mozilla foundation releases fire fox 22 this will change the web forever

Originally posted by jedi_hamster

Originally posted by XPLodER
i design protocols and codecs, do you? and the specific type of protocol used in VP9 in very different than any other codec produced, and i know how it works do you?

as a matter of fact, yes, i've played with designing codes, much more with analysing their internals. i've spend years dealing with different compression methods, coding them, and working with audio and video codecs at work. and i know for a fact that the terminology you're using, is fundamentally wrong. protocol - transmission protocol. codec - a piece of software dealing with encoding/decoding the data being transmitted. audio/video stream compressed using particular codec may be streamable or not - that depends on the codec itself - but the codec itself has nothing to do with the protocol that will be used to transmit the data it generates - as a matter of fact, many different protocols can be used to transmit the same data.

so either post your credentials or shut up, because you lie.

when you are designing a really good codec you can design one for a specific protocol design, that compliments the operation of both protocol and codec, optimized for video for an example.

i owe you nothing,
name calling is trolling behaviour,

xploder

reply to post by XPLodER


datatracker.ietf.org...

show me where does it confirm what you've claimed earlier.
oh wait, you cannot? how surprising.

you're making claims all over the thread that are nothing but a stockpile of BS mixed with outright lies and when caught red-handed, you call it namecalling? it is not namecalling. it is a fact.

Originally posted by jedi_hamster
reply to post by XPLodER

 

datatracker.ietf.org...

show me where does it confirm what you've claimed earlier.
oh wait, you cannot? how surprising.

you're making claims all over the thread that are nothing but a stockpile of BS mixed with outright lies and when caught red-handed, you call it namecalling? it is not namecalling. it is a fact.

i normally ignore trolls, but im in a debating mood,
list your "corrections" and quote my statements they correspond to,
show me lies you have accused me of,

and no name calling will be acceptable from this point forward (if you expect an answer)

xploder

reply to post by XPLodER


i've already pointed out few of your lies in this thread, and the most recent one - involving vp9 - you've tried to turn around, and when that didn't work - you've decided to ignore it. read my previous 2 posts in this thread and reply accordingly, with a proof that what you've said so far, is true.

otherwise all your claims about being 'oh so knowledgeable' about codecs, and presumably everything else (you didn't quite defend some of your earlier claims either, you just tried to 'smear out the BS'), are just plain trolling, and in fact you don't know that much, yet you post BS on ATS while trying to look smart, putting lies in the heads of not so tech-savvy people. nice try, kid. now go post crap somewhere else.

Enough of the personal jibes please.

Courtesy Is Mandatory - This is not a request, this is a requirement for membership.

Go after the ball, not the player.

Originally posted by jedi_hamster
reply to post by XPLodER

 

i've already pointed out few of your lies in this thread, and the most recent one - involving vp9 - you've tried to turn around, and when that didn't work - you've decided to ignore it. read my previous 2 posts in this thread and reply accordingly, with a proof that what you've said so far, is true.

otherwise all your claims about being 'oh so knowledgeable' about codecs, and presumably everything else (you didn't quite defend some of your earlier claims either, you just tried to 'smear out the BS'), are just plain trolling, and in fact you don't know that much, yet you post BS on ATS while trying to look smart, putting lies in the heads of not so tech-savvy people. nice try, kid. now go post crap somewhere else.

i gave you a chance to list my errors and why they are wrong (in your opinion)
and all you had to do was quote me and point out where you believe i was wrong
you did nether and continued to call me out.

make your points of correction,
and
show some manners
else

now go post crap somewhere else

follow your own advice ^^^^^^^^^^^^^

xploder

Opera.. always since 95… and it will stay ..!

Originally posted by XPLodER




sounds like something more than just FF crash bug,
try updating to todays stable release 22

if that does not work try doing a CCLEANER.exe

xploder

i do not blame the browser


the crashes come from the site your visiting or their servers that can't traffic so much.... FF then crashes ~not because of Its internal code~

imho

reply to post by St Udio


i recently downloaded opera,
as the changes made recently made me interested in seeing how well it works now,
must say im impressed

xploder

Originally posted by XPLodER

Originally posted by AfterInfinity
reply to post by Swills

 

I don't think it's Firefox that's giving you problems. As a matter of fact, I literally have 15 tabs open right now. I think it's your computer.

sounds like something more than just FF crash bug,
try updating to todays stable release 22

if that does not work try doing a CCLEANER.exe

xploder

Whatever you do do not do a ccleaner without tabbing through the auto checked boxes and unchecking loads of stuff or you will ruin your computer

reply to post by XPLodER


i've said this already, and i'll repeat myself so you can finally understand - transmission protocol has nothing to do with the codec itself. video encoded with vp9 or any video codec supporting streaming for that matter, can be transfered using http, prioprietary protocols like skype (just an example, i know that skype is not using vp9), and many others. you've made claims that you are knowledgeable about codecs - support your claims. i've posted a link to vp9 datastream description already, and i've asked you to prove me wrong and to prove that what you claim ('i design protocols and codecs, do you? and the specific type of protocol used in VP9 in very different than any other codec produced, and i know how it works do you?'), is true. you've ignored that. is it all just a wordplay to look smart? do you know anything at all? do you?

can you prove that what you've said is true, or will you ignore such request once again, backing up to personal attacks? because so far it seems that's all you've got.

Who here has read Jaron Laniers "Who Owns the Future?"?

I suggest you read it.

I'm by no means a technology buff, but I'm interested in the progress of technology and its wider socioeconomic implications. It seems to me - I agree with Laniers main argument - that this "open" web concept is intrinsically unsustainable.

The main problem is the trend towards depersonalization online, in tandem with the growing influence of what Lanier calls "siren servers" i.e. massive corporations like Google, Microsoft, etc, that essentially owns our information.

And don't think the latest Mozilla wont have a "do you agree" option before you download and upgrade this latest edition of firefox. They WANT your information just as much as google and others do; they make money off it.

Although Laniers suggestive remedy to this problem is likely to ruffle the feathers of many "open web" supporters, it does seem to be the preferable solution. He suggests that each person lose his online anonymity. We become who we are in real while online. This, it might be argued, is the next logical step in our social system. Right now, more and more jobs are being lost to automization. Can we escape this? Of course not. The fact is, todays job's will be performed by machines and computers tomorrow. So what can we do to stave this off, to help people live and thrive in a prosperous future economy? The information that is taken from REAL people and fed into algorithms needs to end; people need to be accredited i.e. payed, for the information they contribute to the success of whomever makes use of it.

It's either this, or the inevitable result of our current technological progress. Lanier acknowledges the problems involved in losing online anonymity which civil libertarians will no doubt notice. He hopes we can establish protections against any potential abuse.

In any case, make no mistake, more and more jobs will be lost because of the internet. Unless we figure out a way to make the internet profitable for everyone without creating massive disparities in wealth between those at the top - the siren servers - and everyone else, this problem will undoubtedly end up in socialism.

reply to post by Astrocyte


it's an interesting idea, and actually a pretty good one. i would sign up for that with both hands. unfortunately, most people aren't that happy to give up their anonymity. they mistake anonymity and privacy, while those are two different things.

who you are is one thing, what do you carry in your pocket or details about your private life, are another thing. people share their private information mindlessly, because they think they're anonymous, and because of that they won't give up that anonymity, because that would heavily compromise them. they behave on the internet like they're wearing masks.

do we wear masks on the streets? of course not.

it wouldn't be that hard to implement either. all ISPs would just have to update their router software to block all traffic that isn't signed by a key issued by public authority, and every person would need such a key to access the internet. every person would have to keep said key to themselves if they wouldn't want their online identity to be compromised.

imagine that all files posted online - documents, photos - are signed with personal keys. all emails, websites - signed. no more fraud, no more intelectual property theft, no more lies. of course, you'll get whole bunch of so called 'freedom fighters', or what i call them 'open everything and bend over maniacs', that will argue that it's an attack on their freedom. unfortunately, the idea of freedom as defined by GPL and as envisioned by people like Stallman, has been distorted beyond recognition, all to manipulate the masses. divide and conquer.

Originally posted by TheMagus
reply to post by Swills

 

Originally posted by AfterInfinity
reply to post by Swills

 

I don't think it's Firefox that's giving you problems. As a matter of fact, I literally have 15 tabs open right now. I think it's your computer.

actually the problem was plugin container.exe
had the same problem swills had a few versions ago
but it's seems to be fixed as of 2or3 versions ago

Actually, I've had problems with this too, and I've also identified the "plugin container" as potentially part of the issue.

I can have more than a couple tabs open. In fact, I can have quite a lot of tabs without major issues. However, certain websites seem to make it more unstable than others. In my case, basically anything using flash. Aside from the "plugin container.exe" I get sometimes one, sometimes two or more (rarely more than two, but it's happened) instances of "flashplayerplugin*****.exe". Sometimes even just a couple / few web pages using flash will begin to hang / freeze firefox really bad. Killing these processes in task manager will sometimes slightly resolve the issue, but not always.

I have also noticed the "memory" listed for the firefox process in task manager will slowly and steadily creep up and up and up until it does the same thing (even in the absence of flash).



Contrast this with a few years ago. A few years ago (and several firefox releases ago) I never had this problem. And I mean pretty much ever. I could run 150 tabs (pretty much literally-- I've done it) 50-75 of which had flash, with no issues. I don't even remember seeing these extra processes (plugin container, flashplayerplugin) several years ago. Perhaps I just didn't notice them because they weren't problematic.) But I did notice, with the last couple / few releases of firefox, the lagging, hanging, and crashing seemed to happen increasingly more often. To the point where now I'm getting angry at firefox on a daily basis.


And it could partly be my computer. It's not the awesomest sometimes. But it ran several previous versions of firefox without these issues-- so I don't get it.


I noticed others have had these issues as well. It's been brought up in the firefox forums. But they don't seem to have any solutions for us. I can tell you that if I can't resolve this issue, I may be trading in firefox for another browser, after nearly 10 years as a (mostly) satisfied user... I'm just almost as frightened about the alternatives...

reply to post by iwilliam


sandboxing of the flash player can be disabled - but that's not recommended. that process is in firefox for a reason - to separate flash from it, so when the plugin misbehaves, it won't bring down the whole browser with it immediately, so there will be no data loss.

it's not a browser to blame, but the flash player itself - i recommend using something like flashblock and activating flash on demand. flash has grown less stable in the past years, because it's using more and more hardware acceleration, so its stability depends on the latest gpu drivers.

and if your gpu is more or less old, you can check if the browser itself uses hardware acceleration for website compositing (about:support), and perhaps try forcing it off and see if that helps, especially when there are no newer drivers for your gpu.

Originally posted by jedi_hamster
reply to post by XPLodER

 

i've said this already, and i'll repeat myself so you can finally understand - transmission protocol has nothing to do with the codec itself. video encoded with vp9 or any video codec supporting streaming for that matter, can be transfered using http, prioprietary protocols like skype (just an example, i know that skype is not using vp9), and many others. you've made claims that you are knowledgeable about codecs - support your claims. i've posted a link to vp9 datastream description already, and i've asked you to prove me wrong and to prove that what you claim ('i design protocols and codecs, do you? and the specific type of protocol used in VP9 in very different than any other codec produced, and i know how it works do you?'), is true. you've ignored that. is it all just a wordplay to look smart? do you know anything at all? do you?

can you prove that what you've said is true, or will you ignore such request once again, backing up to personal attacks? because so far it seems that's all you've got.

edit on 22-7-2013 by jedi_hamster because: (no reason given)

its obvious that you are not up to date with new protocols,

A dash of algebra on wireless networks promises to boost bandwidth tenfold, without new infrastructure.

The practical benefits of the technology, known as coded TCP, were seen on a recent test run on a New York-to-Boston Acela train, notorious for poor connectivity. By increasing their available bandwidth—the amount of data that can be relayed in a given period of time—Medard and students were able to watch blip-free YouTube videos while some other passengers struggled to get online. “They were asking us ‘How did you do that?’ and we said ‘We’re engineers!’ ” she jokes.

www.technologyreview.com...

now if you know how this new protocol works, you can understand that it allows for a new type of codec to run over top of the protocol, one that can be specifically designed with packet loss mitigation (at the protocol layer) to allow the codec to function without resends. because the packet loss is mitigated at the protocol layer, the codec is not required to be loss resistant (software layer)

this means (if you know what your doing) you can design a new codec that exploits the fact that packet loss mitigation can avoid resends and therefore no RTT delay will effect goodput.

so what does this mean for the codec?
well usually people use UDP for low latency streaming applications,
you can now use the new TCP,
the new TCP is assured delivery and low latency combined.

when you look at the design of VP9,
it can be optimized for the new TCP.

that is why i look at the two (codec and protocol) as co operative.

here is a google video explaining VP9


if you know how VP9 works,
you will realise that some of the same design elements from the new TCP (protocol) is used in VP9 (codec)

i will not get into low level detail, you have not been polite enough for me to help you.

xploder

reply to post by Swills


You may want to upgrade your ram then. I'm running an AMD Phenom 1100T and have 8 Gigs of ram. Usually I have 5 or 6 tabs up at once (have 2 monitors) and I never have a problem using firefox.

Originally posted by XPLodER

Originally posted by jedi_hamster
reply to post by XPLodER

 

i've said this already, and i'll repeat myself so you can finally understand - transmission protocol has nothing to do with the codec itself. video encoded with vp9 or any video codec supporting streaming for that matter, can be transfered using http, prioprietary protocols like skype (just an example, i know that skype is not using vp9), and many others. you've made claims that you are knowledgeable about codecs - support your claims. i've posted a link to vp9 datastream description already, and i've asked you to prove me wrong and to prove that what you claim ('i design protocols and codecs, do you? and the specific type of protocol used in VP9 in very different than any other codec produced, and i know how it works do you?'), is true. you've ignored that. is it all just a wordplay to look smart? do you know anything at all? do you?

can you prove that what you've said is true, or will you ignore such request once again, backing up to personal attacks? because so far it seems that's all you've got.

edit on 22-7-2013 by jedi_hamster because: (no reason given)

its obvious that you are not up to date with new protocols,

A dash of algebra on wireless networks promises to boost bandwidth tenfold, without new infrastructure.

The practical benefits of the technology, known as coded TCP, were seen on a recent test run on a New York-to-Boston Acela train, notorious for poor connectivity. By increasing their available bandwidth—the amount of data that can be relayed in a given period of time—Medard and students were able to watch blip-free YouTube videos while some other passengers struggled to get online. “They were asking us ‘How did you do that?’ and we said ‘We’re engineers!’ ” she jokes.

www.technologyreview.com...

now if you know how this new protocol works, you can understand that it allows for a new type of codec to run over top of the protocol, one that can be specifically designed with packet loss mitigation (at the protocol layer) to allow the codec to function without resends. because the packet loss is mitigated at the protocol layer, the codec is not required to be loss resistant (software layer)

this means (if you know what your doing) you can design a new codec that exploits the fact that packet loss mitigation can avoid resends and therefore no RTT delay will effect goodput.

so what does this mean for the codec?
well usually people use UDP for low latency streaming applications,
you can now use the new TCP,
the new TCP is assured delivery and low latency combined.

when you look at the design of VP9,
it can be optimized for the new TCP.

that is why i look at the two (codec and protocol) as co operative.

here is a google video explaining VP9

if you know how VP9 works,
you will realise that some of the same design elements from the new TCP (protocol) is used in VP9 (codec)

i will not get into low level detail, you have not been polite enough for me to help you.

xploder

thank you for giving me a laugh

first, read the coded TCP paper at MIT website. it doesn't guarantee no resends, it optimizes them via optimization of ACKs and preventing window size to get smaller. it actually doesn't guarantee roundtrip time to be small, quite the opposite - with increased window size and thus increased throughput you can be sure to get increased latency as well - but that doesn't matter in case of video streaming, at least not in one-way streaming like in case of youtube.

second, all codecs that are streamable need - and have - some error concealment built in. and you've worded it like the fact that video being transmitted via coded TCP doesn't need to use error resistent codec, is a matter of coded TCP. it is NOT. it's the same case with regular TCP - of course, you'll get lower speed, but it will work - and it does, youtube is just fine so far.

third, you just said that vp9 'can be optimized for new tcp' - sure, you can remove error concealment from every single codec. the question is, what's the point of it if it won't help with anything and will actually break things for those willing to use it over UDP.

they can't even optimize frame sizes in the codec for coded TCP because the window size may change and video data is usually encoded prior to being transmitted, not on the fly and certainly not in response to ACKs for previous video frames, because for that you would need insane cpu speeds - quantum cpu probably.

will be continued.

so:
- you've claimed that 'new protocol' was used in vp9 and that you know how it works - now you said that 'vp9 can be optimized for it' and that 'some design elements were used' - both BS and backing up from earlier lies
- you claim that coded TCP guarantees low latency - a lie
- you claim that coded TCP provides packet loss mitigation - another lie

what's written in the article you've linked, isn't entirely true - it is copied from different article, which is wrong in the same way, including the header, and it was pointed out in the comments. if you would actually read the paper from MIT (which can be easily found within first page of google search results), you would understand that. copy-pasting random stuff from all over the web that you think you know what it means, just to look smart, won't cut it here, * snip * .

summary:
- removing error concealment from the codec just because it isn't needed for coded TCP, makes no sense and it won't happen, not in any reasonable codec. if the opposite would be true, it would happen long time ago - because it isn't needed for regular TCP as well. but guess what, UDP still has its advantages and it's still widely used.
- coded TCP does not guarantee low latency - because of the higher throughput and increased window size, quite the opposite - while that doesn't matter for one-way streaming, it matters for video call applications
- coded TCP does not provide packet loss mitigation - it optimizes the way retransmission works, but doesn't prevent it from happening.

codec - codes and decodes data
protocol - transmits data

protocol deals with signaling ('i received the last packet, send the next one'), codec does NOT. vp9, as a codec, has NO way in hell to deal with signaling - (new) TCP deals with that (or there is no signaling when you're using UDP and then error concealment in codec kicks in). video is encoded with codec before being transmitted, and codec can't even fluctuate specific frame parameters to optimize the datastream for specific protocol, because protocol parameters fluctuate as well on the fly as network conditions change during transmission - which happens long after the video was encoded.

and * snip * , new protocol? that MIT paper is 2 years old already.

bottom line:
you're using a bunch of tech-words copied from websites, thinking you know how stuff works, making claims that are nothing but lies, and when you reach the point where you don't know what to say - you try to imply ('i know but i won't tell you. do you know? blah blah blah. you have not been polite enough - blah blah blah').

* snip * ,

reply to post by jedi_hamster


i can have debates even with people i disagree with,
with completely different view points,
but not with people who are rude,
i have explained my position, and why i believe what i do,

i owe you nothing more.
xploder