I haven't posted these to the announce forum for a while, but they still come out each week! If you aren't subscribed yet, there's an rss link on the page or you can follow me on Twitter, where I post most of them: https://twitter.com/adamdruppe


This Week in D has the argument over export/Object.factory, the if(array) thread, dmd codegen, and the official switch to ddmd!

http://arsdnet.net/this-week-in-d/aug-23.html

Next week, we have another interview planned. In previous weeks, I've summarized DConf, written tips, and highlighted interesting projects.

So if you want to keep up with D but don't want to read all the long threads, take a look at these. I don't cover everything and sometimes my bias will influence the summaries... but it still should give you a more digestable look at the community.

On Monday, 24 August 2015 at 16:09:46 UTC, Adam D. Ruppe wrote:
> I haven't posted these to the announce forum for a while, but they still come out each week! If you aren't subscribed yet, there's an rss link on the page or you can follow me on Twitter, where I post most of them: https://twitter.com/adamdruppe
>
> [...]

You have a stray $(P about halfway through the thread discussion section.

On 08/24/2015 06:09 PM, Adam D. Ruppe wrote:
> the if(array) thread

> But, this can be surprising if you aren't used to it, since an empty array is not necessarily null:
> [] is null passes, since the literal avoids allocating for nothing,  but [1][1..$] is null fails,
> despite the array being empty, because the pointer is then set to the end of the input array.

void main(){
    assert([1][1..$] is null);
}

You forgot to take into account constant folding.

On Monday, 24 August 2015 at 16:09:46 UTC, Adam D. Ruppe wrote:
> </snip>
> This Week in D has the argument over export/Object.factory, the if(array) thread, dmd codegen, and the official switch to ddmd!
>
> http://arsdnet.net/this-week-in-d/aug-23.html
> </snip>

Speaking of giving the ldc and gdc recognition. Wouldn't it be neat to have statistics for them to in This Week in D?
IMO it would also be interesting to graph the statistics to be able to read the trends.

On 25 August 2015 at 22:42, NVolcz via Digitalmars-d-announce < digitalmars-d-announce@puremagic.com> wrote:

> On Monday, 24 August 2015 at 16:09:46 UTC, Adam D. Ruppe wrote:
>
>> </snip>
>> This Week in D has the argument over export/Object.factory, the if(array)
>> thread, dmd codegen, and the official switch to ddmd!
>>
>> http://arsdnet.net/this-week-in-d/aug-23.html
>> </snip>
>>
>
> Speaking of giving the ldc and gdc recognition. Wouldn't it be neat to
> have statistics for them to in This Week in D?
> IMO it would also be interesting to graph the statistics to be able to
> read the trends.
>
>
I don't think this would work as well with the less active compilers. Partly because (gdc) only really goes through a major overhaul/change once every six months, depending on how long the next release of DMD has been in development.  Also the whole process is less driven by dealing with bug reports and more driven by feature/optimization topics that I'm sure would fly over most people's heads.

I'm sure no one cares that:
- C++ support has been backported from 2.067.
- Reducing GC.malloc calls from 7861 to 11 in D EH routines.
- Strict LTR evaluation order honouring has been turned off until spec
better defines what should be done  (op= operations may be reordered as
optimizations see fit).
- Updated to be in sync with gcc-6-20150816 snapshot.
- 16 files changed, 1854 insertions, 269 deletions.
- New gdc-4.9.3 and gdc-5.2.0 binary releases are available.

Well, maybe the last one is of interest, but I hope there's a point. :-)

Iain

On Tuesday, 25 August 2015 at 21:14:39 UTC, Iain Buclaw wrote:
> On 25 August 2015 at 22:42, NVolcz via Digitalmars-d-announce < digitalmars-d-announce@puremagic.com> wrote:
>
>>[...]
> I don't think this would work as well with the less active compilers. Partly because (gdc) only really goes through a major overhaul/change once every six months, depending on how long the next release of DMD has been in development.  Also the whole process is less driven by dealing with bug reports and more driven by feature/optimization topics that I'm sure would fly over most people's heads.
>
> [...]

The work done on GDC is well appreciated, GDC's codebase is much cleaner now than it was before the refactoring.

On 25 August 2015 at 23:25, rsw0x via Digitalmars-d-announce < digitalmars-d-announce@puremagic.com> wrote:

> On Tuesday, 25 August 2015 at 21:14:39 UTC, Iain Buclaw wrote:
>
>> On 25 August 2015 at 22:42, NVolcz via Digitalmars-d-announce < digitalmars-d-announce@puremagic.com> wrote:
>>
>> [...]
>>>
>> I don't think this would work as well with the less active compilers. Partly because (gdc) only really goes through a major overhaul/change once every six months, depending on how long the next release of DMD has been in development.  Also the whole process is less driven by dealing with bug reports and more driven by feature/optimization topics that I'm sure would fly over most people's heads.
>>
>> [...]
>>
>
> The work done on GDC is well appreciated, GDC's codebase is much cleaner now than it was before the refactoring.
>

True, and it will only get more cleaner as each section is rewritten.  But no one personally congratulates you on refactoring code (I have been spearheading a push to remove all dmd-backend-isms from gdc.  It took about 3 months work to make expression (toElem) codegen to be stateless, and remove the dmd-specific 'backend IR state' (IRState) struct from the codebase.  And that is barely 1/8 of what needs to be done to prepare the move to 2.067)

http://wiki.dlang.org/GDC/CurrentReleaseTasks

On Tuesday, 25 August 2015 at 21:43:11 UTC, Iain Buclaw wrote:

>> The work done on GDC is well appreciated, GDC's codebase is much cleaner now than it was before the refactoring.
>>
>
> True, and it will only get more cleaner as each section is rewritten.  But no one personally congratulates you on refactoring code (I have been spearheading a push to remove all dmd-backend-isms from gdc.  It took about 3 months work to make expression (toElem) codegen to be stateless, and remove the dmd-specific 'backend IR state' (IRState) struct from the codebase.  And that is barely 1/8 of what needs to be done to prepare the move to 2.067)
>
> http://wiki.dlang.org/GDC/CurrentReleaseTasks

How can we make it easier for people to show their appreciation?  I appreciate very much your work on GDC, and I know that there is a general problem that people tend to focus on what's visible and not necessarily what's hidden but important.

Everyone knows Steve Jobs.  Wozniak less so.  Who outside of the programming world is familiar with Dennis Ritchie?  So part of that is just a question of awareness, and so this is an interesting development (whatever you think of the person or his contributions - that is not my point):
https://www.patreon.com/esr?ty=h

On 26 August 2015 at 00:55, Laeeth Isharc via Digitalmars-d-announce < digitalmars-d-announce@puremagic.com> wrote:

> On Tuesday, 25 August 2015 at 21:43:11 UTC, Iain Buclaw wrote:
>
> The work done on GDC is well appreciated, GDC's codebase is much cleaner
>>> now than it was before the refactoring.
>>>
>>>
>> True, and it will only get more cleaner as each section is rewritten.
>> But no one personally congratulates you on refactoring code (I have been
>> spearheading a push to remove all dmd-backend-isms from gdc.  It took about
>> 3 months work to make expression (toElem) codegen to be stateless, and
>> remove the dmd-specific 'backend IR state' (IRState) struct from the
>> codebase.  And that is barely 1/8 of what needs to be done to prepare the
>> move to 2.067)
>>
>> http://wiki.dlang.org/GDC/CurrentReleaseTasks
>>
>
> How can we make it easier for people to show their appreciation?  I appreciate very much your work on GDC, and I know that there is a general problem that people tend to focus on what's visible and not necessarily what's hidden but important.
>
>
Maybe by coming to Berlin next year? :-)

Actually, perhaps this might be something worth discussing in the D foundation thread?

Iain

On 25/08/2015 23:55, Laeeth Isharc wrote:
> https://www.patreon.com/esr?ty=h
>

Ha, nice one, didn't know about that. I've signed up as Patreon - ESR is ok in my book, he's a FOSS proponent, but not a FSF zealot.


-- 
Bruno Medeiros
https://twitter.com/brunodomedeiros