Perl 6 - the future is here, just unevenly distributed

IRC log for #metacpan, 2014-02-26

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

All times shown according to UTC.

Time Nick Message
00:02 neilb joined #metacpan
02:10 klapperl_ joined #metacpan
02:11 Zoffix joined #metacpan
02:14 Zoffix oalders, hey. On TPM you said "I'll be participating via Google Hangout." How can I do the same? I don't feel like sitting in 2hr of traffic after a 6 hours meeting at work, for my first ever Perl Mongers meeting... so I can't make it this month, but I'd love to participate electronically.
04:15 oalders Zoffix: just add us to your Google+ circles and we'll add you back https://plus.google.com/+TorontoPerlMongers/posts
04:15 dipsy [ Toronto PerlMongers - Google+ ]
04:15 oalders then you'll get an invite when the Hangout has started
04:15 oalders where are you located?
04:28 Zoffix Brampton
04:30 Zoffix k, added. I'm "Peter Evans" on Google+, because it's too anal to accept Zoffix Znet as a name :)
04:30 * Zoffix leaves to bed.
04:32 oalders :)
06:27 kentnl Man, Why couldn't ElasticSearch just done what everyone else does: Bolt a 2 on the end.
06:29 mst kentnl: for what?
06:29 kentnl mst: re "ElasticSearch appears to have started being called Elasticsearch because they changed the API"
06:30 mst oh. right. yeah, I really do not understand why clinton let them make him do that.
06:37 kentnl Just the question really is, will it actually get better, or will it be a perpetual pain. I'm waging on the latter.
06:41 ether_ joined #metacpan
06:51 mst I'd generally wager on the latter for any software question :)
06:54 ribasushi this is still a young module (the new API i like 4 months old)
06:54 ribasushi killing the new thing off the index is still an option
06:55 ribasushi he just doesn't want to do it for ... histerical raisins
06:55 ribasushi (/me imagines there is a client involved they don't want to piss, so breaking everything else seems like a saner choice)
06:58 jayallen joined #metacpan
07:17 kentnl its just every day it sits there un-fixed, the "too big to fail" problem gets bigger :/
07:18 jayallen joined #metacpan
07:31 [Sno] joined #metacpan
07:38 alnewkirk joined #metacpan
07:47 neilb joined #metacpan
07:50 klapperl joined #metacpan
07:52 dpetrov_ joined #metacpan
08:03 neilb joined #metacpan
08:08 neilb joined #metacpan
08:51 neilb joined #metacpan
09:07 klapperl_ joined #metacpan
09:16 bowtie_ joined #metacpan
09:31 alh_ joined #metacpan
09:36 dpetrov_ joined #metacpan
11:01 chansen joined #metacpan
12:35 oalders i *think* it's because they want to keep the naming of the Elasticsearch client consistent across the different languages
12:36 oalders this was building up to the 1.0 release of ES itself as i understand
12:36 oalders i'm not defending it, though :)
12:36 mst the problem is that OS X developers now can't actually have both
12:38 oalders that's a problem and the fact that it's hard to scan the difference between Elasticsearch.pm and ElasticSearch.pm in chat
12:38 oalders even worse in spoken language
12:39 oalders btw, we're now looking for a student who would like to participate in OPfW http://www.perlfoundation.org/outreach_program_for_women
12:39 dipsy [ The Perl Foundation ]
12:39 oalders so, GSoC + OPfW opportunities here
13:18 kentnl "because they want to keep the naming of the Elasticsearch client consistent" # Not necessarily possible all the time anyway
13:18 kentnl sometimes somebody will get in first on your namespace of choice, and then you're plain out of luck without a NS takeover
14:04 ribasushi oalders: name needs to stay consistent but api can change per version...
14:04 ribasushi <boggle>
14:29 tobyink ribasushi: given that Perl doesn't have any decent solution for multiple versions of the same module to be installed side-by-side, I'd rather that completely rewritten APIs resulted in new module names. Witness JSON.pm 1.x -> 2.x.
14:30 ribasushi tobyink: you are preaching to the wrong choir I think...
14:37 kentnl my favourite options are doing JSON2 or similar, like JSON::v2 , its Ugly, but Ugly > Broken
14:38 ribasushi tobyink: does any dynamic rival out there have a decent solution for multiple versions of the same namespace?
15:15 oalders kentnl, ribasushi: https://github.com/elasticsearch/elasticsearch-perl/issues/5
15:15 dipsy [ Very unfortunate choice of module name · Issue #5 · elasticsearch/elasticsearch-perl · GitHub ]
15:15 oalders that was the discussion on the naming
15:19 kentnl I don't understand how that conclusion is a "sortout"
15:20 kentnl Its basically saying "to keep using the old stuff, you have to rewrite everything"
15:21 oalders yeah, it's going to be a pain point
15:21 oalders on an unrelated note, there's a discussion on what sites are or aren't good fits to link to from user profiles
15:21 oalders https://github.com/CPAN-API/metacpan-web/issues/1104
15:21 dipsy [ Underwhelming profile affiliation options · Issue #1104 · CPAN-API/metacpan-web · GitHub ]
15:21 oalders please comment
15:22 oalders specifically re: Grindr and Chaturbate but would be good to have thoughts in general
15:29 kentnl hell, at this point it would be _better_ if he'd just released a newer version with an API incompatibility with the same name
15:30 kentnl at least it would be obvious what the problem was, and what the solution was.
15:30 avenj kentnl++
15:31 oalders i could live with that
15:37 tobyink ribasushi: Perl6 provided a pretty nice solution for dealing with multiple versions of a library, even within the same process.
15:41 ribasushi tobyink: haven't looked because as far as I know it isn't actually fully implemented yet
15:41 ribasushi and as such the spec is fluid
15:41 ribasushi also - *currently available* dyn languages :)
15:43 kentnl P5 /could/, I imagine, one day support polyversions, just it would require some serious infrastructure changes.
15:44 kentnl it would obviously need some alternative filesystem layout mechanism
15:44 kentnl and some kind of alternative loading semantics
15:45 kentnl and possibly support for multiple, memory-isolated VMs with native marshalled IPC between instances.
15:45 ribasushi kentnl: pass the pipe
15:45 tobyink The trick that Perl6 uses is secretly encoding $VERSION and $AUTHORITY into the namespace.
15:46 ribasushi tobyink: you keep talking about p6, I am asking is there a dyn language with a sizeable external library that solved this
15:47 kentnl https://metacpan.org/pod/Safe::World # This looked interesting in regards to multiple isolated VMs
15:47 tobyink So if you load JSON.pm version 2.51 by MAKAMAKA, the real namespace the class gets stored in the symbol table as something like JSON~2.51~MAKAMAKA.
15:47 dipsy [ Safe::World - Create multiple virtual instances of a Perl interpreter that can be assembled together. - metacpan.org ]
15:47 kentnl it seems implemented in terms of threads, each vm being a high-level perl thread
15:47 kentnl I just haven't tried it to see how it really works because cbf building perl with threads
15:47 tobyink Then if you call JSON->new you get whatever is the latest loaded version of JSON. But you can specifically instantiate a class by a particular version or authority.
15:49 tobyink ribasushi: certainly I agree that Perl6 does not have widespread usage. I'm merely trying to show that it's not a technically insurmountable problem.
15:49 kentnl ribasushi: If I really wanted to pack a bong, I'd be asking for a Perl module that spawned a virtualbox/qemu session slave with a perl in it. ;)
15:49 ribasushi tobyink: everything "can be done" when half of the details are handwaved away
15:50 ribasushi I was asking about a specific implementation in widespread usage that didn't fall apart under its own weight
15:50 ribasushi tobyink: mainly your statement <tobyink> ribasushi: given that Perl doesn't have any decent solution
15:50 kentnl ah, the 2 kinds of apathy
15:50 kentnl "Don't worry, it will be useful eventually"
15:50 ribasushi sounded like perl has some catching up to do with something everyone figured a long time ago
15:50 kentnl and "Its not useful yet, so cbf using it and making it useful"
15:51 ribasushi kentnl: nah, I am just larting tobyink
15:51 ribasushi kentnl: p6 is useful for some stuff, just not for what I am mainly interested in
15:51 tobyink There's C of course. It's not a dynamic language, but libraries are dynamically loaded.
15:51 ribasushi tobyink: are you kidding me - C doesn't even have proper namespacing, left alone versioning :)
15:52 kentnl I saw a dynamic interpreter for C somewhere
15:53 tobyink kentnl: you may be thinking of TinyCC.
15:54 kentnl is that a JIT compiler that can essentially do 'eval $C_CODE' at runtime?
15:54 tobyink You can use it in a shebang.
15:58 kentnl http://root.cern.ch/drupal/content/cling # Here's one for C++, but I was sure I'd found one for C
15:58 dipsy [ Cling | ROOT ]
15:59 kentnl http://en.wikipedia.org/wiki/Ch_%28computer_programming%29
15:59 dipsy [ Ch (computer programming) - Wikipedia, the free encyclopedia ]
17:08 ether [6~/scrollback end
17:43 [Sno] joined #metacpan
17:45 neilb joined #metacpan
17:47 * ether killed that conversation
18:09 kentnl ether: don't worry, as expert practicioners of a dead language, we have the necromancers and the technology, we can rebuild and vivify it!
18:10 klapperl joined #metacpan
18:21 metacpan joined #metacpan
18:21 metacpan [metacpan-web] rwstauner pushed 1 new commit to master: http://git.io/SE5U-w
18:21 metacpan metacpan-web/master 04af92e Randy Stauner: Get pod_lines in module search to help identify pod files...
18:21 metacpan left #metacpan
18:21 dipsy [ Get pod_lines in module search to help identify pod files · 04af92e · CPAN-API/metacpan-web · GitHub ]
18:23 neilb joined #metacpan
18:49 neilb joined #metacpan
19:51 klapperl_ joined #metacpan
19:52 metacpan joined #metacpan
19:52 metacpan [metacpan-web] ranguard pushed 1 new commit to master: http://git.io/gg_wOw
19:52 metacpan metacpan-web/master 7c8cf6a Leo Lapworth: Add thanks to 'The Perl Foundation'
19:52 metacpan left #metacpan
19:52 dipsy [ Add thanks to 'The Perl Foundation' · 7c8cf6a · CPAN-API/metacpan-web · GitHub ]
20:05 oalders ranguard++
20:05 oalders we still need to add EPO and Booking.com as well
20:08 * ranguard will do so now
20:10 ribasushi can we add a mini-icon of abigail with a thumbs-up and "we are hiring!" underneath? :)
20:17 rwstauner who's hiring?
20:18 * rwstauner needs a job
20:19 ribasushi rwstauner: booking is *always* hiring
20:20 oalders but rwstauner isn't relocating
20:20 rwstauner ah, right, booking
20:20 rwstauner thanks
20:20 ribasushi oalders: they make exceptions now (they are that desperate ;)
20:21 oalders i did not know that. i thought it was just one guy in mexico or something
20:21 rwstauner :-)
20:50 ether I'd consider working for booking if I could do so remotely
20:50 ether I'm not sure if I can ever work in an office environment again, now
20:56 metacpan joined #metacpan
20:56 metacpan [metacpan-web] ranguard pushed 1 new commit to master: http://git.io/tM4xaw
20:56 metacpan metacpan-web/master d23fd0b Leo Lapworth: Add more sponsors and move their logos into 'static/images/sponsors/'
20:56 metacpan left #metacpan
20:56 dipsy [ Add more sponsors and move their logos into 'static/images/sponsors/' · d23fd0b · CPAN-API/metacpan-web · GitHub ]
20:57 ranguard oalders: ^^ please update the copy as needed
20:58 oalders ranguard++
20:58 oalders looks good
21:28 neilb joined #metacpan
21:55 jayallen joined #metacpan

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