Camelia, the Perl 6 bug

IRC log for #mojo, 2013-06-05

| Channels | #mojo index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:21 sri sooooo, Mojolicious 5.0 will depend on Perl 5.20 :D
00:21 sri with sub signatures that's a done deal, no discussions!
00:32 good_news_everyone joined #mojo
00:32 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/0fz9wA
00:32 good_news_everyone mojo/master c2752ca Sebastian Riedel: small optimizations
00:32 good_news_everyone left #mojo
00:36 sri adding Mojo::IOLoop::TLS seems pretty easy, only part that requires some thought is SSL_error_trap
00:38 tristan joined #mojo
01:38 shmuel joined #mojo
01:38 inokenty joined #mojo
01:40 * jberger wonders if sri is excited about the signatures :-P
01:41 sri \o\
01:41 sri /o/
01:43 jberger like I said, it wasn't loud, but in that kind of a venue, that's a pretty big announcement
01:44 jberger I wonder what will become of all the cpan modules
01:44 jberger they all are basically implementing the p6 signature spec
01:47 jberger so if there is a native implementation, there isn't much to be gained from any of the external ones (i.e. no better syntax)
01:51 d4rkie joined #mojo
01:57 Meiermann joined #mojo
02:05 sri jberger: from what i hear the parser is pluggable
02:06 sri so there can be more advanced ones on cpan benefitting from the new xs api
02:06 jberger yeah, I was just reading peter martini's talk proposal and was starting to understand the same thing
02:07 sri there are a few p5p discussions
02:10 jberger peter must have infinite patience!
02:40 asarch joined #mojo
02:44 moltar joined #mojo
03:01 whitebook joined #mojo
03:09 jamesw joined #mojo
03:27 mdmonsen joined #mojo
03:38 whitebook joined #mojo
03:46 good_news_everyone joined #mojo
03:46 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/MBSmFg
03:46 good_news_everyone mojo/master 65e605a Sebastian Riedel: a few more small optimizations
03:46 good_news_everyone left #mojo
03:57 hotsnow joined #mojo
04:35 bayashi joined #mojo
05:10 d4rkie joined #mojo
05:17 rem_lex| joined #mojo
05:31 tba joined #mojo
05:49 Britzel_ joined #mojo
05:52 MU joined #mojo
05:54 basiliscos joined #mojo
06:16 suy joined #mojo
06:19 arpadszasz joined #mojo
06:22 dpetrov_ joined #mojo
07:02 Dandre hello,
07:03 Dandre How can I add a renderer globally for a new data type (for instance xml)?
07:08 d4rkie joined #mojo
07:12 denisboyun joined #mojo
07:18 Dandre I don't undestand how content rendering is done
07:22 Dandre I have tried to put a route and I'd like to call the proper rendered based on the format requested.
07:24 Dandre I want to globally provide a xml rendered for all routes that request this format
07:33 dod joined #mojo
07:43 nic Dandre: Do you just want to return the xml, as in 'text'?
07:43 Dandre yes
07:44 Dandre I have a hash and I just want to return it in xml instead os json if the client requests this format
07:45 nic How about 'render(text =. $self-.generate_xml)' ?
07:45 nic (ubuntu is eating all my > hence the '.')
07:45 Dandre I haven't tried
07:46 nic (in the above, generate_xml would be some helper you write yourself)
07:46 Dandre but I wnt some global setting
07:47 nic maybe you want a pre-hook that intercepts all requests and checks if format eq 'xml'
07:50 Dandre I though I just had to register a custom renderer
07:51 dod joined #mojo
07:58 Vandal joined #mojo
07:59 jzawodn joined #mojo
07:59 dod joined #mojo
08:13 vifo joined #mojo
08:14 vifo Hey folks! Is it possible to have multiple captures in a route? Like /graphs/:name/last_(:duration)_(:magnifier)?
08:18 Dandre I have added a renderer using http://mojolicio.us/perldoc/Mo​jolicious/Renderer#add_handler but it is never called
08:24 maxhq joined #mojo
08:26 kwa vifo: Sure.
08:28 kwa Dandre: You've registered your own custom XML handler, and then loaded the plugin?
08:28 Dandre yes
08:29 kwa How are you now trying to call it?
08:29 kwa respond_to?
08:29 Dandre yes
08:29 kwa Then you're loading /someroute.xml?
08:31 Dandre yes
08:38 Dandre kwa: http://pastebin.fr/27635
08:39 Dandre when I GET testlist/10.json I get the json result as expected
08:39 Dandre GET /testlist/10.xml I have a notfound 404 return code
08:41 Dandre if I comment the xml line in the testlist route and uncomment the other xml I get the xml result as expected but this is not what I want because I don't want to set rendered for each route but globally
08:43 fhelmber_ joined #mojo
08:51 alnewkirk joined #mojo
08:56 ver joined #mojo
08:57 kwa hmm, not sure
08:58 Dandre in my XmlRenderer I have change:
08:58 Dandre sub register { $_[1]->renderer->add_handler(xml => \&_xml_renderer) }
08:58 Dandre to
08:58 Dandre sub register { $_[1]->renderer->add_handler(text => \&_xml_renderer) }
08:59 Dandre and I get the result so I guess that Rendered class never calls my renderer registered as xml
09:04 shmuel_ joined #mojo
09:16 Dandre in Renderer class, I have added this piece of code:
09:16 Dandre if (defined($self->handlers->{$options->{format}})) {
09:16 Dandre $self->handlers->{$options->{format}}->($self, $c, \$output, $stash);
09:16 Dandre return $output, $options->{format};
09:16 Dandre }
09:16 Dandre just before #Data ...
09:16 Dandre for test purpose.
09:17 Dandre My XmlRenderer gets called by I don't know whether there is some side effects
09:17 Dandre kwa: ?
09:17 Dandre sri: ?
09:18 kwa You've put that in the core renderer?
09:19 kwa If so, run the mojolicious test suite, it'll tell you if something has broke.
09:20 Dandre run the mojolicious test suite,
09:20 Dandre how?
09:21 moritz perl Makefile.PL && make test
09:27 Dandre as expected :-( :
09:27 Dandre Failed 3/82 test programs. 9/9155 subtests failed.
09:27 Dandre so How can I register a custom rendered without patching core?
09:28 Dandre This has never been done earlier?
09:33 kwa I'm just a hacker mate. I don't know if what you've found is a bug, or just not coded properly. :/
09:38 Dandre noone else can help me?
10:59 hotsnow joined #mojo
11:18 whitebook joined #mojo
11:20 mire joined #mojo
11:21 whitebook joined #mojo
11:23 nicomen joined #mojo
12:02 denisboyun joined #mojo
12:08 human39 joined #mojo
12:10 jbeimler joined #mojo
12:18 maxhq joined #mojo
12:33 mkrull Dandre: http://mojolicio.us/perldoc/Mojolicious#renderer
12:34 Dandre I've read it but that doesn't  help me
12:39 mkrull And i guess you took a look at the epl render plugin then also.
12:40 Dandre yes but as I have understood, it is template based and (not sure of that) handled properly because the template things are dealt later in Renderer class
12:41 Dandre I don't intend to do a template based renderer for xml
12:42 mdmonsen joined #mojo
12:46 mdmonsen joined #mojo
12:51 Dandre My simple test code in the pastebin just ouptput a simple valid xml response to any xml request. It is never called and I don't see in Renderer class where it could be called
12:51 mkrull Hm. I guess it would be best to do it the same way json rendering is done
12:52 Dandre ok but json rendering is coded in Renderer, not in a plugin
12:52 Dandre I don't see how to do that in my plugin
12:52 mkrull inherit and add xml rendering. then set is as application renderer
12:53 mkrull *it
12:53 Dandre I didn't thought of that
12:55 Dandre but this is only a cleaner way for my quickly patch Renderer
12:55 mkrull yes
12:55 Dandre Myy  reqirement hasn't been seen before?
12:56 Dandre sorry for my typos
12:57 moltar joined #mojo
13:04 rihegher joined #mojo
13:04 d4rkie joined #mojo
13:10 mkrull i don't know. i guess json is the data exchange format of choice for the web. so it is built in
13:11 Dandre xml too for rest application. Not as widely as json though
13:11 bc547 sri: about the wrk test yesterday: http://pastie.org/8009858  on a dual 6-core X5670 with hyperthreading enabled (24 cores in total)
13:16 bc547 sri: seems it does not really stress the machine to its max, although request/sec is more than triple the 8 core result of tianon yesterday
13:18 denisboyun joined #mojo
13:24 whitebook joined #mojo
13:42 nicolaas joined #mojo
13:42 mire joined #mojo
13:46 denisboyun joined #mojo
13:47 sri bc547: thanks, what matters is that it distributes the load evenly between all cores ;)
13:47 sri bc547: i think with a higher concurrency setting you could stress it more
13:47 sri but i'm more than happy with the result
13:48 sri scaling up to 24 cores is no problem \o/
13:48 sri Dandre: don't ever paste code into this channel please
13:49 bc547 sri: if you're happy, I'm happy too :-)
13:49 jberger_ joined #mojo
13:53 kwa joined #mojo
13:53 r0b3rt joined #mojo
13:57 Dandre sri: sorry but that was just 4 lines
13:57 sri Dandre: i don't care, it's rude
13:58 Dandre ok non problem
13:58 Dandre have you read my pastebin about custom renderer?
13:58 good_news_everyone joined #mojo
13:58 good_news_everyone [mojo] kraih pushed 1 new commit to master: http://git.io/YC01sw
13:58 good_news_everyone mojo/master 4d33623 Sebastian Riedel: better example for text encoding
13:58 good_news_everyone left #mojo
14:13 gryphon joined #mojo
14:15 bc547 sri: however some quick testing seems to indicate a scaling issue somewhere... best result I can get is 23k req/sec with approx 1% idle cpu. which would be too low considering I got 21.6K with 25% idle
14:16 sri bc547: i wasn't expecting it to scale linearly
14:16 bc547 sri: k
14:16 sri bc547: if you really care you could also make sure EV is using epoll
14:18 Vandal is there a module which strips html tags from text exept the custom list of tags?
14:18 bc547 sri: for me it does not matter since I don't have such usage scenario. But maybe it is important to you :-)
14:18 sri bc547: LIBEV_FLAGS=4 would be epoll
14:19 ynonp joined #mojo
14:19 sri if you want to try it :)
14:19 sri the results could be interesting for me too
14:20 sri and i'd try a concurrency of at least -c 1000
14:21 bc547 sri: current test: http://pastie.org/8010143
14:21 sri ok 200 is much too low to notice it
14:21 sri not even 10 connections per worker :)
14:22 bc547 testing now with -c 1000 -t 4 gives Requests/sec:  23569.57
14:22 sri interesting
14:22 bc547 and cpu is 2 to 3 % idle avg
14:23 sri i'm certain we could make it scale a little better, but overall i'm very happy with the result
14:23 sri bc547++
14:23 bc547 I have to go now, but if you want further tests, I'll be back tomorrow :-)
14:23 bc547 sri: np :-)
14:24 sri o/
14:24 tempire whoa
14:24 tempire http://www.washington.edu/news/2013/​06/04/wi-fi-signals-enable-gesture-r​ecognition-throughout-entire-home/
14:24 sri :O
14:25 gtodd tempire: google/gov't spying will now be complete and total!
14:25 moritz nah, they can't do this with standard wlan components
14:25 sri i guess we'll all be watching the sub signature talk later :)
14:26 moritz it needs to be wifi sender with an external clock, otherwise you can't easily detect the run times and thus position
14:30 jberger_ sri: of course
14:32 gtodd jberger: just wondering ... do you use zoidberg as a default shell?
14:32 btyler joined #mojo
14:32 asarch joined #mojo
14:36 jberger_ I don't actually
14:36 jberger_ not because it doesn't work, but because I forget to set it up honestly
14:36 jberger_ I should more often
14:37 * jberger_ fires up zoid
14:38 jberger_ gtodd, in truth, I'm still learning it
14:38 gtodd zoidberg is addictive ... I find it discourages you from writing tiny perl scripts for things ..  that you want to do ... (you end up just doing it in the shell like you would with awk) ... but then ... lack of history and completion etc etc
14:38 gtodd and you go back to zsh :)
14:38 jberger_ I'm maintainer because I fixed the failing tests (and two really odd bugs)
14:38 gtodd jberger hah
14:39 jberger_ no history?
14:39 gtodd well no tricked out history
14:39 gtodd ESC-. Ctrl-R /search term/ etc
14:39 jberger_ and there should be completion
14:40 gtodd yeah but ... not like zsh :)
14:40 marty sub signatures -- perl 5.20  -- Mojolicous 5.0 --- life is gwoooood
14:40 jberger_ but maybe it doesn't pick up bash-completion, which means it won't interface
14:40 jberger_ :/
14:40 gtodd perldoc [TAB] etc
14:41 gtodd mind you I think I have that turned off everywhere but my monster workstation where it is fast
14:41 sivoais joined #mojo
14:41 jberger_ yeah, those are all bash things
14:41 * gtodd gtodd's workstation is actually no longer a monster but ...
14:41 jberger_ I don't know how they could be included
14:41 jberger_ then again, zoid does have pluggable completion (which I have no idea how to use)
14:41 jberger_ so it could be added
14:42 gtodd I sort of don't want to learn more about zsh ... I want zoid to be like zsh so while learning zoid I also polish my perl
14:42 gtodd :)
14:42 jberger_ indeed
14:42 gtodd cross training :)
14:42 jberger_ thats kinda why I picked it up, back when I became maintainer
14:42 jberger_ I have hopes of cleaning it up
14:42 jberger_ a Moo-ifying would help the internals massively
14:43 jberger_ and if we could remove the job handling (selectively) it should run on windows
14:43 jberger_ so you could have a CPAN installable cross-platform shell
14:43 gtodd jberger perhaps even if it does not immediately evolve to displace zsh  or someother interactive shell it will make for a great interactive shell for perl apps needing shells
14:43 jberger_ but I don't know how to do it :-)
14:44 gtodd jberger yeah ... windows users yearn for shells ... if zoidberg could be that shell ... :-)
14:44 jberger_ I wonder if job-handling could be emulated with coroutines, does coro work on windows?
14:44 gtodd hmm Moo faster/lighter than Moose
14:45 gtodd with CGI gone Moo should be in CORE :-)
14:45 gtodd jberger with your influence just one blog post would be all it would take :)
14:46 jberger_ I'm not sure I have that much pull
14:46 jberger_ :-)
14:46 gtodd jberger_: If I recall the beauty of coro (rather than threading etc) is that it does work on Windows (checking)
14:46 gtodd hehe
14:46 jberger_ core Moo would be cool, but it isn't really needed
14:47 gtodd true
14:47 jberger_ its packable and small
14:47 jberger_ a core Moose would be cool (or mop of some kind)
14:47 jberger_ Moo is small because it doesn't have the power of moose
14:48 gtodd if Moose was in core (I guess it almost sort of is in practical terms)  ... Moo would be the packable small version of that
14:48 jberger_ I can't really do this for example: https://metacpan.org/module/MooseX::Types::NumUnit
14:48 jberger_ hmmmm, I wonder if I can actually
14:48 jberger_ Moo coercions always run
14:49 gtodd so if people build objects using Moose ... you can almost mostly use them with Moo instead of Moose  ?
14:49 jberger_ hmmmmmmmm, interesting
14:49 jberger_ gtodd, the difference is that Moo objects don't have a meta object, which is where you can do really crazy things
14:50 jberger_ most people don't need it, so Moo is enough
14:50 gtodd I mean if your app uses Moo you can access stuff that has been built to use Moose if you are careful about meta
14:50 jberger_ until one of the Moose features is needed, then Moose is required
14:51 gtodd right ... or ... perl6 :P
14:51 gtodd j/k
14:51 jberger_ and if a class is built with Moose that really only needs Moo, then Moose is still required, which is why there is a big push to downgrade lots of CPAN modules from Moose to Moo
14:51 jberger_ many really big projects have done so already
14:52 gtodd the Moose  section in Modern Perl book is very good at boiling it down ...
14:52 jberger_ does it talk about Moo?
14:52 gtodd the relationship between Moose Moo Mouse MooX (??!?)  is unclear in many minds
14:52 gtodd TIMTOWTDOO
14:52 jberger_ I didn't realize chromatic++ has updated it for Moo yet
14:52 gtodd :-\
14:53 gtodd jberger not
14:53 jberger_ oh
14:53 gtodd well not in my copy
14:53 gtodd I wish it was
14:53 * jberger_ checks online
14:53 jberger_ I know he updated
14:53 jberger_ Mouse and Any::Moose should not be used anymore
14:54 gtodd ah ok good
14:54 jberger_ Any::Moose is actually deprecated
14:54 * gtodd crosses those off
14:54 jberger_ have you read mst's article about Moo etc
14:54 * jberger_ finds it
14:54 gtodd hmm maybe I did but in the wrong order :)
14:55 gtodd MP book succeeds in converting one away from classic perl OO
14:55 jberger_ http://shadow.cat/blog/matt-s​-trout/moo-versus-any-moose/
14:55 gtodd to errm "bless" the Moose approach
14:57 jberger_ I'm starting to recommend Moo+Type::Tiny
14:57 gtodd was it chromatic that tried to get a patch that improved syntax of OO in some way (the CLASS patch??) into CORE of a perl version at some point?
14:57 jberger_ you get almost all of Moose that you need
14:57 jberger_ yeah
14:57 jberger_ it was
14:57 gtodd hmm so people are thinking of it ...
14:57 jberger_ the new `package BLOCK` helps though
14:57 gtodd ah ok
14:58 jberger_ its not quite `class` but ya know
14:58 jberger_ it helps with readability
14:58 jberger_ actually I quite like Mojo::Base (getting back to topic)
14:58 jberger_ super fast
14:58 jberger_ always lazy
14:59 jberger_ but I miss BUILD and triggers (and coercion)
14:59 gtodd right :-)  ... /topic :-D   ... anyway if some improved faster variant of Moose (Moo + Type::Tiny or whatever) ended up in CORE ... then ...
14:59 gtodd errm
15:00 jberger_ then people could use it without fear that it bloats their dependency chain
15:00 jberger_ users would already have it
15:00 gtodd right .... I get the sense that Moose so widely used that it can't change too much anymore
15:00 gtodd not that it has to but ...
15:01 nicolaas joined #mojo
15:01 gtodd I'm more surprised by the experimentalness of ~~ and given ... when :-\
15:02 gtodd than removal of CGI.pm :-|
15:02 jberger_ http://stackoverflow.com/a/16930984/468327
15:02 sh4 joined #mojo
15:02 gtodd perl should have something that let's it out of the box do  like     python3 -mhttp.server
15:03 gtodd :)
15:03 jberger_ http_this
15:04 jberger_ or a tiny Mojolicious script
15:04 gtodd yeah ... in core darn it all :)
15:05 gtodd hmm ikegami says ~~ will be fixed and not removed ...ok good :-D
15:05 jberger_ as sri has already said, the web is too much of a movable target
15:05 gtodd sheesh   Smart::Match
15:05 * gtodd looks
15:06 gtodd jberger true ... but some sort listen to this protocol on port
15:07 gtodd whatever ... with a collection of  improved ways of doing that available on cpan :)
15:07 gtodd is better than trying to replace CGI.pm
15:08 jberger_ CGI.pm does not provide a server
15:08 jberger_ CGI is a protocol
15:08 jberger_ CGI.pm is a "framework" that make use of that protocol
15:08 jberger_ but you still need apache or something to use it
15:09 jberger_ what you are suggesting really is to include say plackup
15:09 jberger_ which would be handy indeed
15:10 jberger_ Plack is generally a good candidate for core, its not meant as a framework (as miyagawa++) often reminds people
15:10 jberger_ its just a set of tools for PSGI
15:10 jberger_ and plackup is a PSGI server
15:13 gtodd yes and it's not too hard to make old CGI.pm  apps run with CGI::PSGI
15:13 gtodd so everyone would remain happy and backwards compatible
15:13 gtodd I think we got bugzilla to run under plack
15:14 jberger_ but if you can install CGI::PSGI and plackup, you can install CGI.pm
15:14 jberger_ the reason to remove it is that its not needed in core
15:14 jberger_ CGI.pm has been in core since before CPAN clients (maybe since before CPAN)
15:14 jberger_ by having it in core, too many people use if for new projects
15:15 gtodd anyway if stuff like bugzilla is being killed off ... many smaller lesser CGI apps will live on with CGI.pm installed from cpan
15:15 jberger_ how is bugzilla being killed off?
15:16 gtodd I guess development is still happening but AFAIK mozilla doesn't use it any more ...
15:16 gtodd and they were big on dogfood etc.
15:17 jberger_ well I don't know anything about that
15:17 jberger_ but CGI is not going away, Mojolicious runs under CGI
15:18 dod1 joined #mojo
15:18 jberger_ the CGI.pm module will just no long get undeserved blessing
15:18 jberger_ people will choose from the best available frameworks, rather than pick the worst because it is the most available
15:27 sri hmm, i don't understand a word from issue #500 :S
15:27 jberger_ neither do it
15:27 jberger_ I
15:28 jberger_ wai
15:28 jberger_ wait
15:28 jberger_ I think I got
15:29 * jberger_ can't type today
15:29 gtodd jberger only that bugzilla is/was a high profile user of CGI.pm ... and I *think* I recall reading how they no longer use bugzilla via CGI.pm :)  a bugzilla installation can get very busy
15:30 jberger_ he is having a problem with the pipe (ie | ) in the query string
15:52 Dandre jberger_ Have you seen my issue with a custom rendered?
15:53 jberger_ I was tinkering with it
15:54 jberger_ I was hoping to get a little further, but for now, try adding `handler => 'xml'` to your resepond_to xml stash
15:54 jberger_ but it seems like there should be a better way
15:54 Dandre ok I'll try
15:56 jberger_ http://pastie.org/8010597
15:57 inokenty joined #mojo
15:59 delias joined #mojo
16:01 whitebook joined #mojo
16:01 sri jberger_: there was, it was unpopular and got removed (before_render hook)
16:03 delias joined #mojo
16:07 sri which was basically a hook for render arg rewriting
16:08 sri app->plugins->hook(before_render => sub { my ($self, $args) = @_; $args->{handler} = 'xml' if exists $args->{xml} });
16:13 SmokeMachine joined #mojo
16:16 Dandre sri: is there another solution?
16:19 denisboyun joined #mojo
16:23 sri jberger_: i still don't get issue #500
16:23 jberger_ I think he is not using a properly encoded query string
16:24 jberger_ he is inserting a | in the string and the parser fails
16:24 sri perl -Mojo -E 'a(sub { $_[0]->render(text => r $_[0]->req->url->query->params) })->start' get '/?str=79AD8A6C36932F6E848​86C1%3E|33D7DBE9E&ka=007'
16:24 sri i don't see the parser failing
16:24 jberger_ in  the browser it fails
16:25 sri how can it be a mojolicious bug then?
16:25 jberger_ I don't think it is
16:25 sri oh... i thought you found a bug since you didn't close it
16:26 jberger_ no, I didn't close it because I wasn't certain that it was the browser, perhaps I should have
16:27 jberger_ I couldn't replicate it any other way
16:27 jberger_ I can close it
16:29 sri well, had you not answered i would have closed it as incomprehensible ;p
16:29 jberger_ it is very badly written
16:47 gtodd joined #mojo
16:59 delias joined #mojo
17:02 delias joined #mojo
17:07 whitebook joined #mojo
17:08 dod joined #mojo
17:37 sh3 joined #mojo
17:40 dod1 joined #mojo
17:55 btyler joined #mojo
17:59 ynonp joined #mojo
18:01 basiliscos joined #mojo
18:03 KindOne joined #mojo
18:22 delias|w joined #mojo
18:24 ynonp joined #mojo
18:28 rem_lex|pivo joined #mojo
18:50 marty did I miss the sub signature talk?
18:50 sri it's in an hour i believe
18:50 marty great. Thank you.
18:51 sri but the livestream has dvr functionality you can use to rewatch the whole day anyway
18:51 sri (just click on the timeline)
18:53 mire joined #mojo
19:00 d4rkie joined #mojo
19:01 sh4 joined #mojo
19:11 sh4 joined #mojo
19:14 navi joined #mojo
19:27 sh3 joined #mojo
19:48 beast joined #mojo
19:57 sri sub signature talk is up \o/
20:02 jberger_ watching!
20:02 jberger_ and promoting him on #yapc
20:09 Adura joined #mojo
20:09 mdmonsen joined #mojo
20:10 whitebook joined #mojo
20:25 whitebook joined #mojo
20:26 whitebook1 joined #mojo
20:27 mattastrophe joined #mojo
20:35 rihegher left #mojo
20:44 delias joined #mojo
20:51 jberger_ sri: so nothing much new to learn
20:52 jberger_ but as I have said in #yapc, his composure after all the bs is really amazing
20:52 jberger_ something to learn from and emulate (I'm speaking to myself)
20:56 whitebook joined #mojo
20:58 sri indeed
21:04 jberger_ the first few times I had a publication rejected, my instinct was to write the editor and gripe that the reviewers were obviously idiots who didn't understand
21:05 jberger_ I learned (by example) from my advisor that you must calmly address each concern, because either they were wrong in their understanding, in which case I must be more clear, or occaisonally, I was wrong, and that had to be addressed
21:06 jberger_ as a side effect, you are actually seen as being conciliatory, and that credibility goes a long way
21:09 gtodd sri: jberger where is the talk?
21:09 gtodd is it public or ...
21:10 jberger_ it just ended
21:10 jberger_ and since the stream ended (that room is done for the conf) you can't dvr and go back
21:10 jberger_ I'm sure they will post these
21:12 gtodd oh ok
21:13 gtodd just listening anyway since youtube is behaving badly for me anyway (chrome upgrade busted html5)
21:13 gtodd by the time I get around to viewing all the talks chrome will have patches that fix that so I'll save the best for last
21:14 gtodd :)
21:28 whitebook joined #mojo
21:34 lukep joined #mojo
22:04 achromic joined #mojo
22:56 btyler joined #mojo
23:24 jnbek joined #mojo
23:28 mire joined #mojo
23:56 inokenty left #mojo

| Channels | #mojo index | Today | | Search | Google Search | Plain-Text | summary