<div dir="ltr">On Tue, Sep 19, 2017 at 8:41 AM, Adrian Johnson <span dir="ltr"><<a href="mailto:ajohnson@redneon.com" target="_blank">ajohnson@redneon.com</a>></span> wrote:<br><div class="gmail_extra"><div class="gmail_quote"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><span class="">On 19/09/17 21:02, Matthias Clasen wrote:<br>
> On Tue, Sep 19, 2017 at 7:04 AM, Adrian Johnson <<a href="mailto:ajohnson@redneon.com">ajohnson@redneon.com</a><br>
</span><span class="">> <mailto:<a href="mailto:ajohnson@redneon.com">ajohnson@redneon.com</a>>> wrote:<br>
><br>
><br>
> On 18/09/17 01:03, Matthias Clasen wrote:<br>
> > On Sun, Sep 17, 2017 at 1:12 AM, Adrian Johnson <<a href="mailto:ajohnson@redneon.com">ajohnson@redneon.com</a> <mailto:<a href="mailto:ajohnson@redneon.com">ajohnson@redneon.com</a>><br>
</span><span class="">> > <mailto:<a href="mailto:ajohnson@redneon.com">ajohnson@redneon.com</a> <mailto:<a href="mailto:ajohnson@redneon.com">ajohnson@redneon.com</a>>><wbr>> wrote:<br>
> > I've attached a patch that refactors the float parsing code from<br>
> > cff-subset into a new function: _cairo_strtod(). So you can now use this<br>
> > for a C locale strtod.<br>
> ><br>
> ><br>
> > Yay, thanks.It might still be nice to apply the strtod_l patch on top,<br>
> > and only use that code as fallback.<br>
><br>
> -1. Just use the cairo_strtod all the time. We don't use non ANSI C<br>
> functions with #include guarded fallbacks in cairo. It adds to source<br>
> code bloat and one side of the #if won't get regularly tested.<br>
><br>
><br>
> I disagree, at least in this case, but your call.<br>
<br>
</span>We don't do it that way in cairo. If the fallback works just use it all<br>
the time.<span class=""><br></span></blockquote><div><br></div><div>Fine. Easy enough to drop that commit.</div><div> <br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><span class=""><br>
</span>I did some more testing. With my system fontconfig (2.11) the test<br>
failed (same result) when using the ttf font. FC 2.11 did not recognize<br>
the otf font.<br>
<br>
I tested fontconfig 2.12.5. The test crashed with both the ttf and otf<br>
fonts due to a divide by zero error.<br>
<br>
What version of FC are you using and which font (otf or ttf)?<br></blockquote><div><br></div><div>I have fontconfig 2.12.5 and freetype 2.8 on my system, and I am using the ttf version of the font.</div><div>Behdad, does my test work on your system ?<br></div><div><br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
Is there some printfs you could add to help debug this?<div></div></blockquote><div><br></div><div> I'll think about it. <br></div></div></div></div>