[Tm4j-developers] Post 0.7 JDK support

[Kal A. <ka...@te...>]

OK, I promise that this is the last email for this evening.

Now that the dust has settled on JDK1.4, I would like to gather people's=20
opinions on what JDK version the release after 0.7.0 (1.0 ? ;-) should=20
require. My feeling is that we should move with the times and go up to=20
JDK1.4. My reasons are:

1) TM4J has never supported development of applets that run in a browser =
VM=20
(i.e JDK 1.1.8). So there is nothing to be broken there.
2) JDK1.4 would enable us to ditch Log4J - reducing the number of additio=
nal=20
jars we need to bundle up by one.
3) Most importantly, JDK1.4 has much better URI handling than JDK1.3.x - =
my=20
impression is that with JDK1.4 we could remove the org.tm4j.net package=20
entirely. JDK1.4 URIs are serialisable, so there should be no probs with =
the=20
Ozone implementation. Moving to JDK1.4 would also mean that developers wo=
uld=20
be able to use a familiar URI object interface rather than having to lear=
n=20
the TM4J way. To be honest, I never really saw all the network address=20
handling as being a fundamental part of TM4J - I just did it because I ha=
d=20
to...and now I'm quite glad that Sun have thrown me a lifeline to get out=
 of=20
having to maintain it (and implement URNs ;-)

However, there may be some arguments in favour of sticking with our curre=
nt=20
JDK level (1.3.1) - for example, I am kind of assuming that Ozone is happ=
y=20
with JDK1.4 (I can't think of any reason why not...I guess I should go tr=
y it=20
:-).  Finally, remember I am *not* talking about the 0.7.0 release which,=
=20
when it happens, will still be JDK1.3.1 compatible.

If you have any comments on this, please share them!

Cheers,

Kal
--=20
Kal Ahmed, tecqhuila.com
XML and Topic Map Consultancy

e: ka...@te...
p: +44 7968 529531
w: www.techquila.com