errors

If it fits in memory, then the slowness is likely due to other coding errors causing a bottleneck, not the SVM training.

Because, for a company that's supposed to have done all tue research, they've made a couple of errors.

Mostly the errors are expecting that a calculation is using signed arithmetic when it's not, and getting a large positive value where you expected a small negative value.

They found a whole bunch of dna errors in cancer, but they don't know which ones matter.

It's another to have errors so severe that they make the site permanently unusable for some people.

Even without progressive photon mapping you can choose a photon radius that won't cause visible errors.

Neither were perfect, and made errors every so often, but neither drastically more than the other.

And you can dismiss the numerous non-style related errors that one finds because you can claim a flaw in poster's personality.

They don't even know which are DNA mutations, and which are systematic sequencing errors.

Compile and runtime errors are clearly pointed out, usually with helpful error messages.

And it's 5 characters to prevent a whole class of errors for free (erroneous assignments to method parameters), so why not do it?

I'm kicking myself for halting the execution, because I made order of magnitude errors in my projections.

You can easily dismiss the numerous non-style related errors that JSLint finds because of the claim that JSLint "opinionated".

The "nearby cliff" is a good thing, in that it makes errors come out during testing rather than a month after you ship.