Re: Fastcode Voting - Conditional Compilation

Hi Pierre

> Agreed, but there is a big difference in the amount of work involved.

Not really. I have done Fastcode functions for 3 years without using a
single compiler directive.

> Changing a conditional define and recompiling as not nearly as much work
as
> creating many variations of the same function (especially while you're
still
> debugging/improving it and you have to maintain all of them.).

Finish your work and then make all the versions before release.

> We want as many people to take part as possible, and restricting the use
of
> conditionals would be just another barrier to entry.

We want to create high quality fast code. I am not willing to sacrifice
quality to get more people contributing. If they are interested in
contributing they have to be willing to work as needed.

>Dennis does not have to
> do all the benchmarking work - there are many people willing to run
> benchmarks if asked (me included).

They do not have the target PC's. They are here. Believe me Pierre I have
tried to have other people running benchmarks for me and this is a bad
solution. I will have to beg in this group for them to be run and they are
never ready when I want to release. On top of this consistency is suffering.
I have just collected donations to be able to build all the needed target
PC's. I have also invested money myself and now we have one good PC for each
target. You ask me to throw all this work away just because of your personal
needs for conditional compilation.

You also ignore the name confusion issue. Function ID is very important.

If people are not willing to make the work needed for making high speed high
quality code then they are free to not participate.

We must do whatever work is needed to reach our quality goals.

Why do you suggest that we should reorganize the entire workings of the
project just because you personally find it difficult to live without
conditional compilation. It seems to me that you have a hard time accepting
our quality standards and what is needed to achieve it. I am 100% certain
that you understand the techniqalities, and it is very hard to me to
understand why you cannot accept to work in a group under strict rules.

I do not want to waste all my time discussing the same issues over and over
again.

Regards
Dennis


.

Relevant Pages

  • Re: Quad Core PC for less than a Mac Mini
    ... are the same quality as Intel. ... Let's keep in mind that 'quality' also includes performance in my ... That means system benchmarks, which Eddie still hasn't ... you state again that Gigabyte is the same level of quality as ...
    (comp.sys.mac.advocacy)
  • Re: Fuzzy Still Pictures
    ... > For best quality when using still images with Movie Maker. ... > 2) Resample each clip so that its width matches the width of your target ...
    (microsoft.public.windowsxp.moviemaker)
  • Re: Fastcode Memory Manager 1.1.1 Summary posted
    ... >> far as I know) the full set of benchmarks has only been run on the ... >> compiler since validations on multiple compilers began. ... >> I argue that this will impact quality and verifiability. ... >> I certainly don't intend to ignore community decisions. ...
    (borland.public.delphi.language.basm)
  • Re: DivX software suggestions?
    ... DivX bundle that I have - in other words there are no user ... configurable options for quality other than setting the target size. ...
    (alt.video.dvd)
  • Dolphin "Special Care" Bathrooms
    ... They target the Elderly and Infirm. ... Their bathrooms are expensive, but ... their Quality, Service and Reputation are very poor. ...
    (uk.misc)