hkaiser changed the topic of #ste||ar to: STE||AR: Systems Technology, Emergent Parallelism, and Algorithm Research | stellar-group.org | HPX: A cure for performance impaired parallel applications | github.com/STEllAR-GROUP/hpx | This channel is logged: irclog.cct.lsu.edu
K-ballo has quit [Quit: K-ballo]
diehlpk has joined #ste||ar
ruchipakhle has joined #ste||ar
diehlpk has quit [Quit: Leaving.]
hkaiser has quit [Quit: Bye!]
ruchipakhle has quit [Quit: Client closed]
Yorlik has joined #ste||ar
Yorlik has quit [Read error: Connection reset by peer]
Yorlik has joined #ste||ar
K-ballo has joined #ste||ar
Yorlik_ has joined #ste||ar
Yorlik has quit [Ping timeout: 250 seconds]
hkaiser has joined #ste||ar
ruchipakhle has joined #ste||ar
ruchipakhle has quit [Quit: Client closed]
diehlpk has joined #ste||ar
diehlpk has quit [Client Quit]
hkaiser has quit [Quit: Bye!]
diehlpk has joined #ste||ar
diehlpk has quit [Client Quit]
diehlpk_work has joined #ste||ar
<gonidelis[m]> K-ballo: why is a sign extended char create undefined behavior when used in std::isspace? https://github.com/fenbf/cppstories-discussions/issues/82#issuecomment-1105851141
<K-ballo> > The behavior is undefined if the value of ch is not representable as unsigned char and is not equal to EOF.
<gonidelis[m]> hmm...but why?
<K-ballo> why? because anything bigger than a char can't possibly be a char
<K-ballo> as such, it doesn't even make sense to ask the question: is this absolutely-not-a-char a space char?
<gonidelis[m]> ...
<gonidelis[m]> how can a ch be not representable as unsigned char?
<gonidelis[m]> unsigned char is big enough to include all possible chars, no?
<K-ballo> a sign extended char won't fit in an unsigned char
<K-ballo> whether plain char is signed or unsigned is implementation defined
diehlpk_work has quit [Remote host closed the connection]
hkaiser has joined #ste||ar
<Yorlik_> hkaiser: YT?
Yorlik_ is now known as Yorlik
<hkaiser> hey Yorlik
<Yorlik> Hello!
<Yorlik> I think I need a little help with a linking error
<hkaiser> Yorlik: I'm in a meeting, could we talk a bit later?
<Yorlik> Sure.
<Yorlik> Just poke me - I'm around for quite a while.
<Yorlik> I guess it's just trivial issue with the macros.
<hkaiser> I hate those macros :/
<hkaiser> gnikunj[m]: I really need your solution for macro-less actions
<Yorlik> :D
<Yorlik> Macros are an invention from Lucifer the beast himself.
diehlpk_work has joined #ste||ar
diehlpk_work has quit [Ping timeout: 250 seconds]
<K-ballo> macro-less actions as in auto template params? or something more fundamental
<hkaiser> auto template params
ruchipakhle has joined #ste||ar
<K-ballo> I remember implementing that, way before hpx required 17
<hkaiser> K-ballo: do you still have that?
<K-ballo> probably, but won't be of any use
<K-ballo> -// Macros usable to refer to an action given the function to expose
<K-ballo> #define HPX_MAKE_ACTION(func) hpx::actions::make_action<&func>
<K-ballo> #define HPX_MAKE_DIRECT_ACTION(func) hpx::actions::make_direct_action<&func>
<hkaiser> nod
<K-ballo> should even work without the &
<hkaiser> we should be able to do async<&func>(targetid, ...)
<hkaiser> without explicitly defining the action
<K-ballo> implicit actions, nice
<K-ballo> interface-wise I'd personally prefer something like async(constant_fun<func>)
<hkaiser> yah, that would work as well
<K-ballo> or just async(constant<func>)
<hkaiser> or even async(constant_<func>(targetid), ...)
<hkaiser> would go with the rest nicely
<Yorlik> I kinda start liking the Unreal Engine approach of annotating with meaningless macros, but they require code parsing and custom tools.
<Yorlik> When done under the hood and it just works it's beautiful.
<Yorlik> E.G. You declare a class with a macro "UObject()" before its declaration and it automagically becomes serializable, garbage collected and whatnot.
<hkaiser> Yorlik: you need a separate tool for that
<Yorlik> Yes, thats what they do.
<Yorlik> UBT - Unreal build tool.
<hkaiser> let's not go there
<Yorlik> Not suggesting that. Not at all.
<K-ballo> well, since we're tied to cmake....
<Yorlik> It's just about the UX you want.
<K-ballo> no, not seriously
<Yorlik> Actually CMake together with vcpkg are my friends now.
<hkaiser> good
<Yorlik> Using manifest mode together with a custom vcpkg registry is really fun.
<Yorlik> E.g. I made an Unreal plugin, which automagically pulls our custom registry from git, drops some dependency declarations and the magic happens.
<Yorlik> I just have to copy some files afterwards.
diehlpk_work has joined #ste||ar
ruchipakhle has quit [Quit: Client closed]
<hkaiser> Yorlik: ok, I could talk now
<Yorlik> hkaiser: Great - I'm around
<hkaiser> so what's up?
<Yorlik> I'm getting this linker error
<Yorlik> Sec.
<Yorlik> This one: "unresolved external symbol hpx_exported_plugins_list_hpx_factory"
<Yorlik> hkaiser: Thanks a ton for the help: $<$<CONFIG:DEBUG>:HPX_DEBUG> :D
<hkaiser> :D, most welcome
<gnikunj[m]> hkaiser: my semester ends on 9th. Would it work if I get it done that week right after?
<gnikunj[m]> It should be trivial now that we require C++17 by default
<hkaiser> gnikunj[m]: sure, no pressure ;-)
diehlpk_work has quit [Remote host closed the connection]
hkaiser has quit [Quit: Bye!]
hkaiser has joined #ste||ar