>The wording of the LSVCOST questionnaire suggests that "new releases
>only" could leave one with an inoperable version of L-Soft software and
>no recourse for correction until the next "new release".
This is correct, from a lawyer's perspective. Category III licenses
(that's their official name) give you access to new versions at release
boundaries, and to HIPER fixes. However, for any given installation there
is a possibility that some little detail or other will be considered
critically important, even if it doesn't mean much to most sites. Any of
the fixes I make is a potential HIPER to one particular site with some
very specific requirement, and there is nothing much that can be done
about it. If the customer could decide what is and is not a HIPER, we
would be giving category II service for the price of category III (35%
cheaper). The difference in price is due to the fact that, on the
average, I have been spending 1h helping people and fixing problems for
every 2h spent writing code.
Maybe I should explain the rationale a bit better. Category II is the
"normal" license and corresponds to the level of service that had been
provided free of charge up to Sep 1. Category III was created as a kind
of licensing tool for large accounts, rather than as a final product. For
instance, EARN sites receive the equivalent of cat III service with a
right to receive between-release fixes from one specific site designated
by EARN, which has a category II license. This is a usable final product,
which costs less to L-Soft than a cat II license for everyone, and does
not cost much more to EARN as they already have a full-time LISTSERV
expert. Another illustration is a network that would decide to offer
"free LISTSERV access" to all its members (meaning just that, free access
to the code). This network could license cat III service for everyone,
and the individual members would license cat II upgrades.
The reason there is no "category 2.5" with between-release fixes but no
customer support is that it would never work in practice. People don't
write to ask for "fix number XYZ23", they say "Since yesterday LISTSERV
no longer tlops the spolt in three as usual, and the associate VP for the
XYZ department is very upset, what should I do?". It takes several
exchanges to get a description of the problem with all the necessary
details and figure out what is actually happening. One third of the time
it is a "stupid" error - someone who XEDITs the LIST files directly and
typed "Subscribe=" instead of "Subscription=", for instance. In order for
a cat III license with between-release support to be significantly
cheaper (to L-Soft and thus to the customer) than cat II, one would need
a mechanism to eliminate this overhead, or get compensation from the
customer when the problem is a user error or the description was not good
enough to permit quick identification. This is a nasty can of worms, and
I doubt many sites would be willing to sign that kind of contract. It is
not a good solution for L-Soft either. Joe makes mistake, Joe complains,
L-Soft identifies mistake, Joe can solve his problem, L-Soft sends bill,
Joe screams because the boss will be upset and claims this is really
because the name of the configuration variable is confusing, or the
documentation is inadequate, and we have a big fuss for nothing. None of
this is worth the trouble for a difference of $581/year.
Eric
|