Celestron handheld USB microscope

Since I go through so much trouble whenever I want to shoot macro, I decided to take a look at maybe a cheap microscope and try to figure out how to shoot it with my camera. I was delighted to find out that you can now buy a 150x USB microscope for less than $70. The imaging sensor is rudimentary when compared even to my phone's camera, but in this case the important part is the lens and focusing.

This is a Celestron 44302 handheld digital microscope. It is mainly designed for hobbyists that want to inspect things like coins, stamps and other collectibles, and professionals doing quality assurance inspections.

Installation was simple: pop-in the CD, install the application and then plug it into USB. There, it's done.

The microscope has a shutter that can be triggered by the software, and a focusing wheel. There are no markings or any kind of indicator to tell you your zoom level, and focusing is a pain in the ass. It took me maybe one hour of screwing with it until I more or less figured out how to get perfect shots at the lowest resolution and I can almost get a sharp shot at 150x 75% of the time. The main (only) problem is that the focusing wheel is really hard, and I don't know if it is going to loosen up with usage.

The image quality is decent, it does work. The problem is that at least until the focusing wheel is a bit easier to turn, the stand is by all means worthless: you can't keep it in position and focus at the same time. I took tons of shots of coins, memory cards, etc. It is definitely better than using the camera for straight macro because even handheld it is much easier to aim than the camera.

The Mythical Five-Minute Change

(I posted this originally at http://insomniaccoder.tumblr.com/post/229899519/dilbert-comic-strip-for-11-01-2009-from-the , reposting this for the benefit of one of my coworkers that couldn't find the original post)

From: Dilbert comic strip for 11/01/2009 from the official Dilbert comic strips archive.

This is why there is no such thing as a 5-minute change. Here's the real sequence of events for your "5-minute" change:

1. Somebody finds that a 5-minute change needs to be performed. 5 minutes.
2. Somebody asks for authorization to make the 5-minute change. 5 minutes assuming the explanation is simple.
3. Somebody needs to be selected to make the 5-minute change. 5 minutes if the work schedules are clear and up to date.
4. Somebody needs to explain the programmer the 5-minute change. Good luck explaining it in less than 10 minutes.
5. The programmer makes a change. 5 minutes.
6. Quality assurance checks the work. 5 minutes but it has a mistake.
7. The programmer fixes the mistake. 5 minutes.
8. Quality Assurance re-tests. 5 minutes.
9. Work is deployed to testing environment. 5 minutes.
10. Customer tests the change. 5 minutes. Luckily, he likes it as delivered. Staging is approved.
11. Deployment to staging environment for regression testing. 5 minutes.
12. Quality Assurance does internal regression testing. 5 minutes (in your dreams).
13. Customer does regression testing. 5 minutes (yeah, right). It passes, approved to deploy to production.
14. Work is deployed to production, assuming no outage is needed. 5 minutes.
15. Quality Assurance checks the production environment. 5 minutes.
16. Customer does production acceptance testing. 5 minutes.

That's 16 steps, 5 minutes each. In theory that's 80 minutes, but that's not realistic because there's at least one email between each step. This means no less than 15 emails, with a separation of 15 minutes. Your total is 16 x 5 minutes + 15 x 15 minutes = 305 minutes.
Your "5-minute" task is going to burn at least 305 minutes of billable work, OVER FIVE HOURS OF WORK.
It gets better, this process includes all players getting copied in all emails, so in reality that 5-minute task probably eats a few more hours, just based on all of the people that will be reading those 15 emails.
Isn't project management awesome?