nullprogram.com/blog/2010/09/21/
Ever since I learned
this neat trick from Luke a few months ago I've been using it
almost weekly. If you have Python installed then you have a miniature
web server at your fingertips.
python -m SimpleHTTPServer
Bam! If I need to transfer large files directly to someone over the
Internet, I'll throw up one of these and give them the address. Each
of my home computers has an 8xxx port forwarded to it in my
router's NAT configuration, so they're all ready to do this any time I
need it.
When I needed this before, I would use my
own Emacs web server, but the Python
solution above is even more convenient most of the time. Plus it does
directory listings, which I never bothered to add to my web server.
The reason I bring it up right now is because it finally saved me a
lot of time at work. I was performing a half dozen Ubuntu installs for
a system we're building, but none of these computers have network
access, except to each other. I was using the Ubuntu DVD, which
includes a larger software selection than the CD.
Well, it seems no one ever actually uses the DVD like this, because it
doesn't work right now with the current DVD. You see, apt
doesn't treat the DVD as just any repository. It has its own special
cache for them, keeping track of what packages are on what CDs and
DVDs. Well, the problem is that the CD is named differently than the
DVD, and apt (and apt-cdrom) keeps looking
for the CD when it should be looking for the DVD, foolishly ignoring
any advice I give in the configuration. Python SimpleHTTPServer to the
rescue!
I mounted the DVD and ran the HTTP server at the DVD's root. Then I
added my localhost server to the sources.list. Worked
like a charm! apt had no idea it was pulling packages off
the DVD.