Re: SQL Strangeness

On 1/8/2014 4:01 PM, Mark Murphy/STAR BASE Consulting Inc. wrote:

Yes, you are missing something. Likely because I wasn't clear enough about why I want this. I have multiple environments that I am developing for, usually concurrently. Each environment needs a different library list. When I open a source, I can indeed select the connection I want to use, and if I pay enough attention to do that up front, then I am ok, but I don't have to do it every time, and usually it doesn't matter which environment I am compiling for because all of my environments have a single core set of libraries. But each has its own customization library. Usually the files in the customization libraries are different, and if I try to compile in the wrong one it is more evident, but now and then I have the same file in two environments which are slightly different, and it matters which one I am compiling in. If I didn't make sure I got the connection correct up front, I am doomed because by the time I get to compiling I either explicitly set the connection and it is corre

ct, or I didn't look at the connection and I don't know what it is set to. And now I can't find out either.

Until I read this thread, I had no idea that knowing the connection was
important. I'm not diminishing the frustration you're feeling! I was
caught by surprise that such a big thing could be lurking under the
covers. Let me try to explain.

I never use Ctrl-Shift-A. I'm sure there's a reason to use it, but I
personally haven't encountered it. I always use an RSE connection (I
have many) and a filter (I have many of them too). My brain might be
stuck in 5250 world: session A is my development library list, session B
QA, session C production. Connection A is development, connection B QA,
connection C production. For these connections I have very broad
filters - filters that cover entire source files. So if I get an array
index error in a production job, I can open the production connection,
drill down to the production QRPGLESRC and open the production source
member. If, heaven forbid, I need to compile the thing directly into
production, the library list is all ready to go.

When it comes to development items, I create a new connection - one for
each project. The library lists are set accordingly. The filters are
narrow - encompassing only the source members I'm actually working with.
It's a poor man's change management system. Ish. Yeah, OK it's
horrible, but all the source members I need are easily seen together.
It works for me, mostly. When I need to compile something, the library
list is all ready to go.

When I used to keep test and development 5250 sessions, it was painfully
common for me to be in the wrong session and try to compile a source
member. Library list wrong, compile falls over, madness ensues. Now
that I have RSE connections that are named for the projects they belong
to, I almost never do that.

Ctrl-Shift-A (and the table view) take away the safety net that I've
come to like so much.

If the Javadoc were available, I bet one of us could write an extension
point that we could insert at the 'compile a source member' command.
I've looked at the open source RSE though, and I haven't a clue where
I'd find the connection information. Wishful thinking...
--buck