Why was the .h dropped in std headers?

[Derek]

What was the rationale for dropping the .h suffix for standard
headers (e.g., <iostream> vs <iostream.h>)? Just curious.

(I already asked in comp.std.c++, but the moderators there work
like molasses -- my post from yesterday still hasn't appeared.)

 

[Mike Wahler]

What was the rationale for dropping the .h suffix for standard
headers (e.g., <iostream> vs <iostream.h>)? Just curious.

You'd have to ask the standards committee members to be sure,
but I suspect at least part of the reason is to reduce the
possiblity of conflicts with the inherited C headers (e.g

(I already asked in comp.std.c++, but the moderators there work
like molasses -- my post from yesterday still hasn't appeared.)

Moderated groups are 'slow' by nature. I don't think it's
legitmate to call the moderators 'slow as molasses', as their
job is an unpaid volunteer activity. I'm sure they have their
own lives to deal with too.

-Mike

 

[Jeff Flinn]

What was the rationale for dropping the .h suffix for standard
headers (e.g., <iostream> vs <iostream.h>)? Just curious.

(I already asked in comp.std.c++, but the moderators there work
like molasses -- my post from yesterday still hasn't appeared.)

Your post and responses appeared there yesterday! Prehaps it's your news
server that's slow as molasses.

Jeff

 

[Victor Bazarov]

What was the rationale for dropping the .h suffix for standard
headers (e.g., <iostream> vs <iostream.h>)? Just curious.

(I already asked in comp.std.c++, but the moderators there work
like molasses -- my post from yesterday still hasn't appeared.)

I read it and the replies to it yesterday. Or was it somebody else's
question? I saw 4 replies, IIRC, one from Andrew Koenig.

 

[Jonathan Mcdougall]

What was the rationale for dropping the .h suffix for standard
headers (e.g., <iostream> vs <iostream.h>)? Just curious.

Because there was a de facto standard that most
programmers were using and the names of headers
was part of it. C++98 took these well-known
names, modified them a bit and tried to make their
content (if only the names used, such as cout and
cin) as similar as possible to what people were
used to. Finally, to avoid having two
<iostream.h> on the same machine working
differently, the .h was dropped.

Stroustrup also says the .h is redundant because
#include <> specifies a header, but that's a bit
far fetched.


Jonathan

 

[Derek]

I read it and the replies to it yesterday. Or was it somebody else's
question? I saw 4 replies, IIRC, one from Andrew Koenig.

Thanks for the heads up. It looks like my news reader/service is
not picking up recent posts from comp.std.c++. I guess those
volunteer moderators are faster than I gave them credit for.

 

[Derek]

Moderated groups are 'slow' by nature. I don't think it's
legitmate to call the moderators 'slow as molasses', as their
job is an unpaid volunteer activity. I'm sure they have their
own lives to deal with too.

First, it's my news service that's slow, not the moderators --
apparently I received replies yesterday, though I still can't see
them through my usual means. Second, I wasn't knocking the
moderators and I know they have lives. I only made that
comment as a preemptive strike against the "ask in comp.std.c++"
replies that I was expecting.

 

[=?ISO-8859-15?Q?Juli=E1n?= Albo]

First, it's my news service that's slow, not the moderators --
apparently I received replies yesterday, though I still can't see
them through my usual means. Second, I wasn't knocking the
moderators and I know they have lives. I only made that
comment as a preemptive strike against the "ask in comp.std.c++"
replies that I was expecting.

May I ask what is the urgence in that type of question?

 

[Derek]

May I ask what is the urgence in that type of question?

There was no urgency except my own curiosity.