|
Boost : |
From: Terje Slettebø (tslettebo_at_[hidden])
Date: 2002-09-09 18:50:05
>From: "David B. Held" <dheld_at_[hidden]>
> "Terje Slettebø" <tslettebo_at_[hidden]> wrote in message
> news:020e01c2584e$01151f10$60fb5dd5_at_pc...
> > [...]
> > > You also have to forbid leading spaces after tabs (for instance, to
> indent
> > > a partial tab...6 spaces, on a 4-space tab).
> >
> > No, that's ok, as long as the space is used for alignment, not to denote
> > indentation level.
>
> That's a maintenance nightmare.
>
> > [...]
> > Here's a solution ("*" = space):
> >
> > class foo
> > {
> > <tab>*************foo() { }
> > <tab>void********bar() { }
> > };
> > [...]
>
> Maybe I'm not a perfect coder, but during the course of editing code,
> I break lines in the middle and re-indent them, combine lines, and all
> other manner of combination.
I understand your point. Also, as Shannon Barber (who uses tabs and spaces)
pointed out in a mail, editors typically don't know about this way of doing
things, so if you are at a line and press enter, it will fill as many tabs
as it can, and the rest with spaces, if auto-indentation is used. Thus, you
may have to manually replace those tabs with spaces.
> It would be a major PITA to have to go
> and recheck every line to make sure it had the proper combination of
> leading tabs and spaces.
As has been pointed out in this thread, there are consequences for both ways
of doing things. I've just done a 100+ files project converted from using
spaces, to using tabs and spaces. I used "tabify" for each file, then
changed tab size, so any alignment showed up, as it became wrong, and fixed
that, by converting it to spaces. I only needed to do that a handful of
places. So for me, this was not a problem. The advantage is now that
everybody can use their preferred tab size for it.
For you, as you say, depending on how the code is, it might be a PITA to use
this convention, as it would mean a lot of checking and fixing. For others,
like my project, that isn't the case.
Regards,
Terje
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk