|
Boost : |
From: Andrei Alexandrescu (andrewalex_at_[hidden])
Date: 2002-10-25 12:13:29
"David Abrahams" <dave_at_[hidden]> wrote in message
news:uy98mo7ii.fsf_at_boost-consulting.com...
> "Andrei Alexandrescu" <andrewalex_at_[hidden]> writes:
>
> > I don't claim I'm a guru, but I am trying to drive this issue to the
max. I
> > am mildly surprised that some people pretend that that's not a problem.
As
> > the many recent posts on the subject prove it: the issue can't be tuned
out
> > of existence.
>
> The many recent posts, up until John Maddock finally posted some hard
> data about the library's real impact on program size, proved
> nothing. Many were full of hyperbole and lacked rigor; some people
> were sketching redesigns of the regex library before even knowing if
> there was really a problem. I thought it was important to counteract
> alarmist FUD before people started posting links to the thread in
> support of the argument that most Boost libraries are bloated
> behemoths whose designers ignore issues of code size. If you're trying
> to imply I was pretending program size is not a serious issue, please
> take it back.
There's no reason to be agitated. We're all in the same boat, and working
towards the same goals. There's also nothing that needs to be taken back.
There was also no alarmist FUD posted (by me at least).
Why all the defensiveness? I think the one thing that came out is that some
people do prefer a shared library to static linking, and that 500 KB is too
large a level of granularity. I understand there's a ton of functionality in
there for wide and narrow chars, locales, and all that. It would be nice,
then, to break the library in several subsets (for example, char and
wchar_t). This way, people would include either or both.
Andrei
-- All new! THE C++ Seminar: Oct. 28-30 in Vancouver, WA. http://www.thecppseminar.com/
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk