[alblue]

The APT was rushed into service before it was ready and suffered from initial glitches which ultimately led to bad press killing it off.

On the other hand I use the descendent of the APT (the Virgin Pendelino) to commute into work and back. The tilting mechanism is really sublime and incredibly comfortable. It was ahead of its time.

[moron4hire]

I'm sorry, I hate to be a pain, but it's a pet peeve of mine. They weren't "glitches". They were defects.

Glitches are errors that happen outside of the design parameters of the system. For example, when you touch the antennae on an AM radio and you hear a pop or buzz from your own electrical field in the radio.

Defects are errors that occur within the design parameters. So a train that can't handle some snow on the tracks, or a national health insurance enrollment site that falls over after only a few users show up, those are defects.

Bugs are a type of insect. Using the word "bug" to describe a defect implies that the "bug" showed up, an external factor that got in and gummed up the works or something. But that doesn't happen in software. The defects were always there.

I think it's important to use the terminology correctly and make a habit of using it correctly because I think it puts the emphasis on the fact that we create the defects, and the defects were always there, they didn't just develop, they just had to be found. That's also why I don't like the term "software maintenance". When you have to take a site down for several hours every 3 months for "routine database maintenance", that's a defect in design, not just "changing the oil".

It was our fault. Calling them "bugs" or "glitches" or "maintenance issues" diminishes that.

[alblue]

The reason "bug" is used is because the first computers were programmed through hard wiring and mechanical switches, and they were actually insects that were present. This is from Grace Hopper in 1947:

http://www.computerhistory.org/tdih/September/9/

Convention has stuck which is why we have software debuggers, and the word bug in computer science has come to mean any unintended result, not just the presence of insect wildlife.

(I myself debugged a floppy drive in '87 which had an earwig stuck behind the read head backstop resulting in the read head being misaligned to the sectors)

[jcl]

Use of "bug" to mean "defect" goes back a lot further than the computer age:

http://www.etymonline.com/index.php?term=bug&allowed_in_fram...

At the time of Grace Hopper, the term was used in engineering contexts to refer to defects in general, quite unrelated to insects. Per your link, that's why she wrote "first actual case of a bug being found" -- it was amusing and unusual to find a bug (defect) caused by a bug (insect). If this were not the case, there would be no need to use the words "first" or "actual", or even note the event.

[smoyer]

Well ... unless they were caused by stray transient electrical spikes changing the state of the system. These could be a defect but it might also be caused by improper input signal conditioning (which would be a defect elsewhere in the system). In either case, we've traditionally called these defects glitches and that lexicon has been broadened to more generally mean "unintended behavior".

[zumatic]

As well as the glitches, there's a story that the tilt was initially set to entirely cancel the lateral acceleration from the curve. The press on the demo train was well lubricated with free booze -- result: headlines about the new train making them sick.