Automation

[rowyn]

My department has a task that has required a person to spend a few hours every week pressing a button, waiting for the computer to process it for 30 seconds to a couple of minutes, and then pressing the next button. Repeat 76 times or so.

We have been trying to get the system folks to automate the button pushing. For a year. I kinda thought this was a job for a batch file. They apparently feel otherwise. It got booted around for a while.

Last week, they came back with their final offer: "we'll spend dozens of hours coding it so specific pieces of data are available in another system. Then you can spend dozens of hours rebuilding all your absurdly complicated reports in a way that wil not quite mimic their current form in the new system. That allows scheduling."

I whinged at bard_bloom about this. Bard pointed out that automating the reports should take like 50 lines of VBA code. If you do it badly. I whinged more.

My boss told me to write up all the details of all the data that we need so we could start the hundred-hour process system support wanted us to embark on, that they claimed was the only option.

I looked at this. I though, I bet it would take less time to figure out how to code the automation in VBA than it will to tell them everything we need, never mind the rest.

Three hours of searching, coding, and experimenting later, I have my proof-of-concept. Heh. I can do this after all!

...

I should've done this two years ago. -_-

Comments

[level-head.livejournal.com]

By the way, I completed a very time-sensitive project this morning with a $1.5 million client proposal at stake. And the federal website the proposal got uploaded to broke, spewing Javascript errors all over the place and revealing details of code, system, version and access that were never intended to be seen by users (and would be of great use to hackers).

It's a busy site: more than 400 proposals were uploaded between Friday night and Monday morning, and who knows how many of them got damaged by the problem. After some hours, we worked though the issues and the upload was pronounced bully before the deadline. But it was tense for a bit.

This relatively straightforward system used by every proposal-writing and research writing scientist in the country (and all their institutions) is prone to break, a lot, despite a huge development and maintenance budget. I've put together systems with 1,500 screens and 400 database tables (with 10,000 separate columns) that never had such fragility. On less than 1% of their budget.

This is part of what taxpayers are buying; the federal government agencies don't even have your employer's motivation to keep costs down and make things work, as modest as that motivation can seem to be. In fact, your attitude of wanting to make things work, and the skill to do it, would be considered rather dangerous.

But I like it!

===|==============/ Keith DeHavelle

Yeah, but private industry is incompetent too.

[jorrocks-j.livejournal.com]

I've been dealing with USAJobs, which to be fair sucks a lot less than it did even two years ago. Still sucks though.

But then its private equivalent is Taleo. And Taleo is like the Kim Kardashian of recruiting software: it doesn't work and everybody hates it but somehow, you just can't avoid it.

Re: Yeah, but private industry is incompetent too.

[level-head.livejournal.com]

Private industry can get large, and when it does so, it can start to take on the attributes of a government. Specifically, a bureaucracy, where the preservation of turf and the resistance to threats to the bureaucracy become more important than the overall effectiveness of the work.

Robert Heinlein once described a committee as "a creature with six or more legs and no brain." But a bureaucracy is more "many bellies and no brain" as its appetite must be figured into the equation.

I ran a company for decades that reached the "multi-hundred employees" level. It was difficult to avoid the bureaucratic effect, though I think I was largely successful at that. Well, thinking back, there was one caveat. One part of this, the larger part, I grew from scratch starting in 1976. (I still have an image created for the company by Lady Rowyn for the company's 28th anniversary.) The second part of the company I acquired — and that was a poisonous, bureaucratic bunch of corporate politicians that I had to get rid of quickly. I thought I could reform them, but that would have been a new career and we had work to do.

Companies that get successful must fight these effects. It is tough. Big successful companies did not get that way by acting the way they do now! I admire CEOs like Jack Welch who could keep a good spirit going in a hundred-thousand-employee company.

===|==============/ Keith DeHavelle

[rowyn]

Well, bear in mind I did spend two years whinging "why won't someone do it for me?" first. :/

[level-head.livejournal.com]

I've seen you tackle other things in the meantime. And there's a good reason your co-workers consider you the go-to person for this stuff.

===|==============/ Keith DeHavelle