aserio changed the topic of #ste||ar to: STE||AR: Systems Technology, Emergent Parallelism, and Algorithm Research | stellar.cct.lsu.edu | HPX: A cure for performance impaired parallel applications | github.com/STEllAR-GROUP/hpx | Buildbot: http://rostam.cct.lsu.edu/ | Log: http://irclog.cct.lsu.edu/
<aserio>
so the "get_shared_state" function is what does the trick
<aserio>
I assume it returns a pointer to the server
<K-ballo>
no, it doesn't
<aserio>
damn
<K-ballo>
it returns a pointer to a raw shared state that will eventually hold the id of the component
<aserio>
ah, so the shared state of a future?
<K-ballo>
yes, a future<id_type>
<K-ballo>
*yes, a future of id_type type
<K-ballo>
and that id will "point to" the component at the remote server
<aserio>
Cool!
<aserio>
Thanks K-ballo!
<heller>
patg[[w]]: Charm++ is in close doe proximity as well.
<patg[[w]]>
heller: hi how are things going with you? Dissertation etc.
<heller>
aserio: new<client>(...) does everything you need
<patg[[w]]>
heller: at least they have C++ on their radar
<heller>
patg[[w]]: slow but steady...
<heller>
Urbana-champaign is the recruiting ground for Argonne, isn't it?
<patg[[w]]>
heller: please let me know when you are done I keep expecting that type of news hopefully soon
<patg[[w]]>
heller: I'm new to all this DOE stuff you may be correct
<heller>
patg[[w]]: me too... (Regarding the news)
<heller>
patg[[w]]: look at the geographics ;)
<patg[[w]]>
Anyway C++ is on the radar so someday they may just have to look at HPX
<zao>
Ignorance is bliss.
<patg[[w]]>
lol
<heller>
They don't have to
<heller>
Or well, they do already
<heller>
Here's the main reason that people keep repeating: the main problem of hpx is that it isn't developed by hpc people
<diehlpk_work>
heller, Why do the say that?
<heller>
And in the end, it doesn't really matter what it is called ;)
<zao>
It clearly can't be made by HPC people. HPC people think MPI is the best thing since sliced data.
<heller>
diehlpk_work: would you like to admit that some non mathematician came up with a better way to solve PDEs? (over exaggerating, of course)
<heller>
As in: how should those people know what we need?
<patg[[w]]>
Nice talking to you (and stirring up problems) got to go
<heller>
You have to recognize that mpi was codesigned, from the top, if what physicists/mathematicians needed, and from the bottom of what the vendors could offer
<heller>
s/if//
<heller>
And you won't find a lot of hpc people without a strong math, physics or eei background (for exactly that reason)
<heller>
Especially the "white beards" in control
<heller>
But that's just me ranting ;)
<diehlpk_work>
heller, Ok, I udnerstand what the problem is
<heller>
There's certainly a momentum at the moment towards AMTs
<heller>
But HPX is certainly not the only solution
<heller>
But then, there's more than HPC ;)
aserio has quit [Ping timeout: 252 seconds]
aserio has joined #ste||ar
<diehlpk_work>
Does anyone know when the OpenSuco paper will be published?
Smasher has joined #ste||ar
hkaiser has joined #ste||ar
patg[[w]] has quit [Quit: Leaving]
aserio has quit [Quit: aserio]
aserio has joined #ste||ar
hkaiser has quit [Quit: bye]
aserio has quit [Client Quit]
eschnett has quit [Quit: eschnett]
Smasher has quit [Remote host closed the connection]