Le dim. 3 févr. 2019 à 22:36, Mikael Djurfeldt <mikael@djurfeldt.com> a
écrit :

> It was a great experience and joy for me to meet some of you at FOSDEM
> 2019. Thank you all!
>
> Now a piece of advice.
>
> Everyone who works with Guile knows that it's crap and look with envy at
> projects like Chez and Racket, right? Jim Blandy thinks that GNU should use
> Python as its scripting language. Chris Webber (probably rightly) thinks
> "look how much you can accomplish with Racket so quickly".
>
> I've been there also. I have to confess that I have now and again regarded
> Guile as crap since perhaps 1995 and there has been multiple occasions
> where I have thought that basing the community effort on some other scheme
> would make much more sense, and I have also always looked with envy on Chez
> and mzscheme/Racket.
>
> Yet, it is *amazing* to me how much progress Guile has made since I left.
> I, for example, *love* the new language and compiler infrastructure.
>
> But note now that Racket looks with envy on Chez and intends to base
> Racket on Chez while Andy Wingo thinks that he can beat Chez performance.
>
> My advice is this:
>
> Idiots go around thinking that their own code is the best thing around.
> Sensible people have a natural, and actually productive, tendency to be
> critical about their own work. That is all good, unless it hurts the sense
> of meaning and joy in your work.
>
> Remember now first that we are all irrational creatures. We maybe *think*
> sometimes that we are rational, because what we think looks rational in our
> view. The problem is that the view is usually very limited, with, for
> example, a limited set of presumptions.
>
> For example: Guile is a piece of software, right? Wrong! It is a plant,
> growing over time. Now, if we look over the fence into the other garden,
> the plants there look much greener. But what will determine the ultimate
> fate is not only the shape of it in the present moment, but also the genes
> it carries, the quality of the soil, the amount of sunlight and the skills
> of its gard[i]ners. We could have quit before we got GOOPS, or before we
> got the native threading, or before the compiler tower, without which there
> would be no chance to beat Chez.
>
> If you look at one combination of some Guile features:
>
> * easy embedding in applications
> * support for multiple languages
> * the compiler tower
> * OO system with polymorphic dispatch and metaobject protocol
> * nice, friendly and open community
>
> I think it is pretty strong and impressive, and I wouldn't like to live
> without it. It's especially important to look at Guile as a good breeding
> ground for new amazing work.
>
> That said, we should steal and collaborate all we can!
>
> All the best,
> Mikael
>

Very well said.

I am waiting an official announcement of the ROADMAP for GNU Guile 3 before
giving my point of view.