The forked elephant in the room
Mike Shah
mshah.475 at gmail.com
Tue Jan 16 01:11:16 UTC 2024
On Tuesday, 16 January 2024 at 00:14:50 UTC, GrimMaple wrote:
> On Monday, 15 January 2024 at 23:33:23 UTC, Walter Bright wrote:
>> BTW, has anyone else tried to get changes into another
>> language? I have. It's a years long process, and is extremely
>> difficult to get it through.
>
> Just because it's bad somewhere else doesn't justify leaving
> things as is here
>
>> P.S. We also have some contributors who have a long track
>> record of superb contributions. They also have a consistent
>> commitment to promptly fixing anything that goes wrong with it
>> down the road. Yes, they get a lot less scrutiny. They've
>> earned it.
> Care to give a few names?
>
>> In the spirit of that, I won't be interfering with anybody's
>> forks. I made no attempt to interfere with Tango, Amber, nor
>> any other forks.
>
> Did that work out well for you, or D, in the end? It's very
> interesting how you often use "lack of manpower" as an excuse
> to not get things done, yet you are more than willing to just
> let this "manpower" go without even an iota of interest or any
> attempts to make peace
>
>> I can't help but think a preferred solution already has a lot
>> of buy-in for it, and so needs less justification?
>
> It's not the need of justification that people hate, it's the
> fact that you stop useful contributions based on some
> bike-shedding level issue, whilst allowing yourself to push
> fundamentally flawed features (ImportC for instance). If you
> showed the same level of concern for your own features -- there
> wouldn't be so much backlash. Instead, you can't even be
> bothered to __read__ the string interpolation that works well.
> And, on the other hand, you push in @live (that doesn't work),
> ImportC (that doesn't work), and the list goes on.
>
> I know that you like to say on hackernews that "D has
> ownership" and "D can ImportC", but I'm sorry, it only has
> ownership in an isolated dmd test suite, and it all works
> really poorly in the real world. Same goes for ImportC. If you
> can't compile 99% of C code (I'll give you 95%, fine) it's
> useless in practice.
There's been some proposals I've seen that have taken 10+ years
(std::mdspan) in C++ to get in. Reasons can be for backwards
compatibility, ease of implementation, need in the language
versus simply a library feature, not understanding the feature,
etc. Sometimes it's just a climate change in programming -- so
many more folks leaning into features common in functional
languages now as an example!
I think the C++ committee process is also evolving over time.
Though they have the same (sometimes intense) debates I've
observed here and other communities in regards to features.
I personally think importC is a very important feature for the
language and it's growth/adoption. @live I haven't used much, but
I like that it's there and hope it continues developing.
Getting back to the original topic -- I appreciated reading
Atila's response --thanks Atila! Hoping for the best for all
projects, I'll contribute where I can!
More information about the Digitalmars-d
mailing list