Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6-lwp-gsoc, 2014-06-16

| Channels | #perl6-lwp-gsoc index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
01:13 FROGGS_ joined #perl6-lwp-gsoc
06:53 FROGGS_ joined #perl6-lwp-gsoc
07:41 sergot FROGGS_: oooh :(
07:42 FROGGS_ can you give me commit bit? :o)
07:42 sergot FROGGS_: I've just added you to collaborators
07:42 FROGGS_ thanks
07:42 dalek http-response/header_parses: 742d794 | (Tobias Leich)++ | lib/HTTP/Response.pm6:
07:42 dalek http-response/header_parses: strip .parse method, we will not need it
07:42 dalek http-response/header_parses: review: https://github.com/sergot/http-response/commit/742d794ff9
07:42 sergot thank you
07:43 FROGGS_ so, .parse might be still be needed for HTTP::Message, since HTTP::Request still has a .parse and does a nextsame
07:43 FROGGS_ but I'm not sure it has to parse the headers this way
07:44 sergot Hm, why are you not sure?
07:44 FROGGS_ because it was late :o)
07:44 sergot ok :)
07:46 FROGGS_ I need to implement more stuff for chunked transfer encoding, since it only fetches the first chunk atm
07:46 FROGGS_ but it already fetches the .gz file as a buf8, and the text file as utf8
07:48 FROGGS_ btw, do you think it makes sense to merge all HTTP::* modules into a single repository?
07:48 sergot that's awesome, thank you FROGGS_
07:49 FROGGS_ because if you change something, you usually change >three modules, which means you have to chdir and git diff a lot
07:50 sergot FROGGS_: yes, will modules.perl6.org work with this?
07:51 FROGGS_ it would be a single distribution then
07:52 sergot FROGGS_: is there any github's mechanizm to do that?
07:52 FROGGS_ I mean, does it makes sense to keep a distribution 'HTTP::Response' that is kinda useless on its own and will always need a specific version of HTTP::Message?
07:52 FROGGS_ no, don't think so
07:52 FROGGS_ it woul mean to copy ofer the files and add them
07:52 FROGGS_ you loose the history that way...
07:53 FROGGS_ moritz: do you know a better way? (because I don't)
07:53 sergot hmm, we can do this if it's neccessary
07:53 FROGGS_ would make working on it easier for me at least... dunno how you see it
07:54 sergot Yes, it\d be easier :)
07:54 FROGGS_ also, when somebody reports a bug, what is the right bugtracker? if one can choose from like 7 or 8 then that is too much
07:55 FROGGS_ since we already use basically just one queue, HTTP::UserAgent
07:55 FROGGS_ okay, will be in a conference for the next one or two hours...
07:55 sergot I like the idea, we should do this.
07:56 FROGGS_ awesome :o)
07:58 sergot thank you :)
07:59 sergot oh, midterm after this "biweek"
07:59 sergot :)
08:42 FROGGS_ ahh, DateTime::Parse should stay in its own repo of course :o)
08:43 FROGGS_ do we use Net::HTTP right now at all?
11:51 sergot We dont use it
11:51 sergot :)
11:55 FROGGS_ nice
11:56 FROGGS_ btw, I going to push something in a minute
11:56 FROGGS_ beware!
11:57 sergot \o/
12:15 dalek http-useragent/header_parses: d06b9b4 | (Tobias Leich)++ | lib/HTTP/UserAgent.pm6:
12:15 dalek http-useragent/header_parses: handle chunked transfer encoding
12:15 dalek http-useragent/header_parses:
12:15 dalek http-useragent/header_parses: This involves fetching more and more chunks depending on the chunk
12:15 dalek http-useragent/header_parses: size field every chunk starts with.
12:15 dalek http-useragent/header_parses: review: https://github.com/sergot/http-useragent/commit/d06b9b47d2
12:15 FROGGS_ time perl6-m bin/http-download http://www.cpan.org/MIRRORED.BY
12:15 FROGGS_ real0m3.678s
12:16 FROGGS_ (wget takes 0.9s)
12:24 sergot o
12:24 sergot FROGGS_: what is 'Blob'?
12:25 sergot FROGGS_: you did a lot of work.. thanks!
12:26 FROGGS_ Blob is the role that Buf and utf8 and friends do
12:26 FROGGS_ being a piece of memory (a list of Ints/unit8s) acting a bit Stringy
12:27 FROGGS_ sergot: well, I need it for my CPAN stuff :o)
12:27 sergot FROGGS_: { @a[(state $i = -1) .. $i++ + @b] ~~ @b } what does this mean?
12:28 FROGGS_ it is searching for @b being in @a
12:32 sergot ok, I read the code and it is clear for me I think :)
12:33 sergot when can we merge those branches? :)
12:41 FROGGS_ right away
12:42 FROGGS_ we just need to check the tests
12:43 FROGGS_ because at least the tests about the removed method .parse do fail
12:46 FROGGS_ and then you can also merge all used HTTP::* modules into HTTP::UserAgent :o)
12:53 sergot ok :)
12:53 FROGGS_ sergot++
12:53 sergot FROGGS_++
12:54 FROGGS_ :D
12:57 sergot thanks for your help :)
12:58 FROGGS_ pleasure :o)
13:42 FROGGS_ joined #perl6-lwp-gsoc
14:41 sergot FROGGS_: would you merge those branches now or shall I?
14:48 FROGGS_ don't care
14:48 FROGGS_ I an do that in 5mins
14:49 sergot Ok, doing it right now :)
14:57 dalek http-headers: bcd0647 | (Tobias Leich)++ | lib/HTTP/Headers.pm6:
14:57 dalek http-headers: add grammar to parse headers
14:57 dalek http-headers: review: https://github.com/sergot/http-headers/commit/bcd06479a4
14:59 dalek http-message: 6a18014 | (Tobias Leich)++ | lib/HTTP/Message.pm6:
14:59 dalek http-message: strip .parse method, we might not need it
14:59 dalek http-message: review: https://github.com/sergot/http-message/commit/6a18014499
15:00 dalek http-response: 742d794 | (Tobias Leich)++ | lib/HTTP/Response.pm6:
15:00 dalek http-response: strip .parse method, we will not need it
15:00 dalek http-response: review: https://github.com/sergot/http-response/commit/742d794ff9
15:01 dalek http-useragent: c8bab4c | (Tobias Leich)++ | lib/HTTP/UserAgent.pm6:
15:01 dalek http-useragent: two step response message parsing, to not guess encoding
15:01 dalek http-useragent:
15:01 dalek http-useragent: We fetch the response as a buf8, split its status line, header and body
15:01 dalek http-useragent: and then find out the encoding before actually treat the body as text.
15:01 dalek http-useragent: Since the body also can be a binary blob, and therefore wants to stay
15:01 dalek http-useragent: an buf8.
15:01 dalek http-useragent: review: https://github.com/sergot/http-useragent/commit/c8bab4c264
15:01 dalek http-useragent: d06b9b4 | (Tobias Leich)++ | lib/HTTP/UserAgent.pm6:
15:01 dalek http-useragent: handle chunked transfer encoding
15:01 dalek http-useragent:
15:01 dalek http-useragent: This involves fetching more and more chunks depending on the chunk
15:05 sergot FROGGS_: why did you remove .parse method from HTTP::Message? :)
15:07 sergot another thing - what you did in http-useragent (decoding stuff) can't be done in .decoded-content in HTTP::Message?
15:07 FROGGS_ sergot: about .parse, that was what we talked about earlier :o)
15:08 FROGGS_ HTTP::Message does not know about the connection, that's why it is in HTTP::UA...
15:08 FROGGS_ but I am happy if you move it to a better place
15:09 sergot but what do you think about my idea?
15:09 sergot moritz: [Coke] maybe? :)
15:09 sergot and.. what should I do with the parse method then? :)
15:10 [Coke] sergot: sorry, what now?
15:11 sergot [Coke]: I think that
15:11 sergot this: https://github.com/sergot/http-useragent/blob/master/lib/HTTP/UserAgent.pm6#L80
15:11 sergot can be here:
15:11 sergot https://github.com/sergot/http-message/blob/master/lib/HTTP/Message.pm6#L22
15:12 sergot And it'd be great if you tell me what do you think about this idea.
15:15 dalek http-message: 3b2e83d | sergot++ | lib/HTTP/Message.pm6:
15:15 dalek http-message: Revert "strip .parse method, we might not need it"
15:15 dalek http-message:
15:15 dalek http-message: This reverts commit 6a18014499113847a9356bc6688cc3d2d7c2a1ba.
15:15 dalek http-message: review: https://github.com/sergot/http-message/commit/3b2e83dc05
15:15 sergot FROGGS_: i've just reverted your commit, just to make it works, okay?
15:16 sergot and I'll work on this later
15:17 sergot is it ok? :)
15:18 [Coke] sergot: it makes sense to me that the decoding would go in the message; it's part of the protocol. I don't have a strong feeling, though.
15:18 [Coke] so, +1 from me
15:20 sergot [Coke]: thanks for your opinion.
15:20 sergot FROGGS_: could you take a look at HTTP::UA's tests? Something is wrong I think.
15:28 FROGGS_ k, will do
15:29 sergot FROGGS_++
15:32 FROGGS_ sergot: where do you have Test::IO::Capture from?
15:33 FROGGS_ ahh, IO::Capture::Simple perhaps
15:34 FROGGS_ sergot: it passes on my box
15:36 FROGGS_ ==> Successfully installed HTTP::UserAgent
15:37 FROGGS_ it also passes when it is precompiled... (using panda to test and install)
15:38 FROGGS_ sergot: btw, I said why the logic for chunged transfer encoding is in HTTP::UserAgent currently, and that I'm fine if it could be moved somewhere else
15:39 FROGGS_ and about that parse method: do we need to parse the headers for a request? if not, that we don't need the .parse method in HTTP::Method

| Channels | #perl6-lwp-gsoc index | Today | | Search | Google Search | Plain-Text | summary