Thanks Chris, I was hitting the wrong URL by mistake. Didn't think an extra / will make all that difference :)
On Tuesday, March 10, 2015 at 2:12:13 PM UTC+2, Chris Angelico wrote: > On Tue, Mar 10, 2015 at 10:55 PM, Steven D'Aprano > <steve+comp.lang.pyt...@pearwood.info> wrote: > > Miki Tebeka wrote: > > > >> Greetings, > >> > >> $ time curl -I https://pypi.python.org/pypi/ > >> HTTP/1.1 200 OK > >> Date: Tue, 10 Mar 2015 10:24:30 GMT > >> ... > >> Content-Length: 9870689 > >> curl -I https://pypi.python.org/pypi/ 0.02s user 0.00s system 2% cpu > >> 12.271 total $ > >> > >> Note the long time (for comparison hitting python.org takes 0.209 total) > >> and the size (Content-Length). > >> > >> Anything gone wrong or am I missing something? > > > > I don't know. Am *I* missing something? What makes you think this is a > > problem? > > > > You're downloading via https, which has more overhead than http. There's a > > certificate that needs to be checked, the content can't be cached, and > > there's the cost of encryption. > > Those costs don't factor in here; the content-length is what the > server announces as the number of bytes of oncoming payload. > > > Do you have some reason for thinking that the content-length should not be > > 9870689 bytes? Should it be less or more? > > > > Have you tried opening the URL in your browser? > > I just tried it now, and it's comparable in a browser. 9MB is a lot > for a landing page, and while I don't have a specific record, I have a > vague recollection that it wasn't quite this big before (which would > mean that the landing page was paginated instead of having the entire > list right there). > > Oh wait. What I was remembering was https://pypi.python.org/ without > the extra pathing on it. And yes, that page _is_ short and fast. So > that's an appropriate landing page. > > To the OP: You're downloading the entire list of packages, which is > why it's taking so long. > > ChrisA -- https://mail.python.org/mailman/listinfo/python-list