www.digitalmars.com         C & C++   DMDScript  

digitalmars.D - D 1.x and 2.x where x != 0 - will there be such a thing?

reply "Stewart Gordon" <smjg_1998 yahoo.com> writes:
It seems that there are only two versions of the D language - 1.0 and 2.0. 
Versions of the compiler, OTOH, have various sequences of digits after the 
'.'.

There have been many changes to the language spec over time.  Nonetheless, 
UIMS, none of them incremented the version number of the language, except 
for the change from having no version number (?) to 1.0 and the change from 
1.0 to 2.0.

Are there going to be any other 1.x or 2.x versions of D?

If so, what level of advance in the language would be right to warrant such 
a number assigned to a version of the language, as opposed to the compiler?

If not, why are the versions referred to as 1.0 and 2.0, rather than simply 
1 and 2?

Stewart.

-- 
My e-mail address is valid but not my primary mailbox.  Please keep replies
on the 'group where everybody may benefit.
Oct 28 2008
next sibling parent reply "Tim M" <a b.com> writes:
At the moment if I write something in D i'm either riting in d1.0 or d2.0  
but most probably d1.0. I would hate it to be like D  
1.02.6.4-3-<someNickName-build>-rc4. Anyway if some code actually compiles  
and runs ok is down to the compiler which could change the way it works at  
any time. 1.036 at time of writing.

On Wed, 29 Oct 2008 06:24:22 +1300, Stewart Gordon <smjg_1998 yahoo.com>  
wrote:


 It seems that there are only two versions of the D language - 1.0 and  
 2.0. Versions of the compiler, OTOH, have various sequences of digits  
 after the '.'.

 There have been many changes to the language spec over time.   
 Nonetheless, UIMS, none of them incremented the version number of the  
 language, except for the change from having no version number (?) to 1.0  
 and the change from 1.0 to 2.0.

 Are there going to be any other 1.x or 2.x versions of D?

 If so, what level of advance in the language would be right to warrant  
 such a number assigned to a version of the language, as opposed to the  
 compiler?

 If not, why are the versions referred to as 1.0 and 2.0, rather than  
 simply 1 and 2?

 Stewart.
Oct 28 2008
parent "Stewart Gordon" <smjg_1998 yahoo.com> writes:
"Tim M" <a b.com> wrote in message news:op.ujrup1k5jdp9fl tim-laptop...

 At the moment if I write something in D i'm either riting in d1.0 or d2.0 
 but most probably d1.0. I would hate it to be like D 
 1.02.6.4-3-<someNickName-build>-rc4.


What would a build of a language be?

I don't think we'll need any version numbers as deep in a hierarchy as that. 
But it could've been done as a three-level versioning system.  So, for 
example, DMD 1.02.06 would be the 7th release of the compiler for D 1.02, 
which is the third version of the D1 spec.  This would have enabled 
developers to target particular versions of D to a greater extent than is 
possible under the current system.


 Anyway if some code actually compiles and runs ok is down to the compiler 
 which could change the way it works at any time. 1.036 at time of writing.


<snip top of upside-down reply>

You mean because of bugs coming in and out?

Stewart.

-- 
My e-mail address is valid but not my primary mailbox.  Please keep replies 
on the 'group where everybody may benefit.
Oct 30 2008
prev sibling parent reply =?ISO-8859-1?Q?Anders_F_Bj=F6rklund?= <afb algonet.se> writes:
Stewart Gordon wrote:


 There have been many changes to the language spec over time.  
 Nonetheless, UIMS, none of them incremented the version number of the 
 language, except for the change from having no version number (?) to 1.0 
 and the change from 1.0 to 2.0.
 
 Are there going to be any other 1.x or 2.x versions of D?


Not sure if the version numbers are any good indication, at least
it hasn't been for DMD... Sometimes a feature does a "backflip"
from a later version and gets re-released as an earlier version.

At least that is what happened with the "string" alias and the
two variants of DMD 1.015 available, I'm sure it could happen again.
So the remaining workaround is to detect at runtime, like autoconf...

And of course, the language spec is still missing a lot from GDC.


 If so, what level of advance in the language would be right to warrant 
 such a number assigned to a version of the language, as opposed to the 
 compiler?
 
 If not, why are the versions referred to as 1.0 and 2.0, rather than 
 simply 1 and 2?


I blame Web 2.0 for that. But thought they were mostly "D" and "D2".

Afaik, the original language release was D 1.000 anyway, and not 1.0

--anders
Oct 29 2008
parent "Stewart Gordon" <smjg_1998 yahoo.com> writes:
"Anders F Bj�rklund" <afb algonet.se> wrote in message 
news:ge9a6d$2rbn$1 digitalmars.com...
<snip>

 Afaik, the original language release was D 1.000 anyway, and not 1.0


D didn't have a version number until it reached 1.0 in January 2007.  Only 
DMD did.

D didn't have a version 1.000.  There was, however, DMD 1.00, which was the 
first compiler for D 1.0.  Subsequent versions of DMD had three digits after 
the '.'.

Stewart.

-- 
My e-mail address is valid but not my primary mailbox.  Please keep replies 
on the 'group where everybody may benefit.
Oct 30 2008