It's pretty unrealistic to assume that a new language, developed by a small group of developers located around the world, will have the consistency and polish of a commercial language that has been in development for about 30 years with significant resources behind it. It's mind boggling how well the language has turned out considering the relatively tiny resources available. As has been stated, if you come across specific inconsistencies in design, etc, then you should take it upon yourself to either fix it or propose specific solutions. Rome was not built in a day...
On Wednesday, April 29, 2015 at 11:01:32 AM UTC-4, François Fayard wrote: > > On Wednesday, April 29, 2015 at 4:51:00 PM UTC+2, Stefan Karpinski wrote: >> >> Just to clarify my position, I'm all for better and more generic APIs, >> but until we come up with such and have implementations for them, legacy >> names like sprandn aren't hurting anyone – they get the job done and >> they're what many people using sparse matrices are familiar with. If you >> want to improve the state of affairs, the best approach is not to complain >> about it, but to design something better, get some feedback, and implement >> it. >> > > As some people have noticed, I am a new user, and I don't have an overview > of the language. I only have an experience with Mathematica ans the > solution they use for their system which is closed to multiple dispatch. If > Mathematica has such a nice consistency, it's mainly because someone has > enforced a very strict policy (my guess is that Stephen Wolfram is this > guy). > > I can help on designing numerical algorithms which is why I've coded a few > lines for ODEs. That's my background. I don't have experience in designing > guidelines, but my point is that I need them to feel comfortable in > developing in Julia. >
