The Market Ticker
Rss Icon RSS available
You are not signed on; if you are a visitor please register for a free account!
The Market Ticker Single Post Display (Show in context)
User: Not logged on
Top Login FAQ Register Clear Cookie
User Info Be Scientifically Right, Get Fired; entered at 2017-08-11 11:28:57
Posts: 149433
Registered: 2007-06-26
It was one of those do or die moments. We either succeed or we were out of business. That was in 2000. We are still in business!

Sometimes time is NOT a luxury.


MCSNet was successful because I worked those weekends and overnights. There were others who did too, when the need arose.

My two previous jobs? Both places hired me with the full disclosure that I'd be expected to put in extreme hours a good part of the time and on a continuing basis. The first real (coding) job I had came with the same requirement -- disclosed in advance, and THAT ONE required on-site travel also, often to hit the ground running when the wheels came down.

The entire reason I can run this place today is because I did that. I took it on and succeeded. But there was utterly no way for me to discharge my responsibilities without that sort of time and effort.

It was a trade-off that I willingly took on. There was nothing particularly physically demanding involved (other than humping a router here and there on occasion, which can be a damned heavy piece of gear) but the emotional and intellectual demands are another matter entirely. You either had that capacity or you did not. If you couldn't show up almost-literally every single day and work 12+ hours six days a week and sometimes ALL SEVEN you simply couldn't do the job. Migraines, hangovers, got a cold, allergies, whatever? As long as you can still do the work and be productive, ok -- but if you have call out, and it doesn't matter why -- you're done. There's simply no margin for you failing to pull your weight EVERY DAY. If I had been in ANY way emotionally fragile and unable, as a result, to do that job EVERY SINGLE ****ING DAY I would have failed -- HARD.

When I ran MCSNet at one point we outgrew the original customer management software that basically ran the joint. For security reasons there were two phases of each day's operation; the billing was run and then an export file was used to authenticate the credit card charges on a separate machine. The results from that run were then put back against the operational system. (This was why we never had our CC database broken into; it was literally NOT THERE on the mainline system.) But -- it was 5 years old and built for a time when computers were 1/10th the power they had grown into.

We didn't have the available skilled people to do a "fire team" sort of thing, never mind that giving four or five people each a quarter of the job would have taken a month to design and decide, then months to implement. We didn't have the time and we didn't have the available talent. There was only one way to get it done: I had to do it.

So I told my #1 (Marcus) that absent the FBI raiding the place or a fire in the building knocking on my door was a very bad idea and it was his job to keep people from being fired, which would happen if I was interrupted. I then barricaded myself in my office on a Thursday.

The following Monday morning we had a working replacement, fully-integrated and online. It consisted of several thousand lines of "C" backed by a Postgres database and system software modifications to the FreeBSD OS that actually ran all the user-accessible "things" on the network, along with a custom RADIUS backend plug-in to handle dial-in authentication. NOT ONE LINE OF THAT CODE EXISTED, INCLUDING THE CHARACTER-BASED WINDOW MANAGEMENT FUNCTIONS AND THE SCHEMAS FOR THE DATABASE, WHEN I STARTED. I HAD A FUNCTIONAL FRONT-END RUNNING IN 48 HOURS AND THE ENTIRE SYSTEM UP AND RUNNING IN 96 HOURS -- FROM ZERO, AS ONE MAN.

I think I slept perhaps six hours over the entire four days and I have no idea how much coffee got consumed. I then spent Monday training Liz (Customer Service head), Marcus and a couple of others on how it worked and had them go out on the floor and show everyone how to use it, as it was live and active.

It wasn't flaw-free by any means and it certainly wasn't the best-documented or even "best architected" code I've written but it was working, and over the next month or two the relatively minor bugs that needed to be sorted, were.


Now try that when someone says "I can't; I got a migraine" or "I can't, my 6 month old kid has nobody to watch him/her."

Nope, nope and nope.

Last modified: 2017-08-11 11:35:51 by tickerguy

2017-08-11 11:28:57