Thread overview
State of -preview=rvaluerefparam
Jun 25
Tejas
4 days ago
Manu
June 17

Is -preview=rvaluerefparam going to be part of the language or has this course been abandoned?

For all other preview switches, I have the impression that they’re going to be part of the language once either the deprecation period is over/we get editions, or the feature is ironed out. In the following table, Bugfix means the former, Feature means the latter.

Preview Switch Category
dip25 Feature
dip1000 Feature
dip1008 Feature?
dip1021 Feature
bitfields Feature
fieldwise Bugfix
fixAliasThis Bugfix
rvaluerefparam ???
nosharedaccess Bugfix
in Bugfix
inclusiveincontracts Bugfix
fixImmutableConv Bugfix
systemVariables Feature

When conceptualizing or writing DIPs, it is crucial to take the prospected state of the language into account. For rvaluerefparam, it seems this is an abandoned path. It’s neither an extension of something the language can express nor is it a bugfix in the sense that if we could go back in time, we’d make it like this from the beginning.

June 18

On Monday, 17 June 2024 at 13:40:34 UTC, Quirin Schroll wrote:

>

Is -preview=rvaluerefparam going to be part of the language or has this course been abandoned?

I might be misremembering, but I thought is was superseded by -preview=in.

June 18
On 6/18/24 10:18, Nicholas Wilson wrote:
> On Monday, 17 June 2024 at 13:40:34 UTC, Quirin Schroll wrote:
>> Is `-preview=rvaluerefparam` going to be part of the language or has this course been abandoned?
> 
> I might be misremembering, but I thought is was superseded by `-preview=in`.
> 

Well, that cannot supersede it because `in` implies `const`, and `const` is transitive.
June 25

On Monday, 17 June 2024 at 13:40:34 UTC, Quirin Schroll wrote:

>

Is -preview=rvaluerefparam going to be part of the language or has this course been abandoned?

For all other preview switches, I have the impression that they’re going to be part of the language once either the deprecation period is over/we get editions, or the feature is ironed out. In the following table, Bugfix means the former, Feature means the latter.

Preview Switch Category
dip25 Feature
dip1000 Feature
dip1008 Feature?
dip1021 Feature
bitfields Feature
fieldwise Bugfix
fixAliasThis Bugfix
rvaluerefparam ???
nosharedaccess Bugfix
in Bugfix
inclusiveincontracts Bugfix
fixImmutableConv Bugfix
systemVariables Feature

When conceptualizing or writing DIPs, it is crucial to take the prospected state of the language into account. For rvaluerefparam, it seems this is an abandoned path. It’s neither an extension of something the language can express nor is it a bugfix in the sense that if we could go back in time, we’d make it like this from the beginning.

What about Dip 1040?

June 26

On Tuesday, 25 June 2024 at 00:59:34 UTC, Tejas wrote:

>

On Monday, 17 June 2024 at 13:40:34 UTC, Quirin Schroll wrote:

>

When conceptualizing or writing DIPs, it is crucial to take the prospected state of the language into account. For rvaluerefparam, it seems this is an abandoned path. It’s neither an extension of something the language can express nor is it a bugfix in the sense that if we could go back in time, we’d make it like this from the beginning.

What about Dip 1040?

What about it? Its status is Community Review Round 1, it has no implementation, and it does not propose a preview switch. It’s completely unrelated as far as I can tell.

4 days ago
Right, they are unrelated.

FWIW, -preview=rvaluerefparam is just kinda... sitting there. And probably because I stopped aggressively pushing for progress. I used it plenty and it's fine, but I can't release libs or whatever that uses it while it's just a preview.

The question about 1040 though, isn't that DIP already like 3-4 years old? It had community reviews, why is it stalled too? It's something D really needs to not continue to die a slow death... which I fear has become inevitable on the basis of progress failure.


On Thu, 27 Jun 2024 at 09:27, Quirin Schroll via Digitalmars-d < digitalmars-d@puremagic.com> wrote:

> On Tuesday, 25 June 2024 at 00:59:34 UTC, Tejas wrote:
> > On Monday, 17 June 2024 at 13:40:34 UTC, Quirin Schroll wrote:
> >> When conceptualizing or writing DIPs, it is crucial to take the prospected state of the language into account. For `rvaluerefparam`, it seems this is an abandoned path. It’s neither an extension of something the language can express nor is it a bugfix in the sense that if we could go back in time, we’d make it like this from the beginning.
> >
> > What about [Dip
> > 1040](
> https://github.com/dlang/DIPs/blob/a9c553b0dbab1c2983a801b5e89b51c5c33d5180/DIPs/DIP1040.md )?
>
> What about it? Its status is *Community Review Round 1,* it has
> no implementation, and it does not propose a preview switch. It’s
> completely unrelated as far as I can tell.
>