Re: Setting a list of values

[Brian Schott via Digitalmars-d-learn]

On Monday, 2 May 2016 at 10:15:04 UTC, Marc Schütz wrote:

This check can be done purely by looking at the tokens.


In other words it's trivial for D-Scanner to warn about this.

https://github.com/Hackerpilot/Dscanner/issues/341

Re: It's alive! D building D building D, all on Android

[Rory McGuire via Digitalmars-d-announce]

On 08 May 2016 02:21, "Joakim via Digitalmars-d-announce" <
digitalmars-d-announce@puremagic.com> wrote:
>
> On Saturday, 7 May 2016 at 09:58:14 UTC, Johan Engelen wrote:
>>
>> Fantastic news!
>>
>> I hope we can find a good way to integrate automated testing of github
branches/PRs for Android.
>
>
> There is beta support for Android on Travis, perhaps it can be used to
run the druntime and phobos tests in an emulator.
https://aws.amazon.com/device-farm/

Re: So, About That Official Blog...

[Jack Stouffer via Digitalmars-d]

On Thursday, 5 May 2016 at 18:22:50 UTC, Jack Stouffer wrote:
Please let me know what you think! I believe this is a good 
first step in the right direction.


In an attempt to get more interest in this, I have sent several 
invites to people in the core team who have blogs or have written 
articles in the past. If you're interested in the concept of a 
dlang blog, I would appreciate it if you give me some feedback 
after you've looked at the interface a bit.

Re: Rewriting zlib in Phobos

[rikki cattermole via Digitalmars-d]

On 08/05/2016 3:08 PM, Walter Bright wrote:

On 5/7/16 3:33 PM, Andrei Alexandrescu wrote:

Another library discussed as even more ripe for replacement is zlib:
* We compile the C source code, which is extra aggravation to the

build scripts

* The D wrappers are ancient and clunky - we should have beautiful,

efficient > range streaming

* I think the license allows us to translate the source code to D, so

we have > a viable path of replacement

* Just got word (thx Rikki) the D wrappers use GC in a laissez-faire

manner, > so no way to use the library tightly


So if anyone wants to look into this, it would be a good project.


Before anyone gets too deep into this, the zip file format is
functionally the same as the various .lib and .a formats read and
written by dmd, and there are many other such formats.

Any redesign should be a package based one, such as:

std.archiver.zip
std.archiver.libomf
std.archiver.libmscoff
std.archiver.libmach
std.archiver.libelf

(those are just the ones we have code for)

and they should use a common interface.


Assuming it uses allocators, I'd be happy.

Although I like the idea of replacing what is in dmd with this.
But perhaps for the binary formats derivation that they should be fully 
import/exportable with headers and all? It would step 1 of making a linker.

Re: The end of curl (in phobos)

[Walter Bright via Digitalmars-d]

On 5/7/2016 6:33 AM, Andrei Alexandrescu wrote:

Another library discussed as even more ripe for replacement is zlib:


Started a new thread for this.

Rewriting zlib in Phobos

[Walter Bright via Digitalmars-d]

On 5/7/16 3:33 PM, Andrei Alexandrescu wrote:
> Another library discussed as even more ripe for replacement is zlib:
> * We compile the C source code, which is extra aggravation to the build 
scripts
> * The D wrappers are ancient and clunky - we should have beautiful, efficient 
> range streaming
> * I think the license allows us to translate the source code to D, so we have 
> a viable path of replacement
> * Just got word (thx Rikki) the D wrappers use GC in a laissez-faire manner, 
> so no way to use the library tightly

>
> So if anyone wants to look into this, it would be a good project.

Before anyone gets too deep into this, the zip file format is functionally the 
same as the various .lib and .a formats read and written by dmd, and there are 
many other such formats.


Any redesign should be a package based one, such as:

std.archiver.zip
std.archiver.libomf
std.archiver.libmscoff
std.archiver.libmach
std.archiver.libelf

(those are just the ones we have code for)

and they should use a common interface.

Re: The end of curl (in phobos)

[Adam D. Ruppe via Digitalmars-d]

On Saturday, 7 May 2016 at 20:50:53 UTC, Jonas Drewsen wrote:
But std.net.curl supports not just HTTP but also FTP etc. so i 
guess that won't suffice.


We can always implement ftp too, it isn't that complicated of a 
protocol.


Though, I suspect its users are a tiny minority and they might 
not mind depending on a separate curl package.

Quotes from DConf 2016

[Joakim via Digitalmars-d]

I've gone through about seven sessions in the archived livestream 
videos and here are some quotes that stood out for me.  Share 
your memorable quotes, for those deciding what to watch.


"Normally, I'd have somebody in the company say, 'OK, what can I 
do in D that I can't do in C++?'


Every time, I'd tell them, 'You can save time.'

That's the big thing with it, and a lot of the code you can just 
set up and it just works behind the scenes and requires no 
programmer maintenance _at all_ and no programmer boilerplate 
code is just insane compared to C++."
Ethan Watson at 1:04:30 mark in 
http://www.ustream.tv/recorded/86359657


"CTFE is really a hack.  You can see that it's a hack, it's 
implemented as a hack.  It is the most useful hack that I've ever 
seen, and it is definitely a hacker's tool to do stuff that are 
like magic.  But to be fast, it would need to be heavily 
redesigned, reimplemented, possibly executed in multiple threads, 
because it is used for stuff that we could never have envisioned 
when it was invented.


Like, it was more like this accidental thing, 'Yeah, you have 
this mixin and you can do stuff with strings' and this feature, 
CTFE, just turned out to be so useful and so heavily used, which 
was probably never anticipated.  So what should be made better 
about CTFE is _all of it_.  We need to reimplement it properly 
and then it will be really great."
Stefan Koch at 2:13:35 mark in 
http://www.ustream.tv/recorded/86364066

ggplotd - curve colour

[brocolis via Digitalmars-d-learn]

How do I set the color of a curve with ggplotd?
Thanks.

[Issue 16002] New: Add __traits(isModule) and __traits(isPackage)

[via Digitalmars-d-bugs]

https://issues.dlang.org/show_bug.cgi?id=16002

  Issue ID: 16002
   Summary: Add __traits(isModule) and __traits(isPackage)
   Product: D
   Version: D2
  Hardware: All
OS: All
Status: NEW
  Severity: enhancement
  Priority: P1
 Component: dmd
  Assignee: nob...@puremagic.com
  Reporter: blm...@gmail.com

While working on a set of bindings to a C++ library, I found that it would be
useful to have __traits to determine whether a symbol represents a module
and/or package. I've implemented a solution that appears to work; see PR #5290.

https://github.com/dlang/dmd/pull/5290

--

Re: So, About That Official Blog...

[Nick Sabalausky via Digitalmars-d]

On 05/05/2016 05:25 PM, Jacques Müller wrote:

On Thursday, 5 May 2016 at 18:22:50 UTC, Jack Stouffer wrote:

blog: http://officaldlang.tumblr.com/


That's nice! But it seems like you misspelled the blog title.



It's a feature, it keeps with web 2.0 tradition! :)

Re: [OT] Re: How are you enjoying DConf? And where to go next?

[Timon Gehr via Digitalmars-d]

On 07.05.2016 23:49, deadalnix wrote:

On Saturday, 7 May 2016 at 16:50:10 UTC, Russel Winder wrote:

On Sat, 2016-05-07 at 15:17 +, Adam D. Ruppe via Digitalmars-d wrote:



[…]

Besides, zero is just as arbitrary as thirty-two (which, btw, is a
power of two*) and easy to remember if you use it anyway! I

[…]

0 is a power of two as well.



It is 2 power how many ?



void main(){
  assert(2^^32==0);
  assert(2^^-double.infinity==0);
  assert(2^^(new class{int opBinaryRight(string op)(int x){ return 0; 
}})==0);

}

Re: [OT] Re: How are you enjoying DConf? And where to go next?

[Andrei Alexandrescu via Digitalmars-d]

On 5/7/16 6:15 PM, Adam D. Ruppe wrote:

When people complain about the D forums, they aren't complaining about
infrequent OT stuff.


Most related complaints I got at DConf were about off-topic posts. -- Andrei

Re: How are you enjoying DConf? And where to go next?

[dewitt via Digitalmars-d]

On Friday, 6 May 2016 at 14:13:35 UTC, Andrei Alexandrescu wrote:
The atmosphere here is great, and I'm curious how it feels for 
those who are watching remotely. Is the experience good? What 
can we do better?


Also: we're talking about the DConf 2017 location. Please share 
any initial thoughts!



Thanks,

Andrei


Pretty sure everyone will agree that the next logical location 
would be Phoenix.  Am I right?

[Issue 16001] Lambda syntax: forbid use with FunctionLiteralBody: (x) => {assert(x);}

[via Digitalmars-d-bugs]

https://issues.dlang.org/show_bug.cgi?id=16001

greenify  changed:

   What|Removed |Added

 CC||greeen...@gmail.com

--

[Issue 15989] Win32 optimizer bug

[via Digitalmars-d-bugs]

https://issues.dlang.org/show_bug.cgi?id=15989

Jack Stouffer  changed:

   What|Removed |Added

 CC||j...@jackstouffer.com
   Severity|major   |blocker

--- Comment #1 from Jack Stouffer  ---
Raising to blocker as this is blocking
https://github.com/dlang/phobos/pull/4286

--

Re: [OT] Re: How are you enjoying DConf? And where to go next?

[deadalnix via Digitalmars-d]

On Saturday, 7 May 2016 at 19:11:50 UTC, Mathias Lang wrote:
In any case it should be accessible. As nice as the Salt Lake 
City location was, it was not easily accessible for everyone. 
When looking at the flight price (in November or December), it 
was twice as expensive to fly to SLL (from Berlin, using Air 
France) than to fly to LA / SF. So I'd favor New York over 
Boston for example.




Fly tips: Buy about 6 week in advance, this is when prices are 
the lowest. Buy outside of business hours.

Re: [OT] Re: How are you enjoying DConf? And where to go next?

[deadalnix via Digitalmars-d]

On Saturday, 7 May 2016 at 16:50:10 UTC, Russel Winder wrote:
On Sat, 2016-05-07 at 15:17 +, Adam D. Ruppe via 
Digitalmars-d wrote:



[…]
Besides, zero is just as arbitrary as thirty-two (which, btw, 
is a power of two*) and easy to remember if you use it anyway! 
I

[…]

0 is a power of two as well.



It is 2 power how many ?

Re: [OT] Re: How are you enjoying DConf? And where to go next?

[deadalnix via Digitalmars-d]

On Saturday, 7 May 2016 at 15:15:23 UTC, Adam D. Ruppe wrote:
On Saturday, 7 May 2016 at 13:10:37 UTC, Andrei Alexandrescu 
wrote:
Please curb off-topic discussions and mark them with [OT], 
yadda yadda yadda. Thanks! -- Andrei


When people complain about the D forums, they aren't 
complaining about infrequent OT stuff. Random tangents about 
the weather naturally fizzle out after five or six posts and 
don't leave hard feelings.




You already made more than 5 or 6 OT post yourself.

Re: The end of curl (in phobos)

[yawniek via Digitalmars-d]

On Saturday, 7 May 2016 at 09:46:36 UTC, ikod wrote:

What do you mean under re-licensing?


you added GPL to requests, this is incompatible with phobos and
your code can not be included. could it be changed to boost 
licence?

Re: The end of curl (in phobos)

[Jonas Drewsen via Digitalmars-d]

On Friday, 6 May 2016 at 13:25:55 UTC, Adam D. Ruppe wrote:

On Friday, 6 May 2016 at 13:12:31 UTC, Jonathan M Davis wrote:

[...]



I have an HTTP client lib, I'm pretty sure Vladimir does too, 
and I think vibe wrote one for their framework. I'm sure we're 
not the only ones. I also have an XML lib, and again, I'm 
pretty sure Vladimir does too...


[...]


But std.net.curl supports not just HTTP but also FTP etc. so i 
guess that won't suffice.

Re: The end of curl (in phobos)

[Suliman via Digitalmars-d]

https://github.com/ikod/dlang-requests


+1 for dlang-requests. Thanks for it! I already use it's in my 
project, and it's very easy to use.

Re: The end of curl (in phobos)

[Steven Schveighoffer via Digitalmars-d]

On 5/7/16 3:33 PM, Andrei Alexandrescu wrote:

On 5/6/16 11:32 AM, Robert burner Schadek wrote:

As discussed yesterday at DConf, curl in phobos must go.

The plan is as follows.

1. undocument everything curl related in may 2016
2. deprecate everything curl related in may 2017
3. delete everything curl related in may 2018
3.1 move curl stuff to undead

PR: https://github.com/dlang/phobos/pull/4283


Another library discussed as even more ripe for replacement is zlib:

* We compile the C source code, which is extra aggravation to the build
scripts

* The D wrappers are ancient and clunky - we should have beautiful,
efficient range streaming

* I think the license allows us to translate the source code to D, so we
have a viable path of replacement

* Just got word (thx Rikki) the D wrappers use GC in a laissez-faire
manner, so no way to use the library tightly

So if anyone wants to look into this, it would be a good project.


I agree, I could not use the d wrappers for zlib when I made iopipe.

I think Stefan Koch has created some zlib programs that are fully CTFE 
and D.


-Steve

Re: [OT] Re: How are you enjoying DConf? And where to go next?

[Mathias Lang via Digitalmars-d]

2016-05-07 17:15 GMT+02:00 Adam D. Ruppe via Digitalmars-d <
digitalmars-d@puremagic.com>:

> On Saturday, 7 May 2016 at 13:10:37 UTC, Andrei Alexandrescu wrote:
>
>> Please curb off-topic discussions and mark them with [OT], yadda yadda
>> yadda. Thanks! -- Andrei
>>
>
> When people complain about the D forums, they aren't complaining about
> infrequent OT stuff. Random tangents about the weather naturally fizzle out
> after five or six posts and don't leave hard feelings.
>
>
That's your POV. And in my POV, the forums are useless for working on D
because of the amount of digressions.
I'm not opposed to a section to discuss whatever off-topic subject is
trendy, but I wouldn't read it. When I  a thread about Fibers, lambda or
templates, I don't expect nor am looking for a comparison of metric /
imperial system, why esperanto failed to take off or the history or the
history of LISP in the 80's.

That being said, I would love the next DConf to be somewhere in Europe,
potentially Berlin (where I live).
I believe Europe is where most of the community is anyway. The question is
do we want the conference to be at a fixed location so planning is easier
and success is more predictable or do we want to favour diversity and move
it around ?

In any case it should be accessible. As nice as the Salt Lake City location
was, it was not easily accessible for everyone. When looking at the flight
price (in November or December), it was twice as expensive to fly to SLL
(from Berlin, using Air France) than to fly to LA / SF. So I'd favor New
York over Boston for example.

Also, people travelling far (and unsponsored) are likely to take the
occasion to spend a couple of days around. For the 2015 edition I flew in
the week-end before, and finding an hotel room was more challenging than
expected since it was memorial day.

Re: How are you enjoying DConf? And where to go next?

[Johan Engelen via Digitalmars-d]

On Friday, 6 May 2016 at 14:13:35 UTC, Andrei Alexandrescu wrote:


Also: we're talking about the DConf 2017 location. Please share 
any initial thoughts!


This was my first DConf, and it was great to meet people that I 
had only seen online (sometimes only by handle). On dconf.org I 
read: "DConf is the main face-to-face event for everyone and 
everything related to the D language and environment." So that 
kind of aligns with what I think its goal should be. Imvho, the 
goal is not to market D to people who don't know about D / are 
not (planning to be) involved already.
With that in mind, I think it'd be good to select a location 
based on where folks in the D community will be able/willing to 
travel to.


(for me personally, northern Europe is best. Japan I would try to 
join too ;-))

[Issue 15939] GC.collect causes deadlock in multi-threaded environment

[via Digitalmars-d-bugs]

https://issues.dlang.org/show_bug.cgi?id=15939

Илья Ярошенко  changed:

   What|Removed |Added

   Assignee|nob...@puremagic.com|ilyayaroshe...@gmail.com

--

Re: Reproducible builds of D compilers

[Johan Engelen via Digitalmars-d]

On Saturday, 7 May 2016 at 16:22:34 UTC, Vladimir Panteleev wrote:


https://blog.thecybershadow.net/2015/05/05/is-d-slim-yet/


Thanks for repeating the link to that blog article. I was 
reminded of it at DConf. Would be great if results from GDC and 
LDC could be added to the graphs, plus more tests!

Re: [OT] Re: How are you enjoying DConf? And where to go next?

[Dmitry Olshansky via Digitalmars-d]

On 07-May-2016 19:50, Russel Winder via Digitalmars-d wrote:

On Sat, 2016-05-07 at 15:17 +, Adam D. Ruppe via Digitalmars-d
wrote:



[…]

Besides, zero is just as arbitrary as thirty-two (which, btw, is
a power of two*) and easy to remember if you use it anyway! I

[…]

0 is a power of two as well.


0 is not a power of two. Any power of non-negative number is > 0.




* I really think that's the main difference between American
units and the others - base two vs base ten. Base two is superior
in basically every way, but base ten is more newbie friendly for
doing irrelevant conversions.


They are not American units, they are Imperial units. Well except that
Americans give the same name to different values of the dimension.




--
Dmitry Olshansky

[Issue 15810] Store historical docs

[via Digitalmars-d-bugs]

https://issues.dlang.org/show_bug.cgi?id=15810

Andrej Mitrovic  changed:

   What|Removed |Added

 CC||andrej.mitrov...@gmail.com

--- Comment #1 from Andrej Mitrovic  ---
That's not a bad idea. Other than the compiler, a project itself could be
several versions behind using the reference compiler (or even a whole company
for example).

--

Re: [OT] Re: How are you enjoying DConf? And where to go next?

[Russel Winder via Digitalmars-d]

On Sat, 2016-05-07 at 15:17 +, Adam D. Ruppe via Digitalmars-d
wrote:
> 
[…]
> Besides, zero is just as arbitrary as thirty-two (which, btw, is 
> a power of two*) and easy to remember if you use it anyway! I 
[…]

0 is a power of two as well.

> * I really think that's the main difference between American 
> units and the others - base two vs base ten. Base two is superior 
> in basically every way, but base ten is more newbie friendly for 
> doing irrelevant conversions.

They are not American units, they are Imperial units. Well except that
Americans give the same name to different values of the dimension.

-- 
Russel.
=
Dr Russel Winder  t: +44 20 7585 2200   voip: sip:russel.win...@ekiga.net
41 Buckmaster Roadm: +44 7770 465 077   xmpp: rus...@winder.org.uk
London SW11 1EN, UK   w: www.russel.org.uk  skype: russel_winder

signature.asc
Description: This is a digitally signed message part

[Issue 14984] Make it illegal (or at least a warning) to modify the iteration variable in foreach

[via Digitalmars-d-bugs]

https://issues.dlang.org/show_bug.cgi?id=14984

greensunn...@gmail.com changed:

   What|Removed |Added

 CC||greensunn...@gmail.com

--

[Issue 16000] [Reg 2.072~master] linking issues on OSX with size_t, extern(C++), i386, and clang

[via Digitalmars-d-bugs]

https://issues.dlang.org/show_bug.cgi?id=16000

Martin Nowak  changed:

   What|Removed |Added

Summary|linking issues on OSX with  |[Reg 2.072~master] linking
   |size_t, extern(C++), i386,  |issues on OSX with size_t,
   |and clang   |extern(C++), i386, and
   ||clang

--

Re: C#7 features

[Nick Treleaven via Digitalmars-d-announce]

On Friday, 6 May 2016 at 23:51:59 UTC, Steven Schveighoffer wrote:


Of COURSE D supports local ref variables:

struct RefVar(T)
{
  private T * var;
  this(ref T v) { var =  }
  auto get() { return *var; }
  alias this get;
}


ref get() return {...

Which is unsafe even if the ctor is marked trusted, the struct 
could be moved to a higher scope. But like in another thread, you 
can have a ref property in @safe code:


T v;
@property ref myRef(){return v;}

So why are ref locals disallowed? Now we have the return 
attribute on functions that take myRef by ref, can it still 
escape?


ref myRef = v;

Re: Setting a list of values

[sigod via Digitalmars-d-learn]

On Monday, 2 May 2016 at 23:41:39 UTC, Steven Schveighoffer wrote:

On 5/2/16 6:00 PM, sigod wrote:

On Monday, 2 May 2016 at 10:15:04 UTC, Marc Schütz wrote:

On Monday, 2 May 2016 at 08:46:31 UTC, Ali Çehreli wrote:

[...]


Warning (better: disallowing altogether) about `=>` directly 
followed
by `{` should be enough to cover all cases. To express that 
you really
want a lambda returning a lambda, it can be rewritten either 
as:


(x) => () { assert(x); }

or as:

(x) => ({ assert(x); })

This check can be done purely by looking at the tokens. 
Should we
someday introduce tuples with `{}`, the check needs to be 
done after
the node starting with `{` has been parsed to distinguish 
between

delegate and tuple literals.


It's good idea. I myself stumbled into this before.


Agree.

-Steve


https://issues.dlang.org/show_bug.cgi?id=16001

[Issue 16001] New: Lambda syntax: forbid use with FunctionLiteralBody: (x) => {assert(x);}

[via Digitalmars-d-bugs]

https://issues.dlang.org/show_bug.cgi?id=16001

  Issue ID: 16001
   Summary: Lambda syntax: forbid use with FunctionLiteralBody:
(x) => {assert(x);}
   Product: D
   Version: D2
  Hardware: All
OS: All
Status: NEW
  Severity: enhancement
  Priority: P1
 Component: dmd
  Assignee: nob...@puremagic.com
  Reporter: sigod.m...@gmail.com

As discussed here:
https://forum.dlang.org/post/thgyqyarccinzuqhc...@forum.dlang.org

It's very common for people from other languages to write something like

list.each!(x => { /* use x */ });

Which, of course, doesn't work as expected.

--

Re: Reproducible builds of D compilers

[Vladimir Panteleev via Digitalmars-d]

On Friday, 6 May 2016 at 08:07:30 UTC, Pjotr Prins wrote:
At Dconf we had a discussion on creating reproducible builds of 
the D compilers. One thing that is required is bootstrapping 
the build. In GNU Guix we start from a working C compiler which 
is a reasonable starting point (it could have been LISP instead 
;)


For D this means that even if all components of the compilers 
are written in D we should have a (chain of) compilation steps 
that can be initiated from a C or C++ compiler. This is still 
the case today, but I want to highlight this point here so we 
do not lose that facility.


GNU Guix can be used as a reproducible reference build system 
where we capture the bootstrapping process nicely and test the 
builds on the build farm(s) whenever something changes.


Hi,

Digger can already bootstrap D from the last C++ compiler.

It also shares a lot of the same design ideas as Guix (storing 
build results as a hash of code and dependencies). It uses git as 
the store, which allows e.g. storing every buildable commit of D 
ever in about 5 GB.


On Windows, D needs proprietary software (Visual Studio) to 
build. I'm not sure whether that's something possible with Guix. 
What Digger does is that it downloads only the Visual Studio 
components that are necessary, verifies their integrity, and 
unpacks them into a directory (without requiring administrator 
rights) and uses it from there.


https://github.com/CyberShadow/Digger
https://blog.thecybershadow.net/2015/05/05/is-d-slim-yet/

Re: Setting a list of values

[sigod via Digitalmars-d-learn]

On Wednesday, 4 May 2016 at 04:56:54 UTC, Joel wrote:

On Sunday, 1 May 2016 at 05:42:00 UTC, Ali Çehreli wrote:

[...]


This seems to work the best:

arr.each!(a => { writeln(a); }());


And the ugliest. And probably slowest.

[Issue 15966] [REG 2.071] {public,protected} imports in base class ignored on symbol lookup

[via Digitalmars-d-bugs]

https://issues.dlang.org/show_bug.cgi?id=15966

--- Comment #2 from Martin Nowak  ---
We definitely shouldn't break code, but I also don't see a reason why we should
ignore public imports in base classes. Maybe we can find the prior forum
discussion that Walter mentioned.

--

[Issue 16000] New: linking issues on OSX with size_t, extern(C++), i386, and clang

[via Digitalmars-d-bugs]

https://issues.dlang.org/show_bug.cgi?id=16000

  Issue ID: 16000
   Summary: linking issues on OSX with size_t, extern(C++), i386,
and clang
   Product: D
   Version: D2
  Hardware: x86
OS: Mac OS X
Status: NEW
  Severity: regression
  Priority: P1
 Component: dmd
  Assignee: nob...@puremagic.com
  Reporter: c...@dawg.eu

Undefined symbols for architecture i386:
  "DtBuilder::repeat(dt_t*, unsigned int)", referenced from:
  Initializer_toDt::InitToDt::visit(ArrayInitializer*) in dmd.o
 
_D4ddmd4todt11membersToDtFC4ddmd9aggregate20AggregateDeclarationC4ddmd7backend9DtBuilderPS4ddmd4root5array41__T5ArrayTC4ddmd10expression10ExpressionZ5ArraykC4ddmd6dclass16ClassDeclarationPPPS4ddmd6dclass9BaseClassZv
in dmd.o
  Type_toDt::TypeToDt::visit(TypeVector*) in dmd.o
  Type_toDt::TypeToDt::visit(TypeSArray*) in dmd.o
 
_D4ddmd4todt8toDtElemFC4ddmd5mtype10TypeSArrayC4ddmd7backend9DtBuilderC4ddmd10expression10ExpressionZv
in dmd.o
  "OutBuffer::write(void const*, unsigned long)", referenced from:
  toObjFile(Dsymbol*, bool)::ToObjFile::visit(VarDeclaration*) in
glue.a(toobj.o)
ld: symbol(s) not found for architecture i386


Both seem to happen b/c of a size_t vs. unsigned int/long mangling mismatch.
This is breaking the nightly build since 2 days.

--

[Issue 15999] New: Inline assembly incorrect sign extension instead of error

[via Digitalmars-d-bugs]

https://issues.dlang.org/show_bug.cgi?id=15999

  Issue ID: 15999
   Summary: Inline assembly incorrect sign extension instead of
error
   Product: D
   Version: D2
  Hardware: x86_64
OS: All
Status: NEW
  Keywords: accepts-invalid, iasm, wrong-code
  Severity: normal
  Priority: P1
 Component: dmd
  Assignee: nob...@puremagic.com
  Reporter: safety0ff.b...@gmail.com

Taken from issue 6459:
> I also notice if you do:
>
> and RAX,0x
>
> the compiler accepts it but generates:
>
> and RAX,0x
?
> because, of course, there is no imm64 mode for and/or etc. instructions.
> Same applies for all other immediate constants that can be encoded into 8 
> bits.

I have a fix incomming.

--

[Issue 15999] Inline assembly incorrect sign extension instead of error

[via Digitalmars-d-bugs]

https://issues.dlang.org/show_bug.cgi?id=15999

safety0ff.bugz  changed:

   What|Removed |Added

   Assignee|nob...@puremagic.com|safety0ff.b...@gmail.com

--

Re: [OT] Re: How are you enjoying DConf? And where to go next?

[Adam D. Ruppe via Digitalmars-d]

On Saturday, 7 May 2016 at 15:15:23 UTC, Adam D. Ruppe wrote:
Think how much more productive the forum would appear if people 
- including you and Walter! - actually talked about ongoing 
projects on it instead of primarily just proposals.


Oh, one more important point: I sometimes I want to randomly blog 
D changes here, but I don't feel it fits in, so I limit myself to 
pretty rare announcements (especially since my code is more of a 
living document than discrete releases).


If you guys posted little things ongoing, the length need be no 
more than a tweet, I'm sure others would too, and then we'd know 
of more projects and might even work together.

Re: [OT] Re: How are you enjoying DConf? And where to go next?

[Adam D. Ruppe via Digitalmars-d]

On Saturday, 7 May 2016 at 13:30:28 UTC, Daniel Kozak wrote:
The zero is a point when water became an ice, so anything below 
zero means snow and anything above means rain :)


Except it doesn't quite work that way you can get snow when 
it is a above freezing if the humidity is low enough. In 
Fahrenheit, anything in the thirties is borderline, and twenties 
is snow territory.


In Celsius, the snow line is more like 3 degrees than zero.

Besides, zero is just as arbitrary as thirty-two (which, btw, is 
a power of two*) and easy to remember if you use it anyway! I 
know a lot of people who don't know that water boiling point is 
around 212 F at standard pressure (interestingly, and not 
entirely coincidentally, 180 degrees away from freezing) - but if 
they don't know it and never found a need to, doesn't that 
indicate that it isn't important information to know?



* I really think that's the main difference between American 
units and the others - base two vs base ten. Base two is superior 
in basically every way, but base ten is more newbie friendly for 
doing irrelevant conversions.

Re: [OT] Re: How are you enjoying DConf? And where to go next?

[Adam D. Ruppe via Digitalmars-d]

On Saturday, 7 May 2016 at 13:10:37 UTC, Andrei Alexandrescu 
wrote:
Please curb off-topic discussions and mark them with [OT], 
yadda yadda yadda. Thanks! -- Andrei


When people complain about the D forums, they aren't complaining 
about infrequent OT stuff. Random tangents about the weather 
naturally fizzle out after five or six posts and don't leave hard 
feelings.


The bigger problem we have is that things get discussed without a 
plan of action that actually gets followed through on. How can we 
forget the `virtual` debacle? That's the real spam on the forum: 
a long thread where things are discussed, people believe a 
decision is reached, code gets implemented then all 
mysteriously dropped.


On the other hand, new things pop up from apparently nowhere. We 
have these roadmap documents twice a year but they don't seem to 
be very meaningful.


What I think would help on the forum is if actual work is 
discussed in real time, something like the main contributors to 
check in every other day and tell us what they're working on and 
what they're working toward. No need to discuss it at length, 
that'd get counter-productive, but just take a few minutes for 
you and Walter and others to say "I'm working on big-O right now 
which will help with containers, and plan to continue that 
throughout the week." or "I decided to shift gears and play with 
64 bit exception handling because it will help C++ interop."


Like the agile stand up meeting idea, though even less formal. 
Just so the community has *frequent* updates on the short-term 
direction to complement the longer term direction laid out in the 
2016 H1/H2 documents.


Not so anyone gets to input on it, but just so we feel inspired. 
My review of merged pull requests at the end of the week sort of 
gets there, but it is hard for me to get past the surface, I can 
only see a summary after-the-fact with little indication of why 
or how it fits into your development plans.


Think how much more productive the forum would appear if people - 
including you and Walter! - actually talked about ongoing 
projects on it instead of primarily just proposals.

Re: How are you enjoying DConf? And where to go next?

[Nick Sabalausky via Digitalmars-d]

On 05/06/2016 10:13 AM, Andrei Alexandrescu wrote:

The atmosphere here is great, and I'm curious how it feels for those who
are watching remotely. Is the experience good? What can we do better?



The livestream videos requiring flash was a bit of an issue.


Also: we're talking about the DConf 2017 location. Please share any
initial thoughts!



Cleveland!

[OT] Re: How are you enjoying DConf? And where to go next?

[Adam D. Ruppe via Digitalmars-d]

On Saturday, 7 May 2016 at 01:19:43 UTC, Steven Schveighoffer 
wrote:
Yes, my wife went for an interview there out of college :) The 
snow banks were like 6 feet high!


hehe, yeah, we had some beast winters recently too. This last one 
was pretty mild in comparison, but the one before was one of the 
coldest on record - I don't think it got above freezing once in 
the entire month of February, and the snow really piled up. The 
year previous had a moderate ice storm followed by record 
snowfalls (and significant cold, though it did thaw a couple 
times, thankfully).


The snow banks indeed got very tall, typical corner was like five 
feet, and where driveway plows shoved it, some were like eight 
feet up. (Popular intersections aren't as bad since the city will 
send trucks to dump it in the river, but that's just not feasible 
to do everywhere.)


And this is why I prefer riding my bike. People call me nuts for 
this, but my eyes sit higher on the bike than almost any motor 
vehicle, so I can see over more, are like 18 inches off the front 
wheel instead of several feet like in a car, so I don't have to 
creep out so far to see around things, and my ears still work, no 
engine noise to jam them up.


As a result, on the bike, I'm far more aware of surroundings than 
in the car; those advantages more than make up for the lack of 
mirrors. Moreover, the bike is a lot easier to stop than a car, 
there's a lot of advantages.


It is just cold.


P.S. if it were to be a serious consideration, I'm all for it!


Well, we couldn't literally use my house, it isn't that big, but 
I do know people with larger spaces and there are business class 
hotels up here... and the little Watertown airport is pretty nice 
to use (it only offers service to Philadelphia, but from there 
you can change planes to get wherever).


We also have easy access to Canada!

But it'd prolly be completely unaffordable for the Europeans 
anyway :(

Re: Web page listing all D compilers (and DMDFE version!) on travis-ci

[Nick Sabalausky via Digitalmars-d-announce]

On 05/07/2016 05:44 AM, Johan Engelen wrote:

What I mean is: currently the Name column says e.g. "ldc2-0.17.1", but
in the travis.yml file you must specify "ldc-0.17.1" to get it (without
the "2").


Ahh, you're right, I hadn't noticed that.

[OT] Re: How are you enjoying DConf? And where to go next?

[Daniel Kozak via Digitalmars-d]

Dne 7.5.2016 v 14:54 Adam D. Ruppe via Digitalmars-d napsal(a):


On Saturday, 7 May 2016 at 07:29:28 UTC, Russel Winder wrote:
Oh, of course, you are in the USA and are still using weird 
temperature scales, unlike the rest of the world. ;-)


I really don't understand the appeal of the Celsius scale. Zero 
degrees Celsius is like "put on a light sweater". That's not what zero 
should mean! Zero degrees Fahrenheit, on the other hand, means your 
nose freezes shut instantly upon exposure. Now *that* is worthy of the 
number zero.


(Zero Kelvin is a hardcore zero, but it is just too rare in everyday 
life to be interesting.)



Celsius scale make much more sense to me :). The zero is a point when 
water became an ice, so anything below zero means snow and anything 
above means rain :)

Re: The end of curl (in phobos)

[Andrei Alexandrescu via Digitalmars-d]

On 5/6/16 11:32 AM, Robert burner Schadek wrote:

As discussed yesterday at DConf, curl in phobos must go.

The plan is as follows.

1. undocument everything curl related in may 2016
2. deprecate everything curl related in may 2017
3. delete everything curl related in may 2018
3.1 move curl stuff to undead

PR: https://github.com/dlang/phobos/pull/4283


Another library discussed as even more ripe for replacement is zlib:

* We compile the C source code, which is extra aggravation to the build 
scripts


* The D wrappers are ancient and clunky - we should have beautiful, 
efficient range streaming


* I think the license allows us to translate the source code to D, so we 
have a viable path of replacement


* Just got word (thx Rikki) the D wrappers use GC in a laissez-faire 
manner, so no way to use the library tightly


So if anyone wants to look into this, it would be a good project.


Andrei

[OT] Re: How are you enjoying DConf? And where to go next?

[Andrei Alexandrescu via Digitalmars-d]

On 5/7/16 3:54 PM, Adam D. Ruppe wrote:

On Saturday, 7 May 2016 at 07:29:28 UTC, Russel Winder wrote:

Oh, of course, you are in the USA and are still using weird
temperature scales, unlike the rest of the world. ;-)


I really don't understand the appeal of the Celsius scale. Zero degrees
Celsius is like "put on a light sweater". That's not what zero should
mean! Zero degrees Fahrenheit, on the other hand, means your nose
freezes shut instantly upon exposure. Now *that* is worthy of the number
zero.

(Zero Kelvin is a hardcore zero, but it is just too rare in everyday
life to be interesting.)


Please curb off-topic discussions and mark them with [OT], yadda yadda 
yadda. Thanks! -- Andrei

[Issue 2659] Remove the comma operator

[via Digitalmars-d-bugs]

https://issues.dlang.org/show_bug.cgi?id=2659

Mathias Lang  changed:

   What|Removed |Added

 CC||mathias.l...@sociomantic.co
   ||m

--- Comment #12 from Mathias Lang  ---
https://github.com/dlang/dmd/pull/5737

--

Re: CPU discussion post-DConf

[Kai Nacke via Digitalmars-d]

On Friday, 6 May 2016 at 17:45:52 UTC, Mithun Hunsur wrote:
After my talk, I had a few people come up to me and express 
interest in talking about CPUs and projects related to them - 
unfortunately, I've got a flight to catch (I'm actually at the 
airport now!) and couldn't stick around.


If you're one of those people, or you'd just like to talk about 
the topic in general, please do reach out to me. I'd love to 
talk shop :)


Hi Mithun!

You should look at the two following talks:
https://fosdem.org/2016/schedule/event/llvm_aap/
https://fosdem.org/2016/schedule/event/llvm_aapsim/

This can help you with how to use LLVM. The talks about a 
somewhat similar project.


Regards,
Kai

Re: How are you enjoying DConf? And where to go next?

[Adam D. Ruppe via Digitalmars-d]

On Saturday, 7 May 2016 at 07:29:28 UTC, Russel Winder wrote:
Oh, of course, you are in the USA and are still using weird 
temperature scales, unlike the rest of the world. ;-)


I really don't understand the appeal of the Celsius scale. Zero 
degrees Celsius is like "put on a light sweater". That's not what 
zero should mean! Zero degrees Fahrenheit, on the other hand, 
means your nose freezes shut instantly upon exposure. Now *that* 
is worthy of the number zero.


(Zero Kelvin is a hardcore zero, but it is just too rare in 
everyday life to be interesting.)

[Issue 15799] Misleading error message against the contract followed by semicolon in interface

[via Digitalmars-d-bugs]

https://issues.dlang.org/show_bug.cgi?id=15799

--- Comment #3 from github-bugzi...@puremagic.com ---
Commits pushed to master at https://github.com/dlang/dmd

https://github.com/dlang/dmd/commit/078067a0222440d4bd2b833f8e0d73bb06730b20
fix Issue 15799 - Misleading error message against the contract followed by
semicolon in interface

https://github.com/dlang/dmd/commit/0dfaaecac67e7cb94fbb41eda410751039b64e15
Merge pull request #5571 from 9rnsr/fix15799

Issue 15799 - Misleading error message against the contract followed by
semicolon in interface

--

Re: Reproducible builds of D compilers

[Pjotr Prins via Digitalmars-d]

On Saturday, 7 May 2016 at 08:53:10 UTC, Thomas Mader wrote:
GNU Guix can be used as a reproducible reference build system 
where we capture the bootstrapping process nicely and test the 
builds on the build farm(s) whenever something changes.


I like the idea.
I thought about that too but would not use GNU Guix for it.
I would suggest to use Nix -> http://nixos.org/
AFAIK GNU Guix uses some parts of Nix and I think the community 
behind it is not as big.

There are also packages for dmd on Nix already.
I also was able to get the ddmd frontent version of dmd running 
on Nix.
Nix would be the perfect system to base all the testing on 
because it's very easy to distribute entire system 
configurations over multiple computers.
So once everything is setup correctly and the knowledge is 
there, it makes administering entire server farms pretty 
comfortable.


Hi Thomas! Good news, Nix and Guix can easily be used on one 
system. I am a GNU Guix person (for several reason; also used Nix 
for years), and I am glad to say that the community is growing 
rather fast. Guix is a pretty young project.


I am committed to supporting the D language on Guix (if only to 
deploy my own working environments), be great to have the same on 
Nix.

Re: How are you enjoying DConf? And where to go next?

[Andrei Alexandrescu via Digitalmars-d]

On 5/7/16 2:47 PM, Arjan wrote:

On Friday, 6 May 2016 at 14:13:35 UTC, Andrei Alexandrescu wrote:

The atmosphere here is great, and I'm curious how it feels for those
who are watching remotely. Is the experience good? What can we do better?

Also: we're talking about the DConf 2017 location. Please share any
initial thoughts!


Thanks,

Andrei


I was not able to watch live much. But the parts I did were OK. Audio
was good enough, video I would have liked to be a bit better, but
do-able. A very good idea to have the slides available upfront!

I watched the most in the made available streams afterwards. Easy to
pick and select the topics of interest. IMO it is great to have the
content available right after/during the current day.

[snip]

Thanks very much! -- Andrei

Re: Pointer top 16 bits use

[Nicholas Wilson via Digitalmars-d]

On Saturday, 7 May 2016 at 09:01:11 UTC, Ali Çehreli wrote:

On 05/06/2016 11:08 PM, Nicholas Wilson wrote:

In Dicebot's DConf talk


That was deadalnix. :)

Ali


Touché.

Re: How are you enjoying DConf? And where to go next?

[Arjan via Digitalmars-d]

On Friday, 6 May 2016 at 14:13:35 UTC, Andrei Alexandrescu wrote:
The atmosphere here is great, and I'm curious how it feels for 
those who are watching remotely. Is the experience good? What 
can we do better?


Also: we're talking about the DConf 2017 location. Please share 
any initial thoughts!



Thanks,

Andrei


I was not able to watch live much. But the parts I did were OK. 
Audio was good enough, video I would have liked to be a bit 
better, but do-able. A very good idea to have the slides 
available upfront!


I watched the most in the made available streams afterwards. Easy 
to pick and select the topics of interest. IMO it is great to 
have the content available right after/during the current day.


The atmosphere was even noticeable in the video stream, for me at 
least. To me it seems like a really great and inspirational 
conference!


Location, personally doesn't matter much to me since I'm most 
probably not able to attend anyway. But maybe Japan/Korea?


Topics and niveau, you did not ask but will give my opinion 
anyway, the diversity is good, the level of the talks I saw were 
for people not to intemit with the language and its eco system 
possibly a bit to high / abstract. I don't know whether or not 
there is a specific group targeted by dconf but when one of its 
(main) purposes is to attract new people from other language- eco 
systems, then IMO there should be given a bit more thought on the 
content topics and niveau.

When this is not the purpose then its all fine.

About attracting people to D, my experience is that beside the 
language zealots which will never change to different language 
eco system, most developers do not look around and select the 
best tools for the job at hand, they just look back into to their 
experience-toolbox and start solving the problem using the tools 
and technologies their familiar with, even though better tools 
are readily available. IMO D offers for a lot, a near perfect fit 
for a lot of problems, but fails in 'marketing' department. Most 
professionals I met never heard of D or never tried it (to much 
hassle). When they try it out it is not only the first 5 minutes, 
but the first project that matters. Most of them never finish 
that project in D

Feedback I get:
- integration in tooling not good enough (VisualD is of help, but 
seems to fail)

- debugging problems(!!!)
- documentation for beginners not easy to find or follow.
- just to much of a hassle

I hope we can improve on this.

Arjan

[Issue 15258] Anonymous const union members don't allow for initialization

[via Digitalmars-d-bugs]

https://issues.dlang.org/show_bug.cgi?id=15258

github-bugzi...@puremagic.com changed:

   What|Removed |Added

 Status|NEW |RESOLVED
 Resolution|--- |FIXED

--

[Issue 15258] Anonymous const union members don't allow for initialization

[via Digitalmars-d-bugs]

https://issues.dlang.org/show_bug.cgi?id=15258

--- Comment #3 from github-bugzi...@puremagic.com ---
Commits pushed to master at https://github.com/dlang/dmd

https://github.com/dlang/dmd/commit/5175be50ba15738cf0f90ed4136f9a0484ecd352
fix Issue 15258 - Anonymous const union members don't allow for initialization

https://github.com/dlang/dmd/commit/87a802f7aa6d91c09cfd7604f6ea77c7d6fd477e
Merge pull request #5242 from 9rnsr/fix15258

Issue 15258 - Anonymous const union members don't allow for initialization

--

[Issue 15897] private base class method not seen through derived class

[via Digitalmars-d-bugs]

https://issues.dlang.org/show_bug.cgi?id=15897

Kenji Hara  changed:

   What|Removed |Added

 Status|RESOLVED|REOPENED
 Resolution|FIXED   |---

--- Comment #16 from Kenji Hara  ---
The new concept https://issues.dlang.org/show_bug.cgi?id=15897#c12 is
completely new thing, so language spec needs to be updated.

--

Re: It's alive! D building D building D, all on Android

[Johan Engelen via Digitalmars-d-announce]

Fantastic news!

I hope we can find a good way to integrate automated testing of 
github branches/PRs for Android.


cheers,
 Johan

Bike rental in Berlin/Neukölln

[Joseph Rushton Wakeling via Digitalmars-d]

A few people asked about bike rental.  I can't vouch for anyone 
myself, but friends have told me these folks have a good 
reputation:

https://www.facebook.com/RentABike44/

They're over on Mahlower Straße 9:
https://www.google.de/maps/place/Rent+a+Bike+44+in+Berlin+Neuk%C3%B6lln/@52.4797042,13.4201783,17z/data=!3m1!4b1!4m5!3m4!1s0x47a84fbee8c28a6f:0xfa3075bd1b7d33df!8m2!3d52.479701!4d13.422367

Probably the best way to get there is to take the U7 from 
Grenzallee up to Hermannplatz, and then the U8 south to 
Boddinstraße; the corner of Mahlower Straße is just next to 
Boddinstraße U-bahn station.

Re: Post-DConf brunch?

[Joseph Rushton Wakeling via Digitalmars-d]

On Saturday, 7 May 2016 at 09:24:13 UTC, Adrian Matoga wrote:
On Saturday, 7 May 2016 at 08:55:50 UTC, Joseph Rushton 
Wakeling wrote:

Hello all,

For anyone who's still in town and fancies grabbing a nice 
late breakfast/early lunch in a nice Berlin food place, I'll 
be doing my usual Saturday morning thing of dropping by this 
cafe/restaurant


Cool!
I'll be there in ca. 20 minutes.


Great!  I'm sitting at a table just inside the entrance.  
Probably easily identifiable by the laptop backpack next to me ;-)

Re: How are you enjoying DConf? And where to go next?

[Daniel Kozak via Digitalmars-d]

On Friday, 6 May 2016 at 14:13:35 UTC, Andrei Alexandrescu wrote:
The atmosphere here is great, and I'm curious how it feels for 
those who are watching remotely. Is the experience good? What 
can we do better?



There were some small issues. First there was no link to 
livestream on official dconf or dlang page (So it took some time 
to find this link on twitter). Another problem is with 
presentations. Because of video poor quality it was hard to see 
what is on slides. So it would be very helpful if there will be 
slides available for each of presentation before it starts.



Also: we're talking about the DConf 2017 location. Please share 
any initial thoughts!



Thanks,

Andrei


I would prefer europe. Brno or Prague would be awesome.

Re: Web page listing all D compilers (and DMDFE version!) on travis-ci

[Johan Engelen via Digitalmars-d-announce]

On Tuesday, 26 April 2016 at 06:42:11 UTC, Nick Sabalausky wrote:

https://semitwist.com/travis-d-compilers

That's an auto-generated listing of all versions of DMD, GDC 
and LDC available on travis-ci.


Very useful, but perhaps you can list the string one should use 
to tell Travis to use that version?
What I mean is: currently the Name column says e.g. 
"ldc2-0.17.1", but in the travis.yml file you must specify 
"ldc-0.17.1" to get it (without the "2").
(If the string is incorrect, the install script will give an 
error message from which it is not clear at all that the compiler 
name is not recognized.)


Thanks,
  Johan

Re: The end of curl (in phobos)

[ikod via Digitalmars-d]

On Friday, 6 May 2016 at 10:27:54 UTC, yawniek wrote:
On Friday, 6 May 2016 at 09:21:43 UTC, Andrei Alexandrescu 
wrote:

On 5/6/16 10:32 AM, Robert burner Schadek wrote:

As discussed yesterday at DConf, curl in phobos must go.

The plan is as follows.

1. undocument everything curl related in may 2016
2. deprecate everything curl related in may 2017
3. delete everything curl related in may 2018
3.1 move curl stuff to undead

PR: https://github.com/dlang/phobos/pull/4283


0. Write curl replacement


https://github.com/ikod/dlang-requests looks very promising, 
maybe it could be re-licenced?


Hello,

Just some notes...
I'm going to add some missed features like cookies, file 
downloads, maybe other high level features. My goal is something 
like python "requests" library.


What do you mean under re-licensing?

Re: Post-DConf brunch?

[Adrian Matoga via Digitalmars-d]

On Saturday, 7 May 2016 at 08:55:50 UTC, Joseph Rushton Wakeling 
wrote:

Hello all,

For anyone who's still in town and fancies grabbing a nice late 
breakfast/early lunch in a nice Berlin food place, I'll be 
doing my usual Saturday morning thing of dropping by this 
cafe/restaurant


Cool!
I'll be there in ca. 20 minutes.

Re: The end of curl (in phobos)

[Lutger Blijdestijn via Digitalmars-d]

I was at dconf, but missed the discussion where this was spoken 
of. I'm curious to know what is wrong with the curl wrapper. I've 
only recently been following D again after a gap of several years 
and remember curl being received quite favorablyt at the time, 
several people used it as an example of good api design.


Its such a useful and simple piece of functionality to have out 
of the box, greatly increasing the amount of interesting things 
you can do with D without installing additional packages. Phobos 
already has lackluster support for common web technologies, it 
would be a bummer if curl is deprecated before a replacement was 
in place.

Re: Pointer top 16 bits use

[Ali Çehreli via Digitalmars-d]

On 05/06/2016 11:08 PM, Nicholas Wilson wrote:

In Dicebot's DConf talk


That was deadalnix. :)

Ali

Post-DConf brunch?

[Joseph Rushton Wakeling via Digitalmars-d]

Hello all,

For anyone who's still in town and fancies grabbing a nice late 
breakfast/early lunch in a nice Berlin food place, I'll be doing 
my usual Saturday morning thing of dropping by this 
cafe/restaurant at about 11:30:

https://www.google.de/maps/place/Papilles/@52.4811843,13.4279886,17z/data=!4m13!1m7!3m6!1s0x47a84fbb7a005065:0x8bfd7619bd24d78!2sPapilles!3b1!8m2!3d52.4811811!4d13.4301773!3m4!1s0x47a84fbb7a005065:0x8bfd7619bd24d78!8m2!3d52.4811811!4d13.4301773

(Papilles, Flughafenstraße 25, for anyone who can't get the map 
to work:-)


If you catch the U7 (or walk) to Rathaus Neukölln, then 
Flughafenstraße should be just on your left, the big street that 
heads uphill past the Neukölln Arcaden.


Best wishes,

-- Joe

Re: Reproducible builds of D compilers

[Thomas Mader via Digitalmars-d]

On Friday, 6 May 2016 at 08:07:30 UTC, Pjotr Prins wrote:
At Dconf we had a discussion on creating reproducible builds of 
the D compilers. One thing that is required is bootstrapping 
the build. In GNU Guix we start from a working C compiler which 
is a reasonable starting point (it could have been LISP instead 
;)


For D this means that even if all components of the compilers 
are written in D we should have a (chain of) compilation steps 
that can be initiated from a C or C++ compiler. This is still 
the case today, but I want to highlight this point here so we 
do not lose that facility.


GNU Guix can be used as a reproducible reference build system 
where we capture the bootstrapping process nicely and test the 
builds on the build farm(s) whenever something changes.


I like the idea.
I thought about that too but would not use GNU Guix for it.
I would suggest to use Nix -> http://nixos.org/
AFAIK GNU Guix uses some parts of Nix and I think the community 
behind it is not as big.

There are also packages for dmd on Nix already.
I also was able to get the ddmd frontent version of dmd running 
on Nix.
Nix would be the perfect system to base all the testing on 
because it's very easy to distribute entire system configurations 
over multiple computers.
So once everything is setup correctly and the knowledge is there, 
it makes administering entire server farms pretty comfortable.

Re: It's alive! D building D building D, all on Android

[calex via Digitalmars-d-announce]

On Thursday, 5 May 2016 at 08:17:07 UTC, Joakim wrote:
After a sleepless night of trying to build the latest ldc 
master branch 2.070.2 on my Android tablet a couple nights ago, 
almost the full druntime/phobos standard library test suite 
passes (only one assert in std.conv) and the same for the dmd 
test suite, with a handful of failures across three modules 
exclusively testing C/C++ ABI compatibility.


[...]


Very nice.
Thanks for sharing, Joakim.

I will certainly be using this later— whenever it is that I 
eventually emerge from my cave.

Re: How are you enjoying DConf? And where to go next?

[Russel Winder via Digitalmars-d]

On Fri, 2016-05-06 at 20:59 +, John Colvin via Digitalmars-d wrote:
> On Friday, 6 May 2016 at 14:57:59 UTC, Chris wrote:
> > On Friday, 6 May 2016 at 14:53:15 UTC, Jonathan M Davis wrote:
> > > [...]
> > 
> > Facebook have their European headquarters in Dublin. Maybe 
> > they'd be willing to sponsor DConf2017 (there are loads of 
> > other tech companies in Dublin). Flights from the U.S. to 
> > Ireland and back are very frequent.
> 
> Dublin has cheap direct flights from quite a few places in the 
> US, as well as pretty much everywhere in europe.

Obviously I will suggest London, but I guess it is a non-starter. The
obvious venue would be Skills Matter's CodeNode, but the cost of
accommodation in London is horrifying, and commuting is a non-starter.
Well except for me.

There are venues such as Coventry, the Techno Park has a great venue
for up to about 200 people and fairly cheap and very local
accommodation. (PyConUK has been using this venue for the last few
years very successfully, it has only moved to Cardiff for a change and
because of the problem of outgrowing the venue.)  

I have done the Gatwick → Dublin → (Gatwick | Birmingham) a few times,
but on expenses so used taxis from Dublin airport to either Pearse
Street or E Wall Road – the hotels shall remain nameless.

What venue would be used in Dublin? I guess we could ask the PyConIE
folk what they do, one of the hotels I think.
 
-- 
Russel.
=
Dr Russel Winder  t: +44 20 7585 2200   voip: sip:russel.win...@ekiga.net
41 Buckmaster Roadm: +44 7770 465 077   xmpp: rus...@winder.org.uk
London SW11 1EN, UK   w: www.russel.org.uk  skype: russel_winder

signature.asc
Description: This is a digitally signed message part

Re: How are you enjoying DConf? And where to go next?

[Russel Winder via Digitalmars-d]

On Sat, 2016-05-07 at 01:04 +, Adam D. Ruppe via Digitalmars-d
wrote:
[…]
> 
> Oh, come on! It was actually pretty nice this week, we are in the 
> 60's right now. Finally, April was kinda miserable.

But people die when it gets over 50.

Oh, of course, you are in the USA and are still using weird temperature
scales, unlike the rest of the world. ;-)








PS Yes I do know UK still uses miles for road distances. Jokes can be
made about that as well.

-- 
Russel.
=
Dr Russel Winder  t: +44 20 7585 2200   voip: sip:russel.win...@ekiga.net
41 Buckmaster Roadm: +44 7770 465 077   xmpp: rus...@winder.org.uk
London SW11 1EN, UK   w: www.russel.org.uk  skype: russel_winder

signature.asc
Description: This is a digitally signed message part

Re: Pointer top 16 bits use

[Benjamin Thaut via Digitalmars-d]

On Saturday, 7 May 2016 at 06:08:01 UTC, Nicholas Wilson wrote:
In Dicebot's DConf talk he mentioned that it is possible to use 
the top 16 bits for tagging pointers and the associated risks.


Regarding the GC not seeing a pointer if the top 16 bits are 
used, whats to stop us from changing the GC to (or adding an 
option to) ignore those bits when checking?


I imagine that it would cause a lot more false pointers, but 
when we get precise GC that problem would go away.


The problem with this is, that its not future proof. Once the top 
16-bits of a pointer get used D programs will suddenly stop 
working. So i think we shouldn't support that officially. If you 
really want to do that you can simply hack the scan routine of 
the gc in druntime yourself and ship with the hacked version.

[Issue 15913] cannot initialize immutable fixed size array without similar const-code

[via Digitalmars-d-bugs]

https://issues.dlang.org/show_bug.cgi?id=15913

github-bugzi...@puremagic.com changed:

   What|Removed |Added

 Status|NEW |RESOLVED
 Resolution|--- |FIXED

--

[Issue 15913] cannot initialize immutable fixed size array without similar const-code

[via Digitalmars-d-bugs]

https://issues.dlang.org/show_bug.cgi?id=15913

--- Comment #2 from github-bugzi...@puremagic.com ---
Commits pushed to master at https://github.com/dlang/dmd

https://github.com/dlang/dmd/commit/260d3ded4a40f81fc614c9d4fd3020eddc1ae6b4
fix Issue 15913 - cannot initialize immutable fixed size array without similar
const-code

https://github.com/dlang/dmd/commit/fbca5f87d96901ba7e825274ee552a8c06af9b83
Merge pull request #5652 from 9rnsr/fix15913

Issue 15913 - cannot initialize immutable fixed size array without similar
const-code

--

Re: How are you enjoying DConf? And where to go next?

[Meta via Digitalmars-d]

On Friday, 6 May 2016 at 15:16:00 UTC, deadalnix wrote:
On Friday, 6 May 2016 at 14:13:35 UTC, Andrei Alexandrescu 
wrote:
Also: we're talking about the DConf 2017 location. Please 
share any initial thoughts!




I think American east cost is a natural next location. 
Something like New York/Washington/Boston/Montréal .


Is Montreal considered an east coast city now? ;^)

Jokes aside, Boston would be my vote. By far my favourite 
American east coast city.

Pointer top 16 bits use

[Nicholas Wilson via Digitalmars-d]

In Dicebot's DConf talk he mentioned that it is possible to use 
the top 16 bits for tagging pointers and the associated risks.


Regarding the GC not seeing a pointer if the top 16 bits are 
used, whats to stop us from changing the GC to (or adding an 
option to) ignore those bits when checking?


I imagine that it would cause a lot more false pointers, but when 
we get precise GC that problem would go away.