4 ms·

All valid points, but think of all the open source projects, even browsers, that could be improved with this compression method if it wasn't proprietary. Perha

All valid points, but think of all the open source projects, even browsers, that could be improved with this compression method if it wasn't proprietary. Perhaps it's time to attempt to "free" such software via a cooperative effort, e.g. by talking to RAD about a one-time payment for an open source release and organizing that as a Kickstarter project.

vvanders 11y ago
That's the rub though, you're talking about a one-time payment. That isn't a sustainable way to keep a company running. The amount of work that RAD puts into stuff like this, BINK, and their other tools if you wanted to do a OTP you'd be looking at something in the neighborhood of 50-60M. Everyone uses their stuff in games since it's that good. You'd have to replace licensing costs not only for each game but each platform those games ship on.
- lambertsimnel 11y ago
  I wonder whether it would be viable GPLd with a proprietary licences available for a fee. I suppose that might still be costly and would be incompatible with the licences used by some FOSS projects.
- brigade 11y ago
  I can't find the post now, but one of the RAD developers behind this codec (it's their 3rd ANS based codec I believe) has said that they would likely be willing to open source their ANS family for a one-time payment of iirc around $5 million. No need for the entire RAD stack.
  - brigade 11y ago
    Oops, his ballpark was $50 million. I obviously have no idea how much things cost. Still, I think many of the key ideas behind Kraken are out there and not patented. Zstd and lzfse already use ANS variants, for instance. But they're targeting different speed/compression tradeoffs, and aren't implemented as well as Kraken.
    - eln1 11y ago
      Google VP10 and WebP are also planned to be ANS-based: https://chromium.googlesource.com/webm/libvpx/+/fb9186d68d498919b290d7f444560f44a1d8d1df https://chromium.googlesource.com/webm/libvpx/+/fb9186d68d49... https://chromium-review.googlesource.com/#/c/338781/ https://chromium-review.googlesource.com/#/c/338781/
_3u10 11y ago
Think about how browsers could be improved if people were willing to pay $5 for them. Perhaps it's time to attempt to "pay" for such software via cooperative effort, eg. by the publisher setting a price for it, and then people paying $5. Why people will pay $5 for coffee they could make at home for 'free' but not for software is beyond me.
- voltagex_ 11y ago
  A lot of the time people don't understand how expensive software is to develop.
- overcast 11y ago
  It's a tangible thing. People STILL have issues paying for something that isn't physically in their hands.
  - dilap 11y ago
    Nah, I don't think it's that -- people spend crazy amounts of $ for non-tangible assets in F2P games.
    - ferbivore 11y ago
      I don't think you're talking about the same subset of people here.
      - dilap 11y ago
        what, people who use browsers and play games? there's got to be a pretty big overlap :) (though it's true that a large fraction of the people that play f2p games don't pay at all.) i know for me personally, one reason i've been reluctant to pay for software is just that they paying-for-software experience has been so terrible -- you have to clumsily reenter your payment information, you get emailed, sometimes slowly, a registration code, which you inevitably lose and breaks your software when you have to reinstall on a new computer, etc.
  - 11y ago

vvanders 11y ago

That's the rub though, you're talking about a one-time payment. That isn't a sustainable way to keep a company running. The amount of work that RAD puts into stuff like this, BINK, and their other tools if you wanted to do a OTP you'd be looking at something in the neighborhood of 50-60M. Everyone uses their stuff in games since it's that good. You'd have to replace licensing costs not only for each game but each platform those games ship on.

lambertsimnel 11y ago
I wonder whether it would be viable GPLd with a proprietary licences available for a fee. I suppose that might still be costly and would be incompatible with the licences used by some FOSS projects.
brigade 11y ago
I can't find the post now, but one of the RAD developers behind this codec (it's their 3rd ANS based codec I believe) has said that they would likely be willing to open source their ANS family for a one-time payment of iirc around $5 million. No need for the entire RAD stack.
- brigade 11y ago
  Oops, his ballpark was $50 million. I obviously have no idea how much things cost. Still, I think many of the key ideas behind Kraken are out there and not patented. Zstd and lzfse already use ANS variants, for instance. But they're targeting different speed/compression tradeoffs, and aren't implemented as well as Kraken.
  - eln1 11y ago
    Google VP10 and WebP are also planned to be ANS-based: https://chromium.googlesource.com/webm/libvpx/+/fb9186d68d498919b290d7f444560f44a1d8d1df https://chromium.googlesource.com/webm/libvpx/+/fb9186d68d49... https://chromium-review.googlesource.com/#/c/338781/ https://chromium-review.googlesource.com/#/c/338781/

_3u10 11y ago

Think about how browsers could be improved if people were willing to pay $5 for them. Perhaps it's time to attempt to "pay" for such software via cooperative effort, eg. by the publisher setting a price for it, and then people paying $5. Why people will pay $5 for coffee they could make at home for 'free' but not for software is beyond me.

voltagex_ 11y ago
A lot of the time people don't understand how expensive software is to develop.
overcast 11y ago
It's a tangible thing. People STILL have issues paying for something that isn't physically in their hands.
- dilap 11y ago
  Nah, I don't think it's that -- people spend crazy amounts of $ for non-tangible assets in F2P games.
  - ferbivore 11y ago
    I don't think you're talking about the same subset of people here.
    - dilap 11y ago
      what, people who use browsers and play games? there's got to be a pretty big overlap :) (though it's true that a large fraction of the people that play f2p games don't pay at all.) i know for me personally, one reason i've been reluctant to pay for software is just that they paying-for-software experience has been so terrible -- you have to clumsily reenter your payment information, you get emailed, sometimes slowly, a registration code, which you inevitably lose and breaks your software when you have to reinstall on a new computer, etc.
- 11y ago