On Wed, Apr 26, 2017 at 20:42 Hisham <
h@hisham.hm> wrote:
On 21 April 2017 at 09:04, Alex Silva <asandroq@gmail.com> wrote:
> Hallo,
>
> Regarding the curated set of packages something interesting happened in
> the Haskell world. Haskell packages are stored in Hackage but there is
> no way to know which packages play nicely with the others. So, after
> getting tired of facing failing builds, a private company[1] started
> something called Stackage[2]. The packages in Stackage are grouped in
> releases which are frozen and known to work. They also wrote a tool[3]
> to start and build projects in an easier way based off those releases.
>
> Anyone that wants it bad enough, just like FP Complete did, can start to
> curate sets of Lua packages for each Lua interpreter version, and even
> automate the process of starting new Lua projects based off these sets.
> Something in this vein comes up every few years in the Lua mailing list,
> and yet nothing ever happens. So nobody wants it bad enough.
I think the operative word in this example you gave is "company",
which implies financial resources. People may want it bad enough
(that's why they start these projects), but lack the funding to do it.
Money pays for developer time, which produces work done.
-- Hisham
On a related note...
The current market for Lua, which includes all those that use it and all those that should, does not want a standard library very badly, if at all.
The reason to create a foundation and/or libraries is to attract a group of non-users---people that would choose Lua over other options if conditions were different. Therefore, asking existing users if they want this or looking to them to build it will not produce results. The reason to ask is to gauge minority demand and to figure out a way to do it that will not cause harm.
I'm very grateful for the debate in this thread. This topic has an uneven history of decorum and substance and I think this has helped.
-Andrew
