From: Hendrik Schober (SpamTrap_at_[hidden])
Date: 2005-05-25 10:08:49
Jason Earl <Jason_at_[hidden]> wrote:
> You've probably come out with a worst case scenario
Have I? So what did I do wrong to get me into
this? I just downloaded the stuff and followed
the guide. It seems easy to hit the worst case.
> with regards to
> that attempt to compile things, and I do remember a few things being a
> pain initially when compiling for the first time. However, it didn't
> take long for me to make intelligent guesses on what to do to get
> things working, and it was only minutes before I managed to get things
> working smoothly when I last tried. This is certainly a lot less
> painless than a some of the builds I've done on Linux for web stuff.
> I think a lot of Windows users are too used to things just working out
> of the box, and give up if things don't work like that.
If wanted, I'd most likely find out, too. The
original argument was that it takes too much
effort to do this. I second that.
There's too many out there who simply won't spend
that effort (and who don't share your resentiments
against anything but shell typing). I repeat:
Do you want widespread use of boost or do you want
widespread use of boost among programmers who know
their way around with a shell?
> As for understanding what targets are etc, I think most developers will
> know what targets are. [...]
With thinking a few seconds I came up with three
overloaded meanings we use the term for here. I
still don't know what exactly targets are in the
boost build system. And frankly, don't think I
should need to read its documentation in order
to know. There's a lot I have to fit into one
day, it's already tight without reading it.
> Despite all the errors people might get with installing, Boost does
> have this excellent mailing list where people can quickly get the help
> they need.
Do you want all users come to the mailing list
and ask? (Then why have a build system at all?)
> Maybe it would be good to have some control over build verbosity (there
> probably already is some, I just haven't looked yet), Maybe logging the
> build info to a file would be better and then logging errors, messages
> etc to the screen (and maybe a separate) log file.
Yes, I thought of this myself. But the heck --
there's so many libs out there I will never
have the time to look at.
If I wouldn't know boost is more than just
another one of those, I would just forget
about it keep hacking.
> You really can't complain at long error messages.
Why not? (Where would we be hadn't we complained
all the time? <g> )
> After all C++
> programmers should be used to them, they happen all of the time with
> template based code.
True. My compiler smoetimes spits out 3-4k error
messages for templates, too. But it gives me the
convenience of easily navigating among them. (And
I still complain to my vendor that I want this
> On 25 May 2005, at 14:46, Hendrik Schober wrote:
-- SpamTrap_at_[hidden] is never read I'm Schobi at suespammers dot org "Coming back to where you started is not the same as never leaving" Terry Pratchett
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk