logs archiveBotHelp.net / Freenode / #ada / 2015 / July / 16 / 1
nerdboy
don't suppose anyone here would admit to touching iOS dev stuff?
Visaoni
I'd admit to thinking about it... and then crying myself to sleep in shame
although seriously, as far as I can tell all that stuff just seems like a huge, ugly pain
Natacha
I found objective-C to be an interesting take on "OOPing the C" (much more human-friendly than C++), but I never wandered close enough to anything Apple-y to touch it with a 10-foot pole
(except for apple pies, because I love apple pies)
Visaoni
yeah, I've mostly looked at the Android side of things. I don't think I can get past the Java bit though
Natacha
I wouldn't mind trying to write Android stuff with a GNAT that targets android JVM
Visaoni
that would be significantly more tolerable
Knaldgas
dbotton: I believe I found a bug in your Gnoga Snake-demo: X_Range and Y_Range is one too long.
Sometimes "Food" is placed off-screen
Lines 57 and 58 in snake-connection.adb
OCTAGRAM
Natacha: SOM is a better attempt
marino_
anyone here have openbsd access right now?
i have a small fortran test program that I'd like to see the output of
small as in < 10 lines
Natacha
I might have someone fitting that description in my IRC contacts
marino_
Natacha: okay, here's the program: http://leaf.dragonflybsd.org/~marino/hello.f90
we brought in openbsd libm into dragonfly and there were a couple of issues we found
one of them is acosh for complex numbers
when I run that program on FreeBSD, acosh result has signs reversed but asinh and atanh are okay
Natacha
I have a volonteer, who is currently trying to figure out how to run fortran
marino_
yeah, he'll need to link with -lm too
something like gfortran hello.f90 -lm -o hello
Natacha
https://gist.github.com/kAworu/dc9763ab92d9e198c4d7
seems gfortran 4.2 doesn't like the script
marino_
ah, there's a gcc48 from ports I think
Natacha
(my FreeBSD gfortran48 compiles but then I have runtime issues with gcc version)
marino_
(Action) is trying to avoid converting program to c
actually isn't there a f2c converter in ports? :)
Natacha
my volonteers don't seem keen on installing gcc 4.8 :-(
marino_
Natacha: it's okay. i am surprised they don't already have it
iow, who would think they'd want to run gcc 4.2.1 ?
Lucretia
nerdboy: I have
briefly
Natacha
10:17 < semarie> acosh z= ( -1.15012681 , 2.25629473 )
10:17 < semarie> asinh z= ( -1.15012681 ,-0.685498416 )
10:17 < semarie> atanh z= (-0.381870180 , -1.07198477 )
marino_
Natacha: so openbsd has the same bug
Natacha: acosh should be (1.15.., -2.25..)
Natacha
using python's complex math in FreeBSD, I do get the opposite acosh
though it seems fortran's -1.1_4 is not -1.14 but -1.1
marino_
isn't it -1.1000 ?
my fortran is rusty; i could be wrong
Natacha
I have no idea, I know exactly nothing in fortran
I just read -1.1_4 the Ada way without thinking
marino_
Natacha: freebsd's libm is doing the right thing, we found this as a regression
Natacha: did you tell semarie ? :)
Natacha
yes, and he asked whether I had a link to the bug report in dragonfly
marino_
no bug report, i'm working directly with libm maintainer
Lucretia
anyone know how to add more switches in a gpr file?
it's not liking for Default_Switches ("C") use Default_Switches ("C") & ("-O0", "-g");
got it
darkestkhan
Lucretia: take a look at https://github.com/darkestkhan/cbap/blob/master/cbap.gpr
Lucretia
I did it like this : C_Switches := (".."); and then for Default_Switches ("C") use C_Switches & ("-O0", "-g");
thanks
darkestkhan
I generally assign 'sets' of switches into separate variables and then concat variables
basically the same solution I use ;)
Lucretia
cool
thought I'd better sort out this sdl_build.gpr to handle static/shared and debug/release properly
darkestkhan
Lucretia: I'm going to do the same soon with cbap [release, debug, coverage]
but it is basically the same thing I did with CPU switches in that gpr file
Lucretia
so, what are people using for api doc generation for Ada 2012?
(Action) would rather not use sphinx due to the python dep
and installing locally is a bi**h
gentoo doesn't have support for it
and i'm not building my own python
darkestkhan
Lucretia: I'm using .ads files
Lucretia
??
darkestkhan
spec file is APIs documentation, no doc gen involved
honestly, automatic docgen from .ads files is constantly on my TODO list
*honestly speaking
Lucretia
well, i'm asking what people are using for generation of api/html docs
I want to generate it from the sources
might look at updating adabrowse
although there's a git of it on codelabs
kensan: ??
darkestkhan
I prolly should just go and do it at last...
Natacha
I wouldn't do that before I understand the value of such docs over direct read of .ads files
the point of doxygen is still a mystery to me :-(
high-level docs are extremely valuable, but why paraphrasing the .ads instead of reading it directly?
maybe I'm just too young and immature to understand
darkestkhan
Natacha: being able to just click on link to be directed to the definition of type would be nice ... but just like you said - I don't find it particularly necessary
Natacha
(Action) is clicophobic
Lucretia
Natacha: I want people to go to a website, be interested enough to look at the api docs there without having to download the source first.
(Action) is surprised nobody has updated adabrowse to support 2005 and 2012 yet
(Action) has to go to *really* sh*tty work
darkestkhan
SpinTensor: so how is your tensor? is it up or down?
[bad jokes are bad jokes]
specing
Lucretia: must be really Lucrative work to keep doing it
pnoffke
Is Data_Error supposed to be raised when using 'Read on a stream and 'Valid would return false? Or do I need to explicitly check 'Valid myself after calling 'Read?
I get a Constraint_Error if I try to 'Read a 32-bit fixed-point type where the stream data is outside the type's 'Range, but for an 8-bit value (enum, with two values and Size => 8), if I put 16#FF# in the stream, 'Read is happy but 'Valid returns false.
sparre
pnoffke: Have you checked the LRM? I'm pretty sure you'll get an exception if you receive an invalid value, but I can't remember which.
pnoffke
sparre: The only thing I can find is the Ada.Streams.Stream_IO spec, and Data_Error is only just listed there. I don't see anything about when it should be raised. Anyway, I'm using Ada.Streams, not Ada.Streams.Stream_IO. I'm mapping a stream to a memory buffer for a communications protocol.
In my case, 'Read does not raise an exception for the enum.
sparre
Have you checked the description of the 'Read attribute?
nerdboy
moin
pnoffke
sparre: I don't see anything specified in the RM for raising Constraint_Error on scalar types. So I just need to explicitly check 'Valid myself, which is fine.
darkestkhan
pnoffke: or you could override 'Read with your own implementation of it that does check it...
dbotton
Knaldgas thanks!
darkestkhan
!last
« prev 1 2 next »