What happened to the profile?

[Tomaz Canabrava]

On Mon, Oct 7, 2013 at 11:34 AM, Dirk Hohndel <dirk at hohndel.org> wrote:

> On Mon, 2013-10-07 at 11:44 +0200, Robert Helling wrote:
> > On 07.10.2013, at 10:22, Linus Torvalds <torvalds at linux-foundation.org>
> wrote:
> >
> > Linux,
>
> Hehehe :-)
>
> > > On Mon, Oct 7, 2013 at 12:23 AM, Robert Helling <helling at lmu.de>
> wrote:
> > >>
> > >>
> https://www.dropbox.com/s/3fat0kws2isqzkc/Screenshot%202013-10-07%2009.18.23.png
> > >>
> > >> Below is a profile with the binary I compiled on Friday, above the
> current master. Of course both are showing the same file. Is this
> intentional?
> > >
> > > Looks like something rounds to integer values. It's probably the glib
> > > removal code: I suspect ascii_strtod() is buggered. It uses "atof()",
> > > and likely re-introduced the bug that caused us to use ascii_strtod()
> > > in the first place. I'm guessing you have a german locale, and thus
> > > have the decimal point set to ',' and then atof() parses "3.5" as "3".
> > >
> > > That "ascii_strtod()" function really looks completely broken. It
> > > seems to be missing the whole "ascii" point, and instead does
> > > pointless things like "NaN" and "infinity" parsing. I'm guessing Dirk
> > > copied it from some odd source..
> >
> > indeed, that is the source of the problem. Here is a patch/hack that
> fixes it:
>
> I don't think it's a hack... it seems like what I should have done, had
> I spent the time to think through the code that I copied from Ruby. The
> code that I copied was posted in response to the locale confusion, but
> clearly is not addressing that at all, now that I read it more
> carefully.
>
> Oops.
>
> > It replaces atof by atof_l which takes an additional locale argument
> > where I force feed it en_US. If this is considered to be the way to go,
> > this locale for numeric values should be allocated once and for all in
> > some startup code (where?) and not every time we have to parse a number
> > in an xml file as it is now.
>
> That seems like a reasonable approach.
>
> Except that atof_l doesn't appear to exist on Linux. Some googling seems
> to imply that this is a BSD-ism that hasn't made its way to Linux. So
> back to square 1...
>
> Thiago - I'm sure there is this brilliant easy to use Qt function that
> will solve all our problems, correct?
>

>From the Qt docs, on QString.toDouble()

This function tries to interpret the string according to the current
locale. The current locale is determined from the system at application
startup and can be changed by calling
QLocale::setDefault<http://harmattan-dev.nokia.com/docs/library/html/qt4/qlocale.html#setDefault>().
If the string cannot be interpreted according to the current locale, this
function falls back on the "C" locale.


> /D
>
>
> _______________________________________________
> subsurface mailing list
> subsurface at hohndel.org
> http://lists.hohndel.org/cgi-bin/mailman/listinfo/subsurface
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.hohndel.org/pipermail/subsurface/attachments/20131007/89d12d5d/attachment.html>