is BOOST "synonymous with modern C++ development" ?

J

jaialai technology

The inflammatory comment that "not using BOOST means
you are not doing modern C++" was made to me
recently. Pretty inflammatory! But...is it
more true than not?
 
V

Victor Bazarov

jaialai said:
The inflammatory comment that "not using BOOST means
you are not doing modern C++" was made to me
recently. Pretty inflammatory! But...is it
more true than not?

For some reason some people make emphasis on using particular tools (or
components, libraries, etc.) as the sign of good development culture.
They aren't unrelated, of course (I'd be suspect of a development house
that wouldn't use a profiler or a debugger, for example). But tools and
components aren't the ultimate criterion on which creators of software
should be judged. I am speculating, but those Boostophiles might be
substituting appearances for achievements. Does it matter how (or with
what) my software is created as long as it is functioning properly and
competing on the market successfully?

Don't get me wrong. There are situations when the use of available
components can be ultimately a deal breaker, but the statement as you
posted sounds like an over-the-hill generalization. And as we know, all
generalizations are wrong.

V
 
B

Brian

The inflammatory comment that "not using BOOST means
you are not doing modern C++" was made to me
recently. Pretty inflammatory! But...is it
more true than not?

A rabbi friend sometimes says, "You have to chew the
meat and spit the bones." That's my advice with Boost.
Some of it is meat and some of it is bones. I think
a lot of the Boost container libraries are impressive.


Brian Wood
http://webEbenezer.net
 
J

James Kanze

I don't see how this could possibly be measured empirically.
The only thing that's clear, however, is that Boost is not a
mandatory requirement for "doing modern C++".
I earn a living "doing modern C++". By that, I mean using
pretty much all the libraries that make up the standard C++
library, including the localization library, the iterator
library, the numerics library, the algorithms library, and the
rest.
One thing I don't use, though, is Boost. My opinion of that
library, based on what I've seen, is not very favorable.

It depends on which elements---some are very good. (I don't use
it professionally, but that's because I've never really been
able to get it to build in a way that would integrate into our
development environment.)
It is very easy to disprove an absolute statement of the kind
you brought forward: only one counterexample is sufficient.
And, I'm it.

As I read it, the original statement was a definition (an
axiom). As such, it can't be proved or disproved---if you
accept the definition (I don't), then the fact that you don't
use Boost proves that you're not "doing modern C++".

If you don't have any real arguments, postulating an axiom is
the way to go. But in practice, axioms are only useful if there
is a consensus about them. In this case, that is definitely not
the case.
 
K

Kaz Kylheku

The inflammatory comment that "not using BOOST means
you are not doing modern C++" was made to me
recently. Pretty inflammatory! But...is it
more true than not?

It's only inflammatory because you are mistaking ``modern'' for one
or more of the following, or possibly others:

- robust
- correct
- maintainable
- flexible
- rapidly developed
- good

Stop doing that, and take some meds for any remaining inflammation:
ibuprofen, actetaminophen, aspirin ...
 
J

jamm

jaialai said:
The inflammatory comment that "not using BOOST means
you are not doing modern C++" was made to me
recently. Pretty inflammatory! But...is it
more true than not?

You often hear the claim about boost, or STL, that the authors are
'better/smarter then YOU' and that the code is 'better then anything you'd
write yourself'... Meanwhile bugs are consistently found and fixed. I don't
buy it.

If boost saves you time and your boss okays it, then use it.
 
J

James Kanze

jamm ha scritto:

[...]
That is often true. As far as I am concerned, I certainly
could not implement a vector class that beats the std::vector
provided by my compiler :)

Above all, you certainly can't write one which has been tested
as much.
They consistently find and fix bugs in the STL?

There are usually bug fixes from one release to the next. After
a few cycles, however, they really only concern very arcane
uses, which people hadn't thought to try in the earlier
versions.
 

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments. After that, you can post your question and our members will help you out.

Ask a Question

Members online

No members online now.

Forum statistics

Threads
473,744
Messages
2,569,484
Members
44,903
Latest member
orderPeak8CBDGummies

Latest Threads

Top