- From: Bang, Steinar <Steinar.bang@tandbergtv.com>
- Date: Mon, 5 Nov 2001 09:36:56 +0100
- To: "'Joel Young'" <jdy@cs.brown.edu>, Tal Halfon <talh@smarteam.com>, www-lib@w3.org
Joel Young [mailto:jdy@cs.brown.edu] writes: > By the way I am cutting all of my code over to using > these libraries: > > libcurl --> http://curl.haxx.se/ > libxml2 --> http://www.xmlsoft.org/ I just gave libcurl a look. If I understand it correctly, libcurl is blocking? Ie. the request for an URL doesn't return until the data or an error response is received. Is this correct? My primary reasons for using libwww, was 1. its event-driven single thread operation. 2. the stream building response, which let the application parse and use data before they are completely received The FAQ says that libcurl is thread-safe, so blocking calls may not seem as too much of a limitation. But in the case of the program I was working on, the main application itself wasn't multithread-safe, so we couldn't easily have utilized it. Hmm... I'm wondering how they do such things as keeping an HTTP 1.1 connection open across requests? This email, its content and any attachments is PRIVATE AND CONFIDENTIAL to TANDBERG Television. If received in error please notify the sender and destroy the original message and attachments.
Received on Monday, 5 November 2001 03:37:29 UTC