And here is the rebuttle to @jhowell "creation myth" for .kf8.
I have read it before. (Not sure if by him or someone else.)
I always found it incredibly hard to believe.
The argument goes as follows.
PLEASE UNDERSTAND, that we had to take away your ability to understand the format (= open standard, readable, tweakable, citable (citable outside the Kindle framework) - even human understandable for anyone without a reverse engineering toolkit (how many IDA Pro Usrs do you know in your community...?), and replace it with a "binary representation" -- we were forced to take this step -
because we wanted to give you 10ms faster pageturns. And this really was the ideal way of achieving that.
BULLSH*T
In a world where processors and memory (at least tepidly) get faster every year, and this fuels entire industies - the argument, that .kfx had to be created to service that need is and entirely false one.
Its in fact like saying "everyone had to be moved to the Spotify model - because silence in between two tracks could only be made smaller, going with the streaming model" - its missing the entire point, motive, drive, business model (#walledgarden), of an industry - looking at a small detail - and mistaking it for an "evolution" simply because it is new.
I don't care if the .kfx format has some tangible advantages for Amazon ("if it reduces CPU load, we could make our batteries even smaller!") - as long as we use that as justification for having lost the ability to understand the format.
(The loss is structural - because amazon insisted, that this would be the year where they should be the only ones that would be allowed to create their primary book format. Other people that had stakes in creating books in the past were just presented with a "service" to take over that chore fore them - they were basically bought out. Books have finally become "private goods" again.)
I don't - even for a second - believe that Amazons reason for doing so was to ultimately fractualy reduce CPU load on the Kindles.
Oh, the HUGHE CPU load of rendering a plain new text page every 20 seconds - contrasted with the performance specs you have to reach to make the UI navigateable - at all. Or buy books on the device within the store interface.
Books didn't have to become (scrambled) binaries, because of CPU load.
Thats a fairy tale.
Or to use the more technical term - a false dichotomy. A fools choice.
Last edited by notimp; 12-28-2016 at 12:56 PM.
|