Convertabase

I was on a project once that required triggers to be built for about 200 tables, these triggers were dependent on the structure of the table, so any change meant rebuilding them. The team members that were in charge of the database, were projecting later and later completion dates as changes were made to both the trigger’s output and the tables. (At this point, they were projecting they would complete about a month after the product shipped.) I was on another part of the project and was ahead of schedule.

The manager over the project assigned me to help them “get back on track”. I was less than welcome because they didn’t want some fancy pants know it all to show them up. They had been building all the triggers by hand. I asked for a backup of the current DB, and was told no, so I asked for access so I could try a few things… I got my backup copy.

I went back to my desk and wrote a Trigger writing program. The next day, I went over to their area and asked them to back up their database, they were reluctant, but I threatened to run my test on their live, un-backed up database, and they backed it up.

I ran my program and with a single button click dropped all the triggers they had created so far. Then selected all the tables from a list and generated the triggers with another click. Poof, they were now “caught up”. I showed them how my tool worked and handed the lead for the group a CD with the source. Then I asked, so, do you guys need me anymore? (They were glad to tell me no.)

At the meeting later that day, the database team reported that they had caught up. The manager looked at me and said what happened (and the DB team was glaring at me) I said they had a brilliant idea, and they didn’t need me after all. My manager looked suspicious, but let it go, the DB team looked shocked that I had not taken “credit” and made them look bad.

Later the Manager said, “You had the idea didn’t you”. I responded “Prove it, besides do you really want them to resent my help whenever you offer it?” He said “No… Never mind”. The DB lead came over and asked for some advice on building tools when he was handed a tough assignment later. Success and the DB team didn’t end up hating me after all, success X 2.

I have worked with a large multinational corporation, and as their projects would slip, they would have meetings where they would redefine the meaning of “success” on a regular basis to protect everyone’s bonuses.

I was on the team for a migration to a new version of the software, and spent two weeks on “crash overtime” trying to find a million dollars that had disappeared from the report the CEO relied on, only to finally discover that no one had told him the new report had moved a line item from page one to page three.

This kind of report “misunderstanding” is more commonly an emergency when you do a one time “Conversion” instead of a data migration where the CEO and others have a chance to run their reports before the “drop dead” date. If the CEO in my story had been able to see the reports in advance, telling him the item was on page three would have been easy. Instead, when the company has made the jump and there is no way back, he is focused so tightly on the bottom line, that no one wanted to be the one to say ”are you reading the report right?

There was a large multinational fast food company, that had an ERP (Enterprise Resource Planning) system that they had refused to upgrade for years because the customizations they had done would have to be done all over to the new system.

This system ran their just in time deliveries world wide.

A new version of their ERP software was released by the vendor that finally pushed their version into the sunset provisions of the vendor’s contract. They could not buy support at any price; suddenly the ROI was there for them, and in spades.

They asked the vendor how to get from where they were now, to the newest system. They were told buy every interim upgrade and run them back to back htey were told “you should be down six months tops.”  Obviously, this was unacceptable to them.

I was called in, and ran my tools (http://www.convertabase.net) to get a look at my tools) to get an idea of the problem, and quoted them $250,000 and four months to jump, two whole major revisions to the most recent version. They asked for a formal proposal, which I produced and sent them the next day (automated documentation is so cool…). I never heard back from them, and they would not take my calls. I wondered what I had done wrong, so finally, I contacted a friend who was doing long term contract work for them. He told me the VP in charge had received an identical proposal from a competing consulting firm the day after mine had been delivered, the only thing different was the cover sheet. His son worked at the “Competitor” and they decided to go with them. Now I could have tried to make a fuss, but I knew I could not “win” so I decided to see what would happen when this “Competitor” tried to do with “one off code”, what my tools could have done in less than four months. I kept asking my friend how things were going there and finally over two years and two and a half million dollars later, they had a Friday night conversion with the attendant Monday morning regret. Conversions are non repeatable, one way and since you can’t test like you do with a migration, are not as good of a choice, but I digress. The VP made a decision based on relationship that really hurt his company, but the whole time people were saying what if… he had to say the other guy would have had the same problems… and it’s just not true. Fortunately this company was big enough that they are still around, and going..