Re: v3.1 TODO and deadline

On tis, 2007-10-02 at 23:04 -0600, Alex Rousskov wrote:

> Sure, but releasing v3.1 with a few _more_ features than v3.0 seems like
> a step in the right direction. I doubt we have or will have the capacity
> to implement all 2.x features in a single 3.x release in the foreseeable
> future. Some features will be missing, possibly forever.

And I doubt we even want to be fully complete with respect to Squid-2.
Better to implement the features people actually ask for, and to be
clear on that.

> We cannot please all the users, but can please an increasing number of
> them, which should generate an increasing interest in testing and
> sponsoring new Squid development.

Indeed.

And the two main features of Squid-3.0 is what I put my hopes for

  - ICAP, to attract new proxy users

  - ESI, to attract interest from content providers by allowing them to
solve the content delivery problem in a more structured manner.
Unfortunately a little on the side burner at the moment, but it's there
and likely to improve if people use it..

And for 3.1 my main items on the feature wish list is

  - IPv6, to attrackt even more users of both kinds. Esp in Asia.

  - compression. To attract ISPs with expensive bandwidth or limited
bandwidth applications (i.e. mobile networking).

  - plus the most frequently asked features of Squid-2.

There is also a great need for better performance and cleanups, but
that's always the case. Needs to be done, and what can be done
incrementally in that area with reasonable efforts should be done, even
during a STABLE cycle for that matters.

But the quite intrusive changes needed to be done real proper it's not a
target for 3.1 I think. Remains to be seen what that roadmap looks like.

With this in mind I also think the time schedule of beginning of 3.1 in
early 2008 is quite realistic.

Regards
Henrik