Perl 6 - the future is here, just unevenly distributed

IRC log for #metacpan, 2016-12-09

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

All times shown according to UTC.

Time Nick Message
01:28 karjala perlmodules.net now offers basic tracking of a distro's dependencies
02:35 bouncy joined #metacpan
04:25 haarg karjala: tracking dependency changes is probably the most useful thing that perlmodules.net could provide, but i'm not sure an automated solution would be useful for me.
04:25 haarg i have enough dynamic dependencies that it won't cover, and i don't care about the configure deps in this case
04:27 haarg i also don't care about developer prereqs, and would want to include the deps recursively
05:41 oiami joined #metacpan
05:52 metacpan joined #metacpan
05:52 metacpan [metacpan-api] mickeyn force-pushed mickey/mapping_updates from 6b06a90 to 14e8ce8: https://git.io/v1VvK
05:52 metacpan metacpan-api/mickey/mapping_updates 14e8ce8 Mickey Nasriachi: script/mapping: improved index cloning
05:52 metacpan left #metacpan
07:19 karjala Haarg: what if dynamic dependencies could be added manually, and you had the option to exclude certain dependencies or categories of dependencies? Recursive inclusion sounds difficult though
07:21 haarg why would recursive deps be hard?
07:21 haarg filters on modules/categories and explicit inclusions could work
07:22 karjala It would require keeping track of deps of all distros, for a start
07:22 karjala Ok
07:24 karjala Recursive isnt impossible
07:24 haarg keep in mind i'm just describing what would get it close to something i could see using.  i always watch the module release feed, so i don't really need another feed with basically the same information.
07:24 karjala O
07:27 haarg having the changelog in the feed is nice though.  if there was a feed for everything, i might switch to that over mc.org/recent
07:33 karjala Would you subscribe to the rss feed or subscribe for updates via email?
07:34 haarg rss feed
07:35 karjala Then i guess you would need a single feed fir all your modules' dependencies, right? A feed for each of your modules wouldn't do
07:36 karjala Each feed item should say: this release affects this and this module
07:38 haarg well, i already have a good idea of what all the dependency relationships are.  i watch the feed of all cpan releases and check things that are of interest.  if the changelog was included in that feed, i wouldn't have to click on as many things.
07:39 haarg i tried setting up a feed on perlmodules.net for all of the dependencies of Moo, and it's overall pretty nice
07:39 haarg but i'm unlikely to switch to something like that because i'm already used to watching the full feed.
07:40 haarg this is just my personal use case.  i don't know how well any of it transfers to others.
07:43 karjala Maybe someone should just create a recent_with_diffs feed in mc
07:43 karjala I can do that also in pmnet
07:45 karjala That'd be closer to what you want I think
07:47 haarg yeah
07:54 karjala I have a question: how should it handle the most recent releases that have a lower version number than the distro's max? Should i just not treat those as the latest, and not show changelog diffs about them? What if the previous max release however has been deleted from cpan (I don't have that information)? Should i still ignore the newer
07:54 karjala less-than-previous-max release?
08:04 karjala I mean, if "latest" status can never be achieved by the "less-than-max" releases, then I'm safe ignoring those releases from the diff mechanism
08:05 karjala but they have to be unable to achieve that status even if the previous max release has been deleted. Is that the case in metacpan?
08:06 haarg "latest" version numbers can decrease, but it is very uncommon
08:08 karjala Then I think I will present the less-than-max release in the feed, but in the place of the changelog diff it would say this release has a smaller version number than a previous release
08:08 neilb joined #metacpan
08:13 nakiro joined #metacpan
08:20 nakiro joined #metacpan
08:29 ranguard karjala: if you want to build endpoints in MC API to make things easier for our site (and give everyone the benefit of such end pointt) I'm sure we'd be open to that
08:31 karjala Would I have to learn Catalyst and ElasticSearch?
08:31 karjala (sounds a bit hard)
08:54 karjala haarg: there's your "recent with diffs" feed - linked to by the little RSS icon on the perlmodules.org homepage (also at https://www.perlmodules.net/feed/recent )
08:54 [Tux] Failed to fetch 'fastapi.metacpan.org/author/HMBRAND': Internal Exception at /pro/lib/perl5/site_perl/5.22.0/MetaCPAN/Client.pm line 190.
08:55 [Tux] with MC::C installed from git
09:11 osfabibisi joined #metacpan
09:11 Relequestual joined #metacpan
09:18 neilb joined #metacpan
09:29 ranguard karjala: what an opportunity... to learn Catalyst and ElasticSearch ;)
09:55 karjala But isn't Catalyst superseded by Mojolicious?
10:00 ranguard karjala: nope.. Mojolicious is great for cutting edge features, but does not try to be backwards compatible in the same way catalyst does
10:02 * ranguard can't talk in detail about it, but Dancer / Mojolicious and Catalyst all have their place depending on your overall needs
10:02 karjala ok
10:03 * ranguard uses Catalyst at work, and plays with the others from time to time at home for example
10:03 karjala oh
10:03 ranguard but other companies
10:04 karjala Until now I thought that Catalyst was used only by companies who already had projects before Mojolicious came to be
10:04 ranguard use Dancer and others use Mojolicious... all depending :)
10:05 ranguard it's all just software, we should hate all of it :)
10:08 karjala There's a lot of "edge cases" that need to be taken care of, if diff is to be implemented in mc.org
10:08 karjala like what happens if a different user releases a distro with the same version
10:08 karjala it's a nightmare
10:08 karjala on perlmodules.net I am allowed to not care (because I'm not mc.org)
10:09 karjala So I would have to discuss all of them here
10:56 ranguard karjala: fair enough :)
11:12 [Tux] ranguard, I keep getting «Failed to fetch 'fastapi.metacpan.org/author/HMBRAND': Internal Exception at /pro/lib/perl5/site_perl/5.22.0/MetaCPAN/Client.pm line 190.»
11:12 [Tux] what did I miss?
11:34 neilb joined #metacpan
12:09 zamanf joined #metacpan
13:00 zamanf joined #metacpan
13:21 neilb joined #metacpan
13:45 jnap joined #metacpan
14:33 neilb joined #metacpan
14:40 neilb_ joined #metacpan
14:47 zamanf joined #metacpan
15:45 metacpan joined #metacpan
15:45 metacpan [metacpan-api] oalders pushed 1 new commit to oalders/06perms: https://git.io/v1V5m
15:45 metacpan metacpan-api/oalders/06perms c6b18ba Olaf Alders: Set absolute path to 06perms.txt
15:45 metacpan left #metacpan
16:02 metacpan joined #metacpan
16:02 metacpan [metacpan-api] mickeyn force-pushed mickey/wip_mapping_deploy_simplified from f4c7e13 to 5ae393a: https://git.io/v1lwq
16:02 metacpan metacpan-api/mickey/wip_mapping_deploy_simplified 250ef3d Mickey Nasriachi: Deploy indices/mappings without ElasticSearchX::Model...
16:02 metacpan metacpan-api/mickey/wip_mapping_deploy_simplified e0ab180 Mickey Nasriachi: module.name,documentation --> analyzed
16:02 metacpan metacpan-api/mickey/wip_mapping_deploy_simplified c343733 Mickey Nasriachi: added suggester to fix autocomplete
16:02 metacpan left #metacpan
16:40 ranguard [Tux]: seems ok now?
16:40 ranguard I saw a blip from my logging as well, this seems to be our server related rather than fastly though
17:39 Guest69 joined #metacpan
20:04 oiami joined #metacpan

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