10g: still no whoami or linenum functions, sigh.

Question

It is a fairly well-known fact that PL/SQL lacks the ability to introspect and report on itself, items of great interest when logging and debugging, like:
whoami: satisfies "what is my name?" - expecting function, procedure, package.proc/func, or type name
and
linenum: satisfies "what is the current line number?" - expecting just that, the current line number within the program, with the option of reporting the line before the linenum call, or before the exception block was entered
I could, but won't, write a short novel on everything I've tried to work around these limitations. In the end, the best option to get at the logging and debugging info you need, is to set local variables to record the current routine name, and to record the current line number or action before every statement. Very kludgey. Very aggravating.
Is there anyone that monitors this forum that knows how to communicate to the PL/SQL Developers to push this through? I've been waiting for 9 years for these very basic utilities that other compiled and scripted languages have.
I've even emailed and spoken with Bryn Llewellyn before about this issue. He absolutely agreed, but somehow seemed powerless to get it in the next release.
Anybody with ears? Anybody with power?
[and please don't refer me to Tom Kyte's attempts to replicate this missing functionality. Tom's my hero, but his examples, and my own attempts, quickly break down in all but the most simple scenarios]
- bill c.

Answer 1

Padders,
I had no idea you haunted this forum as well! How do you keep up? Do you require no sleep?
This is actually a follow-on to a largish thread we had going under Error Messages, or some similar topic, in the Quest PL/SQL Pipelines. Unfortunately, that thread has been deleted (I miss the old WebBoard where they kept more online for longer; we're really doing ourselves a disservice by not asking Quest to keep all that in a searchable archive.)
I was pretty explicit there and gave examples how Tom's functions didn't cut it. The WHOAMI, as you noted, doesn't work well in situations where all your code is package based and the point in the call stack is several layers (packages) deep. The function to determine line number (can't remember Tom's name for it), mine is env.prg_line_num, also had problems parsing the call stack. The latter was particularly useless if you used exception handling. As I recall, the primary point of the whole thread was my beef with how you lost context if you choose to use exception handling, but if you remove exception handling, then your call stack could show exactly where the point of the original error was.
Anyway, thanks for replying. I looked through 10g's new features and didn't see that. Thanks for digging through the docs and finding that!