Jeff:

I don't believe phantoms use a seat by default. The solution you advocate is one that has been occurring in the PICK market for years; if you want to do something reasonable then get off of PICK to another product that doesn't use a very expensive "telnet" licensing paradigm, especially for small firms. Many people have taken that route, much to the chagrin of all of us in the MV market-space.

Personally I agree with David that the current MV licensing "scheme" is outdated, and counter-productive, and must be re-thought before there are no MV environments left; Raining Data learned what happens when customers are "squeezed" for every dime possible. I'm presently running into this problem and am beginning to wonder why I've stuck with MV so long.

This is IMHO of course.

Bill

------------------------------------------------------------------------
Jeff Schasny said the following on 1/31/2011 7:32 AM:
Phantoms use a seat. Always have. It really has nothing to do with what the phantom process is doing. It invokes a Universe session. I would suggest that if you don't want to use a Universe seat to accomplish an HTTP read you utilize cURL, Wget or some other OS level command line tool to perform the retrieval of the data. I do it all the time. If you are on a Unix/Linix based system you can even fire off your shell script as a background process just like a phantom.

David Wolverton wrote:
Thanks for the feedback folks.  The reason I was doing this query was to
review the 'reason' for the recent license changes that make CallHTTP
'consume' a seat if it is used within a 'phantom' process. I didn't say WHY
I was interested get 'genuine' usage comments.  Like many of you, I use
CallHTTP to get a piece of data from a remote machine (in my case, a
UniVerse server is validating a code from a UniData machine). But with a recent update to UniVerse, we started having weird 'failures' -- turns out it failed when all the 'seats' on the UniVerse machine were in use, and the
Phantom attempted a CallHTTP lookup. Blam! Dead phantom!

I read all the uses people posted, and unless I was mistaken, no one was
seriously using CallHTTP for the purpose of serving multiple 'logical
users'. It appears everyone is using CallHTTP as a way to gather a piece of data that could have just as easily been in a file on the local disk drive
if the machine you have could have limitless resources.  In my use, and
apparently most of yours, to call CallHTTP 'interactive' would be the same
as calling a disk read 'interactive'.

Here is the link for the 'business case' for making CallHTTP 'eat a seat'
when used in a Phantom.  I wanted to see if the logic made sense for the
CallHTTP feature.  My point to Rocket will be that someone could make a
phantom into a 'multi-user' server by using READ/WRITEs from Text Files --
yet those are 'allowed' -- so trying to 'lock down' the server against a
POSSIBLE misuse of the license terms by removing needed features seems
counterproductive.  UNLESS, that is, you're going to lock down EVERY
POSSIBLE way to misuse the system - Meaning, phantoms should not be able to
READ or WRITE at all. Heck, phantoms should not even EXIST since their
existence could lead to license misuse!
https://u2tc.rocketsoftware.com/rsp-portal/rsp/solutionDetail.asp?id=0002370 1?sterm=iphantom&exact=&searchAction=doSolutionSearch.asp&catFilter=02n40000
000Tqmn&oType=

Am I out on a limb here saying that CallHTTP should probably not cause a
Phantom to go iPhantom? I mean, Rocket can do whatever the heck they want,
it's their sandbox after all and we really have no choice but to suck it
up... But is the logic they employed flawed as I think it is? Or am I just
a loon?  (Hmmmm.. really, the two questions are not mutually exclusive I
guess... But you get the point... )  I'm interested in comments on the
topic, if any.

DW

_______________________________________________
U2-Users mailing list
[email protected]
http://listserver.u2ug.org/mailman/listinfo/u2-users

Reply via email to