Of course, you're still waiting for the first "perfect" release of:
Windows
Windows NT
Solaris
FreeBSD
OpenBSD
Linux
IRIX
AIX
to mention a few... along with the first perfect release of the various
bits o' software for each of those operating systems.
In any code more complex than the "Hello World" samples that we've all done
in various languages, there are going to be bugs - and I think it's
unreasonable to expect Lucent, or anybody else, to release only perfect
code.
More appropriate language in this debate might be: Lucent has released
beta code, which is certain to contain bugs, since beta versions always do.
We're all waiting for those bugs to be fixed, and when they are, Lucent
will release a version that is more stable and supports more modems than
the current beta code. Lucent could, perhaps, do a better job of letting
us know where they are in the process.
Certainly there are those on the list (and I'm sure many other who aren't
on the list, but who call Lucent on the *gasp* telephone) who insist that
they want a target date. Those of us who've been around a while understand
that target dates, especially those associated with complex code, are never
met (and if they are met, the release is generally fraught with problems).
I don't expect perfect code from Lucent... or anybody else. My practice,
where the PM3 is concerned, is to watch this list after the release of new
code (beta or otherwise) and see how well it works. If feedback on the
list is good, I'll put the code on one of my PM3s and see how it does in my
environment. I spent 2 weeks just watching when 3.8.b13 came out -
everybody who tried it reported good results, and fewer disconnect
problems, etc. So I put it on one PM3. After that box had run beautifully
for 3 weeks or so, I put b13 on the rest of them.
In our situation, we can't afford to be bleeding edge if that bleeding is
serious (as would be the case with b15 for us). So I'm waiting patiently
for the bugs to be identified, fixed and the fixes tested. Then I'll go
through the process of wait/watch, then load and test locally (if watching
doesn't turn up immediate problems).
v.90 is important to us - our primary competition is Mindspring. I've had
customers call, ask about v.90, and we handle it by explaining that all
v.90 code has the potential to have problems, and that Lucent is testing
theirs to make sure it's good, with as few problems as possible. I explain
to customers and potential customers that we do not deploy code until it
has been shown to be stable - and I point out that we do this so that they,
the customer, has the best service possible. Ya know, guys, I have never
had a customer leave us after that explanation. So, yes, v.90 is important
- but an interoperable v.90 is vastly more important than giving a few USR
users the chance (and it is just a chance - there's no guarantee) to
connect at a higher rate.
The "when it's ready" answer may not be the one you want - but has been
pointed out time and time and time again, it's better than "March 31,
1998," then on March 30, the date is pulled because serious bugs were
found. Even Q4 1998 as a target may slip by - testing might turn up
serious trouble at the last minute when some tester runs the code in a
novel situation.
I know we're all running businesses here... and some of you feel you have
to have firm dates, since you think you live and die by the calendar. But
maybe I can point something out here: today, you have customers. Tomorrow
you will have customers. If you have fewer customers, maybe you should
take a long, hard look at why - because (and we're not alone in this), if
you treat the customer as important, and you explain to them that you're
not able to offer whatever service they're asking about right now because
you want to be sure it's a quality service, you won't lose the customer.
The "I have to have v.90 right now!" mindset translates to, at least in my
mind, "I'm just in this to make as many $$ as possible as quickly as
possible, and anything that makes me miss a couple of $$ is unacceptable."
With that attitude, you'll never build customer loyalty - which goes a
LOOOOONG way toward keeping those customers for months and months and months.
Sure, we lose customers every day - but we're in an Air Force town, and 90%
or more of those who cancel do so because they're being transferred.
We don't have v.90, and won't until we're satisfied that to upgrade won't
cause more problems than it will solve - yet we have a net gain of 70 to
100 customers per month. Hmmm.... wonder why that is?
-
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/>