16:13 <GitHub53> [rubinius] chuckremes pushed 3 new commits to ffi-io: http://git.io/vcWvW

16:13 <GitHub53> rubinius/ffi-io d597e70 Chuck Remes: add logic to load and detect posix_fadvise constants and enable feature detection

16:13 <GitHub53> rubinius/ffi-io dd5d806 Chuck Remes: add with_feature detection to skip these specs on platforms without posix_fadvise

16:13 <GitHub53> rubinius/ffi-io 5c5905c Chuck Remes: convert usage of posix_fadvise from C++ to Ruby code

18:45 <GitHub56> [rubinius] chuckremes pushed 2 new commits to ffi-io: http://git.io/vcWb7

18:45 <GitHub56> rubinius/ffi-io 5097d35 Chuck Remes: add select function signature; may need to change when we learn more about implementation

18:45 <GitHub56> rubinius/ffi-io 9a852dd Chuck Remes: continue fixup of IO.select; handle coercion and choosing max FD

19:08 <RageLtMan> hmm, looks like the IO work we discussed a few years back is being done by chuckremes. thank you chuck

19:30 <cremes> RageLtMan: please join us on gitter. i’m pairing with tak1n now to work on that branch.

19:30 <cremes> the more the merrier!

19:31 <RageLtMan> the hell is gitter?

19:38 <pipework> It's a reinvention of a wheel because IRC is too hard for some people.

19:39 <pipework> I'm not fond of closed source solutions to perceived open source problems that are centralized.

19:41 <RageLtMan> cremes: i fixed metasploit a few years back to work on RBX, and currently working on unfucking the IO abstractions we have in Rex/Msf namespaces (UDP is a travesty there). I was hoping to do exactly what you're doing to overhaul how Rex handles sockets if RBX is in play

19:42 <cremes> RageLtMan: great, glad to hear it

19:42 <RageLtMan> i did actor abstractions for the threadfactory there, thought of using concurrent-ruby with libuv bindings, but brixen had some decent architectural points as to why that might not be a great idea

19:42 <cremes> pipework: sharing inline code and other stuff in gitter makes it a better tool for certain things than irc

19:42 <cremes> like pairing on some code

19:44 <pipework> cremes: I'm a fan of screenhero for pairing.

19:44 <pipework> And for sharing code, I'm a fan of gist

19:44 <pipework> And IRC has a documented protocol and has many clients, and is easy to develop clients and bots for.

19:44 <cremes> pipework: ok, that’s all good.

19:45 <pipework> Some clients even detect code and turn it into a gist or paste for you in the client without any other interaction.

19:45 <cremes> uncle! so gitter sucks. :)

19:45 <pipework> Egh, it's just preferential things.

19:45 <cremes> heh

19:45 <pipework> I just dislike how it's centralized and closed source more than anything.

19:46 <pipework> The other points could be fixed if the source was open.

19:46 <cremes> well, it doesn’t own me or anything i use it for. if they are dicks or the tool sucks, irc will always be here.

19:46 <cremes> but while it works as well as it does, i’ll use it

19:47 <pipework> Sounds pretty good and fair.

19:47 <pipework> I'd adopt something better that's open source, decentralized, and made extension and adoption as easy as possible.

19:48 <|jemc|> yeah, I tend to agree

19:48 <|jemc|> the "IRC is too hard" problem can be solved with a friendly web client

19:48 <|jemc|> (for IRC)

19:50 <|jemc|> and as far as not missing messages while you're offline, there's public loggers for that

19:50 <|jemc|> in fact, http://irclog.whitequark.org/rubinius/2015-09-30

19:52 <|jemc|> and the to be frank the gitter UI feels rather slow, clunky, unpolished

19:53 <|jemc|> but at least they have the IRC bridge, which I use for all my gitter needs :)

19:55 <pipework> |jemc|: There's even ways to integrate the logs into the client for what you've missed.

19:56 <pipework> And some clients exist, but aren't free, that keep the session alive if you've closed the tab, and just reattach and replay history when you connect again.

20:21 <|jemc|> there are lots of ways to not miss messages, but I think part of the complaint was that if IRC clients themselves are too hard to set up, then catching history while offline is even harder

20:21 <|jemc|> so I was just demonstrating that the public logs serve a nice purpose for that with no setup by individual users

21:26 <GitHub41> [rubinius] chuckremes pushed 1 new commit to ffi-io: http://git.io/vc8cC

21:26 <GitHub41> rubinius/ffi-io 056ae04 Chuck Remes: generate and save a string defining the struct timeval as a FFI struct