||Jun 29, 2000 at 05:17 UTC
(20 years ago)
||Jan 24, 2020 at 13:09 UTC
(39 weeks ago)
|Location:||SF Bay Area|
Oct 19, 2020 at 23:36 -07
|For this user:||Search nodes|
I played the Zork-based game Adventure on a PDP-11 when I was 4 and became addicted to computers in general.
I played Zork too, but I was 12/13 by then.
I learned Basic at 7 using an Apple II. I got my own Commodore 64 at 11. I learned DOS at 13 and Pascal at 15 and C at 16. I worked myself into a management position at 20 that allowed me to play with Foxpro/Mac - ew.
My father (also an Oracle DBA/developer for years) somehow led me to a Berkeley Oracle class, and I really liked how organized and accessible and stable Oracle was - enterprise via Unix vs. MS/Foxpro crap.
I landed a job as an Oracle Developer and got more into Forms/Reports/Menu Oracle apps stuff and excelled.
After a year working professionally, our SysEng
encouraged me to focus more on learning perl as I had a seriously solid understanding of HP-UX at the time, including ksh, regex, sed, ed and sysadmin (HP kernel tuning is fun,) X/CDE dev, etc...
I happened to seek Perl's use to reverse-engineer Oracle Forms (3.0/4.5 at the time) due to the fact that I had taken my vi/sed/ed experience to their respctive limits. I was looking for an editor that could replace multiple lines of code w/any other data I provided. I'll even admit that sed is *supposed* to be able to handle multi-line regex's, but HP-UX/Solaris/AIX seds all seem to have a bug which prevents you from matching/replacing multiline strings - at least in the time I spent on it. Hence my perl use from day one.
I wanted to reverse-engineer Oracle b/c it would mean thousands of clicks across 500 forms for each date field it encountered. It involved opening each form in the IDE separately at the time. This was for a 2k project so that it would convert all Oracle char(2), varchar2(2), number(2) values into date values. The other developers ended up taking the converted forms I created and all they had to do was reposition date fields in the GUI screen since the fields were now two chars longer. I was looking into mathematically going through all the forms and reposition them based upon their position and the relative position of "near" fields, but doing that manually ended up being slightly faster - not much. It's funny how often I run into that scenario.
During the last few years I've applied Perl to so many aspects of my job, it is sickening. Anytime I've needed something new, Perl either already had the module included or there has been a module there for me. Everything from ACLs to zlib.
How cool is that?!
A coupla years ago, I translated my sar data into Excel spreads into charts using Win32/OLE and perl. The fact that I'm so opposed to anything MS was greatly influence due to the fact that I was using perl. That was pretty exhilerating running a perl script that popped up dozens of worksheets showing our cpu load w/appropriate date headers and all...
I was last looking for a perl/NetSNMP solution to monitoring processes on Sun equipment... It seems to me that even Sun is trying to subvert open source by creating agent text-MIBs that are DO NOT want to be compiled and/or do not return entries. This is in reference to Sun's SMC (Sun Mgmt. Center.) <*update>It all ended up OK as soon as I chose to study/use NetSNMP's own native handlers - doh. I now have complete, secure monitoring info from my servers including df's, ps's, device info, ifconfig's and netstat's.
That last paragraph was N years ago and I've learned even more since then...
There are so many perl rock stars on here. I'm always intrigued and amazed after reading posts here. There is a definite community that this site attracts... And it's a brilliant one...
Thanks All :D