[c-nsp] Ode to the old days

[Saku Ytti]

On 8 December 2016 at 23:14, Gert Doering <gert at greenie.muc.de> wrote:

> And vendors that actually cared what their users asked for (at least
> if you offered some money as bait).  And understood protocol specs.
>
> Nowadays?  "No other customer ever asked for that"
>
> Yes, there might have been more bugs in basic stuff, but people cared
> about making that stuff work.  Now, all you get is "more shiny" (and
> more insane list prices so you could get more impressive discounts).

Don't experience any of these differences, but hard to quantify. I've
heard 'no other customer ever asked for that' as long as I can
remember. Perhaps easiest to quantify is the list price:


2004:
    WS-C2950G-48-EI      Catalyst 2950, 48 10/100 with 2 GBIC slots,
Enhanced Image      D     4,135

2016:
 $4195 - WS-C2960X-48TS-L     [Catalyst 2960-X 48 GigE, 4 x 1G SFP, LAN Base]


This one anecdote at least does not seem to support price inflation.
It seems entry level 48 port switch is cheaper now than it was in 2004
considering yearly inflation. It probably should cost like 5200-5300
now to cost the same.

I remember vendors having clueless and clueful people as long as I've
worked with them. Sadly I can't neither say that I think software
quality has changed, I seem to open same amount of TAC cases as ever.

If there are significant changes, they are escaping my capacity to
quantify them. One thing I can say is, there is lot more red tape
today as Internet is lot more important, I remember when all testing
was done in production and maintenance windows didn't exist, you did
what you needed when you felt like it should be done. But that is
mostly to have better excuses and less responsibility on your failures
than actually having higher quality or fewer outages.



-- 
  ++ytti