Hi Tim, Thanks for those words which I do appreciate. I was obviously not trying to push people to do my bidding, but just to find people who thinks those things do matter.
I want to work on my own version of Ode.jl. I have some solvers that I wrote myself in C++ and that I want to adapt to Julia. That would be a good way to work on an interface that I find clean so people can throw tomatoes at it ;-) I'll have a look at #9493. Thanks On Wednesday, April 29, 2015 at 8:23:07 PM UTC+2, Tim Holy wrote: > > François, perhaps we should emphasize some points we agree on: > 1. There are inconsistencies in julia, both in the language and with the > "guideline" > 2. It's probably fair to say that a majority of people actively working on > julia value code above guidelines > 3. As you've noted from the lack of prompt action on your complaints, > there do > not appear to be hordes of developers who are breathlessly waiting for > some > genius on the mailing list to finally point them towards a project they > can > work on :-). > > This is just life in open source. When you fully accept the implications > of > point 3, you'll come to accept that you don't have a tool for bludgeoning > unpaid developers to do your bidding. But that's what you seem to be > asking > for, or at least that's how it may read to some folks. Volunteers do what > they > do because of their own internal priorities, and it's actively > counterproductive to try to convince them to abandon those in preference > for > your own. After all, _you're_ not willing to put your money where your > mouth > is, so clearly it can't be that important. > > If you do decide to jump in, there are issues like #9493 that are a great > way > to get your feet wet. Or, start renaming functions and writing > deprecations to > improve consistency. _Many_ people have done this before you, including > "new" > users, and so there is no reason why it should be beyond your means. > Finally, > don't be so sure that people wouldn't appreciate efforts to improve > process--- > if you'd followed the growth of julia's testing (first Travis, then > AppVeyor, > then Coveralls, hopefully soon rr), I think you'd be a little less worried > and > more inclined simply to help speed things along. > > Bottom line: time is precious, and responding to long threads on the users > list is a huge waste unless it results in change. Make it happen! > > Best wishes, > --Tim >
