We take the credibility issue quite seriously. Boardwatch is somewhat
unique in that it is virtually the only magazine in the industry owned by
the editor. We have no investors, no boards, and no court of appeal. That
has some downside aspects to it, but it does pretty much assure that
editorial calls the shot - not the ad guys. This involves a continuous and
ongoing financial penalty that we endure cheerfully. But after enduring
it, we think it gives us the ground to not deal with it either.
Unfortunately, this is not ubiquitously true across the publishing
business. And we occasionally get painted with an appropriate brush but on
an inappropriate canvas.
On the topic of statistics, methodologies, and outcomes, it is a very
different matter. We are always subject to question there, and as anyone
who has been in testing and electronics can tell you, it's almost
impossible to DO a clean test. But wait, it gets worse. We specifically
eschew lab tests and almost everything you will ever see from us is an
attempt at real world testing. In that arena, you simply cannot control
for all variables. We give it our best shot, and publish what we see. The
backbone tests, the dialup tests just completed, in fact almost anything we
would bother to do with respect to testing, will inherently be rough, and
contain the exposure and risk of being indeed wrong. We think they're
valuable, and are worth the risk. If I have to fall on my sword in public
over this one, it actually wouldn't be the first time, and it wouldn't
prevent us from doing them again.
We think these are both interesting and good tests. Otherwise we wouldnt'
publish them. We're not in process editoriallly. The stuff is at the
printer. You WILL see what we did, why we did that, and what came back
with regards to results. You'll also see our analysis of it, and some of
it very blatantly wanders into conjecture and is pretty clearly labelled as
such.
We've been clobbered by what we're seeing. We started with the ASSUMPTION
we were doing something wrong and have backed into what look like
inescapables from there. It looks like new information. It looks like a
broad view of a very varied network that does provide a result very
different from what we've seen done or published. After some of the e-mail
exchanges, I think it presents a very different view from what ISPs see
from their equipment room perspective and what they CAN see.
But it is really a client-side view. Livingston and Portmasters aren't even
mentioned in the articles and we've rather been sucked into a non-sequitur
conversation sideways.. I have always held Livingston in the highest
esteem, and it has little to do with personal dealings with them. We talk
to a lot of ISP's, and THEY hold Livingston in the highest esteem. I
inherit this view. I am personally familiar with a long history dating
from 1992 or so when Portmasters where HOW you COULD do dial-up to the
Internet, and from Radius which was the only way you could tell you DID do
dial-up to the Internet. From my personal perspective, Livingston invented
it. Today they have to compete with other vendors and a march of
technology that offers little recompense to those who should by rights own
it. We all get to earn our stripes brand new each day, regardless of all
the good things done in the past. But Livingston remains and is regularly
reported by ISPs' to be a stellar company to do business with and with
solid equipment. Period.
This conversation has drawn down into hopeless detail, with no context. It
is my opinion that a possibility exists that there are significant
differences in the performance of K56flex and x2, and that they will
persist into V.90. But lost in all of this is the fact that it is a
client-side issue and will become more so, not less so. Under V.90,
Portmasters, and Max TNT's and Total Control systems will all have to
deliver data pretty much as instructed by the client modems with regards to
the constellation map used to deliver data over various digital topologies.
This could lead to a situation where one client modem works better than
another. But I would be very surprised if it carried to central site
equipment anyway. Portmasters will have to take calls from 3COM V.90
modems as well as Zooms. The issues of proprietary code at that level will
fall on the client side - not the central site. There ARE obviously things
that can be done on the serving side, and there may emerge some
differentiation there obviously at some point. But we have no visibility
into it and probably won't for many months. It would require a very
different test setup to even examine.
We are publishing the data. It is ALWAYS open to question, and
particularly this test, in the back of my mind I actually do think there
has to be SOMETHING to explain this that we're missing. So far we can't
find it. I will in person show the results and take questions at the KARL
DENNINGER ISN'T HERE session at ISPCON on Wednesday afternoon and all
questions about the data, the methodology, and so forth are fair game.
Discussions of future tests that SHOULD have been run or should be run in
the future are not only fair game, but my favorite topic of conversation
these days. We'll talk frankly about what we did and what we almost did
but decided not to, and why you think we should have done C instead of A or
B either one. My position will be to defend it, and your position will be
to take it apart, and something will come out of it and hopefully something
to be learned by all. Then gin.
Whether 3COM bought us isn't. They didn't. They can't. And I would be
comically and hysterically surprised if they attempted to. Knowing a great
deal I know that you all don't, and I can't tell, it is in this case a
particularly ironic accusation. It goes quite to our bottom line.
Fortunately, I don't know exactly at any moment what our bottom line is,
and in any event we'll make it up later somewhere else down the road in the
long run. Eleven years in publication gives you a kind of long term
perspective where some things are important, and others are just annoying
for the moment.
Doubly ironic is the fact that from what I can tell, Livingston is actually
the one current equipment vendor that takes a similarly "long term"
approach to a market that suffers dramatic changes on almost a weekly
basis. I have no idea how this is affected by the Lucent acquisition. But
for most of the Livingstonoids I would know, and for me, V.90 is Spring
Modem Fashions - 1998, the latest in a long string of such steps that make
110 bps connections go much faster now, thankfully. There'll be some more
along in a year or two. And reputations are built over time. The first
rule of winning in a competitive market, is to be in it. And the quick
buck artists tend to come and go with the seasons. I wouldn't be so bold
as to suggest this same philosophy for those in the ISP business, but on
the other hand, I wouldn't suggest otherwise.
I rather gather, and agree with Mr. Sasek, that while entertaining, we've
taken this discussion about as far as possible prior to the publication of
the tests where we have some common numbers and things to shoot at, or in
my case, defend. So I'm going to try to go to quiet for a few days and
hope other topics perhaps more germain to portmaster-users prevail. Karl's
been great fun as well, but we've about run out all the possible
combinations of name calling I can do creatively at this point. A number
of you have expressed some angst to me privately at the apparent animosity
between us. I would assure you that it all looks worse in e-mail than it
actually is. This sort of thing has gone on for many years, and we're both
rather prone to engage in it. Don't fall into a dead faint or go into some
sort of spasm if you see two guys drinking Martini's at the end of the KARL
DENNINGER ISN'T HERE session, and one of us isn't even actually there
technically speaking. Stranger things have happened.
Jack Rickard
----------
> From: Joe Sasek <sasek@livingston.com>
> To: portmaster-users@livingston.com
> Cc: Jack Rickard <jack.rickard@boardwatch.com>
> Subject: (PM) Boardwatch and my previous email
> Date: Monday, February 23, 1998 11:41 AM
>
> Portmaster-Users,
>
> Last post from me on this I promise ;-)
>
> Last Thursday evening around 9:30PM reading email at home I got an email
> containing information regarding a soon to be published edition of
> Boardwatch Magazine. As any of you who have been here long enough would
> realize, I've been here 4 years now (employee 21), and take criticism of
> our products very seriously (some might say too seriously ;-)
>
> Jack Rickards input in the email to which I'm referring implied that the
> Portmaster "sucked". After careful consideration of my response and
opinion
> of Jack relative to Boardwatch being subject to advertising driving
> editorial opinion was way off base.
>
> Just as emotion sometimes takes over when someone attacks one of your
> family, that is exactly what happened last Thursday. I took a low shot
for
> which I am sincerely apologetic. I spoke with Jack at length this
morning,
> and believe that his ethics are beyond reproach and in fact believe they
> are very similar to Livingston/Lucent's.
>
> I _am_ concerned about the supposed methodology of the soon to be
published
> Boardwatch testing of X2 and K56Flex, which promises to be an interesting
> event ;-) I am very interested in the facts being told accurately (as is
> Jack), and hope that we can work together to insure that this is exactly
> what happens!
>
> Regards,
>
> Joe Sasek
> V.P. of Sales and Mktg.
> Lucent Remote Access Business Unit
>
>
>
>
> -
> To unsubscribe, email 'majordomo@livingston.com' with
> 'unsubscribe portmaster-users' in the body of the message.
> Searchable list archive: <URL:http://www.livingston.com/Tech/archive/>
-
To unsubscribe, email 'majordomo@livingston.com' with
'unsubscribe portmaster-users' in the body of the message.
Searchable list archive: <URL:http://www.livingston.com/Tech/archive/>