IF (big if) we have a signal/slot framework in C++ we should get that first. Then let third party libraries do the POSIX translation. After a few years we can look at the libraries and decide what/if the correct way to add POSIX to our system is.

POSIX adds a bunch of rules that I don't want in the generic signal handler framework. Though I'm open to be convinced otherwise if those rules really are worth the cost, but worth the cost wouldn't be to support POSIX, it would be [insert something I can't think of here]

-- 
  Henry Miller
  hank_at_[hidden]
On Tue, Jul 27, 2021, at 10:23, Phil Bouchard via Std-Proposals wrote:
> Yeah if we don't abstract POSIX signals, people will come up with a million different messy alternatives. The costs / benefits are there I think and it's trivial to implement anyways.
> 
> On 7/27/21 5:37 AM, Antoine Viallon via Std-Proposals wrote:
>> No, but this is really interesting. 
>> I needed a clean way to handle signals.
>> 
>> And, for reference, new code bases don't have signal handling code. And I really think POSIX signal handling helpers in the STL would make everybody's life easier.
>> 
>> Antoine Viallon
>> 
>> 
>> *De :* Edward Catmur via Std-Proposals <std-proposals_at_[hidden]>
>> *Envoyé :* 26 juillet 2021 22:44:35 GMT+02:00
>> *À :* std-proposals <std-proposals_at_[hidden]>
>> *Cc :* Edward Catmur <ecatmur_at_[hidden]>
>> *Objet :* Re: [std-proposals] Signals & Slots
>> 
>> On Mon, 26 Jul 2021, 21:32 Thiago Macieira via Std-Proposals, <std-proposals_at_[hidden]> wrote:
>>> On Monday, 26 July 2021 12:54:07 PDT Phil Bouchard wrote:
>>> > ... or to create some "process singleton" already emitting these converted
>>> > signals. On 7/26/21 3:51 PM, Phil Bouchard via Std-Proposals wrote:
>>> > One thing that could be done though is to provide a static conversion
>>> > function called explicitly that will convert these POSIX signals into the
>>> > new safer signal types. Just an idea...
>>> 
>>> That is what I call "new use-cases".
>>> 
>>> No, thanks. Whoever needs to deal with POSIX / Unix signals already has the 
>>> code. I don't really care to make their lives easier.
>> 
>> Hopefully everyone is aware of asio signal_set? https://www.boost.org/doc/libs/1_76_0/doc/html/boost_asio/overview/signals.html
>> 
> -- 
> 
> *Phil Bouchard*
> Founder & CTO
> C.: (819) 328-4743
> Fornux Logo <http://www.fornux.com/>
> -- 
> Std-Proposals mailing list
> Std-Proposals_at_[hidden]
> https://lists.isocpp.org/mailman/listinfo.cgi/std-proposals
>