Re: failure notice

From: Kevin Menard (kmenar..ervprise.com)
Date: Fri Oct 07 2005 - 22:24:45 EDT

  • Next message: Andrus Adamchik: "Re: failure notice"

    On Fri, 07 Oct 2005 21:48:46 -0400, Andrus Adamchik
    <andru..bjectstyle.org> wrote:

    > So I am starting to re-evaluate our "pragmatic" approach. We are a
    > framework provider, and unlike end-user applications, it does matter
    > what is under the hood. Fewer dependencies is better; zero dependencies
    > is ideal (though unrealistic at this point). Look at HSQLDB - a single
    > jar that does everything (they even have their own HashMap, presumably
    > for 1.1 compatibility). As a result, the only possible conflict it can
    > cause is with another version of self.
    >
    > Again, I am not calling for a rewrite of Cayenne, just wanted to point
    > out to our priorities.

    Out of curiosity, what are our priorities? Everyone strives for a full
    set of features, ease of use, easily adaptable codebase, performance,
    etc. As with all things, a focus on one normally means a trade-off on
    another. Clearly once you start phasing out external implementations in
    favor of internal ones, you increase the workload all around. Taking JDOM
    as an example . . . presumably the years of development that went into
    that are going to be more efficient than something that I spend some free
    time on. Then there's issues with transfer of knowledge (will I be the
    only one that knows how this XML implementation works?). There are a
    bunch of other potential issues I'm sure everyone here is aware of.

    Don't get me wrong, reducing dependencies is an admirable goal. I
    personally like being more pragmatic about things, but then again, I've
    never really written a publicly used framework. To me, this shift seems
    to appease a group of users that are using what I would consider to be a
    broken app server. I can't believe IBM expects people to only use their
    supplied beta version of a rather popular library. How large is the Web
    Sphere market? How many of them are using Cayenne or considering it? How
    many other people use Cayenne? Are we going to be doing a lot of extra
    work to help a small minority at the expense of the larger community? Is
    this really a larger problem than the current Web Shere issue?

    These aren't questions meant to direct us down any one path. They're just
    thoughts that come to mind and they're genuine questions that I don't know
    the answer to, but would like to.

    Actually, come to think of it, I'm not even sure what the goals of 1.2
    were. I like what I'm seeing for sure, but I'm not sure what the
    destination is.

    Anyway, a little goal clarification might be in order.

    -- 
    Kevin
    



    This archive was generated by hypermail 2.0.0 : Fri Oct 07 2005 - 22:24:23 EDT