RE: [DiaCanvas] Me again.

[<arj...@he...>]

> Arjan:

Hi Phlip,

> I have polished my hard drive and installed Red Hat.
>=20
> CompUSA just didn't have Mandrake 9. Slackers.
>=20
> But this means I now have a golden RH8, with Gnome2,
> to try diacanvas on.
>=20
> The 0.9.0 tarball crashes at the exact same place.
> Deep inside of GTK constructing an object.

Even when you compile all without PIC (also see the README file):

	$ configure --without-pic
	$ make clean all
	$ make check

> g++ sez it's version 2.3 20020903
>=20
> But the reason I'm back is because of the unit tests I
> showed you how to right. I'm writing a book called
> /Test First User Interfaces/.=20

Ahh... That's nice ;-). Those unit tests did actually save me quite a =
bit of work. Did you know I wrote the constraint solver (DiaSolver, =
DiaConstraint, DiaExpressrion, DiaVariable and DiaStrength) in less than =
one weekend and wrote unit tests for them too. As you can see in CVS =
that code hasn't changed since!

> I'd love to add a C chapter.

Good idea. A lot of stuff is still written in C, expecially mission =
critical (and performance critical) stuff.

> So if we can possible create a version of diacanvas
> that does not detect my body's magnetic resonance
> signature, or whateverthefuck this bug is, then
> diacanvas goes into that C chapter.

Alright.

I'm aware that not all of DiaCanvas2 is tested, but the most critical =
functionality has.

> Deal?

Okay.


-------------------------------------------------------------------------=
-------