so which one would you want? :)

[By-Tor]

@spwolf:

it is still eventual possibility… however, while we researched jpeg compression last year, our first version was slightly better compressing than WZ Jpeg and 3x faster on dual core machine… and that could be done even better with 20% better compression than wz and up to 3x faster speed on dual core machines :).

so which one would you want? :)

Why not both? :p

[spwolf]

it takes 2 months for engineer to implement each, so during those 2 months he could do:PDF, MP3, PNG instead etc, etc.

it is very time consuming and demanding project, which is why nobody but PA can read these archives currently.

[By-Tor]

Ah. From the sound of it, I thought the implementations were close to complete. I guess the first thing I’d like to see then is the wz jpeg compression, just for compatibility. I’m less concerned with a unique format right now as portability issues would be a hindrance. For another unique archive format to catch on, I think you’d need to release an open source portable command line version at the very least before it would be truly useful among other users.

[spwolf]

@By-Tor:

Ah. From the sound of it, I thought the implementations were close to complete. I guess the first thing I’d like to see then is the wz jpeg compression, just for compatibility. I’m less concerned with a unique format right now as portability issues would be a hindrance. For another unique archive format to catch on, I think you’d need to release an open source portable command line version at the very least before it would be truly useful among other users.

compatibility with only WinZip though, no other utility, and certainly no free or open source utility… so what is the difference really between PAF, IZEHRBLAH, and ZIPX, when only two products support extracting ZIPX archives.

Problem is that ZIP itself, due to old nature of format, will never be anywhere as good as modern format, and whatever extensions are added to ZIPX not only that they cant be as good as new format, they will also not be compatible with other products either.

Who knows, maybe PAF will be open format and maybe there will be free extraction tools, eventually anyway :)

[By-Tor]

WinZip has a large userbase, regardless of whether we like it or not. 7-zip can handle most of ZIPX, with the exception of jpeg compression, but, Igor has stated he would implement some form of it if he saw a need.

ZIP by its nature is just a container format. So, you’re right, it doesn’t matter much what you call it. I’m just thinking about trying to capture the largest audience possible.

I’m by no means anti-PAF. heh. I’d love to see a better format. After seeing so many archive formats come and go, I know it takes time for any format to catch on if at all.

[spwolf]

@By-Tor:

WinZip has a large userbase, regardless of whether we like it or not. 7-zip can handle most of ZIPX, with the exception of jpeg compression, but, Igor has stated he would implement some form of it if he saw a need.

ZIP by its nature is just a container format. So, you’re right, it doesn’t matter much what you call it. I’m just thinking about trying to capture the largest audience possible.

I’m by no means anti-PAF. heh. I’d love to see a better format. After seeing so many archive formats come and go, I know it takes time for any format to catch on if at all.

yeah, but there are much bigger differences to be made with new, proper format… you cant just put something into zip and get good results, it doesnt work that way.

compress folder with zipx lzma and with 7zip lzma, and notice that the difference in size could be up to 50%, simply because zip doesnt and will not ever have proper solid compression.

point with paf is that we can make something unique with it that nobody else has - compression for many popular formats… weather it becomes next biggest thing is something else, we plan to use it in our next backup utility as well, where nobody does any of these things and space savings will be significant.

[spwolf]

but pretty much, everything on that list takes at least a month to implement for single engineer… thats why nobody has done it before (expensive) and why all the new formats are very much alike (variations of old).

[By-Tor]

@spwolf:

point with paf is that we can make something unique with it that nobody else has - compression for many popular formats… weather it becomes next biggest thing is something else, we plan to use it in our next backup utility as well, where nobody does any of these things and space savings will be significant.

I totally agree. I wasn’t clear on how far reaching you wanted things to be.

[spwolf]

@By-Tor:

I totally agree. I wasn’t clear on how far reaching you wanted things to be.

it is massive undertaking though, and it takes a while…

but it will be worth it for instance for backups, where we will be able to do 50% more efficient backups that rest of the current utilities :)

[spwolf]

keep in mind we have been “researching” this for past 2 years, so we are not staring from 0 here…

things we did so far:

• ver 1 jpeg compression - 3x speed of wz jpeg with slightly better compression. Aim in PAF: maintain similar super speed while adding extra 20% compression compared to WZ Jpeg.
• ver 1 differential and versioning system - compared to leading Mozy backup service, we had 30% smaller differential backups (which means 30% faster, 30% more effcient, 30% less costly for bandwith and storage space), without strong compression implemented… Goal is to tie it together into PAF and enable stronger compression that would give us 50% gain total.
• Various multicore research into improving speed for other operations during compression, not just compression codec, that should add extra 20-30% speed improvements over current formats (with similar codecs).
• modification of lzma2 codec into pa-lzma, to fit our format better, and better overall performance (to be released as open source), this is 80% done as it is…