[ODE] ODE and the STL

[Alex Hetu]

I don't think people are being too stubborn or just plain dismissing STL
for the sake of dismissing it. Jon came up with a lot of excellent
arguments and I totally agree with him. You can't assume that STL will
behave the same way on different targets, it just won't. Like Jon said,
it is great for a self-contained application, but please oh please don't
impose it on others. STL can do magic it's true, but it can also do some
great voodoo. 

Alex

-----Original Message-----
From: ode-bounces at q12.org [mailto:ode-bounces at q12.org] On Behalf Of
Daniel K. O.
Sent: Monday, December 25, 2006 10:13 PM
To: Jon Watte (ODE)
Cc: ode-list
Subject: Re: [ODE] ODE and the STL

2006/12/26, Jon Watte (ODE) <hplus-ode at mindcontrol.org>:
> What, "I've tried that three times, and each time, it hurt like hell,
> for the following reasons..." isn't a solid argument? If that isn't, I
> don't know what is.

Maybe "solid" wasn't the right word. Abstract, that's how I should
qualify the arguments. I'm NOT insinuating that any of you are too
incompetent to use STL, or that the vendor that supplied the STL lacks
implementation skills, or that the "Linux community" (whoever they
are, be the GCC, libc, libstdc++, kernel, etc) are too stubborn,
whatever. It's just that it leaves a big margin for any possibility. I
don't think that I'm currently able to do better than any of the
people involved in each relevant part (inside and outside ODE); but I
think that dismissing a possible approach for those reasons isn't
enough.

That is, as you are certainly much more experienced developers than
me, I believe you are making the right decision; I fail to see the
proper reasoning, that's why I was being so insistent. But it'll more
likely be a "figure it yourself" matter; I'll get back at this issue
after doing some research and a lot of tests.

Sorry for bothering you so much.

--
Daniel K. O.
_______________________________________________
ODE mailing list
ODE at q12.org
http://q12.org/mailman/listinfo/ode