[double-ended] where is it?

classic Classic list List threaded Threaded
5 messages Options
Reply | Threaded
Open this post in threaded view
|

[double-ended] where is it?

Boost - Dev mailing list
I remember going through this review. It was an active, lively and
constructive review. Ever since I've been dying to use this new-fangled
vector. The repo is dead, no dev branch being worked on.

I am happy to implement this container starting from scratch, using the
discussions we've had and the valuable comments as a base. I can give it a
try, but it will (by the time it gets out of the sausage machine) be a
C++20 lib. If that is not acceptable, then no-can-do.

degski

🕔https://mega.nz/C!AlJBkaLI

_______________________________________________
Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost
Reply | Threaded
Open this post in threaded view
|

Re: [double-ended] where is it?

Boost - Dev mailing list
On 11/06/2020 19:32, degski via Boost wrote:
> I remember going through this review. It was an active, lively and
> constructive review. Ever since I've been dying to use this new-fangled
> vector. The repo is dead, no dev branch being worked on.
>
> I am happy to implement this container starting from scratch, using the
> discussions we've had and the valuable comments as a base. I can give it a
> try, but it will (by the time it gets out of the sausage machine) be a
> C++20 lib. If that is not acceptable, then no-can-do.

It was supposed to be integrated in Boost.Container, but the lack of
time from my side and lack of news from the author has stopped it. If
there is interest now, we can try to do a plan. To integrate it in
Boost.Container it has to support all C++ standards, or at least C++11
and I could do the port to C++03.

A feature for deque improvements (selecting the number of contiguous
elements) was incorporated into boost::container::deque:

https://www.boost.org/doc/libs/1_73_0/doc/html/container/configurable_containers.html#container.configurable_containers.configurable_deques

Best,

Ion

_______________________________________________
Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost
Reply | Threaded
Open this post in threaded view
|

Re: [double-ended] where is it?

Boost - Dev mailing list
In reply to this post by Boost - Dev mailing list
On Thu, Jun 11, 2020, 19:33 degski via Boost <[hidden email]> wrote:

> I remember going through this review. It was an active, lively and
> constructive review. Ever since I've been dying to use this new-fangled
> vector. The repo is dead, no dev branch being worked on.
>
> I am happy to implement this container starting from scratch, using the
> discussions we've had and the valuable comments as a base. I can give it a
> try, but it will (by the time it gets out of the sausage machine) be a
> C++20 lib. If that is not acceptable, then no-can-do.
>
> degski
>

The verdict of the review was to merge double_ended into container - a wise
decision. However, it'd require the original c++11 library to be converted
to a c++03 library - I didn't have the taste for that.

Others also argued that the small buffer optimization should be taken care
of by the Allocator, something that would require substantial changes.
Later my attention turned to a different open source library (
http://binlog.org - take a look if you are interested in high performance
structured logging), and didn't have time to finish any of this. Sorry
about that.

Benedek

>

_______________________________________________
Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost
Reply | Threaded
Open this post in threaded view
|

Re: [double-ended] where is it?

Boost - Dev mailing list
On Fri, 12 Jun 2020 at 12:10, Benedek Thaler <[hidden email]>
wrote:

> The verdict of the review was to merge double_ended into container - a
> wise decision. However, it'd require the original c++11 library to be
> converted to a c++03 library - I didn't have the taste for that.
>
>>
I apologize for not having spotted that. You possibly read, what I was
willing to consider to do, yes, a very good reason to drop your tools. In
2020, writing C++14 has become painful in hindsight, i.e. not a very good
invest of time, support C++03, anything other than the latest really.

hand,

degski

_______________________________________________
Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost
Reply | Threaded
Open this post in threaded view
|

Re: [double-ended] where is it?

Boost - Dev mailing list

>> The verdict of the review was to merge double_ended into container - a
>> wise decision. However, it'd require the original c++11 library to be
>> converted to a c++03 library - I didn't have the taste for that.
I'd solve this by declaring Boost.Container to be C++11 and include
Double-Ended in it.
It's almost a decade since its publication so even conservatively taken
it is "old enough" to require it. And library maintainers are free to do
that.
> I apologize for not having spotted that. You possibly read, what I was
> willing to consider to do, yes, a very good reason to drop your tools. In
> 2020, writing C++14 has become painful in hindsight, i.e. not a very good
> invest of time, support C++03, anything other than the latest really.

True. However I'm quite happy with C++14 so far and having to resort to
C++11 then is not that hard unless you need constexpr all they way down
and up.
And IMO Boost should provide extended support for older standards. So
I'd recommend going from C++03 to 11 or 14. The latter might be an issue
with older MSVCs which lack behind in 14 but are quite complete in 11.
But well...




_______________________________________________
Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost

smime.p7s (6K) Download Attachment