Become a fan of Slashdot on Facebook

 



Forgot your password?
typodupeerror
Slashdot Deals: Prep for the CompTIA A+ certification exam. Save 95% on the CompTIA IT Certification Bundle ×

Comment Re:Rotten apple ?!? (Score 1) 196

Because apple, who make a huge amount of noise about wanting to protect their dear beloved users dont disable the storage of and access to the security tokens when their devices are jailbroken?

THATS the story here, they could, however they do not. Hence they have left the apple IDs knowingly open to theft.

Users, for better or worse, have convinced themselves that Apple keeps them magically out of any such trouble, however this is a clear
case where they could, but they do not. Which is a pity.

Come on Apple, the obvious fix is to make the secured data inaccessible once jailbroken.

Comment Re:From TFA: bit-exact or not? (Score 1) 172

I would really REALLY suggest you spend a little more time researching those other compressors you so easily consider to be 'text streams', they are not.
for example, one of them also happens to hold the current record for non lossy image compression..

Its all a matter of feeding them the right models, and I can guarantee that a good PPM or CM set of models will do much better than a weeks worth
of model development - but of course they reason they WILL is because they take care of the downstream details - the work you have done in finding
context is exactly what they do need.

Remember, there are three stages to compression, and using 'state deep within a video decoder that doesn't apply to text streams (like what above-neighbor color presence bits are set)' is the top level - finding context to model. What I would suggest is that the decades of research as to how best to utilise that context
could be of use... then again perhaps you have done better than they can - and that is what testing against the corpus will show.
When it comes to non lossy compression, there is no such thing as a text compressor, there is no such thing as an exe compressor, there are just different
models of data, and different ways of using those models.

You are not the first, or I would suspect the last to look at bitstream detokenisation and recompression in its many forms..

If you dont read up on this, you are missing something that matters, for example:
https://en.wikipedia.org/wiki/PAQ
http://www.squeezechart.com/bitmap.html
http://mattmahoney.net/dc/dce.html
http://www.maximumcompression.com/data/jpg.php
But then perhaps you are aware of that all.

Dont get me wrong, 22% is VERY respectable on jpeg.. but why not try to do better.

Comment Re:From TFA: bit-exact or not? (Score 2) 172

Its good that you understand that bold claims require clear evidence.. Thank you for replying.

It is not surprising you can compress h264 using a 4mb block and token decode/recode, because of course that means you are using more resources than it (as you state) and removing functionality..
I refer you to the following, hopefully you are aware of it..
http://mattmahoney.net/dc/text.html
Perhaps you should try your core modeling/tokenising against that, then consider how the ones that beat you do so.. not as an insult to your systems
but as a guide to current advanced techniques. IF you cannot match them, perhaps you should consider why and if using some of those techniques
would help... (hint: they will)

BTW, by your description your system is not useful for streaming - streaming requires the ability to both recover from errors rapidly and to enter a live
stream at an point withing a small window - that is pretty much WHY h264 has to reset state with great regularity. If you cannot do that then you do
not support streaming.

Towards the end you seem to be talking at odds to your 4MB block.. you claim you only need a single previous frame for decode, and that your memory requirements are small.. If that is so then I would suggest that there is other memory also being used.. or you are not fully utilising your 4MB block.

Just a suggestion, you should compare yourselves with h264 that is extended to use similar resources - that can still be beaten (as it must support streaming and you dont), but you will find its compression goes up significantly - even though you are going 'off book' with respect to its standards.

What you seem to be doing in effect is decoding the h264 token stream and then recompressing that without some of the functional demands that cause h264 to be structured as it is - that works, just be aware of the limitations you create - they are not just there because of 'committee'.

Comment Re:From TFA: bit-exact or not? (Score 0) 172

Sigh, another person who doesnt actually know what arithmetic coding IS.
Your first statement is completely false. Arithmetic coding is demonstrably perfect within 1 bit over the entire stream at optimally representing the token distributions you give it. Of course you are confusing it with the distribution model and tokeniser preceding it.

You second statement is also completely wrong, unless the application is 'a system to compress pi'. If it is a system to compress arbitrary length decimal numbers, then good luck compressing pi.. There have been attempts at algorithmic source derivation compressors.. none have even come close to working.

Your third statement is of course true, however what on earth that has to do with the subject of arithmetic coding I would love to know. It is true only in the context of the tokeniser and statistical model driving the arith coder.

Your fourth statement is just wasting everyones time, as it has nothing to do with the problem at hand.

Let me throw another one in for you, just for fun.
You will notice I have concerns about the runtime (therefore run energy) and memory footprint costs of this full implementation. There is a damn good reason for that which IS related to your last statement. To develop a system that actually saves organisations something, the cost in extra time, energy, and resource must significantly beat the cost of the extra storage required to store it without such treatment.
Yes, reducing lets say dropboxes total storage requirements by 20% would be a saving, but not if it doubled their computational costs..

Welcome to the real world. This has been looked at many times, and the questions that matter are well established.

Comment Re:From TFA: bit-exact or not? (Score 1) 172

You seem to be confused as to what arithmetic coding is..
What you seem to be talking about is the accuracy of the token counts being used to drive the arith coder.. arithmetic coding says nothing about those, except that they have to exist.
Beating a given implementation? of course, there are several ways..
But claiming to have better arithmetic coding itself is silly, what you have is better token distribution figures.

Want to pony up some estimates on performance and memory requirements?

Comment Re:From TFA: bit-exact or not? (Score 1) 172

And just to reply to myself.. it is generally a BAD idea to imply you have an encoding method better than arithmetic (lets hope
the article horrible miss quoted you there..
'yet it is well known that applying an additional arithmetic coder to existing JPEG files brings a further 10% reduction in file size at no cost to the file," he says. "Our Pied Piper algorithm aims to go even further with a more efficient encoding algorithm that maps perfectly back to existing formats."'
As of course it is a numerical impossibility to be more efficient than correctly implemented arithmetic encoding. But of course you know that right?
Its the modeling of the distribution per token before the arith that matters of course.

Comment Re:From TFA: bit-exact or not? (Score 2) 172

OK, As you are the author..
Care to comment to the performance and window length of your encode/decode?

As of course there is an innate difference between algorithms that must run streaming (for example... h264) and ones
that can consider all of the content - the same for computational complexity - for video to be useful it must decode in
real time on 'normal' machines.. Memory footprint for the compression window also matters a lot..

My guess is that your decode overhead is not high, but you need a LOT of memory resource to hold your decode window,
and that encode performance is horrific as you need to search a long way for matches.

If that is true, then its (unfortunately) just a case of not much to see here - as I am sure you know longer windows = better
compression. If it is not true (you are doing this with short windows and low encode overhead, then congratulations, it
may well matter.

jpeg is well known these days as leaving significant lossless rate on the table due to computational limitations when it was
created. h264 does the same because of its need to support reasonably live streaming latency and be implemented in hardware.

Comment Re:I volunteer as tribute. (Score 2, Insightful) 381

No you are wrong, almost certainly because you are addicted and in denial.

Sensible diet and exercise WILL reduce weight in someone who is obese. By definition. It cannot not work.
There are a few factors however that damage such perceptions.

1) a persons metabolic rate determines what is 'enough' food. person A may be able to eat twice as much as person B for the same effect, but people now
expect to be able to eat until they feel full. In fact they practically demand it.
2) people expect to undo YEARS of overeating with a month or twos diet - that is not going to happen (at least not in a healthy way)
3) people lie - mostly to themselves. They convince themselves that today is a special day, so they can eat more - or that today they dont need to
exercise because of something or other.. but of course they forget that.. human nature.
4) food is habit forming and addictive in many forms. exercise is hard, especially if obese. People are getting less and less willing to make the effort.

your 2% figure, which of course we understand is pulled out of your arse, is of course a silly form of self denial.

Comment Re:Done to _gouge_ the customer better (Score 1) 378

I think it should be unlawful

Yeah, because random thoughts should become laws, because you think they ought to.

How about government stop trying to fix businesses making stupid marketing decisions? You, as a customer, can take this information (its free) and use it to find a better alternative. If you can't find a better alternative, then Xerox (in this case) has proven its case and gained a customer.

Fantastic, all for this.
While we are at it I assume we will be reversing copyright extension, DMCA, and anti reverse engineering laws to allow people to return to having their full first sales rights and therefore the ability to reverse engineer, second source, etc to avoid these gouges?
I assume we will also be enacting full free trade laws so that consumers are free to source their most competitive products without government sanctioned protection of manufacturers?
And while we are at it bring in more stringent consumer protection laws, against false advertising, hidden pricing, warranty avoidance, etc?
And while we are at it, how about protecting these protections by limiting any political contributions to money paid by private earning registered voters that they have paid person tax on?

Great, I cannot wait.

Comment string.. (Score 1) 176

A piece of wood or a rock, some string/rope tied to it, job done.
Much easier to 'hit' it, and once the string is tangled, its not going anywhere.
You can even pull it back for another 'shot' if long enough

Of course if its further away, not so easy.. but then its also probably not yours to 'shoot down' to say the least.

Leaving alone somewhat the question of if you SHOULD 'shoot down a drone'. I wonder how many people would walk
up to a stranger, grab their camera and throw it hard to the pavement - its much the same thing really.
They would have to be doing something pretty damn wrong...

Comment Re:And it all comes down to greed (Score 1) 585

Did you ask your English teacher to teach you basic grammar? Just wondering..
I would suggest your education loans and CS degree were at a minimum a waste, in fact I truely wonder how you got one..

Oh,and its interesting that you assume American made stuff is somehow better ;) Its great to claim supporting local products is
supporting local jobs, but dont try and paint that as 'better quality', often it is not.

Comment Re:And it all comes down to greed (Score 4, Insightful) 585

That would have been true about 15 years ago.

These days you seem to have forgotten that most of the truly physically productive jobs have already been outsourced to other countries (China, Mexico, etc), and you are now sliding down the slippery slope of those countries refining their ownership, capital, and trade situations to take advantage of this.

Japan was once the 'cheap labor' for the US.... and yet people never seem to see a trend.

My only suggestion? Stop damn well following the crowd and consuming every little luxury you can convince yourself you deserve on credit!
Learn a practical skill or three (and no, manipulating office politics is not a practical skill).
Live somewhere that is sustainable n what you can actually contribute.
Take a walk in a park on a nice day and revel in the fact that you are alive.. and that doesnt actually cost (well, much).
Oh, and perhaps treat your friends/family with due care and respect, because when shit happens - everyone needs some support.

Oh, sorry, not in line with the American Dream? oh well, good lucky with that.

Comment Re:"...the same as trespassing." (Score 1) 1197

And if he climbed over that fence because some crazed nutter was chasing him with a gun and he was trying to run away?

Or perhaps he is the new neighbor and his kids favorite toy had got over the fence - he had tried to knock on your door (you didnt hear),
so he thought you were out and didnt want to cause a bother and just get it.

Dont worry, his wife and children should be quite happy after you explain it was just an accident.

He: Let's end it all, bequeathin' our brains to science. She: What?!? Science got enough trouble with their OWN brains. -- Walt Kelly

Working...