will python 3000 break my code?

Daniel Berlin dan at cgsoftware.com
Thu Mar 2 01:17:58 CET 2000


>>>>> "IVL" == Ivan Van Laningham <ivanlan at callware.com> writes:

   IVL> Daniel Berlin wrote:
   >>

   IVL> [snip]

   >> God damn it. It's not a feature. It's a bug. A bug. It should have been
   >> an error nit he first place. It's only by the grace of Guido and a slip
   >> of the eye that it wasn't.

   IVL> <now-tell-us-how-you-*really*-feel<wink>>-ly y'rs, Ivan;-)

If your really must know, i feel like someone who was forced to write a Part
of Speech tagger (HMM based) in Perl in one week (handling unknown words and
all), and had to watch it do 6-fold cross validation.
It takes it 118 minutes to tag a 12 meg file in perl, because it sucks for
number crunching.
I rewrote it in C++ while waiting for it to finish the validation, and it
took 38 seconds.
It's not an algorithm issue, either. Perl just sucks that badly at number crunching.
To quote Professor Allen, when I asked if i could turn in the C++ version,
and said that i didn't think perl was the right language to be doing the
number crunching in, "Yes. I really didn't realize it would be this bad"[1].
The kicker?
Mine is the fastest tagger of the class.

So besides my midterms, I have perl angst.
Forgive my anger.

--Dan

Footnotes: 
[1]  I have a feeling this isn't a rare statement about perl programs.





More information about the Python-list mailing list