November 07, 2013
, el  6 de November a las 21:53 me escribiste:
> On Wednesday, 6 November 2013 at 20:11:13 UTC, Aleksandar Ruzicic wrote:
> >versions must be marked with rc, as betas are marked with b
> >"flag". Something like 2.064-rc.1, 2.064-rc.2, ... 2.064
> >(stable/major release), 2.064.1 (patch release), ...
> >
> >This (-rc.xx) is how RC versions should be marked as per SEMVER
> >"standard" (http://semver.org), although I know that D doesn't
> >follow semantic versioning as defined in that standard.
> 
> The D version numbers fail requirement 2 of semantic versioning:
> 
> 2. A normal version number MUST take the form X.Y.Z where X, Y, and Z are non-negative integers, and MUST NOT contain leading zeroes.
> 
> I know that was discussed somewhere, but I don't know/recall why there is a leading zero in the minor version number.

I think because back in the stone age, it was hard to sort versions like this: 1.5 and 1.15. Lexicographically speaking 1.5 > 1.15.

I don't think there is any reason now for leading zero, just historical
reasons. It would be awesome to get DMD follow semantic versioning as
much as possible. Even when is not really a library, I guess the
language specification can be taken as the API. The only problem is from
time to time some tiny non backwards compatible changes are made and
I don't anyone would like to bump the major version because of that. But
I think an exception could be made for that, and I think those changes
appear less and less frequently, so it shouldn't be a big issue.

-- 
Leandro Lucarella (AKA luca)                     http://llucax.com.ar/
----------------------------------------------------------------------
GPG Key: 5F5A8D05 (F8CD F9A7 BF00 5431 4145  104C 949E BFB6 5F5A 8D05)
----------------------------------------------------------------------
All men are born equal
But quite a few get over it
November 07, 2013
Walter Bright, el  6 de November a las 12:01 me escribiste:
> On 11/6/2013 5:16 AM, Jordi Sayol wrote:
> >In "dmd.2.064.2.zip", src/VERSION contains "2.064". Should be "2.064.2"
> 
> I deliberately didn't do that because it would have required rebuilding all the binaries just for that.

And that's bad because.... ?

-- 
Leandro Lucarella (AKA luca)                     http://llucax.com.ar/
----------------------------------------------------------------------
GPG Key: 5F5A8D05 (F8CD F9A7 BF00 5431 4145  104C 949E BFB6 5F5A 8D05)
----------------------------------------------------------------------
"The Guinness Book of Records" holds the record for being the most stolen book in public libraries
November 07, 2013
Walter Bright, el  6 de November a las 11:57 me escribiste:
> On 11/6/2013 4:34 AM, Leandro Lucarella wrote:
> >Also I find strange that the first patchlevel version is 2 and not 1. Was that intended or just an error?
> 
> It was intended. I felt that 2.064 => 2.064.1 would have been confusing, hence 2.064 => 2.064.2

That's funny, I find it very confusing to jump from 2.064 to 2.064.2. 2.064 is implied to be 2.064.0, as version 1 is implied to be 1.0 (and as a floating point number 1 is 1.0, not 1.1).

Every other project out there uses this convention. So I wonder why do you find 2.064 => 2.064.1 confusing.

Looking at previous versions I just noticed you did the same with 2.063, I didn't notice then. But please, could you consider changing that naming scheme and using 2.0XX.1 as the 1st patchlevel (see the relation? :).

Thanks.

And I would also want to thanks for another great release, with a great changelog despite the protests! :D

-- 
Leandro Lucarella (AKA luca)                     http://llucax.com.ar/
----------------------------------------------------------------------
GPG Key: 5F5A8D05 (F8CD F9A7 BF00 5431 4145  104C 949E BFB6 5F5A 8D05)
----------------------------------------------------------------------
A lo que Peperino respondióles: aquel que tenga sabañones que se los
moje, aquel que padece calvicie no padece un osito, no es bueno comer
lechón en día de gastritis, no mezcleis el vino con la sandía, sacad la
basura después de las ocho, en caso de emergencia rompa el vidrio con
el martillo, a cien metros desvio por Pavón.
	-- Peperino Pómoro
November 07, 2013
Jacob Carlborg, el  6 de November a las 22:06 me escribiste:
> On 2013-11-06 20:57, Walter Bright wrote:
> 
> >It was intended. I felt that 2.064 => 2.064.1 would have been confusing, hence 2.064 => 2.064.2
> 
> That's what's happening if you start to add new digits. The first release should have possibly been 2.064.0. BTW, there was a 2.063.1, if I recall correctly.

I also have the impression I saw a 2.063.1. There are certainly posts in
the devel list about that version, there is none with that version in
the download directory:
http://downloads.dlang.org/releases/2013/

Maybe the discussion was about 2.063.1 but then Walter name it 2.063.2, or maybe it was removed from the web server?

-- 
Leandro Lucarella (AKA luca)                     http://llucax.com.ar/
----------------------------------------------------------------------
GPG Key: 5F5A8D05 (F8CD F9A7 BF00 5431 4145  104C 949E BFB6 5F5A 8D05)
----------------------------------------------------------------------
Software is like sex: it's better when it's free.
	-- Linus Torvalds
November 07, 2013
On Thursday, November 07, 2013 00:11:37 Leandro Lucarella wrote:
> Walter Bright, el 6 de November a las 11:57 me escribiste:
> > On 11/6/2013 4:34 AM, Leandro Lucarella wrote:
> > >Also I find strange that the first patchlevel version is 2 and not 1. Was that intended or just an error?
> > 
> > It was intended. I felt that 2.064 => 2.064.1 would have been confusing, hence 2.064 => 2.064.2
> 
> That's funny, I find it very confusing to jump from 2.064 to 2.064.2. 2.064 is implied to be 2.064.0, as version 1 is implied to be 1.0 (and as a floating point number 1 is 1.0, not 1.1).
> 
> Every other project out there uses this convention. So I wonder why do you find 2.064 => 2.064.1 confusing.

Yeah. Going from 2.064 or 2.064.0 to 2.064.1 would be pretty standard. Jumping straight from 2.064 to 2.064.2 is what's likely to confuse most people.

> Looking at previous versions I just noticed you did the same with 2.063, I didn't notice then. But please, could you consider changing that naming scheme and using 2.0XX.1 as the 1st patchlevel (see the relation?

I had assumed that there was a 2.063.1 prior to 2.063.2 but clearly wasn't paying enough attention.

- Jonathan M Davis
November 07, 2013
On 11/6/2013 3:43 PM, nazriel wrote:
> Good job everyone!
> DPaste is already using it

Nice!
November 07, 2013
On 11/6/2013 3:20 PM, Leandro Lucarella wrote:
> Walter Bright, el  6 de November a las 12:01 me escribiste:
>> On 11/6/2013 5:16 AM, Jordi Sayol wrote:
>>> In "dmd.2.064.2.zip", src/VERSION contains "2.064". Should be "2.064.2"
>>
>> I deliberately didn't do that because it would have required
>> rebuilding all the binaries just for that.
>
> And that's bad because.... ?
>

Time, and then wondering what is different when it isn't different
November 07, 2013
On 2013-11-05 23:08, Walter Bright wrote:
> Ok, this is it:
>
> http://ftp.digitalmars.com/dmd_2.064.2-0_amd64.deb
> http://ftp.digitalmars.com/dmd-2.064.2-0.fedora.i386.rpm
> http://ftp.digitalmars.com/dmd-2.064.2-0.fedora.x86_64.rpm
> http://ftp.digitalmars.com/dmd_2.064.2-0_i386.deb
> http://ftp.digitalmars.com/dmd-2.064.2-0.openSUSE.i386.rpm
> http://ftp.digitalmars.com/dmd-2.064.2-0.openSUSE.x86_64.rpm
> http://ftp.digitalmars.com/dmd-2.064.2.exe
> http://ftp.digitalmars.com/dmd.2.064.2.zip
> http://ftp.digitalmars.com/dmd.2.064.2.dmg
> http://ftp.digitalmars.com/libphobos2-64_2.064.2-0_amd64.deb
> http://ftp.digitalmars.com/libphobos2-64_2.064.2-0_i386.deb

The changelog is missing issue 10700. I though that part was automatically generated.

-- 
/Jacob Carlborg
November 07, 2013
how do make that comiler work?
[Issue 11457] New: Cannot compile 64bit apps with Visual Studio 2013

this is a desaster for me. was that release tested? if i amand the pathes and run as admin or not - it will not find the libs (user32)

windoes8.1, visual studio2013, 64bit

November 07, 2013
On Thursday, 7 November 2013 at 08:58:50 UTC, tester wrote:
> how do make that comiler work?
> [Issue 11457] New: Cannot compile 64bit apps with Visual Studio 2013
>
> this is a desaster for me. was that release tested? if i amand the pathes and run as admin or not - it will not find the libs (user32)
>
> windoes8.1, visual studio2013, 64bit

if it compiles with -m32 and fails with -m64 then i think you need manually edit ur sc.ini to add correct paths for windows sdk/kits and visual studio.