From: Don G (dongryphon_at_[hidden])
Date: 2005-04-02 13:02:13
While we are coming from different sides on this, I do want to thank
you for your input on libevent and epoll. It is a positive
contribution to the discussion. Please see below for my replies.
> But I also know that modern monster sites like
> http://www.livejournal.com/ would not be possible
> without epoll. (Well, I may be exaggerating a
I believe that an epoll-based implementation (not an exposed
interface) would be one of the most likely. I'm really not dis'ing
epoll's use, just where it is used.
> Having said that, nowadays epoll support is a
> requirement for any linux which is going to be
> used in any decent networking.
That is good news for the Linux platform, but it doesn't change
things in the general case.
> select bit manipulation is ugly and boring. poll
> is much better. epoll is pretty much as simple as
> one can get.
I agree with the progression of better-ness all the way to "as simple
as one can get". I think it can be much simpler.
> This is what I'm talking about - my problems were
> solved long time ago. I'm happy with a C-style
> callback taking void* in libevent. And it's trivial
> to wrap it to make it accept and callback arbitrary
> function objects. I've been already living in a
> perfect world...
I guess a "perfect corner of the world" would be more accurate :).
libevent is GPL, which for me is a full stop on its use and I cannot
(politely<g>) say just how much I dislike C driving my C++
programming. While I came from a C background, I want to see pure C++
solutions for my problems. They are safer, easier to understand, use
and maintain, but let's not get into a debate on such ethereal issues
in this space. I think most folks using boost would agree: GPL evil,
C bad, C++ good. ;)
Do you Yahoo!?
Take Yahoo! Mail with you! Get it on your mobile phone.
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk