K-ballo 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/
hkaiser has joined #ste||ar
shahrzad has quit [Quit: Leaving]
hkaiser has quit [Quit: bye]
bita_ has joined #ste||ar
bita has quit [Ping timeout: 240 seconds]
bita_ has quit [Ping timeout: 260 seconds]
hkaiser has joined #ste||ar
akheir has joined #ste||ar
bita_ has joined #ste||ar
shahrzad has joined #ste||ar
<gonidelis[m]>
why did we replace `util::invoke()` ??
<K-ballo>
internally you mean?
<gonidelis[m]>
yes
<K-ballo>
util::invoke() is still there to be used
<K-ballo>
the macro creates one less call frame, the better to debug it, and is slightly lighter at compile time
<gonidelis[m]>
huh? why does the compiler hit me then?
<gonidelis[m]>
sec
<K-ballo>
we have not replaced util::invoke, we've replaced some calls in internal implementations
<gonidelis[m]>
do i need to include `functional/invoke.hpp` or `functional/detail/invoke.hpp` in order to invoke the macro?
<K-ballo>
detail for the macro
<gonidelis[m]>
hm... sounds like time to rebase
<K-ballo>
ideally you would rebase early or so, but why does it matter here?
<K-ballo>
how did the compiler hit you into using the macro?
<gonidelis[m]>
the compiler hit me on using `util::invoke()` from a PR that is open since september
<K-ballo>
doesn't sound like something that would be related to the macro
<gonidelis[m]>
it's not
<gonidelis[m]>
it's related to `util::invoke()`
<K-ballo>
ok, *how* is the compiler hitting you on `util::invoke()`?
<gonidelis[m]>
sec
<gonidelis[m]>
libs/parallelism/algorithms/include/hpx/parallel/util/transform_loop.hpp:126:46: error: no matching function for call to `xinvoke(hpx::parallel
<K-ballo>
well then, it's not a macro thing, it's not an include thing
<K-ballo>
sounds like `f` isn't callable with `first`
<K-ballo>
the real compiler message would say
khuck has joined #ste||ar
<gonidelis[m]>
btw any good suggestion on how to read compiler output when on ssh
<gonidelis[m]>
?
<khuck>
is Hartmut completely offline due to the storm?
<K-ballo>
I just copy the whole history and read it locally
<gonidelis[m]>
khuck: he was active a few hours back
<gonidelis[m]>
K-ballo: how do you transfer the text ;p ?
<K-ballo>
uhm right click, copy all to clipboard
<gonidelis[m]>
K-ballo: lol...
<gonidelis[m]>
I mean it's terminal
<gonidelis[m]>
I tried redirecting output to a file
<K-ballo>
the tool I use has a "copy all to clipboard" terminal option included
<K-ballo>
I don't redirect to any files, I let it go to the terminal, then copy all to clipboard
<gonidelis[m]>
tool?
<K-ballo>
I happen to be using putty here
<K-ballo>
traditional terminals have select all / copy options too
<zao>
If you have clipboards sufficiently forwarded, you can pipe into xsel or xclip, or you could use a command line tool to make a private GitHub gist.
<zao>
Or if you trust VS Code, use its remote SSH connectivity to browse files.
<zao>
If you want to see interactive output from a program but also store stdout to file, there's the "tee" command.
<gonidelis[m]>
zao: I am using the ssh through vs code
<zao>
So you could do something like "blargh 2>&1 | tee persistent.log | ..."
<K-ballo>
now that you mention it, I've used scp when dealing with some header tools that would output hundreds of mb of output too
<gonidelis[m]>
but the error lines might be a lot and vs refuses to open the file
<gonidelis[m]>
K-ballo: now we are talking
<K-ballo>
come on, template error messages are bad, but not hundreds of mb of output bad
<gonidelis[m]>
might that be the case
<zao>
If it was local, you could use the terminal in VS Code and pipe into 'code -' which would open the output as a file in the editor, but not sure if that also works remotely.
<K-ballo>
it's not
<gonidelis[m]>
but my errors are not just template errors
<zao>
Shouldn't you have some build system integration in Code anyway?
<gonidelis[m]>
they are rookie tamplate errors
<zao>
Still proud over the 500 MiB of template instantiation error I got out of Spirit Classic back in the early 00s.