Logs for #nikola for 2015-09-07

10:36:28 -GitHub[nikola]:#nikola- [nikola] ralsina pushed 2 new commits to idea-for-1814: http://git.io/vZkYS
10:36:28 -GitHub[nikola]:#nikola- nikola/idea-for-1814 ad24c22 Roberto Alsina: fix
10:36:28 -GitHub[nikola]:#nikola- nikola/idea-for-1814 dd49ef6 Roberto Alsina: merged master
10:44:16 -travis-ci:#nikola- getnikola/nikola#6240 (idea-for-1814 - dd49ef6 : Roberto Alsina): The build is still failing.
10:44:17 -travis-ci:#nikola- Change view: https://github.com/getnikola/nikola/compare/d1bc076f6158...dd49ef6e26ff
10:44:17 -travis-ci:#nikola- Build details: https://travis-ci.org/getnikola/nikola/builds/79101225
11:02:43 <juanjo_> hi
11:02:45 <juanjo_> wird thing
11:03:15 <juanjo_> I'm trying to create a plugin to import from Twitpic, basically i have to read a txt utf-8 file
11:03:24 <juanjo_> in python console I can do
11:03:41 <juanjo_> f = open("tweets.txt")
11:03:47 <juanjo_> chunks = [p.strip() for p in f.read().split("\n\n") if p != '\n']
11:03:50 <juanjo_> and it works
11:03:57 <juanjo_> in the plugins .py
11:04:15 <juanjo_> where, imitating other plugins, i import form future unicode_literals
11:04:19 <juanjo_> i got this error
11:04:29 <juanjo_> UnicodeDecodeError: 'ascii' codec can't decode byte 0xc3 in position 92: ordinal not in range(128)
11:04:41 <juanjo_> does this happened to you?
11:08:25 <juanjo_> Ok, i fixes it as
11:08:25 <juanjo_> chunks = [p.strip() for p in f.read().decode('utf-8').split("\n\n") if p != '\n']
11:08:39 <juanjo_> thansk to this http://stackoverflow.com/questions/809796/any-gotchas-using-unicode-literals-in-python-2-6
12:04:13 <ralsina> yes
12:04:39 <ralsina> juanjo_: the problem is, once you have unicode_literals, '\n'\n is unicode, which means f.read() has to be unicode
12:04:58 <ralsina> juanjo_: so, you get unicode(f.read()), which fails because you are reading bytes
12:05:16 <ralsina> juanjo_: so, your code was buggy, and unicode_literals makes it fail hard enough to notice :-)
12:05:47 <ralsina> juanjo_: which is one of the reasons to have unicode_literals, it makes it work like py3 which makes unicode bugs more obvious
12:09:50 <juanjo_> +1
12:11:58 <juanjo_> What about adding to Nikola site a page with all the sources you can import from?
12:15:41 -GitHub[nikola]:#nikola- [nikola] felixfontein created wordpress-import-fix-tags (+1 new commit): http://git.io/vZkon
12:15:41 -GitHub[nikola]:#nikola- nikola/wordpress-import-fix-tags 6361ec0 Felix Fontein: Added basic tag/category case sanitizing. Offers only two choices (first or lower).
12:16:52 -GitHub[nikola]:#nikola- [nikola] felixfontein opened pull request #2034: Basic tag/category case sanitizing. (master...wordpress-import-fix-tags) http://git.io/vZkoi
12:17:00 <capri> hey, im new to nikola and i want to change the default bootstrap3 theme - i just want to change the colour of the header - instead of gray i want an other color
12:17:15 <capri> in which file i have to change that css data?
12:22:27 -GitHub[nikola]:#nikola- [nikola] felixfontein pushed 1 new commit to wordpress-import-fix-tags: http://git.io/vZk6f
12:22:27 -GitHub[nikola]:#nikola- nikola/wordpress-import-fix-tags 993f63d Felix Fontein: Fixed test.
12:26:50 -GitHub[nikola]:#nikola- [nikola] felixfontein pushed 1 new commit to wordpress-import-fix-tags: http://git.io/vZki8
12:26:50 -GitHub[nikola]:#nikola- nikola/wordpress-import-fix-tags 23b4580 Felix Fontein: Fixed stupid error.
12:27:57 <ralsina> juanjo_: sounds like a good idea!
12:28:20 <ralsina> capri: you can just add a files/assets/css/custom.css and that one is guaranteed to be used
12:28:34 <ralsina> capri: so just put any CSS fixups there and it should just work
12:29:18 <capri> ralsina, thanks - ill try it :)
12:32:25 -travis-ci:#nikola- getnikola/nikola#6245 (wordpress-import-fix-tags - 23b4580 : Felix Fontein): The build was fixed.
12:32:26 -travis-ci:#nikola- Change view: https://github.com/getnikola/nikola/compare/993f63d6ca50...23b45808efe3
12:32:26 -travis-ci:#nikola- Build details: https://travis-ci.org/getnikola/nikola/builds/79112779
12:54:10 <capri> ralsina, custom.css works perfect. I also want to add my own logo into each header - is that possible to configure two? i know i can edit output/index.html, but that one will be overwritten every build
12:54:25 <capri> to configure too -> sorry
12:54:46 <ralsina> capri: check LOGO_URL option in conf.py
12:54:53 <ralsina> as long as you are on bootstrap3 :-)
12:55:05 <capri> yes i am :) - thanks
14:12:10 -GitHub[nikola]:#nikola- [nikola] ralsina closed pull request #2034: Basic tag/category case sanitizing. (master...wordpress-import-fix-tags) http://git.io/vZkoi
14:16:55 -travis-ci:#nikola- getnikola/nikola#6247 (master - dc1b848 : Roberto Alsina): The build passed.
14:16:56 -travis-ci:#nikola- Change view: https://github.com/getnikola/nikola/compare/621f2ee75772...dc1b848fbdd4
14:16:56 -travis-ci:#nikola- Build details: https://travis-ci.org/getnikola/nikola/builds/79129921
14:59:37 <Aeyoun> ralsina: (sitemap stuff) I’m generating an unknown number of tasks based on the result of scan_locs(). I can’t acheive that using a calc_dep. or am I misunderstanding you entirely?
15:00:06 <ralsina> Aeyoun: yes, basically, it will run scn_loc once, and then will run your task if it changes
15:00:29 <ralsina> we already need to run scan_loc to decide if we want to build the sitemap anyway
15:02:35 <Aeyoun> ralsina: https://github.com/getnikola/nikola/pull/2030/files#diff-33521a0ed370ee68cb5929eff2fbb717R303
15:04:57 <ralsina> Aeyoun: and remove the call to scan_locs in 301
15:05:01 <Aeyoun> ralsina: urlset is only populate after running scan_loc. how should I depend on having done this before it’s been run?
15:05:46 <Aeyoun> the `for` on 303 yields one or more tasks depending on urlset.
15:07:27 <ralsina> Aeyoun: it *will* run scan_locs before running your task, since it's a dependency
15:07:53 <Aeyoun> ralsina: that is too late. ;)
15:08:08 <Aeyoun> Oh, I see what you’re saying now.
15:08:33 <ralsina> Then I am not understanding the problem
15:08:33 <ralsina> argh, I have lag
15:12:09 <Aeyoun> ralsina: no, it’s too late. the task generator depends on urlset existing before they’re generated.
15:13:46 <ralsina> Aeyoun: in which case, let's just run scan_locs and not depend on it
15:13:46 <ralsina> what's the worst that can happen? We always run it anyway
15:14:05 <ralsina> The only part that's worth avoiding is writing sitemaps when they would not change content
15:14:31 <Aeyoun> well, scan loc will be run more than once?
15:14:41 <Aeyoun> for the worst thing that can happen category.
15:15:13 <Aeyoun> no, I can avoid that.
15:15:34 <ralsina> well, let's make sure THAT doesn't happen :)
15:16:22 <ralsina> Aeyoun: ahhh!
15:16:43 <ralsina> make sure scan_locs runs after all the tasks that generate files in output
15:17:08 <ralsina> otherwise, the "worst" is sitemap is never uptodate
15:19:13 <ralsina> this was *very* tricky to get right
15:21:20 <Aeyoun> if not urlset: scan_locs(); in scan_locs_task() prevented duplicate runs.
15:21:38 <Aeyoun> so I can just run scan_locs() where I am running it now.
15:25:46 -GitHub[nikola]:#nikola- [nikola] Aeyoun pushed 1 new commit to smaller-sitemaps: http://git.io/vZIzz
15:25:46 -GitHub[nikola]:#nikola- nikola/smaller-sitemaps 32e93e2 Daniel Aleksandersen: Prevent running scan_locs() more than once
15:26:34 <Aeyoun> I think that should be better?
15:27:46 <ralsina> Aeyoun: there's also the problem of scan_locs running too early. I am sure that was a problem in the past but can't find the issue:-P
15:28:10 <ralsina> if you run it when preparing tasks, it will not have any of the new files from posts being added and such
15:28:15 <ralsina> IIRC
15:28:19 <Aeyoun> ralsina: oh, f*. yeah. didn’t think about that.
15:29:14 <Aeyoun> ralsina: no way to generate tasks after  every other task, I presume?
15:29:45 <ralsina> LateTask if it works correctly will *run* after all other tasks, not sure if they are generated after all other tasks
15:30:16 <ralsina> but they need to be generated after the other tasks are already runned. That's why this was tricky in the 1st place :-)
15:30:41 -travis-ci:#nikola- getnikola/nikola#6248 (smaller-sitemaps - 32e93e2 : Daniel Aleksandersen): The build was broken.
15:30:42 -travis-ci:#nikola- Change view: https://github.com/getnikola/nikola/compare/f4257102e002...32e93e2c0bd1
15:30:42 -travis-ci:#nikola- Build details: https://travis-ci.org/getnikola/nikola/builds/79140560
15:31:25 <Aeyoun> I’ve got no idea how to do this then.
15:33:19 <ralsina> Aeyoun: I can't think very hard at this point. I'll suggest something later.
15:33:50 <ralsina> Aeyoun: I don't see why this usage needs different requirements/dependencies than writing a single sitemap
15:34:07 <ralsina> Aeyoun: in both cases, it depends on "the entire content of the site". Right?
15:34:20 <ralsina> We could even have a single task that writes all the sitemaps at once.
15:36:14 <Aeyoun> ralsina: one task would overwrite all the sitemaps every time. a HEAD from Google to any one of the sitemaps would tell Google that the sitemap had changed, even though the change as only initiated for another sitemap file.
15:37:03 <Aeyoun> I thought I could avoid that by splitting up into multiple tasks responsible for writing one file each.
15:37:10 <ralsina> hmmm
15:37:24 <ralsina> they can still all depend on scan_locs
15:37:53 <ralsina> I need to think a bit further :-)
15:38:02 <Aeyoun> ralsina: but I don’t know before scan_locs have run how many tasks will be created.
15:39:00 <ralsina> interesting problem
15:39:20 <ralsina> Anyway even running scan_locs doesn't tell you that if you run it too early
16:37:03 -GitHub[nikola]:#nikola- [nikola] ralsina pushed 1 new commit to idea-for-1814: http://git.io/vZIdm
16:37:03 -GitHub[nikola]:#nikola- nikola/idea-for-1814 b07ef5d Roberto Alsina: fix translation links
16:41:30 -travis-ci:#nikola- getnikola/nikola#6250 (idea-for-1814 - b07ef5d : Roberto Alsina): The build is still failing.
16:41:31 -travis-ci:#nikola- Change view: https://github.com/getnikola/nikola/compare/dd49ef6e26ff...b07ef5dca4d3
16:41:31 -travis-ci:#nikola- Build details: https://travis-ci.org/getnikola/nikola/builds/79151752
17:02:37 <ralsina> Aeyoun: the only difference in html_headstart between base and bootstrap3 is <meta name="viewport" content="width=device-width, initial-scale=1">
17:02:47 <ralsina> base has no initial-scale ... would it be horrible to add it?
17:04:49 <ralsina> I can get rid of a 70-line chunk of complicated HTML copied if I can make those the same :-)
17:12:47 <Aeyoun> ralsina: remove it from bootstrap3 instead
17:12:51 <Aeyoun> 1 is the default value …
17:12:57 <ralsina> awsome
17:13:22 <Aeyoun> adding max-scale would qualify as "horrible."
17:14:29 -GitHub[nikola]:#nikola- [nikola] ralsina pushed 2 new commits to idea-for-1814: http://git.io/vZLvM
17:14:29 -GitHub[nikola]:#nikola- nikola/idea-for-1814 f00cca5 Roberto Alsina: more simplification
17:14:29 -GitHub[nikola]:#nikola- nikola/idea-for-1814 e50f728 Roberto Alsina: jinjified
17:19:23 <ChrisWarrick> Aeyoun: bootstrap calls that tag mandatory…
17:20:03 -travis-ci:#nikola- getnikola/nikola#6252 (idea-for-1814 - e50f728 : Roberto Alsina): The build is still failing.
17:20:04 -travis-ci:#nikola- Change view: https://github.com/getnikola/nikola/compare/b07ef5dca4d3...e50f7289e7d5
17:20:04 -travis-ci:#nikola- Build details: https://travis-ci.org/getnikola/nikola/builds/79156016
17:27:31 <Aeyoun> ChrisWarrick: mandatory how? could be their JS does magic based on it’s existence. I wouldn’t know. On its own, it’s not doing anything.
17:28:18 <ChrisWarrick> Aeyoun: http://getbootstrap.com/css/#overview-mobile 
17:30:34 <Aeyoun> ChrisWarrick: "The initial-scale property controls the zoom level when the page is first loaded." (from spec) Setting it to 1 (default value) is meaningless on its own.
17:30:47 <ChrisWarrick> Aeyoun: and width?
17:31:10 <ChrisWarrick> Aeyoun: is this still valid? https://developer.mozilla.org/en/docs/Mozilla/Mobile/Viewport_meta_tag
17:34:10 <Aeyoun> Aeyoun: width is stil lvalid
17:34:16 <Aeyoun> *valid
17:35:17 <Aeyoun> ChrisWarrick: yes, that link looks accurate. (MDN usually is)
17:35:53 <ChrisWarrick> Aeyoun: kaboom, device width is the problem
17:36:31 <ChrisWarrick> how different is your theme from base, and does it look good on mobile? (you have device-width)
17:36:58 <Aeyoun> ChrisWarrick: looks good on mobile and TV.
17:37:20 <Aeyoun> ChrisWarrick: it’s not very different in terms of layout. one extra container for ads.
17:37:21 <ChrisWarrick> Aeyoun: On TV? How does chriswarrick.com and getnikola.com look there?
17:37:39 * ChrisWarrick has a dumb TV, but one that runs FreeBSD
17:38:11 <Aeyoun> ChrisWarrick: get an Xbox One (IE11) or PS4 (WebKit)
17:38:25 <ChrisWarrick> Aeyoun: #pcmasterrace
17:38:42 <ChrisWarrick> (though my measly PC overheats rather easily and is generally garbage)
17:39:31 <Aeyoun> Wii U? Nintendo Browser is … lets just say it’s an interesting take on a "web browser".
17:40:29 <Aeyoun> ChrisWarrick: my site uses slightly higher contrast colors on TV-like browsers.
17:40:53 <Aeyoun> because they’re usually over-exposed because default TV brightness settings are higher than most PC screens.
17:41:05 <ChrisWarrick> huh
17:41:24 <Aeyoun> other than that, its the same responsive layout.
17:41:43 <Aeyoun> ChrisWarrick: tl;dr: contrast is important on TV because it sucks by default.
17:53:10 <juanjo_> nikola import_twitpic -o posts/twitpic/ dumps/twitpic/
17:53:18 <juanjo_> just run coplete, it's almost ready
17:54:30 -GitHub[nikola]:#nikola- [nikola] ralsina pushed 1 new commit to idea-for-1814: http://git.io/vZLYO
17:54:30 -GitHub[nikola]:#nikola- nikola/idea-for-1814 cad8c28 Roberto Alsina: modified jinjify to account for parent-includes
18:01:34 -travis-ci:#nikola- getnikola/nikola#6255 (idea-for-1814 - cad8c28 : Roberto Alsina): The build is still failing.
18:01:35 -travis-ci:#nikola- Change view: https://github.com/getnikola/nikola/compare/e50f7289e7d5...cad8c28a5ca2
18:01:35 -travis-ci:#nikola- Build details: https://travis-ci.org/getnikola/nikola/builds/79160933
18:07:57 <elopio> Hello
18:08:11 <elopio> I'm saving my pages in the "notebook" directory.
18:08:23 <elopio> now how can I put an index in mysite.com/notebook?
18:21:04 <ChrisWarrick> elopio: how are you saving your pages?
18:21:14 <ChrisWarrick> elopio: You probably want STORY_INDEXES.
18:22:35 <elopio> ChrisWarrick: that's it! thank you.
18:23:25 -GitHub[nikola]:#nikola- [nikola] ralsina pushed 2 new commits to idea-for-1814: http://git.io/vZLCn
18:23:25 -GitHub[nikola]:#nikola- nikola/idea-for-1814 3040fa9 Roberto Alsina: typo
18:23:25 -GitHub[nikola]:#nikola- nikola/idea-for-1814 5ae6926 Roberto Alsina: fix paths on import too
18:27:22 -travis-ci:#nikola- getnikola/nikola#6256 (idea-for-1814 - 5ae6926 : Roberto Alsina): The build is still failing.
18:27:23 -travis-ci:#nikola- Change view: https://github.com/getnikola/nikola/compare/cad8c28a5ca2...5ae69264bed3
18:27:23 -travis-ci:#nikola- Build details: https://travis-ci.org/getnikola/nikola/builds/79163879
18:34:00 <juanjo_> i hate this error!
18:34:02 <juanjo_> ERROR: Nikola: You have tags that are too similar
18:34:32 <juanjo_> i'll change my plugin to instead of adding tags, first check if there is a similar tag already added and use it
18:38:20 <ralsina> juanjo_: check latest commit in the wordpress importer :-)
18:42:41 <juanjo_> mm i think that doesnt help me ralsina
18:42:53 <ralsina> juanjo_: same problem
18:43:37 <ralsina> in any case, yes, get the site's tags first and if your tag is too alike another one, replace it with the already-existing one
18:45:27 <juanjo_> i have tags like @pyconar conflicting PyConAr
18:45:44 <juanjo_> yes, i'll do that
18:46:06 <juanjo_> in the twitpic import i have too many "tags" that are @mentionds or #hashtags
18:46:18 <juanjo_> in nikola I aalready have a better version of that tags
18:46:50 <juanjo_> it will also be useful for the tweitter importer
18:47:47 <ralsina> juanjo_: for those you probably want to just remove the  @ or #
18:47:52 <ralsina> since those make little sense in tags
18:48:13 <juanjo_> yes
18:48:22 <juanjo_> anyway i end with different cases
18:48:31 <juanjo_> so the safer option is use the one in the site
18:48:46 <juanjo_> that could be another option for the wp ones
18:48:48 <juanjo_> one*
18:49:55 <ralsina> juanjo_: yes, you should comment that to felix
18:50:23 <ralsina> but I suspect in the wordpress case getting the tags of the site is not always possible and maybe more trouble than it's  worth
18:50:26 <juanjo_> i may add the code to utils
18:50:49 <juanjo_> yes, generally when you use the wp importer, you have no posts yet
19:05:01 <juanjo_> yay! now tag errors when building after twitpic import!
19:05:02 <juanjo_> ralsina: now this is how I add tags in import_twitpic
19:05:15 <juanjo_> self.tags.append(self.site_tags.get(utils.slugify(tag), tag))
19:07:32 <juanjo_> site_tags is a dict where they key is the sluglify version of the tag and the value is the tag
19:15:49 <juanjo_> nwo i have the "ralsina" tag in my blog http://picpaste.com/Captura_de_pantalla_2015-09-07_a_las_16.10.36-GQPADfOn.png
19:20:11 <ralsina> hahaha
19:28:32 <juanjo_> now i have a --tags
19:28:43 <juanjo_> i think i'm done with features
19:28:51 <juanjo_> now i have to polish it for PR
19:29:23 <juanjo_> --tags allows you to specify extra tags that will be added to all imported posts
19:29:51 <juanjo_> all import_* plugins should have it, what do you think?
20:14:20 <KwBot> [nikola] jjconti opened issue #2035: Adda https://github.com/getnikola/nikola/issues/2035
20:39:58 <gour> 1 /quit
20:41:02 -GitHub[nikola]:#nikola- [nikola] ralsina pushed 2 new commits to idea-for-1814: http://git.io/vZL9q
20:41:02 -GitHub[nikola]:#nikola- nikola/idea-for-1814 4d0d189 Roberto Alsina: undo post.tmpl changes :-(
20:41:02 -GitHub[nikola]:#nikola- nikola/idea-for-1814 59d3e15 Roberto Alsina: undo post.tmpl changes :-(
20:42:01 <juanjo_> example of import_twitpic http://www.juanjoconti.com/etiquetas/twitpic/
20:45:26 <ralsina> juanjo_: yes, the --tags thing sounds generally useful
20:48:13 -travis-ci:#nikola- getnikola/nikola#6259 (idea-for-1814 - 59d3e15 : Roberto Alsina): The build is still failing.
20:48:14 -travis-ci:#nikola- Change view: https://github.com/getnikola/nikola/compare/5ae69264bed3...59d3e1571a14
20:48:14 -travis-ci:#nikola- Build details: https://travis-ci.org/getnikola/nikola/builds/79181530
21:31:33 <ralsina> Anyone wants to ask for early access to a CMS that will work with any SSG? I am sorta all-early-accessed-out and can't take more stuff :-) https://www.reddit.com/r/Python/comments/3jxx1y/static_site_generator_cms/
21:31:57 <ralsina> It's by the bitballoon people, who have a track record of delivering stuff...
23:17:23 <juanjo_> anyone use netlify?
23:36:26 <ralsina> juanjo_: I experimented with bitballoon, their previous thing, was cool