Thread overview | ||||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
|
August 03, 2012 2.060 on reddit | ||||
---|---|---|---|---|
| ||||
... with a few stats. Vote up! http://www.reddit.com/r/programming/comments/xm5y0/d_260_released_d_programming_language/ Andrei |
August 03, 2012 Re: 2.060 on reddit | ||||
---|---|---|---|---|
| ||||
Posted in reply to Andrei Alexandrescu | On Fri, 03 Aug 2012 11:31:27 -0400 Andrei Alexandrescu <SeeWebsiteForEmail@erdani.org> wrote: > ... with a few stats. Vote up! > > http://www.reddit.com/r/programming/comments/xm5y0/d_260_released_d_programming_language/ > > > Andrei Wow, that's much more positive towards D than Reddit used to be. Used to be much more mixed. |
August 03, 2012 Re: 2.060 on reddit | ||||
---|---|---|---|---|
| ||||
Posted in reply to Andrei Alexandrescu | When are allocators going to be ready? On Fri, Aug 3, 2012 at 10:31 AM, Andrei Alexandrescu <SeeWebsiteForEmail@erdani.org> wrote: > ... with a few stats. Vote up! > > http://www.reddit.com/r/programming/comments/xm5y0/d_260_released_d_programming_language/ > > > Andrei |
August 03, 2012 Re: 2.060 on reddit | ||||
---|---|---|---|---|
| ||||
Posted in reply to Caligo | Caligo: > When are allocators going to be ready? Direct experience shows me that once things are in Phobos, it's not easy to fix their interface/API. Andrei fears of breaking changes, so even small API improvements of Phobos stuff written last year are refused. See as example: http://d.puremagic.com/issues/show_bug.cgi?id=8467 So Phobos stuff once created is almost set in stone. This means that in practice the Phobos APIs should be perfect on the first try. So better to not rush things, and do things right. Bye, bearophile |
August 03, 2012 Re: 2.060 on reddit | ||||
---|---|---|---|---|
| ||||
Posted in reply to bearophile | On Friday, 3 August 2012 at 19:44:22 UTC, bearophile wrote: > Caligo: > >> When are allocators going to be ready? > > Direct experience shows me that once things are in Phobos, it's not easy to fix their interface/API. Andrei fears of breaking changes, so even small API improvements of Phobos stuff written last year are refused. See as example: > > http://d.puremagic.com/issues/show_bug.cgi?id=8467 > > So Phobos stuff once created is almost set in stone. This means that in practice the Phobos APIs should be perfect on the first try. So better to not rush things, and do things right. > > Bye, > bearophile "When in doubt, leave it out." says Joshua Bloch http://www.youtube.com/watch?v=aAb7hSCtvGw |
August 03, 2012 Re: 2.060 on reddit | ||||
---|---|---|---|---|
| ||||
Posted in reply to bearophile | On 2012-08-03 21:44, bearophile wrote: > Direct experience shows me that once things are in Phobos, it's not easy > to fix their interface/API. Andrei fears of breaking changes, so even > small API improvements of Phobos stuff written last year are refused. > See as example: One option would be to create a new set of containers, in a new module. These can basically be copies of the original containers with their API's adapted to support custom allocators. It's not pretty but it won't break any code. -- /Jacob Carlborg |
August 04, 2012 Re: 2.060 on reddit | ||||
---|---|---|---|---|
| ||||
Posted in reply to bearophile | bearophile <bearophileHUGS@lycos.com> wrote:
> Caligo:
>> When are allocators going to be ready?
>
> Direct experience shows me that once things are in Phobos, it's not easy to fix their interface/API. Andrei fears of breaking
Go's solution are experimental packages. You can always use them, but you know that they will change and at some time will be in antother namespace/directory.
|
August 04, 2012 Re: 2.060 on reddit | ||||
---|---|---|---|---|
| ||||
Posted in reply to Stefan Scholl | On Saturday, August 04, 2012 02:37:07 Stefan Scholl wrote:
> bearophile <bearophileHUGS@lycos.com> wrote:
> > Caligo:
> >> When are allocators going to be ready?
> >
> > Direct experience shows me that once things are in Phobos, it's not easy to fix their interface/API. Andrei fears of breaking
>
> Go's solution are experimental packages. You can always use them, but you know that they will change and at some time will be in antother namespace/directory.
|
August 04, 2012 Re: 2.060 on reddit | ||||
---|---|---|---|---|
| ||||
Posted in reply to Stefan Scholl | On Saturday, August 04, 2012 02:37:07 Stefan Scholl wrote:
> bearophile <bearophileHUGS@lycos.com> wrote:
> > Caligo:
> >> When are allocators going to be ready?
> >
> > Direct experience shows me that once things are in Phobos, it's not easy to fix their interface/API. Andrei fears of breaking
>
> Go's solution are experimental packages. You can always use them, but you know that they will change and at some time will be in antother namespace/directory.
It's been discussed a time or two that we should have an incubator project for Phobos where potential Phobos modules go to be used and ironed out before actually being reviewed for inclusion in Phobos. But it's never materialized. Someone(s) would have to organize it and manage it, and no one has done so.
- Jonathan M Davis
|
August 04, 2012 Re: 2.060 on reddit | ||||
---|---|---|---|---|
| ||||
Posted in reply to Jonathan M Davis | On 04-08-2012 04:42, Jonathan M Davis wrote: > On Saturday, August 04, 2012 02:37:07 Stefan Scholl wrote: >> bearophile <bearophileHUGS@lycos.com> wrote: >>> Caligo: >>>> When are allocators going to be ready? >>> >>> Direct experience shows me that once things are in Phobos, it's >>> not easy to fix their interface/API. Andrei fears of breaking >> >> Go's solution are experimental packages. You can always use them, >> but you know that they will change and at some time will be in >> antother namespace/directory. > > It's been discussed a time or two that we should have an incubator project for > Phobos where potential Phobos modules go to be used and ironed out before > actually being reviewed for inclusion in Phobos. But it's never materialized. > Someone(s) would have to organize it and manage it, and no one has done so. > > - Jonathan M Davis > I thought this was the idea of etc.* all along... If nobody's against it, we should definitely get the ball rolling. -- Alex Rønne Petersen alex@lycus.org http://lycus.org |
Copyright © 1999-2021 by the D Language Foundation