🚀 go-pugleaf

In article <BsKpn5.GC3@atherton.com> dlw@atherton.com (Remo Williams) writes:
> What do you think of floating per user licenses? The idea is you purchase X  
> number of licenses. As each distinct user fires up the app a license is  
> permanently consumed for that user. This user can use the app from any  
machine  
> on the network. When you want more users to have a license you "buy" them  
over  
> the phone, fax, email, etc. Then you get a new key back which upgrades the  
> quantity of total licenses.

One problem with *floating* licenses is that they can greatly raise the unit  
cost of infrequently  used applications, or applications that do not take a  
long time to run.  The per-user licenses David mentions not only have an  
advantage over node-locked licenses, but they can result in lower unit costs  
for such applications.  Here's how.

Consider an application (like a compiler) that you run that usually only takes  
5 minutes or less to run each time.  Now consider a very large site with say  
500 users of this application. Now let's say that the vendor of the product can  
afford to sell it for $50,000 for the site.  That's a hefty sum and may require  
a vice president's approval (say 3-4 levels above the end user). That's likely  
to take a long time to get and be frustrating both to the end users who want  
the application AND to the vendors who want to sell it.

Now, if you have a node locked version OR a per user copy license that comes to  
only $100 per user, an amount that the lowest level manager or engineer can  
purchase without any additional approval.  (The per user is actually more  
versatile if you don't end up  buying copies for everyone since once you use it  
you can continue to do so no matter where you are on the network, and with much  
less hassle if you upgrade or move machines, etc.)

Now consider the floating license.  Assume a 40 hour/week work week. 

40 hours/week * 60/minutes/hour / 5 minutes/use slot = 480 use slots/ week

But  you only have 500 users, if they each use the product only once a week,  
you need on average only a little over 1 simultaneous use slot if the uses are  
uniformly distributed.  So the equivalent per floating user cost would have to  
be $50,000 / 2 = $25,000 per floating license.  Even if they aren't evenly  
distributed, or if you everyone uses it multiple times a week, you'll only need  
a few additional copies, say 10 total. 

$50,000 / 10 floating user copies = $5000 per copy.

Now, getting approval for a $5000/floating user package is probably going to be  
easier and faster than a single $50,000 copy.  But it may be more difficult  
than for some $100/user copies.   Note that in the end in all these scenarios  
the vendor gets the same total revenue and the client pays the same.  The only  
difference is the approval process.  

Which is really the best? For frequently used, long running applications which  
the person uses all the time (e.g. email, or spreadsheets for financial  
wizards, or word processors for writers) it may turn out the same. But for less  
frequently used or shorter run applications (like compilers) per user may be  
better.