DSXchange

Unlikely that you will, but it doesn't cost anything to try.

I tried going down the official route of getting a Beta and had sent over 10 emails to the official beta email address as well as one direct one to a gent in the beta group and did not receive a single answer.

You need to re-install the DataStage engine on the new machine; a UNIX level copy is not sufficient and will not result in a running DS installation. The DataStage projects might be copied (create the project normally on the new machine and then overwrite the contents from the original) but it makes...

But Ray - that is cheating . All the date OCONV()s produce a separator, the use of two conversion formats just "internalizes" the stripping of the characters.

Yes, the DirectConnect or DataDirect (can't recall what they are called these days) drivers will let you read the .xls file directly using ODBC and the drivers are platform specific, so the driver on LINUX won't work on SOLARIS. Best ask the pricing and portability questions to the vendor directly. ...

Looks like a lengthy statement like IF In.Type="A" and In.Value=122 THEN 'N' ELSE IF In.Type = "A"... is your best best, since there is no discernable logic for the values of Type and Value for setting the flag.

It sounds like a good solution and is simple enough to withstand the test of production. If your routine has lengthy SLEEPs to wait and also has a maximum timeout so that it always will exit sometime then you are set.

Nothing as drastic as looking elsewhere is necessary - I just read my reply and looked at the previous post and couldn't figure out why I said what I had... it took me a minute to realize what had happened and that you had changed the post. It is better to just add a new message to the thread. I hav...

...simply convert the date from one format to other... Unfortunately as there is no OCONV() date format that has no separator, it will always output some sort of character; no OCONV() code will output yyyymmdd or similar formats. You will need to use CONVERT('/','',DateString) to remove forward sla...

I'm not at a DataStage machine now, but you should do 2 things: 1. Search under the DataStage server path for the file "drsdb2.dll ". If you don't find it then you haven't installed DRS correctly. 2. Make sure that your LIBPATH or equivalent setting includes that directory. You can check the actual ...

Sure, but you will need to do this in your own function in BASIC. The math is not too difficult. I looked up what you called "hexadecimal float" and I assumed that it is IBM Floating Point The page describes the 3 components, 1st bit is sign, 2-8 is the exponent and 9-32 is the mantissa (assuming a ...

It is something that certainly does not belong in the FAQ forum, please read this before posting in the FAQ forum and then post your query in the appropriate forum.

I've found the problem - thanks for that :D Hmmm... you edited your post and changed the contents around completely. That is not really nice. The original post stated that you had copied the dataset and in that light my subsequent response made sense; but after you changed it my reply makes absolut...

There are many different approaches and methods to do this. The actual information about a job's runtime statistics is available at the command line using the dsjob program and is also available as API calls in BASIC (mainly DSGetJobInfo() and also such calls as DSGetLinkInfo() and DSGetStageInfo( )...

I can state that the difference in these two is that column "'pt_state" is not null.

How did you copy the dataset from another job? In this new job, how many columns does the dataset have (just the one declared as string[1] or more)?