Burst.com's streaming snake oil

Salon recently ran a two part feature on Burst.com. The feature talks how Microsoft stole Burst.com technology, and positions Burst.com as a David to Microsoft’s Goliath.

Salon’s articles are normally wonderul. The problem with this one is that the author apparently didn’t have the technical background to see through Burst.com’s “technology”.

While I was Apple’s QuickTime Evangelist, I was a magnet for all kinds of folks who claimed to have miraculous codecs and other holy-grail technologies. Burst.com claimed to have a revolutionary way of delivering streaming content. Lossless. Faster than realtime.

Well, golly. You can deliver content losslessly and faster than real time via HTTP and FTP, too. Only Burst.com did this with a magical, proprietary protocol that required a magical, proprietary server that they would be happy to sell to you. The secret of the “secret sauce” that Burst.com CEO Richard Lang mentions in the feature is that there is no secret sauce.

Mr. Lang believes that Microsoft was out to get him. However, the reality is that Burst.com was, at best, a fly to Microsoft’s mountain.

Now Burst.com is suing Microsoft, a move apparently prompted by Windows Media 9’s “Instant On” feature. If you have a really fast conneciton and there are no bottlenecks along the way, it lets you see/hear media almost instantly. It works by putting a huge buffer at the client, and then filling that buffer as fast as possible so that buffering time is minimized.

QuickTime’s “Fast Start” provided much of this functionality with QuickTime 3’s progressive streaming (1998), and QuickTime 6 added the final missing piece (random access) with its Instant-On feature earlier this year. RealNetworks uses a similar method to optimize the viewing experience in RealSystem 9.

Burst.com is one of the last vestiges of the P.T. Barnums that contributed to the terminal over-inflation of the Internet bubble. It’s time that they go away, and I’m rooting for Microsoft as they pound the nails in the coffin. Microsoft’s media monopoly Goliath crushes David

MPEG LA extends AVC "essential patents" submission deadline

MPEG LA has extended the deadline for companies to submit patents neccessary to support AVC, from October 11, 2002 to January 31, 2003. That makes sense, since AVC hasn’t quite reached FDIS (Final Draft International Standard) status, and “essentiality” — whether a patent is essential to AVC — can only be determined against the final text.

MPEG LA licenses “patent portfolios” for technologies like MPEG-2, FireWire, DVB-T and (soon) MPEG-4. In other words, they’re the one-stop licensing shop that companies will go to when they want to use MPEG-4 Visual or MPEG-4 Systems patents.

AVC (a.k.a. H.264, JVT, etc.) is the forthcoming MPEG-4 video codec, which MPEG intends to finalize in December. MPEG-4’s current video codec offers very good quality, but not enough to match the quality of proprietary codecs even with all of its quality-enhancing “tools” enabled. The perception of MPEG-4’s current video quality is not helped by implementations such as Apple’s, which offers quality levels significantly below what today’s MPEG-4 Video codec is capable of.

Qualcomm's new 3G chipset supports MPEG-4

Qualcomm has announced that they’re shipping an important new mobile chipset for 3G (CDMA 2000 1X) devices, and it includes their “Qtv” MPEG-4 decoder.

This is a significant bellwether for MPEG-4. It says that Qualcomm’s customers are demanding MPEG-4 — and as significantly, not Windows Media or RealSystem — in the DNA of their mobile wireless devices. Link

RealNetworks releases open-source client

Last night, RealNetworks announced to Helix Community members that its Helix DNA Client source code is available to registered users.

The jury is still out on whether RealNetworks’ open source are going to make a difference. The Helix mailing lists have been dead quiet, suggesting massive disinterest. The RealNetworks licenses don’t help — for example, anyone who uses the software for anything but research purposes can expect to be hit with licensing fees. If there’s a strategy (other than hoping developers are compelled to add value to the RealSystem platform out of the goodness of their hearts), it’s not apparent to me. Is it to anyone? Link

UPDATE&nbsp&nbsp&nbspI tuned into the webcast, hoping to get an insight into RealNetworks’ strategy. Rob Glaser said that 2,000 developers had joined “Helix Community”, and claimed that that was really good. Well, this little site has around 2,000 regular readers, and that’s with basically no marketing efforts thus far. I’m thinking, “They can’t do better?”

Rob then trotted out pre-recorded clips featuring VIPs from Nokia and Palm Source. The Nokia VIP could’ve been talking about Windows Media or MPEG-4 — he didn’t say anything specific to RealSystem or RealNetworks. The Palm Source VIP read (poorly) from an off-screen card, and strongly encouraged Palm OS developers to, erm, do something with Helix DNA Client for Palm OS.

There may have been more, but just as an OpenWave VIP started to talk my RealOne Player crashed spectacularly, and by the time I was able to fire it up on a different PC they were already into the part of the webcast that told developers how to compile the code. Is a webcast really the best way to tell developers how to compile Helix DNA Client?

As I signed off, RealOne Player alerted me that an upgrade was available. Nice touch, at least until I figured out that “upgrade” actually meant “upgrade to the non-free player”. Sheesh.

Flash for Palm OS

Today, Macromedia demonstrated Flash Player on one of the new Sony Cliés at DevCon 2002. According to the presenter, “Flash Player is built into the device”. This is very significant news for digital media content developers since it suggests that Flash will be the de-facto standard for rich-media on PDAs, and for Palm since Flash playback was previously limited to WinCE devices.

Windows Media Audio reverse-engineered

Microsoft’s very-proprietary Windows Media Audio format has been reverse engineered as part of the FFmpeg project. Codecs from the FFmpeg project are licensed under the LGPL and are used by several other projects.

Now, the only hitch to building Windows Media Audio-compatible players is Microsoft ASF-related patent, which I personally believe is invalid based on prior art. But it seems hopeless for individuals or small companies to challenge it, and not a good business move for larger companies to challenge it. Proprietary formats suck, don’t they?

Get ready for Wi-Fast: 802.11a and 802.11g

Wi-Fi (a.k.a. 802.11b, a.k.a. AirPort) is mainstream. We know this because my mom recently asked me what “wardiving” was. I calmly explained that “wardiving” was a tragic occurance in which Slashdot readers die while looking for open wireless access points at the bottom of the ocean. The good news is that when they wash up on the beach, the gear is top notch.

Wi-Fi is about to shift into 2nd gear with 802.11a and 802.11g. Where today’s 802.11b supports data rates of up to 11 Mbps, 802.11a and 802.11g support data rates at up to 54 Mbps. 802.11g works in the same 2.4 GHz band used by 802.11b, which (as apartment dwellers are painfully aware) is polluted with interference from cordless phones, microwave ovens and (in the future) Bluetooth. 802.11a works in the UNII (Unlicensed National Information Infrastructure) band recently allocated by the FCC.

Although there are some products that support 802.11a — for example, Linksys has an 802.11a access point, and recently announced a dual-band 802.11b/802.11a access point — only now are low-cost wireless chipsets appearing that support 802.11b, 802.11a and 802.11g. This should enable compelling, affordable products by mainstream hardware manufacturers (including Apple and Microsofft) by this summer. Link

Consumers shun copy-protected audio CDs

In a survey recently released by GartnerG2, 77% of respondents said that they should be able to copy CDs for personal use in another device, 60% said they should be able to give copies of CDs to members of their families, and 82% said that they should be able to copy CDs for personal backup purposes.

And why shouldn’t they? I’ll certainly be doing all of these things, whether or not the RIAA says I can.

74% of respondents also said that these AC/CDs — they’re not CDs, so I’m going with “Anti-Copy Compact Disc” for now — should have giant orange warning labels, presumably so that they can avoid them. Link

Wireless data services: "You are here"

A Red Herring article discusses where we are on the high-speed wireless data curve, comparing this period of network upgrades to the analog-to-digital transition of the last decade.

…the industry remains in the midst of its most difficult transition since the migration from analog to digital voice service in the mid-’90s. Back then, the business suffered through a period of substandard growth. Users didn’t want to buy new phones that would operate on the soon-to-be-outmoded analog network, but they also were reluctant to buy digital handsets until the necessary networks were built. Once nationwide digital networks were complete, however, handset sales took off, and the industry boomed.

Although there are useful insights in the article, it’s a bit confused. The article talks about a “transition” from wireless voice to wireless data services, but obviously data services simply supplant voice services. It’s also contradictory, in the subtitle asserting that “communications devices reflect the limitations of their underlying network” and then later declaring that “handsets that exploit the existing network will be far more compelling this holiday season than they were last”. It also claims that “Motorola has the most compelling handset product cycles”, which is completely ridiculous IMHO. Caveat lector. Link

Microsoft's Palladium (DRM) talk at MIT

Arnold Reinhold has posted a summary of the technical overview of Palladium that Microsoft recently presented at MIT.

Palladium is a collection of Microsoft technologies for enforcing system-wide DRM. The technologies fall into four categories: (1) “Curtained memory” ensures that code can’t observe (read) or modify (write) other code’s memory. (2) “Attestation” means that code can “attest” that it’s data was created by it and belongs to it. (3) “Sealed storage” means that only code that created the data (or code it trusts) can get to it once it’s stored to a non-volitile storage medium. (3) “Secure input and output” means that communications to the keyboard, mouse and display are encrypted. Palladium achieves all of this through a virtual PC running it’s own OS (“Nexus”) that communicates to the real PC via agents.

How to prevent the secure channel to the on-screen window from being spoofed is still an open problem. Brian suggested a secure mode LED that lights when that window has focus or having the secure window display a mother’s-maden-name type code word that you only tell Nexus. Of course this doesn’t matter for DRM since your trusting the window is not the issue.

Sheesh. Get ready for your forced move to a gated community where you can’t see what’s in the refrigerator unless you put it there. Link