This content has been marked as final. Show 18 replies
Gussay wrote:Are you referring to the Freeze View icon? If so then I can confirm that changing it doesn't seem to make any difference.
Pinned just means the window is pinned, i.e the little red drawing pin holds the object window in place,
i don't think that is significant but thought I'd mention it.
As far as I can tell: no, no, yes.
Is SQL Developer seeing the Oracle DB as the external application that has made the changes to the body of the package? OR Is SQL Developer seeing itself as the external application that has made the changes to the body of the package? OR Is SQL Developer just seeing a difference and popping up this warning?
There are really two distinct cases for reloading the SQL Developer code editor:
1. Open for edit an external file containing PL/SQL (say you just saved a package body from the code editor to an external .sql file).
2. Open for edit a package spec or body already in the DB.
In case 1, a modification to the external file via an external application will be detected (not sure if this means always, or just when you switch between applications and a test gets triggered). The SQL Developer help text you posted apparently refers to case 1. If you have not changed the internal copy, then Automatically reload with Silently reload will update the internal copy automatically. If Silently reload is not checked, or you have modified the internal copy, then you should get the Reload? prompt (I did not test with Silently reload unchecked).
In case 2, only a difference between the object stored in the DB and your internal editor copy will be detected, but it seems to apply only to changes you make in the editor. There is no awareness of, let's say, some SQL*Plus session or another instance of SQL Developer making concurrent changes to the same object you are already editing.
I suppose the ideal would be to warn always of any external changes. I'm not certain if the default behavior of warning about uncompiled changes to the body after compiling spec changes is even intentional -- it might be just an accidental side-effect.
Apparently not. The first time I tested in 3.0 the Reload? dialog did not appear, but on subsequent trials in the same session it did. The test was simply to add an extra IN parameter to a procedure in a package, first to the body without compiling, next to the spec, then compile.
is the new feature in 3.1?
Hope this helps,
Gary Graham wrote:Any idea why I'm not able to reproduce?
I'm not certain if the default behavior of warning about uncompiled changes to the body
after compiling spec changes is even intentional -- it might be just an accidental side-effect.
In any case, I'm glad I do not see that "side-effect" and I hope it won't become a "feature".
I wish I knew. Maybe some preference setting that unexpectedly affects this area, or even corruption (in a good way ;-)) of your system folder settings. Count your blessings!