On Fri, 2014-01-03 at 14:56 -0800, Dirk Hohndel wrote:
> On Fri, 2014-01-03 at 13:34 -0800, Linus Torvalds wrote:
> > On Fri, Jan 3, 2014 at 1:20 PM, Dirk Hohndel wrote:
> > >
> > > I need to build old versions of Subsurface to find one that still works
> > > so I can get a correct download...
> >
On Fri, 2014-01-03 at 13:34 -0800, Linus Torvalds wrote:
> On Fri, Jan 3, 2014 at 1:20 PM, Dirk Hohndel wrote:
> >
> > I need to build old versions of Subsurface to find one that still works
> > so I can get a correct download...
>
> Hmm. My last Uemis download was last June. And that was with th
On Fri, Jan 3, 2014 at 1:20 PM, Dirk Hohndel wrote:
>
> I need to build old versions of Subsurface to find one that still works
> so I can get a correct download...
Hmm. My last Uemis download was last June. And that was with the gtk
version, because back then the Uemis code wasn't hooked up at a
On sexta-feira, 3 de janeiro de 2014 23:18:35, Lubomir I. Ivanov wrote:
> TMK, the situation on OSX is a bit "complicated" and i would suggest
> against C++11, for the time being.
That may be. Let me do a bit of investigation when I'm back in Portland and
I'll post my conclusions.
--
Thiago Mac
On Fri, 2014-01-03 at 13:14 -0800, Linus Torvalds wrote:
> On Fri, Jan 3, 2014 at 1:06 PM, Benjamin wrote:
> >
> > Diving in hot springs? Dive computer left in the kettle by mistake?
>
> I wish. My last dive at Yellow House was 8°C water.
Yeah, pretty much, IIRC we in reality had 9°C at the surf
On 3 January 2014 22:41, Thiago Macieira wrote:
> On sexta-feira, 3 de janeiro de 2014 11:37:01, Dirk Hohndel wrote:
>> So are we going to switch to C++11? I'm open to it.
>>
>> Thiago, are you sending patches?
>
> I need to investigate a bit more what might break. Especially on Mac, there
> are s
On Fri, Jan 3, 2014 at 1:06 PM, Benjamin wrote:
>
> Diving in hot springs? Dive computer left in the kettle by mistake?
I wish. My last dive at Yellow House was 8°C water.
But maybe it's global warming. Run for the hills.
Linus
___
subsu
On Fri, 2014-01-03 at 23:11 +0200, Lubomir I. Ivanov wrote:
> On 3 January 2014 14:49, Lubomir I. Ivanov wrote:
> > From: "Lubomir I. Ivanov"
> >
> > When a dive contains no cylinders, clicking the '+'
> > button could SIGSEGV if current_dive->dc.model is NULL.
> >
> > Signed-off-by: Lubomir I. I
On Fri, 2014-01-03 at 18:41 -0200, Thiago Macieira wrote:
> On sexta-feira, 3 de janeiro de 2014 11:37:01, Dirk Hohndel wrote:
> > So are we going to switch to C++11? I'm open to it.
> >
> > Thiago, are you sending patches?
>
> I need to investigate a bit more what might break. Especially on Mac,
On 3 January 2014 14:49, Lubomir I. Ivanov wrote:
> From: "Lubomir I. Ivanov"
>
> When a dive contains no cylinders, clicking the '+'
> button could SIGSEGV if current_dive->dc.model is NULL.
>
> Signed-off-by: Lubomir I. Ivanov
> ---
>
> newEditMode=MainTab::NONE) at qt-ui/maintab.cpp:225
>
Diving in hot springs? Dive computer left in the kettle by mistake?
On 3 January 2014 22:35, Dirk Hohndel wrote:
> Haven't bisected this, yet, but I'm rather suspicious of these data:
>
>
>
> /D
>
> ___
> subsurface mailing list
> subsurface@hohndel.
On sexta-feira, 3 de janeiro de 2014 11:37:01, Dirk Hohndel wrote:
> So are we going to switch to C++11? I'm open to it.
>
> Thiago, are you sending patches?
I need to investigate a bit more what might break. Especially on Mac, there
are some considerations due to the way that GCC and Clang are
Haven't bisected this, yet, but I'm rather suspicious of these data:
/D
<>___
subsurface mailing list
subsurface@hohndel.org
http://lists.hohndel.org/cgi-bin/mailman/listinfo/subsurface
On sexta-feira, 3 de janeiro de 2014 11:37:01, Dirk Hohndel wrote:
> > > Yeah, and that totally destroys the point of using "pressure_t" in the
> > > first place (it's there purely for type safety, exactly so that you
> > > can *not* use a plain integer by mistake).
> >
> >
> >
> > Make the constr
"Broken Hacks".
Mentioned, HÁ.
*runs*
On Fri, Jan 3, 2014 at 5:48 PM, Linus Torvalds <
torva...@linux-foundation.org> wrote:
> On Fri, Jan 3, 2014 at 11:31 AM, Thiago Macieira
> wrote:
> >
> > Make the constructor explicit by adding the explicit keyword.
>
> Oh wow. This is classic "No, no, C++
On Fri, Jan 3, 2014 at 11:31 AM, Thiago Macieira wrote:
>
> Make the constructor explicit by adding the explicit keyword.
Oh wow. This is classic "No, no, C++ is fine, you can add more hacks
on top of things".
It's crap, I tell you.
The C++11 syntax makes sense. Why? Because it's the sane one.
On sexta-feira, 3 de janeiro de 2014 10:33:43, Thiago Macieira wrote:
> On sexta-feira, 3 de janeiro de 2014 10:31:40, Thiago Macieira wrote:
> > On quinta-feira, 2 de janeiro de 2014 21:36:01, Linus Torvalds wrote:
> > > This seems to fix it from quick testing. I'd love to make the "mbar"
> > > be
On sexta-feira, 3 de janeiro de 2014 09:19:09, Linus Torvalds wrote:
> C++11 certainly has some improvements. This is likely the only one
> that we'd notice, though. The rest are for things I don't think we'd
> care about in subsurface (template improvements, threading memory
> models etc).
There
On Fri, 2014-01-03 at 17:31 -0200, Thiago Macieira wrote:
> On sexta-feira, 3 de janeiro de 2014 08:54:32, Linus Torvalds wrote:
> > On Fri, Jan 3, 2014 at 4:31 AM, Thiago Macieira wrote:
> > > If you want to pass a default argument of type pressure_t, then that
> > > struct
> > > requires a const
On Fri, 2014-01-03 at 11:23 -0800, Linus Torvalds wrote:
> On Fri, Jan 3, 2014 at 11:04 AM, Dirk Hohndel wrote:
> >
> > If no average depth is given we should create standard rectangular
> > profile or a rectangular profile with safety stop (assuming the dive was
> > long and deep enough for that
On sexta-feira, 3 de janeiro de 2014 08:54:32, Linus Torvalds wrote:
> On Fri, Jan 3, 2014 at 4:31 AM, Thiago Macieira wrote:
> > If you want to pass a default argument of type pressure_t, then that
> > struct
> > requires a constructor for it to work in C++98.
>
> Yeah, and that totally destroys
On Fri, 2014-01-03 at 09:51 -0800, Linus Torvalds wrote:
> On Fri, Jan 3, 2014 at 9:40 AM, Dirk Hohndel wrote:
> >
> > side note: I am amused by how much energy we are spending on figuring
> > out whether we should correctly parse the local translation of an
> > antiquated unit of weight used by r
On Fri, Jan 3, 2014 at 11:04 AM, Dirk Hohndel wrote:
>
> If no average depth is given we should create standard rectangular
> profile or a rectangular profile with safety stop (assuming the dive was
> long and deep enough for that to be reasonable).
Try it. It looks ridiculous for a 145ft dive.
On Fri, 2014-01-03 at 09:41 -0800, Linus Torvalds wrote:
> On Fri, Jan 3, 2014 at 9:27 AM, Dirk Hohndel wrote:
> >
> > What the WHAT?
> > I didn't realize what that meant for people who used to track just max
> > depth (as you do with paper logbook and some phone apps). The profiles
> > look posit
from a Qt perspective, the hability to connect signals with lambdas is
pretty userfull too.
On Fri, Jan 3, 2014 at 4:07 PM, Lutz Vieweg wrote:
> On 01/03/2014 06:04 PM, Dirk Hohndel wrote:
>
>> If it makes our code better, I'm all for it.
>>
>
> If nothing else, it at least makes it less probab
On 01/03/2014 06:04 PM, Dirk Hohndel wrote:
If it makes our code better, I'm all for it.
If nothing else, it at least makes it less probable that code
is (accidently) added that would be incompatible with contemporary C++.
And it's not unlikely 3rd-party code (like Qt) will at some point
requi
On 3 January 2014 19:36, Linus Torvalds wrote:
> On Fri, Jan 3, 2014 at 9:08 AM, Dirk Hohndel wrote:
>>
>>> In particular, remember: if the string is some translation of the
>>> standard unit, matching _that_ doesn't even *matter*, because it will
>>> be matched by default by simply matching the
On Fri, Jan 3, 2014 at 9:40 AM, Dirk Hohndel wrote:
>
> side note: I am amused by how much energy we are spending on figuring
> out whether we should correctly parse the local translation of an
> antiquated unit of weight used by roughly one country in the world...
Hey, it's US and Liberia! *And*
On Fri, Jan 3, 2014 at 9:27 AM, Dirk Hohndel wrote:
>
> What the WHAT?
> I didn't realize what that meant for people who used to track just max
> depth (as you do with paper logbook and some phone apps). The profiles
> look positively ridiculous.
I didn't check it much, but it worked for some ran
On Fri, 2014-01-03 at 09:36 -0800, Linus Torvalds wrote:
> On Fri, Jan 3, 2014 at 9:08 AM, Dirk Hohndel wrote:
> >
> >> In particular, remember: if the string is some translation of the
> >> standard unit, matching _that_ doesn't even *matter*, because it will
> >> be matched by default by simply
On Fri, Jan 3, 2014 at 9:08 AM, Dirk Hohndel wrote:
>
>> In particular, remember: if the string is some translation of the
>> standard unit, matching _that_ doesn't even *matter*, because it will
>> be matched by default by simply matching the locale setting (which
>> will always match the transla
I think the new fake profile is very nice IFF there is an average depth
in the entry that we get. But I think it gets a bit of "squirrel on
crack" when there is no average depth:
/*
* We want to fake the profile so that the average
* depth ends up correct. However, in th
On Fri, Jan 3, 2014 at 9:04 AM, Dirk Hohndel wrote:
>
> If it makes our code better, I'm all for it.
C++11 certainly has some improvements. This is likely the only one
that we'd notice, though. The rest are for things I don't think we'd
care about in subsurface (template improvements, threading m
On Fri, 2014-01-03 at 09:03 -0800, Linus Torvalds wrote:
> On Fri, Jan 3, 2014 at 4:49 AM, Lubomir I. Ivanov wrote:
> > From: "Lubomir I. Ivanov"
> >
> > Post c49d3885f this patch should make the input field
> > in the equipment tab for weight to accept the user
> > translated strings.
>
> No, t
On Fri, 2014-01-03 at 08:54 -0800, Linus Torvalds wrote:
> > Turning C++11 on for Subsurface would destroy support for OS X less than
> > 10.6
> > since only GCC 4.2 can compile for it.
That seems to be wrong - I compiled the latest binaries with clang and
they at least claim to run on 10.5. And
On Fri, Jan 3, 2014 at 4:49 AM, Lubomir I. Ivanov wrote:
> From: "Lubomir I. Ivanov"
>
> Post c49d3885f this patch should make the input field
> in the equipment tab for weight to accept the user
> translated strings.
No, this is broken, for a few reasons.
> - if (!strncmp(end, "kg", 2))
On Fri, Jan 3, 2014 at 4:31 AM, Thiago Macieira wrote:
>
> If you want to pass a default argument of type pressure_t, then that struct
> requires a constructor for it to work in C++98.
Yeah, and that totally destroys the point of using "pressure_t" in the
first place (it's there purely for type s
On 3 January 2014 06:58, Linus Torvalds wrote:
>
> +double string_to_grams(char *str)
> +{
...
> + if (!strncmp(end, "kg", 2))
> + goto kg;
> + if (!strncmp(end, "lbs", 3))
i think this part should be translated, so i sent a patch for it:
[PATCH 1/2] Models: translate th
From: "Lubomir I. Ivanov"
When a dive contains no cylinders, clicking the '+'
button could SIGSEGV if current_dive->dc.model is NULL.
Signed-off-by: Lubomir I. Ivanov
---
newEditMode=MainTab::NONE) at qt-ui/maintab.cpp:225
at qt-ui/maintab.cpp:546
...
---
qt-ui/maintab.cpp | 1 +
1 fi
From: "Lubomir I. Ivanov"
Post c49d3885f this patch should make the input field
in the equipment tab for weight to accept the user
translated strings.
e.g.
'lbs' in Russian is '' and '??' in Bulgarian,
while 'kg' is the same for both languages - '??'.
Signed-off-by: Lubomir I. Ivanov
On sexta-feira, 3 de janeiro de 2014 10:31:40, Thiago Macieira wrote:
> On quinta-feira, 2 de janeiro de 2014 21:36:01, Linus Torvalds wrote:
> > This seems to fix it from quick testing. I'd love to make the "mbar"
> > be a proper "pressure_t" rather than an unsigned int, but setting
> > default va
On quinta-feira, 2 de janeiro de 2014 21:36:01, Linus Torvalds wrote:
> This seems to fix it from quick testing. I'd love to make the "mbar"
> be a proper "pressure_t" rather than an unsigned int, but setting
> default values for arguments with complex types is a fairly recent
> thing, so compilers
I just pushed the 4.0.1 tag
Insomnia can be really productive. Let's hope I got the website edits
right and that the Windows and Mac binaries I uploaded actually
work. :-)
I'm waiting with posting the announcement for a few hours to give the
folks translating the website a chance to add a locali
43 matches
Mail list logo
