x
Page 1 / 7   >   >>
laserbrain 12/4/2012 | 7:44:22 PM
re: The IP Priesthood more like the IP big-leagues. no matter how big the salaries, there's still a shortage of the great ones.

maybe if there were brain steroids as strong as the juice Barry Bonds is on, we'd have a surge in router jocks like we do home run hitters. Would you sacrifice your liver and various, um, important organs for Tony Li power?
beowulf888 12/4/2012 | 7:44:15 PM
re: The IP Priesthood Oh, give me a break! Conspiracy? Are there are thousands upon thousands of top-rate SONET engineers out there? And ATM engineers? The only reason there's dearth of top-rate routing protocol coders, is that TCP/IP has spread like wildfire with the growth of the Internet. The demand has far outpaced the supply. I like Lightreading, but it's sort of a cross between an industry publication and a supermarket tabloid. Well, Elvis is alive and he's a BGP guru...

cheers,
--Beo
skeptic 12/4/2012 | 7:44:13 PM
re: The IP Priesthood
The reason for things being the way they are
is that protocol specifications for BGP et al
are next to useless. Anyone who tries to use
them as a guide to develop a router will fail
utterly.

Why is this the case? Because the IETF as
a standards organization is a fraud. There
is an IETF for the public in terms of the
meetings and the mailing lists, but none of
the real decisions are made there. The
real decisions are made in hotel suites or
in private email discussions.

Why are the standards for BGP and certain other
protocols so bad? You will get three answers
usually:

1. I dont know you, so you are obviously wasting
my valuable time.

2. We have to be "vendor netural" (therefore
promoting an real-world interoprable BGP
standard is somehow seen as favoring cisco
- of course the gated people might actually
still believe this I guess)

3. This isn't an important issue for the IETF.
(they are working the serious issues like
which working group gets TDM over MPLS).

As far as cisco goes, IOS is a mess. Even
cisco doesn't have as much knowledge about
it as people would think. And many of the
people selling themselves as "experts" don't
really know all that many "secrets".

There is plenty of so-called router "talent"
with a miserable track-record of failed startups
behind them. The worst are nothing better
than leeches. There are VCs who are so ignorant
that they think that serving as the chair of
a working group translates into practical
product development skills. Most of the time
it doesn't.

Its not just a question of talent. You have
to hire people who can:

1. Work in a team.

2. Share information with co-workers rather
than playing "I've got a secret".

3. Be willing to write code (rather than
sit on the phone all day and talk to friends).

4. Be able to treat other engineers as
equals. (rather than hanging around the execs
and the marketing people all day).

5. Have some track-record in delivering successful
products. Just because you maintained IOS does
not mean that you can write new routing protocol
software. They are very different skill-sets

6. Are committed to the success of the company.


These seem like simple things, but they are
not.

As far as the people who claim to have all
the secrets, most of them don't. Their ideal
interview is to sit in front of some manager
and play all smart. If you start asking them
the specifics of the "secrets" of IOS that
they know, they will either say nothing
or just collapse.

The example of netstar (in the original article)
is hilarious. Netstar failed not because of
cisco secrets, but because their testing effort
was a joke and they could never get gated to
run stable. They had fundemental problems
like software crashes and inconsistant forwarding
tables. Thats not cisco's fault and that
would not have been fixed by hiring any amount
of high-priced experts.



beowulf888 12/4/2012 | 7:44:10 PM
re: The IP Priesthood IETF standards are free. There's something to be said for not having to dole out beaucoup bucks to read something that will put you to sleep.

--Beo
beowulf888 12/4/2012 | 7:44:10 PM
re: The IP Priesthood Skeptik:
Sounds like somebody needs a cookie and nap ;-).

Standards documents always leave something to be desired, but the IETF has a much better track record for developing implementable protocols than just about any other standards body.

I would highly recommend a book entitled "Open Systems Networking: TCP/IP and OSI" by Dave Piscitello. It's a little out of date, but he examines the standards development process of both those camps, warts and all.

Yes, I would agree with you that the within the IETF there are many decisions "made in hotel suites or in private email discussions". But the characteristic that has made the IETF successful is that there has always been a "make it work" and then "finish the standard based on how we made it work" mentality. With other standards organizations, such as industry groups or government-mandated groups, there seems to be "make a standard" and then "lets make it work to conform to the standard" mentality. The former's cultural perspective no doubt has its roots in the hacker ethic that came out of the universities, while the later seems more beholden to the bureaucratic ethic. Why else has TCP/IP been so successful?

BTW: When is comes to obfuscation, who can beat the ITU?

cheers!
--Wulf
skeptic 12/4/2012 | 7:44:02 PM
re: The IP Priesthood Standards documents always leave something to be desired, but the IETF has a much better track record for developing implementable protocols than just about any other standards body.
--------------

Lets put the historical propoganda aside. What
the IETF may have been years ago, it isn't today.

When the whole organization starts to function
in secret, it ceases to be an organzation where
people can get anything useful from or make
any contribution to. I mean it was one thing
for some of the higher level parts of the IETF
to operate in secret, but what sense does it
make to run working groups that way?

Its a to a point where more information about
what the IETF is doing can be found out from
watching cisco than it can from following the
mailing lists.


--------------

I would highly recommend a book entitled "Open Systems Networking: TCP/IP and OSI" by Dave Piscitello. It's a little out of date, but he examines the standards development process of both those camps, warts and all.

-----------

I lived through parts of that. I don't need
to read about it and lord knows I would never
recommend what the ISO did in standards as a
good thing.

------------

Yes, I would agree with you that the within the IETF there are many decisions "made in hotel suites or in private email discussions". But the characteristic that has made the IETF successful is that there has always been a "make it work" and then "finish the standard based on how we made it work" mentality.
-------------

I don't agree with you on that. And would point
to what happened in diffserv as a classic
example of people who were not interested
in making it work or finishing the standard.

The "senior" people in the IETF, who are so
fearless about tearing people they don't know
apart, lacked the guts (or the interest) to
say a word against the great Van Jacobson no
matter how much damage he did or how many times
he would play "I know best because I know more
than you but I can't tell you what I know".
He may or may not be right, but its impossible
for people to productively work on standards
with this sort of thing going on.

And in the end, the answer always seems to be
to let the group fold and do something else
rather than fix whats wrong. There are people
fighting whiny battles in public at IETF over
nothing who will not step in and fix things
that are clearly off-track or wrong.

If you want another example, look at the so-called
"policy" efforts. Based on the bad ideas of
a few influential people, the whole thing has
been knocked off the rails and re-directed
away from solving useful problems of real people
into something that few (if any) people are
going to want. (whatever year they finish).




I certainly don't mean to say the ITU is a
good approach. What I want is for the IETF
to operate like its theorically supposed to.


















netskeptic 12/4/2012 | 7:43:59 PM
re: The IP Priesthood > Lets put the historical propoganda aside. What
> the IETF may have been years ago, it isn't
> today.

Yes, and the change did happen a long time ago.
At the same time other standard organizations are even worse (e.g. ATM Forum).

So, I suspect that it is going to be a reformation indeed.

Thanks,

Netskeptic


nonobvious 12/4/2012 | 7:43:56 PM
re: The IP Priesthood I think at this point most people realize that the IETF has degenerated into a forum for industry deadbeats to spew endless smart-sounding drivel in support of inflated option packages, resumes, and egos.
Bumper_car 12/4/2012 | 7:43:36 PM
re: The IP Priesthood Designing and building supportable routed networks has for years been and remains more of an art form than a science. The complexities of application/user usage and loading while trying to provide some level of system/circuit failure protection remains too complex for computers to model properly, particularly for large networks. Many of inherent characteristics of the interworkiing between the complex "circuits", the Ethernet LANs, and the routing protocols are undocumented and have to be infered by people with many years of experience in building large networks.

There is a saying about the definition of a data networking professional: Some one that has been around long enough to have "screwed up" and seen enough "screw ups" to know what not to do. The problem with most of the younger people (including those at startups) is that they do not know what to avoid. There are some briliant people out there that are trying to design and code network protocols, they just don't have the experience.

Having a doctorate in data communications, math or some other "science" can and never will substitue for experience. More than anything else, what the IETF is currently suffering from is a lack of experience. They don't know what not to do.
tony1athome 12/4/2012 | 7:43:32 PM
re: The IP Priesthood Would you sacrifice your liver and various, um, important organs for Tony Li power?
------------------------------------

Trust me, it's not worth it.

Tony
Page 1 / 7   >   >>
HOME
Sign In
SEARCH
CLOSE
MORE
CLOSE