Disqus - Latest Comments for gzino

Re: Don’t Make the Internet Safe for Monopolies

gzino — Mon, 24 Apr 2017 22:36:00 -0000

agree, Tom, and nicely put @Daniel Berninger "the prospect of government as the antidote to telco/cableco misbehavior is laughable given the track record you note":

also you could argue the hypothetical ISP behavior of prioritizing their own content (or purposely degrading content from others) would actually help open the door for alternative ISPs and methods.

note: we do have a real need to increase local ISP competition. unfortunately, "net neutrality", regardless of how it plays out, is serving as a distraction to addressing that need.

Re: Sending Stuff To The Wrong People

gzino — Mon, 23 Feb 2015 09:42:32 -0000

gmail mobile upgrade rolled out to two of my Androids this wkd; perhaps related. I see the same (bad) behavior on those, one non-upgraded Android and web.

especially for sharing cloud content (dropbox etc) with groups (as opposed to individuals), will it be much longer before API integrations enable us to choose a group to share with (defined by a name we give it), rather than need to type in list of emails?

Ideally with a 'group DNS' type layer that enables us to define a group as collection of mobile numbers that can then be used by multiple mobile apps (with sufficient authentication layer)?

Re: Dear Internet: Let’s Demo The Slow Lane

gzino — Wed, 07 May 2014 11:34:27 -0000

Agree on the themes however we (US) are still top 10 in the world for retail IP, and most of the rest of the world is not close, so this is not just a US problem (consider China and India alone; e.g. I have employees in India that can't get a good enough retail connection to do a video call with me from home, and that's using a video service that we specifically engineer to deal with bad networks).

That said, agree "real broadband" is critical to our future, it is inexcusable that we don't get there and a slow lane day is a terrific idea for raising awareness.

Re: Is Google Inadvertently Supporting SBCs?

gzino — Wed, 07 May 2014 11:22:02 -0000

I am anti-trapezoid for many use cases, and video transcoding is painfully expensive, but we can't put this all on Google's broad shoulders or even the video codec:

1. Even w/ homogeneous media codec, firewalls, web proxies, security policies and NAT are making most implementations into trapezoids, especially in the enterprise space. Sometimes "just" signaling but often media as well.

2. Many (if not most) developers still favor VP8 as a single MTI.

3. Video is only one part of WebRTC. Even apps that do need video have options for many use cases, e.g. try P2P video but if it doesn't negotiate then try P2P voice or trapezoid video. Of course (1) above may mean that neither P2P voice nor P2P video work...but in that case the issue is not the video codec...

Re: The decline of the mobile web

gzino — Mon, 07 Apr 2014 15:12:10 -0000

Share the concern but not sure if the data is skewed by mobile gaming (and mobile gaming is more gaming appliance replacement than it is web replacement)?

Re: Teaching Computer Science To High School Students On The Way To Work

gzino — Wed, 02 Apr 2014 09:13:11 -0000

Awesome. Hope to see this model extended across other education domains as well, e.g. life sciences, increasing the probability that each student discovers experts and fields that resonate.

Re: Sidecar

gzino — Wed, 19 Feb 2014 10:41:36 -0000

Agree, personal is certainly difficult to build and even harder to scale, but we will see a wave of attempts.

On the demand side, most verticals are in races to the bottom, meaning most people/users/consumers are eating commoditized products/services and developing an appetite for more personal ones.

Meanwhile, on the "supply side" (in a sense), we know mobile/compute/Internet is making it feasible and viable for some companies to feed these appetites in profitable and sustainable manner.

Re: Sidecar

gzino — Wed, 19 Feb 2014 09:50:18 -0000

Love the Etsy-Amazon analogy.

Believe personal will be the next wave in a few verticals. We know personal is hard to replicate, commoditize or automate.

Whereas efficiency can be race to a wide bottom - one Pacific Ocean "winner" per widely defined industry, believe personal can be a climb to the top of many peaks - customized/individualized/narrower scope sub-verticals.

Re: Google’s three Ps

gzino — Fri, 24 Jan 2014 17:59:48 -0000

Great article.

This sentence is interesting:

"For all the technology innovations the company can claim, it has not had one internally sourced business model innovation"

I would classify their initial product - AdWords - as business model innovation. I know Yahoo (and maybe others) dipped their toes in selling ads based on a user's perceived intent (as determined by their search strings), but Y! and others always fundamentally seemed to be basing their business models on selling content? And of course Google wrapped self-serve etc. around AdWords, essentially inventing a business model?

Maybe the argument is that Google's investment in search algorithms essentially caused that business model, but I am not sure if Google would have invested in their search technology (and self-service etc) to the degree that they did if they weren't convinced of the business model itself (and ultimately funded by it)?

These days, I completely agree: one of the world's most powerful labs and we are often at the mercy (good or bad; knowingly or unknowingly) of the experiments.

Re: VC Pitches In A Year Or Two

gzino — Wed, 15 Jan 2014 23:04:17 -0000

"The problem in the USA is too much regulation and artificial monopolies, not a lack of regulation". Perfect line.

And I would add that the outcomes we are most worried about - harming the little guys in order to build their own profits - are not nearly as likely as we fear.

With or without NN regulation, the ISPs would do it if they could do so in a sustainable, profitable manner. 15 years of VoIP experience tells me that they can't.

And, if I am wrong, and they can, then they would find a way to do it around the regulation, w/ the regulation being costly, full of unintended consequences, and hard to unwind when the world changes. See the current PSTN landscape for a perfect example.

Re: Taking To Dos and Moving Up The Y Axis

gzino — Mon, 16 Dec 2013 08:48:03 -0000

Excellent way to articulate that framework.

Probably there (haven't watched the video yet) but imo the key is the steps before the conversations:

1. Hire the people that fit what you need and don't compromise
2. Put those folks in a position to succeed - high expectations but viable expectations
3. Nail the "what you want to accomplish in the meeting" part - focus your "request"
4. Keep eyes and ears open - objectively process the inputs - tweak 1-3 accordingly.

When those elements are in place, very few conversations tend to move up the y-axis to begin with, and the good manager manages the ones that do.

Re: Thanksgiving

gzino — Tue, 12 Nov 2013 17:20:26 -0000

in response to the sm pings...yes, my post was intended to be critical of wal-mart's announcement to start their black friday sales on thanksgiving...sorry i don't do a great job writing these types of snarky posts.

Re: Decoding WebRTC with Erik Lagerway

gzino — Mon, 04 Nov 2013 14:06:50 -0000

Erik makes great points - the API/SDP/signaling are far more important than video codec.

That said, I believe innovative implementations such as Hookflash will make these issues mainly moot in the long-term. It will cover them up from end user perspective and developers will be able to *afford* the extra time, pain and cost.

Telecom is moving from needing to innovate and solve problems in purpose built hardware to software and web solutions. Orders of magnitude faster and cheaper, and ultimately more powerful. That is the disruption of WebRTC (and similar), imo.

Re: VP8 vs H.264 in Real Deployments

gzino — Thu, 31 Oct 2013 14:53:52 -0000

The codec discussion itself is very interesting and gets even more interesting when we consider the rest of the package.

Does the device have hardware echo cancellation? How good are the jitter buffer implementations and packet loss concealment methods? What is the webcam doing or trying to do when it "sees" constraints? Is the client too CPU-intensive? Etc.

This will be a critical issue for the industry as WebRTC helps move communications off of dedicated, purpose-built, integrated hardware/software, and we simultaneously get increased variety in the last mile (3G, 4G, WiFi, asymmetric DSL, etc.).

Re: A Stealthy Video and WebRTC Acquisition

gzino — Thu, 31 Oct 2013 14:44:39 -0000

Great points. Hadn't thought about it in those terms.

If you are a Fidelity-sized shop, and you want to RTC enable your applications in a high QoE, secure, user-friendly (integrated, SSO, etc.) manner, then you need internal Vidtel type skills. Especially with what I assume is a high number of proprietary apps, some of which need revamping to properly integrate RTC.

Almost an aqui-hire?

Re: How Apple iBeacon Will Transform Local Commerce

gzino — Wed, 02 Oct 2013 22:31:51 -0000

Great post. A couple other potential uses:

1. Niche case but one may of us are interested in...auto disable of cell phone while driving:

http://nextblitz.com/blog/i...

2. P2P mesh networks, including ad-hoc. Use iBeacons + multipath TCP (also in iOS 7). iBeacons for power efficient peer/network advertising/discovery/selection. MPTCP to help maintain sticky TCP and TLS sessions during network changes, as well as to aggregate bandwidth across multiple interfaces.

Re: WebRTC and enterprise security

gzino — Tue, 24 Sep 2013 17:37:09 -0000

KJ, great points. I should have included an entire section on app and protocol centered possibilities. To your point innovation in those areas (albeit in a proprietary manner) was a large part of the Skype win.

Not sure NAT goes away, even with IPv6, due to the amount of "carrier NAT" we are seeing and concerns (real and perceived) on direct IPv6 address exposure?

Re: There Are Too Many Choices for Video

gzino — Tue, 24 Sep 2013 09:35:30 -0000

Completely agree. Although the shelves are increasingly stocked with video options, don't they currently look too similar? 55 cloud covered boxes of toasted oats?

I don't see the diversification, yet (variety in targeted niches, GTM, business model, technical, operational, etc).

But do think we end up with app specific video (and comms in general), maybe the initial overstocking of the generic cereal is just a necessary starting point while the market and technology evolves?

Re: TwilioCon - Empowering Software People

gzino — Tue, 24 Sep 2013 09:28:53 -0000

Nice post, glad to see some focus on the underrated cultural side of this developing ecosystem.

Software (and the web) has certainly eaten telecom.

Do you think there will eventually be some pendulum swing back towards self-contained/vertically integrated communications services over time?

I think the API model is here to stay, and the Twilios of the world are the new telcos, but usually when the waves all go one way, there are opportunities left in the wake for other approaches.

Re: The Google and Vidyo Partnership Enters a New Phase: Scalable WebRTC

gzino — Thu, 29 Aug 2013 12:36:34 -0000

If there are more *teeth* in the agreement (or behind the agreement) then meets the eye, then agree this is a great win for Vidyo.

However, could this mainly be Vidyo PR to counteract the negative attention of the announcement that Google is taking Vidyo out of Hangouts (and a favor from Google to support the PR; a favor earned by Vidyo being a good Hangout partner)? For example, there is no guarantee that standards decisions or Chrome implementations will end up in Vidyo’s favor when the dust settles, regardless of what Vidyo agrees to contribute.

All that said, SVC is powerful and if Vidyo can get enough support from Google and others to add it to future WebRTC implementations then that has significant upstream impacts on the overall architecture including signaling, SBCs and MCUs...

Re: Personal WebRTC Portals

gzino — Fri, 16 Aug 2013 13:04:40 -0000

hmmm, good point, enterprises are yet another angle. possibly their experience trying to manage enterprise users within their own app stores will be instructive there, at least for the large enterprises?

Re: Personal WebRTC Portals

gzino — Thu, 15 Aug 2013 16:20:21 -0000

Very interesting.

Possible that we end up with (n) "personal portals" as our current sites add WebRTC capability?

You give good example of how an exciting opportunity for WebRTC is enablement of new comms paradigms, rather than just replacement of old. More of those new paradigms in this post, would love your thoughts:

http://nextblitz.com/blog/l...

Also what are your thoughts on users eventually running our own personal portals? Meaning providers making it easy and secure enough for ordinary users to run on our own domains and home network based web servers?

Re: Google Glass: Twitter's Vine Could Be The Killer App - by Brian S Hall

gzino — Thu, 02 May 2013 21:25:23 -0000

Brian, love it, pocket-to-hand-to-video will quickly seem like work compared to the Glass experience. Two more thoughts for you:

1. The amount of photos/videos generated by Glass will be ridiculous...added to our already tiny signal to noise ratio...huge opportunity for ecosystem innovation assuming Google exposes enough attributes via Glass APIs?

2. Enterprise videoconferencing...yes, stodgy old enterprise video:
http://goo.gl/YxmfF

What do you think?

Re: For $19, An Unlimited Phone Plan, Some Flaws

gzino — Wed, 20 Feb 2013 13:59:35 -0000

Agree, this is glimpse of future.

Over time, we will develop apps that auto choose the network, behind the scenes, e.g. avoid bad WiFi networks for RTC, as opposed to always blindly trying WiFi first. SPs can adjust their pricing accordingly, whereas for the user it just works.

Hopefully, we will also see more unlicensed spectrum and have more than just WiFi and 3G/4G to choose from in future. Multiple access networks with smart apps and platforms gluing them together for the end user on a per use case basis.

Re: The US Needs A New Spectrum Policy

gzino — Tue, 19 Feb 2013 22:48:57 -0000

Is five 9s from one centralized solution the goal of tomorrow?

I would prefer a distributed, diverse set of access options as part of an antifragile (Taleb) system. The type of set best developed in an open environment.

Multiple dumb access options, none individually at five 9s, with smart, agile, continually evolving apps, platforms, glue and interfaces to stitch it together (behind the scenes) for me as a user, even on a per use case basis when sensible.