Eric S. Raymond
esr at thyrsus.com
Wed Nov 8 10:03:57 UTC 2017
Gary E. Miller via devel <devel at ntpsec.org>:
> So then what? Any alternatives? Translating by hand would be a PITA and
> error prone. Rewriting from scratch even worse.
I was assuming we'd do translation by hand with acript assistence.
But that's not as bad as you might think it would be.
The reason is this: when doing this kind of translation, you're going
to screw up most on the boring bits, because they put your brain to
sleep. Your odds of translating a stretch of code erroneously are
thus inversely proportional to how easy it is to do with a regexp
That's sort of the flip side of the 80% wall I talked about. It's
*why* the Go devs wrote a tool that was a dead end - because before
hitting that dead end it automated away a lot of their grunt work.
They elected not to try and document it, and warned people away from it,
because if you publish a tool like that that does 80% of the job and
ovccasionally introduces an error, you have to be prepared for people
who don't realize how hard the last 15% is to flame you mercilessly.
If you don't understabd this kind of teanslation as well as I do it
looks like crappy work.
If we do this, I expect to write my own equivalent of c2go and
make it work about as well as they did. I probably won't
publish it either.
<a href="http://www.catb.org/~esr/">Eric S. Raymond</a>
My work is funded by the Internet Civil Engineering Institute: https://icei.org
Please visit their site and donate: the civilization you save might be your own.
-------------- next part --------------
A non-text attachment was scrubbed...
Size: 811 bytes
Desc: not available
More information about the devel