At 2005-04-18 10:02 -0400, Elliotte Harold wrote:
To be honest, I'd be a lot more interested in doing the work myself or
as part of a team than in paying somebody else to do it. I've been
considering that for a while now. So far it's been easier to fix Jaxen's
XPath and rely on Xalan for XSLT than to fix Saxon, but that may not be
true forever, especially if performance becomes a concern.
But to me Xalan has far more egregious errors with respect to attribute
sets in the import tree than those to be fixed in Saxon. So it isn't just
performance, it is conformance.
Even if we managed to raise the funds to pay Michael Kay to do 6.5.4, what
happens when we need 6.5.5?
I'm not convinced that after so many months of stability in 6.5.3 with so
very few problems that we are going to quickly need any 6.5.5. And if we
did, then it might become a business decision for someone to get a 6.5.5
version created.
I think we'd be much better served by an
active Saxon community than by hiring one programmer, no matter how talented.
Not for my immediate needs, Elliotte, which is why I'm undertaking to do
this right now and as soon as possible. I have a business need to get
attribute sets in the import tree to work, which is why I'm contributing to
an effort to get it done.
I see the problem with "an active Saxon community" is the lack of
accountability. And I think that it is more than Mike's great talent, it
is his intimate knowledge of the details and his assets of regression tests
and techniques that I'm willing to help pay for.
Development communities can be very useful, but I've been let down by being
unable to make a business decision to include an open tool when there is no
accountability for the reporting and handling of problems.
I've privately funded other developers for specific features of open-source
tools in which they are the lead because of the business decision that I am
willing to pay someone for something to have it now and have it work with
confidence because they are they author. If I can bump up a new feature's
implementation with some monetary incentive, then it is worth it to me.
My appeal is just to share the burden and the benefit of some improvements
to an already great tool that many people are using. If I can convince
people to support this upgrade, then as they take advantage of features of
XSLT they haven't used they won't trip over these oversights than can be
addressed in short order.
And, if I cannot raise enough to bring the shared cost to what is
affordable to me, then I'll have to make a business decision otherwise and
abandon the effort to get it fixed.
. . . . . . . . . . . . Ken
--
World-wide on-site corporate, govt. & user group XML/XSL training.
G. Ken Holman mailto:gkholman@xxxxxxxxxxxxxxxxxxxx
Crane Softwrights Ltd. http://www.CraneSoftwrights.com/s/
Box 266, Kars, Ontario CANADA K0A-2E0 +1(613)489-0999 (F:-0995)
Male Breast Cancer Awareness http://www.CraneSoftwrights.com/s/bc
Legal business disclaimers: http://www.CraneSoftwrights.com/legal