Prototype app outperforms and outlasts outsourced production version

Behind every successful company there is that one weird Visual Basic 3 app still running the show

Who, Me? We all want the users of our software to be happy, but how far would you go to fulfill that requirement? For one Register reader, perhaps a bit too far. Welcome to Who, Me?

"Harry" (not his name for reasons that will become clear) was working as a contractor and doing some onsite development. His team had come up with a nifty prototype for a new process, and the users absolutely loved it. Management was also impressed, but Harry and co. would not get to develop a production version. That task would fall to a big outsourcer.

While the Big Guns worked on a production version, Harry and his team kept the users happy by polishing the prototype. Slow bits were rearchitected. Working bits were buffed. User feedback was addressed. The usual sort of thing.

This went on for a few years until "the Big Guns asked for us to stop and freeze the build," said Harry. "They had handed the job to a Big Gun Contractor (you know the shortlist) and it was moving with all the speed and user appeal of a striking slug."

At every meeting the users kept adding to the requirements list and the contractor couldn't keep up. The production beta was in a horrid state and the users refused to go near it. "Could we please give them at least a chance?" was the plea.

Management conceded and declared that putting out new features was forbidden. It would be a sacking offence. The changes had to stop.

Except Harry's prototype was still rolling along in the background. "We had a new build nearing completion," he said, "including some deep productivity features the users were keen on, and also featuring a revised branding because the functionality had by now way overtaken the original concept."

"Could we just roll that out?" wondered Harry. The bosses were told some very slight untruths: "It was purely a maintenance release and had no new functionality, honest!"

"Userland pushed, management acceded. Rollout!"

Except that once logged in, it was clear this was quite a bit more than a maintenance release. "The revised branding involved a total overhaul of the landing pad," admitted Harry, "it looked like we had grafted on a whole new product."

It didn't take long for the first angry email from management to arrive. Sure, the users could be trusted to keep quiet about getting their wishes granted. But if anyone more senior took a deeper dive then the upgrades would be revealed and pink slips issued.

The road to redundancy is paved with good intentions.

What was Harry to do? The answer was to double down. One of his many hats was as a technical author, and he'd been involved in setting up an in-house wiki in which staff and contractors alike were encouraged to post. It had become wildly popular. Could it also save Harry's bacon?

"I now noticed that in his ire, the manager had forgotten to BCC everyone and all the anonymous pointy-haired managers who had obstructed us over the years were revealed in all their sordid glory," said Harry.

"I reckoned it would take any one of them perhaps 10 minutes to steam over it and check their facts before responding to it. That was how long I had. So I hit Reply All and started typing."

Tappity-tap: The old UI had begun crashing. The new one was just a cosmetic makeover. Everything was in the Wiki. Send.

Harry switched to the system's page in the Wiki and hit Edit. More tappity-tapping: A longer version of the email. Copy and paste a list of the maintenance fixes. Upload the new logo. Pop in the overview from the Help manual. Save. Done.

He had just enough time to grab a well-earned coffee when the first reply dropped into his Inbox. A manager had indeed taken a look at the wiki page: "Yes, it's true," went the reply, "it is only a cosmetic thing, I checked the wiki and it is quite clear about that. We can all breathe easy and go back to work."

Lucky nobody thought to check the edit history, eh?

Having gotten the final update out, there was nothing left for Harry's team to do and they went their separate ways, "but at least we still had jobs to go to."

But what of the production system? "It was always ludicrously expensive and unusable," recalled Harry. "Other, newer systems took over piecemeal as changing needs dictated."

In an effort to push the users over the line, accounts in the prototype were disabled "but that just meant the business quality plummeted."

There would be more years of pain before the system was finally decommissioned and the bits of Harry's team's prototype that were not outdated were resurrected.

"It is a warm feeling when your prototype lash-up outlasts the production tool," he concluded.

Ever tried to help your users, but found yourself having to cover your tracks when management took exception to your helpfulness? Or written a temporary bit of code that ended up becoming too permanent? Confess all with an email to Who, Me? ®