Re: [plt] Push #30060: master branch updated

[Matthias Felleisen]

We do in the style file. -- Matthias



On Sep 14, 2015, at 5:24 PM, stam...@racket-lang.org wrote:

> stamourv has updated `master' from dedde2cd60 to d1ad70b7a0.
> http://git.racket-lang.org/plt/dedde2cd60..d1ad70b7a0
>
> =====[ One Commit ]=====================================================
> Directory summary:
> 100.0% pkgs/racket-doc/scribblings/style/
>
> ~~~~~~~~~~
>
> d1ad70b Vincent St-Amour <stam...@racket-lang.org> 2015-09-14 16:23
> :
> | Remove mention of unstable.
> :
> M pkgs/racket-doc/scribblings/style/todo.scrbl | 2 --
>
> =====[ Overall Diff ]===================================================
>
> pkgs/racket-doc/scribblings/style/todo.scrbl
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> --- OLD/pkgs/racket-doc/scribblings/style/todo.scrbl
> +++ NEW/pkgs/racket-doc/scribblings/style/todo.scrbl
> @@ -14,6 +14,4 @@
>
> @item{Find and link to good/bad examples in the code base.}
>
> -@item{Do we need a discussion of life cycles that start in @tt{unstable}?}
> -
> ]

[Vincent St-Amour]

Do you mean that the style guide refers to unstable? A quick grep didn't
find anything.

Or do you mean that the style guide should keep a discussion of
unstable-related workflows? I would say no, since we're now discouraging
these workflows.

Or are you referring to something else?

Vincent

> --
> You received this message because you are subscribed to the Google Groups "Racket Developers" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to racket-dev+...@googlegroups.com.
> To post to this group, send email to racke...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/racket-dev/5C1469D1-0DCB-424C-8CAB-88CCFF9B61D6%40ccs.neu.edu.
> For more options, visit https://groups.google.com/d/optout.

[Matthias Felleisen]

On Sep 14, 2015, at 6:06 PM, Vincent St-Amour <stam...@eecs.northwestern.edu> wrote:

> Do you mean that the style guide refers to unstable? A quick grep didn't
> find anything.
>
> Or do you mean that the style guide should keep a discussion of
> unstable-related workflows? I would say no, since we're now discouraging
> these workflows.

Yes. If you want to keep unstable, the style guide should mention the life cycle of unstable ideas.

[[ May I ask whether Eli was right about unstable? ]]

[Sam Tobin-Hochstadt]

On Mon, Sep 14, 2015 at 6:08 PM Matthias Felleisen <matt...@ccs.neu.edu> wrote:

On Sep 14, 2015, at 6:06 PM, Vincent St-Amour <stam...@eecs.northwestern.edu> wrote:

> Do you mean that the style guide refers to unstable? A quick grep didn't
> find anything.
>
> Or do you mean that the style guide should keep a discussion of
> unstable-related workflows? I would say no, since we're now discouraging
> these workflows.

Yes. If you want to keep unstable, the style guide should mention the life cycle of unstable ideas.

[[ May I ask whether Eli was right about unstable? ]]

The `unstable` collection served a very useful purpose, but now that we have a package system and a split repository, it's not longer needed in the way that it was. Which is why it's great that Vincent has cleaned it up.

Sam

[Vincent St-Amour]

Precisely.

Many of what used to be unstable subcollects now live on as independent
packages, which makes them more discoverable, and removes the specter of
unstability.

A lot of the rest have been merged with the relevant portions of the
standard library, also making them more discoverable, and less "risky"
to use.

Vincent

[Laurent]

On 15 Sep 2015 12:15 am, "Vincent St-Amour" <stam...@eecs.northwestern.edu> wrote:
>
>
> Many of what used to be unstable subcollects now live on as independent
> packages, which makes them more discoverable, and removes the specter of
> unstability.
>
> A lot of the rest have been merged with the relevant portions of the
> standard library, also making them more discoverable, and less "risky"
> to use.

Although I also think unstable is now not as useful as it was, I would argue against  the "more discoverable" argument: I used to have a look at the unstable docs now and then to look for new interesting bindings but now I wouldn't know where to look.

More general suggestion:
Maybe it would be nice to have, on the pkg webpage or at least maybe through raco, the list of new bindings per package and per collection, with a reference point like 1 week, 1 month and 6 months for example.

[Matthias Felleisen]

I think we should actually list the package content on a page that comes with the docs and make it immediately indexable (even if not installed). When I then run

raco doc fancy-data-structure

I may have a fighting change that it says

available as package FancyDataFooBar from package server.

And I go off and install it. -- Matthias

[Laurent]

That would be amazing, even more so if it comes with a one-line description of its purpose (but that may require a lot more space?).

[Robby Findler]

Is this perhaps a service that the pkgs.racket-lang.org could provide
(via pkg-build.racket-lang.org)?

(I feel like there is a lot we can do with the ring-0 concept that
we're not doing yet.)

Robby

> --
> You received this message because you are subscribed to the Google Groups
> "Racket Developers" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to racket-dev+...@googlegroups.com.
> To post to this group, send email to racke...@googlegroups.com.
> To view this discussion on the web visit

> https://groups.google.com/d/msgid/racket-dev/CABNTSaFSwxR341xP5DMw_oB-C3r_Ab9WXNvaU96EL0PpDGj5tw%40mail.gmail.com.