“The promise of HTML5’s video tag was a simple one: to allow web pages to contain embedded video without the need for plugins. With the decision to remove support for the widespread H.264 codec from future versions of Chrome, Google has undermined this widely-anticipated feature. The company is claiming that it wants to support ‘open codecs’ instead, and so from now on will support only two formats: its own WebM codec, and Theora.” Sorely disappointed in Ars’ Peter Bright. Us geeks reviled web developers for sticking to Internet Explorer when Firefox came onto the scene, and yet now, the same arguments we used to revile are used to keep H.264 in the saddle. How us mighty geeks have fallen.
The most ridiculous statement being, firefox support h264 through microsoft plugin, As far as I know google is dropping the native support of h264, support through 3rd party plugin is still possible.
it’s going to be just like chromium supports video codecs – through external library
It is obvious to anyone from my posts that I don’t agree with this guys conclusion. That said, I read his piece with great interest and unlike most of the silliness I usually see posted about this he presented the arguments accurately and rationally, there really isn’t a single factual error in the whole article. I give credit where credit is due.
He is absolutely right the argument against h.264 isn’t about “openness”, it boils down to money. To me though, how you see this depends on which side of the fence you are looking at it from.
If you are a business that distributes video content through multiple channels, there are certainly downsides to Google dropping support for h.264 – at the very least it complicates your options. On the other hand, if you are interested in actually participating in the development and enhancement of video distribution on the internet webm is a great thing – because without it unless you are financially backed by a large corporation you are not allowed to participate in any meaningful way. Sure you can work on x264 or some other encoder/decoder, but you can’t actually productize it – actually implementating your work into any kind of product is simply off the table with h.264 if you don’t want to play ball with MPEG-LA.
He is also right that webm is no more “open” than h.264, they are both open in the sense of being transparent from an implementation point of view. The real difference is the royalties.
This is how it all lays out to me. Google dropping h.264 will more than likely increase the uptake of webm at least to some degree, which will result in hopefully a royalty free standard that has enough legs to at least survive. Best case scenerio is that flash supports webm container format in the near future – this would, for the first time (I think ever) give a royalty free video format an across the board delivery mechanism – <video> tag for those that support webm and flash as a fallback. I think the existence of such a scenario would open the floodgates to large scale participation in the video ecosystem for the “little man” – the gates keepers would lose their power. That imo is more important than the “practical” concerns he expresses in his article.
I don’t expect h.264 to go away, that is a pipe dream being peddled by irrational people – it ain’t going anywhere. Google dropping support for it in their browser is nothing but a glancing blow – and the big players are so enthralled with DRM I don’t see what difference it makes – they were never going to using HTML5 video anyway… If there is a step backwards here for the commercial world, it is a very very tiny step…
I think the most important thing to remember here is what Google did was for the video-tag, nothing else.
If most used thus default or standard codec/container is open (ok, not patent encumbered) then that would be a (very) good thing.
There is hardly any existing content which uses the video-tag. So there is still a lot to win here.
If you want to productize x264 without paying anything to the MPEG-LA, just loan a server in the software patent free area. Almost anywhere will do provided the country does not have software patent laws and that is most of the world: Russia, India, China, The Philippines, etc will do.
Just avoid Australia (the worst one), Japan, south Korea and the US. Europe is mixed. They don’t allow patents on “programs for computers” but they allow patents on “technical solutions”, even if that solution is applied using a computer program. The risk is still high so I would avoid Europe too. New Zealand is software patent free for now but that may change once the drafted bill has passed so I suggest to stay out of it as well.
The downside is that the consumers in software patent areas will have slower access since the server will be far from them. That may be solved by faster intercontinental networks or by just not using h264 indeed. You may use both formats and only mirror the Theora and webM content to servers in the software patent zone.
Edited 2011-01-13 10:12 UTC
There is one other fact that is important, and which everyone seems to be trying to omit.
The reqirement for any codec (or any technology for that matter) which is specified as part of HTML5 is that it must be royalty-free.
This cannot be emphasised enough. Being “open” (i.e. specification is published) is not the only requirement, being “royalty-free” is of primary importance.
H.264 does not meet the primary requirement. WebM does.
Edited 2011-01-13 10:53 UTC
If Google really wants to promote WebM and WebP is by providing a DirectShow coded (that actually CAN encode as well) so that people could rip/convert their DVD/BR/DivX/Xvid into WebM format. Unless WebM/WebP aren’t more widely supported, they’ll be no chance to get them adopted. Same happened for Divx, no one tried to convert their movies in that format until portable player were able to display them. Now some tepid manufacturers do support MKV. If the promised WebM hardware-based support happens to be true, 2011-Q2 will be pretty interresting to say the least…
Kochise
The really killer question is why aren’t Microsoft and Apple supporting WebM or OGG? I mean it’s not like they couldn’t.
Why attack Google for choosing to drop support for H.264 when they are just as guilty of not supporting WebM/OGG?
Until at least DirectShow and Quicktime come with WebM and OGG this is no more than a few companies trying to raise a barrier to entry for web video.
They are supporting it. Microsoft has already announced of an official WebM support for IE9 whereas Safari is able to support it if the user installs the codec for Quicktime.
I don’t remember the situation with Ogg Theora, but I don’t think missing it is going to be anything serious anyways. It’s available as a plugin too, anyways.
Edited 2011-01-13 09:22 UTC
Are you sure? I remember there was a leak that the latest build of Internet Explorer 9 supported Vorbis but I’ve heard nothing about WebM other than the statement that if a WebM ‘Media Foundation’ plugin is provided that Internet Explorer 9 will use it.
Microsoft used an HTML5 audio tag with both MP3 and Vorbis _sources_ (such that it would play on both IE and Firefox). This doesn’t in any way mean that IE would support Vorbis. They pulled it quickly because they don’t want to be seen promoting or using Ogg codecs at all; being part of the MPEG-LA pool, and all.
I highly doubt MS cared how they looked due to their membership within MPEG-LA. MPEG-LA isn’t important, since they don’t own a significant enough portion of the patents for it to be a lucrative business for them. IIRC, just paying for the h264 support that comes included with Windows costs more than they get in h264 royalties as a whole.
More likely, they would have attempted to hush-hush the vorbis business since the IE team have been trying really hard to repair their image in the web browser market. The last thing they’d want is for web devs to start using vorbis, if the IE team haven’t committed to supporting it and/or possibly can’t support it due to licensing*.
*I have no idea how vorbis is licensed and if this creates any issues.
Is there any word on Microsoft supporting WMA/WMV/VC-1 using the video and audio tags? From what I understand the licensing for WMA/WMV/VC-1 is a whole lot nicer when compared to h264 whose license is so horrendously complex that it can reduce a harden lawyer to tears. I mean, worse case scenario there is always WMA/WMV/VC-1 and Microsoft could provide free plugins for alternative platforms if the whole h264 becomes a giant fiasco.
Edited 2011-01-14 02:09 UTC
VC-1 is also handled by MPEG-LA (falls under some of the same patents) and is under similar terms. As for providing codecs for alt platforms, they haven’t so far. Windows remains the only consumer-level platform from which you can encode to VC-1.
I was under the impression that VC-1 as it was based on WMA/WMV was a ground up clean room CODEC developed by Microsoft but from what I have read via wikipedia and a few other sources it is still heavily reliant on MPEG-LA given the origins of its formats before Microsoft joined MPEG-LA. I sometimes wish there was a way that the US government can mark something off to force it to be patent and royalty free under the argument that it is of strategic national importance.
They’re part of the MPEG-LA pool but what they get back from their contribution in terms of patents I doubt they’ll get back in terms of it offsetting licensing costs. So I doubt it has anything to do with their relationship with MPEG-LA and more to do with corporate politics.
Microsoft for example hates relying on third parties hence they’ll prefer getting a hold of the code or creating it in house than being reliant on an external project – hence the accusation by some that they suffer from the NIH (Not Invented Here) syndrome. Apple on the other hand has this idea that there is only one way of doing something – that developers and end users just have to jolly well put up with it. Microsoft will provide multiple frameworks addressing particular scenarios where Apple will develop one framework and tell developers, “here it is, if you want something different tough toe nails”. It might address the 66% who sit in the middle of the bell curve but it ends up excluding an awful lot in the process.
As I said, it’s a corporate culture rather than something nefarious – add to that Apple being hardware and software it means having a much simpler set of technologies to support, reduced support costs thus higher margins and higher profits. Common sense would tell you that maybe Apple should support it (WebM and provide a way for people to add support manually) and let consumers and content creators decide but Apple has this idea that it is up to them to craft the industry by erecting barriers to prevent the supporting of other technologies. We already see it with QtKit and the lack of support for plugins as one example – something that I think isn’t an accident but a deliberate attempt to funnel everyone into using the formats that they deem ‘acceptable’ in the Mac universe.
Edited 2011-01-14 02:06 UTC
Microsoft are usually using their own modular software. It seems likely they are going to use Windows native videoplaying software complete with codecs etc., this also gives them high quality hardware accelerated decoding for free. If they do this they would have to active _block_ WebM to prevent it from working in IE9, and thiat seems unlikely While I can’t remember where, I think they have confirmed that they will be able to play WebM if you download and install the codec yourself.
Apple on the other hand are being much more closed. Not sure if it is because they are not using the quicktime framework in Safari or what.
Edited 2011-01-13 21:24 UTC
Yes you are correct, it was over at the Internet Explorer blog where the developer said that they won’t be providing support themselves but if WebM developers wish to create a ‘Media Foundation’ plugin (Media Foundation replaces DirectShow plus a few other frameworks) then it’ll work with Internet Explorer. If WebM takes off and there isn’t a way for WebM support to be added easily to Safari then it’ll added to my list of reasons why I should reconsider Windows as an option.
My assumption was that Apple uses QuickTime for their video playback – are they bypassing the whole framework altogether and hooking directly to the CODEC itself? I remember that QtKit is in a state of incompleteness hence the reason why Apple, Adobe and a few others use a bridge between their 64bit applications the Carbon 32bit QuickTime framework hence I wonder whether it is a temporary thing until the QtKit framework is complete. What ever the case maybe I think you’ve provided more incite as to why it isn’t as simple as providing a CODEC for Mac OS X if Safari flat out refuses to acknowledge the plugins existence.
Not in a meaningful way they don’t. DirectShow and Quicktime don’t come with any out of the box WebM or OGG support. So for all intents and purpose they only support H.264.
Until they do I call ball$ on their self-serving hypocritical war against dropped support for their favorite codec.
Google could easily pay OEM’s to have it pre-installed on computers shipped just as they do with the Google toolbar, Chrome, Google Earth and other stuff that they make available.
Even with such support on the desktop there is still the issue of handheld devices where neither WP7 or iOS support WebM and there is no way to actually provide support via a codec either. If Google does win the battle of WebM on the desktop they would have lost the war because a huge chunk of the market would have been lost to h264.
Edited 2011-01-14 01:52 UTC
Mobile is only like <2% of traffic. Don’t overestimate its importance at this time.
That is in the developed world – in the developing world the mobile phone is the primary form of accessing the internet so don’t under estimate the long term impact of these trends. You can sit around dismissing it claiming it is all a fad, that it’ll never catch on but we all know of companies who have said similar phrases in the past only to later regret it.
Edited 2011-01-14 02:10 UTC
HAHAHA!
HAHAHA!
http://www.youtube.com/watch?v=eNxygsLGHSQ
I was quite surprised by this article as ususally articles at arstechnica are of high quality but this one is really incredible.
H.264 is a video format made by pooling software from various company, the idea being of paying companies roughly with the number of their patents used, so being in favor of H.264 means *de facto* being in favor of software patents..
So arstechnica is saying that refusing to support software patents is a step backward for openness???
The mind boggles!
I agree
Pure technically he is right, webm has not been submitted to a standards body thus their is no open specification. So it is a step back for open.
But what Google obviously means, which he left out of his article, with open is open to participation for webdevelopers. Which webm does allow webdevelopers to participate freely (yes, without having uncertainty about payments later).
I also don’t agree with his assessment that GIF is still used on the web because it allows animation and PNG does not. That is not true. PNG does allow animation (APNG), IE does not support it though. That is why it never caught on.
Many versions of IE don’t support PNG-transparency either and those that do badly at that.
It doesn’t help that APNG is only supported by Gecko and Presto (Opera), meaning WebKit-based browsers like Chrome and Safari don’t animate it either.
…or that, because it’s a violation of the “PNG header means single frame image” promise in the PNG spec, web developers like me who accept PNG uploads add small code snippets to either refuse the upload or sanitize it by conversion to single-frame PNG.
No I mean, because IE does not properly support it it never got populair which is probably why it never made it into webkit
Ars technica is very much filled with Apple minions 😉
The worst thing you can say about the Ars article is they don’t explicitly label it an editorial piece but rather a ‘feature’. Ars do occasionally post something up that they know will be controversial.
Google is the decision maker in online video, remember that. Second, they ship with flash so no way that playing youtube videos on that player will be impossible – especially as they probably already play all videos in both formats.
Now Firefox and Chrome already captured a huge percentage of browsers – with Firefox even leading over IE in Europe and Chrome, the best browser on the market IMHO, is close to 15% so their choice is more than relevant, being together at around 50%.
I understand your (OSNews) opinion on the topic.. h264 vs openness… Making so much noise about this.. it’s because that allows you to show once again Google as the good soldier ?
In all those discussion about h264 going against web openness, you refuse to bring up the topic about Flash because Android ships with the (not so good & proprietary) Flash player. And Google is “proud” of that. Why not writing several articles pushing Google to abandon Flash and help Apple in encouraging everyone to go HTML5 with no flash ? That’s the real challenge for web openness. For once the “devil” is doing the right thing for openness and has been somewhat successful in it, but you can’t give a single nice word about it..
Enough has been said about Google’s real motivation : money. We all know Google’s strategy about openness and open source is only valid because their main revenue source is the ads and search engine – a monopoly. Have they open-sourced their secret sauce for that ? no way ! They can afford playing that open-source/openness game in about everything else because they have that constant revenue stream with their search/ads business.
Note : I am supporting WebM – choice is always good. But stop saying it’s for openness : h264 somehow helps moving towards HTML5 without Flash. I’d rather have openness advocates spend their time on a that battle against Flash – that has been the one thing really polluting the Web with a real proprietary technology that is not standard at all.
It’s running away from Charybdis to Scylla.
WebM (or similarly-licensed codecs) is the only acceptable video codec. Period. You can say whatever you want about Google, it has no weight on WebM.
As for the “noise”, you should look around: the furor and snark about the drop of h264 by Google is unbelievable, it is even worse than when Apple decided to bar flash from all iOS devices – despite the fact the arguments against WebM are far shoddier.
Wow, that’s esoteric.
The flash problem can be solved with gnash. Some things may not correctly with gnash but the most important ones do. Hopefully gnash will improve over time and implement the remaining bit and all flash content will work. Gnash is a top priority project for the FSF.
I have been keeping away from this discussion enough now.
A lot of you people miss one vital point here:
Flash, as utterly damaging it has been, and is to the web, is FAR BETTER than what we will have with H.264!
I cannot stress this enough…
Indeed.
Freeware VS rapeware.
Are you high? H.264 is the path of least resistance to delivering video through Flash and to iProducts; if anything, supporting h.264 encourages the continued use of Flash for video.
So instead of having web video technology largely controlled by MPEG-LA + Adobe, you’d prefer to just give the MPEG-LA total control? Brilliant! And what’s the rationale? Something like “the situation is already FUBAR, so why bother even trying to keep it from getting worse”?
It’s also damned funny that most of the handwaving rhetoric about Google “forcing” WebM on us boils down to a fear that Google might potentially resort to the same kind of childish playground-politics that Apple has used in Jobs’ little anti-Flash jihad.
And if Google is “evil” for “forcing” WebM on us, then what does that make Apple? Unlike Apple & Flash, Google hasn’t engaged in any sort of propaganda/disinformation campaign against h.264 or approached large users of h.264 & tried to convince them to ditch it.
I have always thought the video tag should support any type of video supported by the host Operating System’s Multimedia components.
I can see browser vendors including their “first-class” video support for H.264, Theora, and/or WebM, but for video formats the browser itself doesn’t recognise, it should see if there is a decoder installed on the system and use it if it’s there.
That puts the responsibility back on the web developer to provide whatever video format will be viewable by most of its’ customers.
If Google wants to use WebM on YouTube, it won’t be any harder to install the codec than it is to install Flash if it isn’t present. Same with Theora, and if they still want it to be Free beer, H.264.
The advantages are there too; if Netflix, for example wanted to serve movies as HTML5 video with some DRM infested codec, they could simply provide the decoder, and it would just work rather than needing to install Silverlight.
I wish companies would look at the big picture, and rather than trying to figure out what they want to support, think about what they have to do so they don’t need to add another redundant component that is already provided by the OS to their browsers.
Problem is, browser vendors can’t fix a hole in a third-party codec so this can become a security threat. That’s the reason why Google provides flash and a pdf reader with chrome. So they can fix security threat in a timely manner.
Too bad I cannot mod you +1 after I’ve posted, but YES!
For the love of God, don’t give us even more third-party security holes. Getting rid of flash is suppose to make the web BETTER. Not worse…
Do NOT let the codec pack of the week handle playback in our browser. Please, I get enough spam as it is!
This I was unaware of:
http://www.nvidia.com/object/tegra-2.html
Tegra 2 supports hardware accelerated VP8 en and decoding. This means that virtually ALL Android tablets shown at CES support VP8, as well as many of the Android phones shown at CES.
So, the lack of hardware argument is even weaker than I thought. More and more the impression arises that people like Gruber and Bright have some sort of vested interest (my my if only Ars hadn’t just launched an online video service for which they probably had to pay MPEG-LA… Oh wait*).
I obviously don’t believe this is the case (I’m not a conspiracy nutter), but still – it’s getting uncanny.
* http://arstechnica.com/site/tv.ars
I highly doubt Gruber has a vested interest in H.264. He just thinks from the perspective of the regular, non-FOSS/opensource fan userbase that just cares if their browsing experience works or not.
You know what the average person is going to think the first time a video doesn’t work because there’s no H.264?
“Wow, this browser sucks. I’m going to use something else”.
I get that “open” is better. I get that being beholden to a patent troll sucks.
However the vast amount of people Don’t. Give. A. Shit. They just use computers. They’re not a part of some emotional movement.
Same goes for Apple leaving Flash out of iOS, or VLC insisting that the player gets removed from the App Store. Sure, it’s perhaps the Right Thing To Do in the interest of freedom (VLC being removed), but it just ends up hurting the general computer-using public.
I think a lot of open source/FOSS folks really over-estimate the average computer/electronics user, the people that actually spend the most money and use most of the devices and browsers in question.
Edited 2011-01-13 15:11 UTC
…And that is how, by looking only at short-term convenience, you end up digging your own hole. Look how the south Koreans are locked into the Microsoft ecosystem because of the choices of their gov & companies decades ago. Look how the web relies too much on flash because everybody though “I don’t care about standards and openness, I just want to make a website the easiest way”.
Let’s make the same mistakes again and again! Five million lemmings can’t be wrong!
This is a (long) summary of my feelings on HTML5 video and H.264:
– H.264 is not something everyone is allowed to implement
– H.264 streaming of paid-for content requires royalty payments
– Decoding and encoding of H.264 requires licensing
How can communication be free (in both the sense of “speech” and “cost”) with H.264 if not everyone is free to implement it? One of the greatest things about the internet is that it is the “great equalizer”… everyone, regardless of means, can have a voice, compete against the biggest players, and have a chance to succeed. Never before has any medium allowed for such wide-reaching open communication and expression by ANYONE.
Can you imagine if a person had to get licenses for every file format type used on the web? For GIF images, that’s one license… for JPEG, that’s another… same goes for PNG, HTML, CSS, Javascript, SVG, etc. With HTML video, I think everyone would like it to become as ubiquitous as those existing formats. By allowing H.264 to become an acceptable format for the web, a worrisome precedent is set. As new formats and types of content appear and the web integrates them, how many more licenses will a person need? How much will need to be paid in royalties? In the long run, even the amount to be paid for licensing and royalties could be minor compared to the complexity of simply making sure you or your organization stay within the letter of the law with all of the content you provide.
Innovation on the internet has been driven by the ability of ANYONE to be able to implement the standards used on the web and the ability of ANYONE to be able to use those standards to share their ideas with. The internet would not be where it is today without that. Firefox could never have existed without that. Chrome probably would never have existed without that. Even Safari probably would never have existed without that.
Something that the Arstechnica article seems to miss however is that the likelihood of VP8 infringing on H.264 patents is just as great as H.264 infringing on VP8 patents. The patent infringement guessing game is something that really can’t be played here. No one seems to be able to say say with any confidence that any given codec doesn’t infringe on any patents. MPEG-LA seems afraid to bring infringement charges against either Theora or VP8 which leads to the conclusion that they don’t really have a case against Theora or VP8, the patents they are claiming Theora or VP8 are infringing on are weak and are likely to be thrown out, that they maybe counter-sued for infringement of VP8 patents, or a combination of any of those things.
Google seems to be receiving a lot of flack about including Flash with Chrome however; people are crying “hypocrite!” at the top of their lungs. Keeping Flash bundled with Chrome however, with the web in its current state, is a pragmatic decision. Only a small fraction of the web is currently using the HTML video tag while Flash is already deeply rooted. Google made their move with the video tag at a time when it was still feasible, while Flash can’t be dropped nearly so easily. The use of MP3 and AAC in the HTML audio tag is however a different matter; this is a case where I will agree with most that if Google is dropping support for H.264 over the matter of openness and freeness, MP3 and AAC should go as well.
Other people are arguing that this move will stop big content providers from supporting the HTML video tag, but they seem to forget a critical thing: big content owners/providers such as Netflix, Amazon, and Hulu INSIST on DRM. There is currently nothing in any proposed HTML standard, let alone any implementation, that meets this requirement. While the debate on if DRM is truly needed is for another time, the fact of the matter is that these content providers will continue to use Flash or Silverlight as their delivery mechanism until they are satisfied that HTML can provide them with the DRM they feel is necessary, so in no way is this move detracting from the openness of the web.
The last group of people who complain about this move are those who are complaining that they will now have to maintain two versions of video files. Don’t worry too much: Adobe has already promised to support WebM in a future Flash release allowing you to still have a single file that is served to the HTML video tag and Flash. It’s obviously not a solution that works right now, but it surely will eventually.
Google is doing us all a tremendous favor by dropping H.264 video from Chrome. Could they pay the needed licenses and royalties? Of course. But I think they realise more than most people how important having standards that can be freely used and implemented are to innovation and diversity on the web. By making this move, and hopefully moving YouTube to WebM, Google is pushing both Apple and Microsoft to allow the web to stay open. Without this move, both IE and Safari would be unlikely to ever support Theora or WebM, although I fully admit that IE and Safari supporting these formats will nearly take and act if God: if they ever include these formats, that’s openly admitting that they believe these formats do not infringe on any patents or that they are OK with the infringements which is unlikely to happen while both Apple and Microsoft are part of MPEG-LA. The main people who are complaining are only looking at the present, where content creation software already has excellent support for H.264, Flash can play H.264, portable devices have dedicated hardware for decoding H.264, and most OSs already have the required codecs installed for playing back H.264. However, it is incredibly shortsighted to allow H.264 to become the de facto web standard. For the long-term health and continued innovation on the web, an open, free to use and implement standard is required.
I would mod you up but I already posted. Nice summary.
Great post, the only comment I have is on “The use of MP3 and AAC in the HTML audio tag is however a different matter; this is a case where I will agree with most that if Google is dropping support for H.264 over the matter of openness and freeness, MP3 and AAC should go as well. ”
I think MP3 is close enough to patent expiration for decoding (The initial near-complete MPEG-1 standard (parts 1, 2 and 3) was publicly available in December 6, 1991 as ISO CD 11172.) that support for it does not need to be dropped, since the monopoly can probably only last until about December 2012.
http://en.wikipedia.org/wiki/MP3#Licensing_and_patent_issues
Well, that may be the case, but until December of 2012, MP3 is still a a format that is illegal to decode without proper licensing and in the Wikipedia article there doesn’t seem to be a reference as to when an encoder would be legal to distribute without a license, which is important as well. If we cannot produce the content freely, get the content to the users for no additional charge, and freely reproduce the content in a human usable form, this format is unsuitable for freely transmitting ideas by everyone regardless of their financial means, therefore it has no place as part of a web standard. Don’t get me wrong, H.264, MP3, and AAC are great formats, but they should not be standards for the web.
I agree with you in your context that MP3 should not be part of the standards for the web until decoding and encoding can be freely done. I think that there is less motivation for Google to remove the MP3 decoder than things like AAC and H264.
I am guessing that once MP3 decoding is royalty free, it should be possible to make a MP3 encoder, but it may be a pretty bad one. The other year that is mentioned is 2017.
The linked article states “But all the while, GIF was widely used across the web, and there’s no real sense that the web was held back by the GIF patents.”
GIF images could be freely decoded, which is a huge difference. Encoding was the only part that was patented. Since decoding was patent free, legal open source decoders could be written. So in the GIF case, legal open source browsers could be written that displayed the images. With H264, decoding is patented, so legal open source browsers cannot be written that display the images.