Posts by austin cheney 20 publicly visible posts • joined 30 Dec 2009
"Jobs is also right regarding performance. It is one of the slowest interpreted languages around. Perl, python, java are all significantly faster."
Actually, JavaScript is considerably slower: http://www.themaninblue.com/writing/perspective/2010/03/22/
How is any of this about open standards if Apple has gone to incredible effort to lock down its IPhone/Pad platforms? There are death row inmates that have more freedom of expression than IPhone app developers. Additionally, there is all this talk of openness and a patent pool at the same time. Clearly, the point of this patent pool is a gathering to find grounds to engage in legal action opposed to protecting their legal ground from action. I want to say this all my Apple-hating cult-defying bias except that Apple did put out a HTML5 demo page that is all about this openness, but was in fact not open.
This talk of openness in the context of Apple's conduct and business direction seems more than an ethical accident and rather a blatant lie. Honestly, I am absolutely astonished people are still drinking from the Apple punch at this stage in the game. If Apple were really about openness they would allow Flash on their platforms and provide their user's a method to disable Flash that is engaged by default. They would also not publish open standard demo pages that are anything but open, as well as a host of other openness defying behaviors.
The technologists, the actual smart people, came up with such a thing as a Web 3.0. It was called the semantic web. It is largely ignored by an industry whose only means of revenue is advertising. In this case all that matters to business is that traffic is being generated. Why risk screwing that up by moving the technologies forward and doing right by the technology?
Mozilla is just as fast to jump on the HTML5 bandwagon as everybody else. They too call it a standard and claim standards support. HTML5 is not a standard. This is what happens when idiots jump the gun. The purpose of standards is conformance in the industry. HTML5 is not a standard so should anybody really be surprised that there is no conformance? Microsoft will only end up making things worse by jumping the gun, even if later in the game than everybody else. This is the kind of circus you get when there are no adults driving the bus and when anybody who disagrees that premature adoption of HTML5 as a holy prophecy gets shut up and flamed.
HTML5 is rapidly becoming a failure and its only going to get worse.
The problem is expensive sucky licensing on one side of the issue and Apple's intent to form a patent pool to troll about the web and work very hard to discover a form of punishment that need not exist. Patents could go away, but these two problems would still exist as in H.264 would still be expensive and Apple would still be an ass to openness.
"Firstly, users of the iPhone (and possibly other embedded devices) have no way of installing an alternative codec"
If you choose to purchase a proprietary device limited to certain limited proprietary software then that is your mistake and yours alone. Don't punish the rest of us for your poor decision.
"so all video sites end up having to encode things twice"
That is the nature of giving your users a choice. Mandatory elimination of such choice is no different than forcing clothing vendors to carry only one brand of clothing or forcing only one means of licensing upon software or forcing consumers to purchase only one brand of automobile or forcing all software vendors to adopt the same single licensing scheme. Choice is a good thing.
"with no guarantees that a particular format will work everywhere"
There has never been the expectation of such a guarantee before. The only difference now are the cost interests of the web services industry. The only group that benefits from the elimination of choice are those who bear competition derived costs in a diverse market. Competition and choice is always good for the consumer even though there are business costs associated with such.
"and secondly the terms of the license prevent anything that is GPLv3 licensed from working with h.264 video due to the patent restrictions"
That is all the more reason to prevent elimination of codec choice.
"This means that it's quite important that an open codec becomes the de facto standard"
If the standard is specified in HTML5 then it is not de facto, and that defeats the entirety of your point.
Plugins don't have anything to do with this as it is about playing video directly in the browser. Furthermore, I never had the problem you speak of since many media players typically have come with many codecs supported in advance except possibly Real and Quicktime. As a result I would simply use my choice as a consumer to never download or play Real or Quicktime media.
Why is it there is so much opposition with giving users a choice? Isn't choice really the only difference between an open format and a proprietary format?
For some faulty reason the browser vendors seem to think there must be a single video codec. One that rules them all. Instead why not allow media to be served in my flavors and let the user decide. They should simply submit an RFC requesting a name to IANA that assigns mime type names to each codec send media annotated by the correct mime type in the HTTP header.
Apparently people have forgotten the early days of video served from the web and there were a massive variety of formats from plain old AVI, DIVX, XVID, ASF, Real, and so on. Various codecs for choice have some how magically never been a problem before until HTML5 came along. Imagine that. Apparently whether or not media displays directly next to browser text or in an external application inspires mental illness.
Where is the option for senior management? Since there is no such option I cannot participate in your poll. Your might need to review your pollster's CISSP credentials a bit more closely.
Just to be clear "The Business" is not the senior management. If all decisions were left purely to business there would be no security and we would be drowning in popups and advertisements.
I have patent pending technology about some minute novelties in a markup language created for email instead of various social deformities of the web. In my experience I have learned that you cannot trust software companies to follow a specification before releasing a software product using something from that specification or in essense of that specification. I suggested my technology be patented, because then I can sue software companies for fucking it up even if I am allowed to give licensing away for free.
Imagine what the web would be like if violations of concerned technology standards resulted in legal damages removal of the faulty software product from the market. I can't count on a scientific calculator how many times I have heard people complain on this website that Flash should die or in other places that constant infantile crying about the evils and trials of IE.
Now imagine if a markup language were created for email and software companies were allowed to deliver crap without regard for harm to the network technology on which that markup language exists. I would rather give the licenses away for free and use patent fear to keep people from half-assed using my technology to destroy email.
I say I partially disagree, because I can't help but admit there are parasites who do nothing more than hold patents in expectation of litigation as a business. Those people should die. If a patent is not used solely to protect the entry of a technology or art to market then it should not be granted the benefit of protection in court.
"If engineering specifications are good enough for contracts between corporations then they should be good enough for patents. Likewise, since almost all patents and inventions are essentially engineering in nature, arbitration in disputes should be done by a panel of engineers rather than a colony of lawyers**."
While that sounds like a fantastic idea in theory, I bet it would prove more faulty than even the current system which I think you described incredibly well. The problem is that engineers do not always understand the inventions of others in fields they are presumed to be expert authorities upon. A case in point is the nature of semantics as a computer science of relational data in the XML, and perhaps even amongst some few enlightened folks of the more archaic SGML field, versus the understanding of semantics to those who value only product usability, such as the HTML5 field.
If a panel of "those" experts were charged with deliberating over my markup patents I would be upset. No, no upset is too soft a term. Homicidal is even far too soft to fully describe the rainbow colors of rage and disgust. The problem is that even engineers disagree upon engineering concepts even when those concepts are defined in engineering specifications written solely by fully qualified engineers.
While it appears to be a fanboys wet dream to see Flash die I fail to see how this change offers users any benefit, but certainly increases harm. It seems in media and advertising based web businesses usability is everything and security is a pain in the ass tack on. Clearly, no thought to security was considered in this migration, but when you sell advertising or offer media based services it seems security is entirely irrelevant. Logically, though, why should they care? The security problems will exist primarily on the client side, typically execute on load, are silent to the end user, and they generate revenue regardless. Afterall, its not as though client-side scripting from the web is a bastion of security or that exploits or rare. Its all that Symantec measures anymore in its annual Internet Security Threat Reports.
Before anybody starts crying that I don't know that I am talking about provide any example of XSS exploits used to maliciously alter swf files. You can't, because swf files are compiled to bytecode. They have to be created with a malicious intent, or they are not malicious. This can never be said of JavaScript that is what makes Canvas interactive.
If somebody ever wrote a Firefox plugin that prompted users any time JavaScript executed potentially malicious code, such as back doors for other malicious code or exportation of personal information via unauthorized AJAX, people would say far different things of such migrations. But, since the security problems don't slap us in the face with obtrusive prompts, they must clearly not exist. Out of sight out of mind, right?
Thats nice, but it sounds like an alphabet soup mess. Instead people could use a simple solution like integrating the behavior of modern spam filters with content formatted by a markup language designed specially for email. Fortunately, somebody built such a thing:
http://mailmarkup.org/
I entirely agree with Mozilla's position on this issue. The problem is not whether a particular technology is open or proprietary but what the conditions of licensing are upon that technology. Mozilla's position is that it could not have entered the market if fees were attributed to common standards and will not condone or support such fees as they would limit future competition from entering the market like wise.
I completely support the idea of proprietary standards as technology standards on the internet so long as fees are not a condition of adoption, distribution, processing, or other associated open use. Open standards are slow to form and slow to adapt to a rapidly moving market, where proprietary technologies are not so burdened.
Additionally, the second part of this problem is that HTML5 is attempting to bind media codecs into its markup language standard for no other reason than to benefit content distributors at everybody else's cost. There are various different video and audio codecs that are preferred by different people. In an open environment people have choice. Since the Internet and the World Wide Web are about openness then it must be media agnostic to not interfere with such media processing choice.
The only appropriate consideration with regard to media codecs is to treat them like how browsers treat fonts. There is not a standard font for the web and I do not see HTML5 pushing a standard font that everybody must use. Instead users are burdened to download and maintain fonts on their own. When a user does not have a certain font that is requested to present online content then the concerned content fails to render as expected. Media codecs should be treated no differently.
Since HTML5 wishes to propose media processing barriers in opposition to openness I cannot conclude that it is an open technology dispite its claims to the contrary.
Seriously, who didn't see this coming. As of this point last year 95% of all security vulnerabilities are the result of some form of client-side scripting from the web and 70% of those are aimed at some form of application memory corruption, such as buffer overflows. The only solution for that 70% is to uninstall the software and the only known mitigation is to patch as often as possible. There is no solution for the remaining 25% of client-side scripting vulnerabilities except to abandon use of client-side script execution entirely from the browser. People generally do not seem to want to accept that mitigations are not solutions and the degree of the problem.
HTML5 fits the very demands the of the current attack vectors. As an emerging technology that is both incompetent and irresponsible. People should seriously consider migrating towards another platform for an application environment since, in this regard, the web is a failure and HTML5 does everything possible to stroke that failure.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
