On Mon, 2007-09-24 at 11:55 +0100, Rupert Smith wrote:
> We also knew that the python one could be run from within the maven build
> through jython. I bet nobody considered how much of a known quantity JRuby
> is, wrt doing the same thing. I can understand that people do not want a
> Java dependency on the C++ code, and might therefore object to Velocity
> (although such a dependency really doesn't seem that bad). Python or XSLT
> seems like the best middle ground dependency for Java and C++ to me.
> 
> I think that more effort has been expended on writing code generators, than
> would have been expended writing the generated code by hand.
> 

As the person who actually expended the effort to write the ruby
generator I am  quite happy that it was a huge net saving in effort. I
believe the others who use it agree. Since it has absolutely no impact
on anybody else, I'm unclear on what concrete problem it is causing.

I accept that python might have been a better choice of language. I
chose ruby because I know it better than python and I already had ruby
code to hand that I could use. We have a huge amount of code to write
for 0-10 and not a lot of time, so we needed a solution quickly.
If someone feels it would be productive to port the ruby generator to
python I'll be happy to help and to use/maintain the python generator
thereafter. 

I accept that a single generator for all purposes would be preferable
*if it fits everyones needs*. The Java generator did not fit the needs
of the C++ project. The ruby generator is a better fit. If others want
to adopt the ruby generator I'll be glad to help. If someone comes up
with an even better alternative that is suitable for all projects I'll
be happy to adopt it for C++. 

> On 23/09/2007, John O'Hara <[EMAIL PROTECTED]> wrote:
> >
> > You know this is just an amazingly bad way to work.
> > It looks like almost everyone who has worked on the code generators just
> > threw away the previous one in a fit of Not Invented Here and started
> > again,
> > with no discussion.

Lets be clear: I did not "throw away" anything. I did discuss it with
all of the developers who are affected by it, and I think they all agree
it was a good move. 

The motivation was not "not invented here", it was "not suitable to the
task at hand". Again can I stress that *all of the people actually
affected* by the ruby generator are happy with it. Its not clear to me
why people who are not affected by it are so upset.

Cheers,
Alan.

Reply via email to