From: Björn Karlsson (bjorn.karlsson_at_[hidden])
Date: 2001-11-27 03:21:13
I recently posted a question about "real-world" use of Boost.
The (surprisingly few) answers I received can be found at the end of the
I think it is safe to say that the usage is more widely spread than the
[Note: I have excluded the names from the e-mails, although not explicitly
asked to do so.]
There were replies from 8 Boosters. It seems that the lightweight utility
classes are the most popular, likely due to the fact that their impact on
existing systems is easily manageable.
Thank you for your answers!
Using some small parts of it.
Selected because it lets you pick and chose things to use from a large set
of (relatively) small components/tools rather than ending up "buying into"
some grand scheme of things. Well designed small interfaces rather than
including the kitchen sink, making it easier for people to pick up/maintain
code later.... Did what I needed to do at the time.... Note that the things
I have elected to use so far have been small and intended to reduce the risk
of mistakes (eg operators, traits) rather than anything vast.
We have just started to use Boost for converting a Python prototype to C++.
The experience is still limited, but the results after a few days of work
are very good.
I'm using Boost heavily in FEMTown (www.femtown.org). FEMTown is an
open-source finite-element framework in development.
We're using Boost because it provides stable and slim functionality that
helps us making FEMTown more efficient, flexible and robust.
(although our choice of C++ for a scientific application is even kinda
At this moment, I'm using the Boost lexical_cast in one of our commercial
software products (Pilot - see
I needed to convert from integers to text for transmission across the
network, then back from text to integer value. I had been using strstream,
but the lexical_cast wrapped all the complexity up for me. I chose Boost
because it's portable (our software runs on Windows NT and on embedded
software using gcc). I knew there is a good chance that lexical_cast will
become a standard keyword.
As I learn more about the various libraries that are available, I will
likely find appropriate places to use them.
My group is currently using Boost on one project as an experiment. We do
cross-platform Macintosh/Windows development and it was valuable to have a
tested library that worked on our two main development platforms.
We have been phasing different parts of Boost in as come across problems
that Boost will solve. As a community I find Boost a source of best
practices that I have tried to implement in our workplace.
We've been using Boost at Zetec for over two years. Initially just
for shared_ptr, but since expanding to lexical_cast, iterator_adaptors,
tokenizer, and others. If you need any other info, let me know.
I am starting a new project using boost. It is a
financial risk-management engine. I cannot tell you a
lot about it yet. We are a team of 8 developers, 4
developing the backend and 4 developing the
The product will hopefully be a big success in the
financial world as it is developed in cooperation with
some the best people of the financial industry.
It is a 12 months development effort using the XP
development process. boost will be used to provide
some utility classes (smart pointers, anys, threads
... ). The engine will run on Solaris and Linux. A
future version on Windows may exist in the future.
We've used small parts of Boost in a digital video distribution back-end
the system was developed by a client of us(us=CMG). If you want names in
I'd have to ask for permissions from them.
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk