adkron_ has quit [Read error: Connection reset by peer]
adkron_ has joined #rubygems
kgrz has joined #rubygems
tbuehlmann has joined #rubygems
tekin has joined #rubygems
yashshah has quit [Read error: Connection reset by peer]
yashshah has joined #rubygems
mikewintermute has joined #rubygems
Paul_G has quit [Ping timeout: 252 seconds]
mikewintermute has quit [Client Quit]
Paul_G has joined #rubygems
mikewintermute has joined #rubygems
vertis1 has joined #rubygems
vertis has quit [Ping timeout: 240 seconds]
erikh has joined #rubygems
<erikh>
so, maybe i'm late to the party, but i'm pretty sure whatever is generating the gem index behind the scenes is broken, and has been for an hour or so.
<erikh>
just figured I'd say something in case someone hasn't seen it (and is up at this ungodly hour to act on it).
<erikh>
also I want a refund
<erikh>
:P
Emily is now known as EmilyAFK
<erikh>
well, to be clear, new gems and their versions aren't getting added -- webhooks still seem to be firing though, and the gems do appear on the website.
<erikh>
but gem install $foo and bundler bits don't think they're there.
<erikh>
ok, that's enough novel from me. stay frosty.
erikh has left #rubygems ["WeeChat 0.3.9.2"]
_br_ has quit [Excess Flood]
_br_ has joined #rubygems
_br_ has quit [Excess Flood]
_br_ has joined #rubygems
yashshah has quit [Read error: Connection reset by peer]
stevenharman has quit [Quit: Leaving...]
yashshah has joined #rubygems
EmilyAFK is now known as Emily
stevenharman has joined #rubygems
Emily is now known as EmilyAFK
tcopeland has quit [Quit: Leaving.]
rondale_1c has joined #rubygems
mose_ has joined #rubygems
michal_papis has joined #rubygems
mikewintermute has quit [Quit: mikewintermute]
Sinjo_ has joined #rubygems
pipework has quit [Remote host closed the connection]
bhaak_ has joined #rubygems
nz has quit [*.net *.split]
drbrain has quit [*.net *.split]
bhaak has quit [*.net *.split]
Sinjo has quit [*.net *.split]
mose has quit [*.net *.split]
rondale_sc has quit [*.net *.split]
mpapis has quit [*.net *.split]
michal_papis is now known as mpapis
stevenharman has quit [Ping timeout: 245 seconds]
drbrain has joined #rubygems
mikewintermute has joined #rubygems
bhaak_ is now known as bhaak
dwradcliffe has joined #rubygems
stevenharman has joined #rubygems
charliesome has joined #rubygems
terceiro has quit [Read error: Connection reset by peer]
terceiro has joined #rubygems
stevenharman has quit [Ping timeout: 245 seconds]
tekin has quit [Quit: Computer has gone to sleep.]
stevenharman has joined #rubygems
mikewintermute has quit [Quit: mikewintermute]
fromonesrc has joined #rubygems
terceiro has quit [Ping timeout: 245 seconds]
kgrz has quit [Quit: Computer has gone to sleep.]
yashshah has quit [Read error: Connection reset by peer]
yashshah has joined #rubygems
tekin has joined #rubygems
jcaudle has joined #rubygems
tbuehlmann has quit [Remote host closed the connection]
pipework has joined #rubygems
yashshah has quit [Read error: Operation timed out]
<hemanth_>
qrush, yes aware of that, https://rubygems.org/profiles/hemanth are few of my silly gems, will parse them and try to build the deb, i'm using gem2deb, will have to try fpm as well
<qrush>
i think we have nagios or something already
<tbuehlmann>
qrush, might have a look at the author name here? https://rubygems.org/gems/sinatra-kittens seems weird. couldn't reproduce it by setting up my own rubygems.org instance
<dwradcliffe>
qrush: on aws? not that I'm aware of...
<ddd>
they exported RC4 before the weaker ciphers as expected for mitigation. ssllabs will still consider them as 'vulnerable' until they drop all *other* TLS1.x and only keep the RC4. But if they do that then they are incompatable with most browsers and client-side 'fetchers'.
ckrailo has joined #rubygems
<ddd>
not really rubygems's goal or place to force fixing of client side issues, just the server
qmx|away is now known as qmx
<[reed]>
offering those cipher is not a client-side issue
<[reed]>
that's most definitely a server issue
<[reed]>
what would be preferred would be something like:
<[reed]>
it's not, though... there is *no* current mainstream browser that wouldn't support a more conservative cipher suite choice, such as the one I gave above.
<swills>
ddd: curious, which clients require the low ssl?
<swills>
[reed]: browser clients are not the only clients.
<swills>
but you're right, something a bit newer should be OK
<swills>
why not !eNULL in there too?
<swills>
where did you get that list from?
<ddd>
[reed]: speak to the SSL stack creators and have them add the option to _only_ allow RC4, which currently no stack does
<[reed]>
swills: you're not doing ALL, and eNULL isn't included in HIGH or RC4, but happy to add as well
<[reed]>
evan: also, instead of explicitly listing SSLv3 and TLSv1, can you just say -SSLv2, so when your nginx + openssl set up supports TLSv1.1 / TLSv1.2 (or newer), it will just work?
<evan>
if nginx supports that, sure.
<ddd>
need to find nginx's setting for SSLHonorCipherOrder on and SSLInsecureRenegotiation off
<ddd>
ah cool. nginx doesn't do TLS renegotiation at all
workmad3 has quit [Ping timeout: 264 seconds]
mockra has joined #rubygems
tenderlove has joined #rubygems
havenwood has quit [Remote host closed the connection]
yashshah has quit [Read error: Connection reset by peer]
yashshah has joined #rubygems
tcopeland1 has quit [Ping timeout: 240 seconds]
huoxito has quit [Quit: Leaving]
whit537 has joined #rubygems
whit537_ has joined #rubygems
whit537 has quit [Read error: Connection reset by peer]
<[reed]>
evan: http://wiki.nginx.org/HttpSslModule#ssl_protocols -- default seems to be what you want, honestly... though, if you do want to explicitly specify it, just do |ssl_protocols SSLv3 TLSv1 TLSv1.1 TLSv1.2;|
<drbrain>
it seemed reasonable on the face of it, but due to module inclusion it breaks
<drbrain>
I'd rather stick with ugly metaprogramming that's been around for two major release cycles (or more?) than pretty metaprogramming with one major bug
<Defiler>
that and I don't see what was wrong with module_function, looking at the diff
<Defiler>
It's hard to really take any version of ruby I know of as fully 'production ready'
<Defiler>
MRI/YARV have the evil GC double-return bug a la joe damato's "broken promises" writeup
<Defiler>
only 2.0.0 supports a modern (as in, post-2005) version of SSL/TLS
<drbrain>
I guess the OPT_TABLE is mildly objectionable
<Defiler>
rubinius is still not quite ready to rock and suffers from the same openssl binding problem
<Defiler>
jruby is solid but not well-suited to some unix programming tasks
<Defiler>
I love the language but it's still amazingly immature in some areas
<drbrain>
assuming Nari presents at RubyKaigi this year, I wonder what amazing new GC thing he will have to show us
<drbrain>
he has been working on parallel GC stuff lately
<drbrain>
it'll take a new C API to get rid of the "broken promises" bug
<Defiler>
I might be spending some more time with Idris this year
<Defiler>
I want those features but in Erlang, honestly
<Defiler>
but idris is promising
martinisoft has quit [Quit: martinisoft]
<drbrain>
what are "cumulative universes"?
<Defiler>
I think that's set-theory nerdspeak for the ability to process a hierarchy of types
tcopeland has quit [Read error: Connection reset by peer]
tcopeland has joined #rubygems
<Defiler>
Used in a sentence: "In order to axiomatize the cumulative universe of sets within the typed language, we introduce certain laws in form of inference rules."