RE: Modern Universe (TESTING)
If you Dont Agree, Prove it... Everybody can Talk.. Where are your Test Results? The problem with performance tests is that there are so many variables. And then if you remove a lot of the variables to perform a lab test, that does not reflect real world. So then you try and make your lab conditions simulate real world - and simply doing a COUNT on a file is not enough - sorry ! Yes, UV could count for 15 mins if say the file was really badly sized. So what does that mean ? It does not mean that UV is crap, it means it's not been tuned right. Send Michael Schumaker round the track with flat tyres - when he does not perform is that the fault of the car ? So my final point is that it takes time and effort to do proper testing - something most of us are not going to dedicate resources to just to prove a point here. Andrew Gissing -- u2-users mailing list [EMAIL PROTECTED] http://www.oliver.com/mailman/listinfo/u2-users
RE: [UV] Ever wondered how something works...
so I have to think the list will still be active after the WHO is completed. Not so. I discovered to my own use that simply using LIST, without file name, will clear the selectlist. And in any event, proof of who below: T C L SELECT CUST 3117 record(s) selected to SELECT list #0. WHO 120 mm From andrew andrew gissing -- u2-users mailing list [EMAIL PROTECTED] http://www.oliver.com/mailman/listinfo/u2-users
RE: UNCLASSIFIED RE: Universe / ODBC
Hey hey ! I've just looked at our readme file and it appears our problem is issue #11264. Extract from readme file: = Problems Fixed in UniVerse ODBC Client Version 3.7.r2 = So this appears to be the same as what you've got, but where I got my numbers from was the windows odbc manager, when looking at a list of drivers the version number appears in one column - that's where the 3.07.02.1084 appears. Regards, Andrew -- u2-users mailing list [EMAIL PROTECTED] http://www.oliver.com/mailman/listinfo/u2-users
RE: Login question for Universe/SB+
One of the prior concerns that R. Bruce Lunt raised was about security - if you set all the SB+ passwords to be the same, you lose that. However, i'd like to add that if you've got users who do not have access to TCL; do not have access to a unix shell; and in effect, once they login can only do things from menu's programmed for them, then this is not an issue. At our site our users login under their own names at unix. All normal users share the same profile which is basically exec uv path uv SETUP where SETUP is our own program. And the purpose of exec is so that when the user finally reaches the last END statement in the chain of programs he runs, not only will the universe session terminate but also the unix session will too. Therefore, no access to TCL, no access to unix shell - does it really matter what the password to SB+ is ? A user won't have an option as to which name they login as - it will be provided from unix ? Their profile could look like cd sb directory ; exec uv path uv, and if VOC entries with DATA statements are present for the upper case version of the user name, all should be well ? This only applies however to normal users. For development staff who do have access to TCL, yes it does allow them to login to SB+ under someone else's name - but then, you've always got ED to hand edit files without using SB+ screens anyway - and that's another kettle of fish. I've written the above from a unix point of view - I don't know how you'd go about doing similar with NT but i'm sure it can be achieved. My .02 Andrew Gissing -- u2-users mailing list [EMAIL PROTECTED] http://www.oliver.com/mailman/listinfo/u2-users
