John Roese’s Blog CTO, Nortel

Trackbacks/Pings

1. […] So what’s Riedel thinking these days and what is Nortel’s strategic focus? Fortunately, Nortel blogger - and CTO - John Roese managed to snag Riedel for a video interview. […]

   January 27th, 2008 at 2:22 pm from All About Nortel » Blog Archive » Nortel’s Mystery Man: George Riedl

Comments

1. John,

   Thank you.

   I think most of the strategy questions asked on your blog are relevant to technology, don’t you? I agree day to day stock price is outside the scope, but the technology, strategy and product decisions are things that are within the scope and ultimately do affect stock price.

   What in your view is the difference between a Chief Strategy Officer and a Chief Marketing Officer? How about between CSO and CTO? How do you affect and influence your product managers and their decisions and how do you affect and influence each others views?

   Transformed Enterprise:
   * So what is planned for enhancing your SOA interface? I can’t find much on it other than that there is one.
   * What is your reaction to the Gartner report that BI growth is slowing? http://www.informationweek.com/news/showArticle.jhtml?articleID=205906963

   Next Gen mobility/convergence:
   * Infrastructure PBT seems to have gone quiet any wins out there.
   * How are your products helping operators deal with new requirements like the PSST?
   * It sounded like you are going after the tier 2 and tier 3 markets and providers, is that a strategy?
   * I agree, the vz announcement is good for you as long as nortel is an equipment vendor.
   * Bumps in the optical road, eh? A gift for understatement. I agree the 40gig announcement is more important than people realize currently.

   Vertical integration:
   * Been successful for nortel in the past, especially since certain market verticals seem to be “early adopters” and nortel can learn a lot about what other verticals will want moving forward. How do you capitalize on what you learn from the “early adopters”.
   * What makes you think your “telepresence” solution is better than the recently heavily advertised cisco product? Why is a partnership with polycom *better* than an end-to-end single vendor solution?

   Nortel *is* a midscale provider in a shrinking market. Perhaps you do still have experience in services integration, but how do you exploit that experience and knowledge while you and your core market are shrinking? I know of some key people in that area that have recently left. How is nortel attracting new technology talent as well as identifying and retaining talent that is (and has been) there?

   January 22nd, 2008 at 5:08 pm by many

2. John - Many has posted a relevant and insightful comment on your conversation with George Riedel. I have a number of comments as well, but I don’t get the impression that you are paying attention to this blog. It feels like a Nortel marketing initiative. You almost never interact with the posters which leads us all to believe that that this is nothing more than a marketing initiative managed by your ‘handlers’. Right? Wrong? Let us know. If you interact, I will too. Otherwise, why would I bother?

   (note to marketing team…these guys obviously aren’t power dressers…c’mon, get it under control)

   January 23rd, 2008 at 12:02 am by Nortel Watcher

3. BTW John,

   If you are looking for topics, this discussion on tech dirt is probably more relevant to the telecom industry that to the video game biz:

   http://techdirt.com/articles/20080116/005211.shtml

   That said, in the future the discussion of the two may meld into one. What is the effect of “hyper-connectivity” on standardization vs. innovation? My take on SIP and IMS is that there are *by far* too many ways to do the same thing. My take on some of the other technologies linked to this “hyper-connectivity” is that they have the same problem.

   To me this slows deployment and acceptance, increases interactions and behavioral problems and generally makes a mess. I don’t think you will reach an “inflection point” until things calm down a bit?

   As an analogy; this sort of reminds me of C++ in the 1980’s where it was possible to write structured understandable code or an undecipherable mess. Then Java cam along and took some of the “best” practices and enforced them. This “standardization (if you will indulge my analogy) gave rise to all sorts of innovation on the Java platform and (I think) was good for the business of software in many ways.

   Opinion?

   January 24th, 2008 at 8:56 am by many

4. Mr. John

   Actually Nortel has a good idea to performing Strategy and Direction for futures Nortel, but Nortel still low profile to broad Strategy Roadmap to your existing and new customer, lets wake up and educate customer and partners.

   Now yours Enterprise products already losses moment to improving and most products line is not relevance with NGN market, Nortel company also has to further rationalize its product line, which observers contend is still confusing and redundant after acquisitions that took place almost 10 years ago.

   So as Nortel partner, we difficult to sell your product and must have “special strategy” to get customers. but in term of products is mostly strong engine and established performance for baseline used.

   I am looking for next product can improve for feature and technology likes Juniper, and provide wide product lines meet customer needs, i am agree with Jim Duffy with article “Five things Nortel must do to complete comeback” at http://www.networkworld.com/news/2007/011907-nortel.html and makes your partner has good perspective in next Year.

   Good luck

   January 24th, 2008 at 12:19 pm by Hariyadi

5. Many, this is a marketing feed, not a blog. Blogs are generally interactive. I value your posts because unlike most they’re insightful, but it’s pointless to address them to John because he doesn’t interact. In an era of hyperconnectivity, he’s a half duplex connection in transmit mode.

   On the other hand, Phil Edholm understands what a blog is and when he has the time to participate it adds a lot of value.

   I support your view of the value Java brought to the s/w industry. In a similar vein, I’ll suggest that the IETF was a mostly ineffective ’suggestion forum’ that couldn’t bring SIP to a practical standard. It was only when the ITU got involved that the principles became ‘hardened’. Having said that, the ITU then ‘wrecked’ the whole idea with the IMS architecture. A dogs breakfast that will never see the light of day.

   January 25th, 2008 at 1:03 am by Nortel Watcher

6. Nortel Watcher;

   I mostly agree about IMS being a “dogs breakfast”, although I think (I hope) all is not lost. Carriers need a next gen platform for services to replace IN. If not IMS then what? To me, one problem is that the carriers are not active enough in driving requirements into the MEF, ITU and IEEE standards bodies (forget the IETF), then engaging with (demanding that) the equipment manufactures 1) conform to at least the basic standards structure and 2) STOP their product managers form trying to pull every Mickey Mouse feature/capability they can’t sell through with the platform. That would go a long way (IMNSHO) to moving things forward and lowering the cost per byte to the customer.

   January 25th, 2008 at 8:23 am by many

7. Many, the Carriers will not replace the IN services model (which is really little more than 800/900, calling card, and calling name). Carriers will even lose their fixed line voice business as broadband proliferates and the new generation comes of age. However, life is not all gloom and doom. The Carrier is a trusted (not necessarily liked) provider of services to the consumer. This long standing relationship can be leveraged in the direction of the consumer and in the direction of the application provider. A wide range of billing services can be offered. Session QoS can be provided with many different SLA models. There is lots of business to be had for the Carriers that can be achieved by swimming with the current instead of against it. Probably more business opportunity than ever before. But it’s expensive and pointless to build IMS as an evolution of today’s Carrier network model. Today’s Carrier network model may not have died yet but it’s already extinct. Too late to evolve it. Reinvention is needed.

   January 25th, 2008 at 1:08 pm by Nortel Watcher

8. Nortel Watcher; here I respectfully disagree.

   What I see is Camel/IN services (and there is so much more than CND ond LNP/800/900 services) will migrate somewhere or else there will need to be new revisions of signalling to support new services.

   I see three possibilities on the near horizon, but I am open to hearing more:

   1) a migration to VoIP and new equipment and all that it entails including abandoning huge amounts of paid for money printing infrastructure, retraining all of its workers, pissing off the customer (because lets face it VoIP still has issues, the services are screwed up, and the regulatory stuff isn’t there either

   2) Status quo till the dust settles. This would be a bad thing ™ for equpment manufacturers.

   3) Migration to a IMS (like) model where the signalling and data network becomes a seperate pick_your_protocol_of_the_month network and uses the existing TDM voice network and it’s superior (and paid for) voice technology for bearer path. This is where IMS could (and I say could) be a good fit.

   I may have already tipped my hand that I don’t see the first option happening I think it will be the last one, unless the equipment vendors continue to be really stupid and force the second option.

   January 25th, 2008 at 5:18 pm by many

9. ‘Respectful disagreement stimulates dialog, which in turn fuels innovation. Nothing wrong with that. At least you didn’t grab my face! (couldn’t resist)

   4) IN/SS7 networks will be frozen and encapsulated, not evolved. IN Trigger events will be passed to/from the SIP and increasingly the XML worlds where useful services are finally being developed. Companies like JNetX and Stratus who are building these ‘trigger gateways’ ought to do well. IMS will evolve to become IMS-Lite and will be focused on session management things like policy control, network border control, and perhaps some opportunities to cost-reduce and integrate some wireless network functions.

   The existing TDM voice network will not be used for voice bearer paths except where the service provider has a geographical lock on the customer (e.g., rural) IMHO unless the operators lower their costs to compete with the new VoIP providers, which they show no signs of doing. The voice quality of my Vonage line is on par with the voice quality that Verizon would like to provide for me at 2.5x the cost of Vonage, even when Verizon bundles voice with broadband and video. People are abandoning the TDM voice services for a combination of VoIP and mobile alternatives.

   We’re going to end up with an Internet that is communications-enabled, not a Telecom infrastructure that is Internet-enabled, which is my impression of the thinking behind IMS. Regardless of the merits of IMS, it just won’t move fast enough because it has companies like Nortel, ALU, etc.. driving it. The world won’t wait so the major problems that IMS is trying to address will get solved in other ways, many of which may be less effective but more responsive than an IMS solution.

   January 26th, 2008 at 12:34 pm by Nortel Watcher

10. Hi,

    Sigtran (ss7 over IP) is being deployed rapidly it will soon replace traditional SS7 a links. IN Trigger events via SIP is the definition of IMS (and Session Border Controllers). BICC can play a role as well. I agree with you assessment of IMS in your first paragraph the question is not whether but when?

    I disagree with your second paragraph. I am guessing your Vonage is over a MSO? (Cable TV provider)? What do you think the underlying technology for the MSO’s cable plant is? If you guessed Sonet, you are correct. Last time I checked Sonet And DWDM were TDM technologies. They are essentially doing what I described, taking you VoIP signaling (likely H.323) using an IP infrastructure (perhaps Ethernet over Sonet, perhaps just switched Ethernet) and sending your RTP packets over Ethernet then over Sonet.

    I suspect in the end we will not know if the internet is communications enabled or the telecom infrastructure is internet enabled. I think this has been a false distinction all along David Isenberg’s “The Rise of the Stupid Network” and “Netheads vs. Bellheads” were landmark papers/articles that flipped thinking upside down and pointed out the economics of the problem. At the end of the day the “internet” has been carried over the telecom infrastructure in one way or the other since ARPANET. Applications that used to be solely in the telecom network are now being split apart and reinvented in ways that enable them to take advantage of and be taken advantage by new capabilities. The problem I see is that it is an implementation “Dodge City” and nothing short of the sorting out of the behaviors and interactions of these implementations will drive wider adoption and deployment.

    This is why I pointed to the link on discussion on “standardization vs. innovation” in the game box industry. On one hand if we had standardization would we have Wii (pun intended) OTOH how many people are delaying purchasing anything until the dust settles?

    Will Session Border Controllers replace IMS functions (they appear to be eating the IMS markets lunch at the moment), but if that happens, how do feature interactions get arbitrated? Is there a policy server per network? How are interactions managed between networks? When roaming? How can the end user self provision their own behavioral preferences? All of these items (and more) are problems IMS has the opportunity to address, areas where the Session Border Controllers are not as agile.

    January 26th, 2008 at 3:00 pm by many

11. Many, it’s getting to the point where we may need a whiteboard soon. I’m familiar with sigtran and it’s role, which I’ll crudely summarize as ‘enabling a softswitch to interwork with the SS7 world’. But that’s really only the signaling path and it only enables call-control interworking. For application level interworking, you need to go beyond that and map the triggers to/from SIP/XML.

    My broadband service is Verizon FiOS (FTTP).

    I like the ‘Dodge City’ implementation description. That’s exactly what’s going on right now, and that’s why IMS ‘proper’ will never happen. IMS is a big honking architecture designed to enable sophisticated applications. But the operators can’t afford large IMS build-outs without the applications to pay for it. Meanwhile, the application developers aren’t interested in waiting for IMS, so they work around it. Life was simple in the days of SS7. 800 service - by itself - generated enough income to pay for the installation of the SS7 infrastructure by the operators. The operators need the equivalent of ‘800 service for IMS’ or it won’t happen.

    Session Border controllers are an example of a solution for a need, versus a piece of a grand architecture. The function is also included in an IMS solution but Dodge City has a fire burning in the saloon right now so they need to use buckets…they can’t wait to install overhead sprinklers!

    Now back to the current vendor context, the vendors (including and maybe especially Nortel) are trailing the operators in their ability to envision and articulate the network of the future. That never used to be the case and it’s happening at a time when the operators are fighting to reinvent themselves in order to fight off the threats from new entrants to their space. This relegates the traditional vendors to competing on price for business that the operators specify, a game that the Chinese will likely win more and more over time. The future of IMS - and Nortel - requires vision and speed-to-market, something which both are lacking at the moment.

    Mike Z brought in a wave of production specialists and in doing so he consciously or unconsciously drove out a lot of the remaining innovative talent at Nortel. It’s likely too late but Nortel needs to hire a leader that can attract top talent and rebuild the thought *and execution* leadership it once had.

    January 27th, 2008 at 11:53 am by Nortel Watcher

12. You are right, Sigtran is only the signaling, which is my point. I agree that it needs triggers (similar to MAP/CAP) to interface with SIP/IMS/Policy/Location/Presence/Identity/Charging - whatever. It would result in the connection layer control protocol, which switches/makes/breaks/holds/merges/transfers/monitors/measures - adnausium the bearer path for voice.

    Vz FIOS would be nice. I wish I could get it here, but they just can’t find a market for it with the cows & horses. I believe that Vz is using L2TP (pseudowire) for the L2 connections, which is a reinvention of the circuit switched network on IP :), I do agree it does not use Sonet though I stand corrected.

    The interesting things that I have seen at the large Service Providers is that Ethernet is harder for them to manage, there is little in the way of “demarcation” which is what they are used to. They have also been slow to roll out IPv6 which is crucial to the proliferation of addresses that will happen and will also solve a lot of issues with NAT/PAT and Security.

    I agree with your comment about Session Border controllers being “bottom up need” vs. top down “grand design”.

    I think it is more complex for equipment manufactures. On one hand they have their installed base which generates (declining) yearly income, on the other the “grand designs” which are built in a semi vacuum and suppose a single vendor implementation. (BTW I see cisco going down this same road). Service providers are suspicious, and they don’t see how these new products are going to be manageable at all let alone make their networks simpler (which was at least part of the sales pitch) Equipment vendors need to wake up, the networks will never be single vendor at best it might be two or three vendors at a particular layer. I agree it is also (currently) a race to the bottom. I am not as optimistic about the Chinese equipment vendors except in their own markets (except at the consumer edge).

    The positive thing is that once this spectrum auction is over with (and the FCC approves at&ts takeover of the aloha spectrum) things might break loose. I am not counting on a telecommunications infrastructure initiative from the Feds anytime soon, but it would be nice.

    I agree nortel has really slowed down, I am currently trying to get a design document out of them and I can barely get an answer to an email without copying a VP. ALU has been very responsive lately, but I think that is just pure hunger.

    Your comment about the current leadership resonates with me when I talk about doing vs. being. Zafirovski and company are very much a “doing” entity. They have process and instructions for everything. However necessary that “doing” is, I see a whole lot of doing and not much being from nortel. Showing up to a meeting is doing, contributing positively and supporting those that contribute is being. Answering an email is doing, actually contributing to understanding and expressing oneself clearly is being. Starting a blog entry is doing, responding to the comments and participation in the blog entry is being

    January 28th, 2008 at 9:13 am by many

13. Many - your last paragraph eloquently captures the malaise that Nortel is suffering from, in my opinion. I’ll also add that I think they have lots of leaders in the wrong roles. Let’s give all of them the benefit of a doubt and assume they’re really good at something or they wouldn’t have risen to the level they’re at.

    Having said that, let me pick on the Carrier president for a moment. His business is the one that is closest to going over a cliff. I don’t see any evidence that he has the depth of understanding or personal insight needed to figure out what’s happening to the carrier market segment and what Nortel can be doing about it. I think he’s an operational guy and likely can’t talk off the charts given to him by the marketing department or by the like minded people he’s attracted to work for him. I don’t think he can add to the conversation when Verizon asks “what should we do?” and he can’t help Verizon paint a picture for their customers, new and old. And that’s bad…it means Nortel is in a corner they can’t find their way out of.

    Contrast this with the Optical president. He grew up in the business, knows it inside out, and I believe he could have a dialog with anyone with or without charts. I think he could explain the advancements that got the industry to where it is today and I think he could articulate the challenges and opportunities ahead. I imagine his biggest challenge is getting internal support because he’s one of the only ‘old guys’ left. Must be almost time for him to go…;-)

    To conclude, and get back to the subject of this thread, I think Nortel suffers from a leadership problem, a ‘doing versus being’ malaise that you nicely articulated, and a critical need to chart a very clear path forward (and act on it). Priority one needs to be to get the right people in the right roles and get out of this ‘running blind’ mode.

    January 28th, 2008 at 11:33 am by Nortel Watcher

14. You guys should have your own blog. At least there would be some interaction and debate

    January 29th, 2008 at 1:05 pm by moosebump

15. Regarding “direction” all Richardson R&D - and I’ll bet some other sites - basically would like to know if its history. People there are turning off lavatory lights to save money. Is this helping?

    January 29th, 2008 at 9:58 pm by Johnny Aerotek

Leave a Reply

Name (required)

Mail (will not be published) (required)

Website

John Roese’s Blog is powered by WordPress.
Copyright © Nortel Networks 1999-2008. All Rights Reserved. View the Nortel Networks Privacy Policy.