2014-05-11 UTC
ss1081, becker_11, luxagraf and grantmacken joined the channel
JasonO, GWG, iangreenleaf, tpinto, tilgovi, luxagraf, gRegor`, squeakytoy, netweb, krendil, caseorganic, glennjones, KevinMarks, catsup, Garbee, bnvk and barnabywalters joined the channel
kbs joined the channel
# 12:55 kbs !tell aaronpk backup looks good :) btw, any chance you might also be able to expose an rsync endpoint?
# 12:55 Loqi Ok, I'll tell them that when I see them next
luxagraf joined the channel
cweiske joined the channel
# 13:44 GWG I turned my Facepile theme code into a Wordpress plugin.
# 13:45 GWG Need to find some testers to comment, as it is a bit rough.
jonnybarnes and kylewm joined the channel
luxagraf, tpinto and kbs joined the channel
# 14:20 Loqi Ok, I'll tell them that when I see them next
bnvk joined the channel
# 15:03 ben_thatmust !tell barnabywalters great, i searched for your site once and now every note pops up in my Google Now cards
# 15:03 Loqi Ok, I'll tell him that when I see him next
# 15:04 kbs thought google now has a way to skip/ignore certain cards?
# 15:10 ben_thatmust well you can swipe them away, and in certain cases you can tell it "No, I don't want to see updates on this any more" but with "updates from sites you read" its all or nothing
# 15:11 kbs ben_thatmust: aah, I see. annoying
# 15:11 ben_thatmust plus I kind of want updates on the articles i don't necessarily want every note
# 15:11 ben_thatmust google just needs a better management so i could switch it to the articles only stream or something
eternicode and anos joined the channel
# 15:28 kbs I was just updating the same page, let me finish that and take a look to see how I can do my 0.02 a little better
# 15:32 kbs rascul: love you've got both http and rsync endpoints. Would be awesome if there's something similar directly from the source website as well
# 15:32 rascul heh i was thinking of making a git repo out of it
KevinMarks joined the channel
# 15:32 kbs yep - wget is basically just as good
# 15:32 kbs I did notice some small deltas between a diff -r of your site, and what I got from btsync
# 15:33 kbs not significant diffs, (basically, index.html files and possibly some btsync-has-not-yet-got-to-it sort of things)
# 15:34 kbs I didn't look very closely - just ran diff -r and eyeballed the diffs
# 15:34 kbs is not a huge fan of btsync despite all its niceities
# 15:34 rascul nginx and rsync are both serving directly from the direcotyr btsync puts it all in
# 15:35 rascul so as long as btsync keeps it updated it should be the latest copy on indiewiki.rascul.io
luxagraf joined the channel
# 15:37 kbs there are a few chains where things can break down - at the moment, it looks like aaronpk is probabbly exporting changes periodically. Then, btsync -> your site etc. I think it's in reasonably good shape all in all
# 15:37 kbs especially once a few more people have some copies.
# 15:39 rascul i think aaronpk was making something to change locally and push changes somehow to the wiki? that would be nifty
# 15:39 kbs dunno. my thought was really something more dumb - eg: the mysql db fat-finger drops a table, export pumps out a subset of files, btsync updates everyone to a smaller portion of wiki etc. Hoping git ought to take care of it to some extent, especially as it's easy to replicate along with all its snapshots.
# 15:40 kbs I much rather use your http endpoint because it's easier to script than btsync (and not proprietary, for what that's worth)
# 15:41 Loqi aaronpk: kbs left you a message 2 hours, 46 minutes ago: backup looks good :) btw, any chance you might also be able to expose an rsync endpoint?
# 15:41 Loqi aaronpk: kbs left you a message 1 hour, 21 minutes ago: another thought (if simpler than opening up rsync) is an http endpoint would also be neat [c.f. rascul and http://indiewiki.rascul.io ]
# 15:42 kbs aaronpk: just starting to set up my random two-bit things to have a copy of your wiki backups - stuff related to that (and, thanks! :)
# 15:46 aaronpk awesome, already 2 copies on significantly different platforms
# 15:46 aaronpk yeah it's a little annoying btsync is proprietary, but not the end of the world
# 15:47 kbs (also http would be just fine from my point of view at least. rsync lets it use its internal incremental update thing, but more a nice-to-have rather than anything else)
# 15:49 kbs has usually found rsync daemons to chew up cpu on the server end
# 15:49 aaronpk incremental updates are good since there are a bunch of photos
# 15:49 aaronpk oh yeah good point. well this server isn't exactly at capacity, so i could give it a shot
# 15:49 kbs ^^ :) best source of what works best
# 15:50 kbs wonder if that's because I run --checksum on rsync (don't trust timestamps quite that much)
# 15:51 kbs I can run a test here if you like, one sec...
# 15:53 kbs hm, wouldn't the checksum be driven by the client? dunno - testin anyway :) making sure I'm giving the magic options
# 15:56 bret aaronpk: re syncing a copy of the wikii
# 15:57 aaronpk i should probably do a blog post about my new outlook on backups
# 15:59 aaronpk kbs: i should use --delete on the rsync command when cloning down, right?
# 15:59 kbs looks up the man page to make sure
# 16:00 aaronpk my rsync client seems to be trying SSH, is there a way to force it to use rsync protocol?
# 16:05 kbs the scripts just keep rolling through a series of folders with rsync. periodically, offline archive a tarball of one or more of the folders as well. Usually handy to go back after I've accidentally rm -r'ed a folder or two
# 16:07 bret aaronpk: are you still going to use btsync?
# 16:07 aaronpk cause btsync will get incremental changes immediately
# 16:08 aaronpk i've never used rsync in rsync protocol mode before. I'm getting what appears to be a directory of the rsync paths i've defined when I try to run it
# 16:09 kbs aaronpk: paths - great, looks good to me
# 16:09 kbs rsync is particularl obtuse about trailing slashes in paths, etc etc
# 16:12 kbs if I remember, the doublecolons are basically named endpoints you can additionally define in your rsyncd.conf
# 16:14 kbs oh, are you trying to use rsync's own authentication mechanism?
# 16:14 aaronpk yeah, is there a better way? I don't want to do ssh auth
# 16:15 kbs well, would probably ease some of your worries to just try to secure one thing maybe
# 16:15 kbs also, maybe [if say you're able to chroot your rsync stuff] could leave it open
# 16:16 aaronpk i guess a no-shell ssh user where i put the password on the wiki could work... but that seems sketchy
# 16:16 aaronpk but it seems less scary to have relatively public rsync access than try to set up semi-public ssh access
# 16:17 kbs only thought then is to see if you are able to jail your rsync process
# 16:19 kbs rascul: thoughts on alternatives?
# 16:19 rascul container it if you want something of that sort
# 16:20 kbs but more and more, I'm also starting to think that maybe starting with just an http endpoint might be the least complicated
# 16:20 aaronpk or I could run your git script and host a git endpoint
# 16:23 kbs thanks rascul - I used to stick with chroot jails till now (on freebsd mainly) but lxc looks neat
# 16:25 aaronpk for some reason there are 5 files that it thinks are different every time I run rsync
quinn_ joined the channel
# 16:27 aaronpk kbs: does your git script handle deleted files too?
# 16:27 kbs aaronpk: it should delete them from the latest commit head, but it will remain in the archive
# 16:30 kbs yay, another eye to squish bugs, happy!
# 16:31 aaronpk ok well... considering the pros/cons of git-over-http vs rsync
# 16:32 aaronpk git-over-http means I don't have to host another daemon and open another port, so in theory is more secure
# 16:32 aaronpk however it's also more error prone as there is an intermediary step of committing the changes to the git repo first
# 16:34 kbs so considering just serving pile of txt files over http? let the clients archive it as they see fit?
# 16:34 Loqi gives aaronpk some sort of incremental thingy
# 16:36 rascul can you make the wiki update comments into git commit messages? that would be nifty
# 16:36 kbs I suspect there's always going to be some step after exporting from your mysql db to get something incremental from your 'pile of txt files'
# 16:37 kbs unless you find think rascul 's idea of wget -rm covers it adequately (ie, timestamps)
# 16:37 kbs I guess you'll now need to make sure your exports are also preserving timestamps, ah well
# 16:38 aaronpk actually part of the export is setting the timestamp on the filesystem to the date the page was last edited
# 16:38 kbs so maybe that's all that's needed to start with
# 16:38 aaronpk i'd need to publish a directory listing of all the files
anos joined the channel
# 16:40 kbs hm, isn't that usually automatically done by the server? Guess that's how ngnix seems to be doing it on rascul's endpoint
# 16:46 kbs robots.txt ... [let me see if I can ignore it]
# 16:47 kbs I should probably put a sleep there to avoid getting it too fast
jedahan joined the channel
brianloveswords joined the channel
# 16:50 kbs Last-modified header missing -- time-stamps turned off
# 16:51 kbs oh wait - I guess that's just for the index listings
erikmaarten joined the channel
# 16:55 kbs okay - great - I'll cross-check with the btsync version when it finishes, and update my scripts to pull from this endpoint. [http is easier to manage for me than btsync, so very happy if this works out.]
# 16:56 kbs I guess what I should do is to an incremental pull mostly, and a full pull every once in a while
# 17:06 kbs I have a feeling wget doesn't have a convenient way to remove deleted files
# 17:06 kbs well, besides the brute force way I guess :) *continues to think*
# 17:06 aaronpk I guess a periodic full-sync would accomplish that
# 17:07 aaronpk I wonder how long until a full download becomes prohibitive
# 17:09 kbs well - this sounds like a client problem too
# 17:10 aaronpk also could you rewrite that with the full name flags? will be easier to look up individual flags that way
# 17:11 kbs (I'm also looking at httrack haven't used it, but seems more specific for this purpose)
# 17:12 kbs timestamp - -m == --mirror -> equivalent to --recursive --timestamping --level inf
# 17:14 GWG I'm trying to figure something out.
# 17:14 GWG If I took ChromeOS off my Chomebook and installed a full Linux OS, is that de-siloing?
# 17:14 bret so rad, a whole copy of the wiki in a few mins
# 17:15 bret lol jk as long as you can get your stuff in and out of it
# 17:15 GWG I just got annoyed with IRC on Chrome.
# 17:17 GWG bret: I tried a few things. I may go back. But for now...I switched it to Fedora
# 17:17 aaronpk kbs: feel free to add a section there for httrack!
# 17:17 bret GWG: if you can hang with linux desktops, its probrably for the better
# 17:19 GWG I tried ChromeOS, and may again, as something I didn't mind leaving my house.
# 17:23 GWG I may be one of the few people who liked the 10-11 inch netbook form factor.
# 17:24 bret aaronpk: say the wiki blows up. is it easy to restore from the .txt backup? or are these mainly for archival purpose?
# 17:25 aaronpk when my site was mediawiki, I had these text files synced locally and I could edit them and it would push the changes back to my site
# 17:25 aaronpk obviously the edit history doesn't come as part of these text files, but that is ok I think
# 17:26 bret this makes the wiki so explorable now, i enjoy the new perspective
# 17:26 aaronpk yeah! side effect is you get an offline copy of the wiki you can read!
# 17:27 aaronpk a weekly epub with only pages that have changed that week
# 17:28 bret it would be interesting to run a dead link check across all articles
# 17:28 GWG I've been playing around with mailing lists, and thinking about non-instant syndication
# 17:29 aaronpk i bet someone could come up with a small shell script to extract all the URLs out of the txt files
KevinMarks joined the channel
# 17:37 kbs (wiki-archive is the btsync version)
jonnybarnes and chrissaad joined the channel
# 17:41 Jeena so it come to this that I am thinking about redesigning my homepage. When I only had a blog there it was a no brainer to only have a list of blog posts there, but now with notes and photos I kind of would like to have them there too.
# 17:45 GWG The question is, what do you want that to look like?
# 17:46 Jeena I'm really not a fan of the composite stream, it is perhaps ok for a feed but it is so confusing on a homepage
# 17:46 aaronpk kbs: fixed! was a rogue entry in the nginx config
# 17:46 kbs aaronpk: cool, one more test here...
# 17:47 Jeena GWG, that is a really good question and I don't really have an answer to that. This was my attempt on my old website: http://jeenaparadies.net/ which I think is quite ok.
# 17:48 Jeena and I'm also not a fan of full content on the index page
# 17:49 aaronpk Jeena: I also don't like full content on my home page! I've always done article summaries
# 17:50 Jeena and I have to say that http://tantek.com/ is one of the bad examples, there is no structure, just a wall of text with the occasional image and a sidebar
# 17:52 GWG Jeena: Tantek claimed when I met him last month he was trying for an 'old' style Twitter vibe.
# 17:53 GWG Jeena: I tried having the Notes a larger font size and different style than the articles. That seemed to work.
# 17:59 kbs aaronpk: the btsync archive has (for example) newbase60.Conflict.txt, while the http archive has newbase60-2.txt -- kosher?
# 17:59 GWG Jeena: You can also use subtle shading.
# 18:00 GWG Jeena: I can't say mine is perfect...still a work in progress, but you are welcome to look.
# 18:00 aaronpk that's usually due to case-insensitivity on the client
# 18:00 kbs ah - filesystem case sensitivity strickes again
# 18:00 aaronpk i'm trying to avoid it but may have missed some edge cases, will check again
# 18:03 aaronpk there we go! you all should see some changes in btsync!
# 18:03 aaronpk specifically check out the wiki/backup.txt file for the new section at the bottom
# 18:05 GWG Jeena: I still think the about stinks.
# 18:06 GWG But you'll see the different styles.
friedcell joined the channel
# 18:07 Jeena I kind of get confused that everything is just in one column, but yeah I can see the differences.
# 18:07 Jeena have you thought about different background colors for different types?
# 18:09 Jeena hm and what is this for, GWG? <link rel="feed" href="http://david.shanske.com/">
# 18:09 Jeena Firefox shows it as one of the Feeds when I click on the RSS button
# 18:10 aaronpk just added the URL at the top of every file, sorry for the full resync
# 18:10 Jeena and what is the difference between a status and an aside
racsul joined the channel
# 18:11 GWG Jeena: h-feed, I believe. Some of the markup needs to be fixed.
quinn_ joined the channel
# 18:12 Jeena hm I don't think h-feed needs a <link> in the head
# 18:12 GWG Jeena: You can see the different types in the top menu.
# 18:12 GWG Jeena: I did a lot of the markup in the wee hours.
# 18:12 kbs aaronpk: haha - thanks for the heads-up :)
# 18:13 Jeena yeah I can see the types but the content seems to be of the same type ;)
# 18:13 GWG Jeena: As for status vs. aside...they are Wordpress conventions. For me... A status is an aside with a profile picture next to it....
# 18:13 Jeena I was just confused about the semantics ^^
# 18:13 GWG Jeena: The types are something Wordpress calls post formats. They are styling differences.
# 18:13 GWG Jeena: They aren't post types, precisely.
# 18:13 GWG Jeena: I have plans for post types, but in the interim, I'm using the format types
# 18:14 Loqi Ok, I'll tell them that when I see them next
# 18:15 Jeena I see, I like what you did with your website
# 18:25 kbs (http endpoint has "_", while the btsync has " ")
# 18:25 kbs (I guess these must not actually be the same txt files?)
# 18:25 aaronpk hm my btsync version shows the underscore version
# 18:26 kbs the http version has the undercores, the btsync version has spaces
# 18:27 kbs also saw some txt files have disappeared (yay for git :-) which let me test the deleting things too
# 18:27 GWG Jeena: It is all Wordpress. I'm working on development. And yes, I know I need more images.
# 18:27 aaronpk i wonder if btsync is using timestamps internally
# 18:27 aaronpk cause the timestamp didnt' change because I force the timestamp to the last edit date
# 18:27 GWG Jeena: I have been splitting off pieces of my design into plugins, which means I'm rewriting code, but it isn't visible.
# 18:29 kbs aaronpk: looks good to me more or less - remaining niggling things are only the case-sensitivity issues
# 18:29 kbs but I'll start switching to the http endpoint and update my scripts (also wiki, etc.)
# 18:32 GWG Jeena: When I'm done, anyone could use the plugins I'm creating on their sites.
KartikPrabhu joined the channel
# 18:33 Loqi KartikPrabhu: kylewm left you a message 19 minutes ago: hey I abstracted/cleaned up the reply-contexts and comment parsing that I use in Red Wind. Take a look if you have a chance :) https://github.com/kylewm/mf2util
jgee, KartikPrabhu and _6a68 joined the channel
# 18:58 aaronpk it just occurred to me that I should probably include the last modified date in the txt file as well
_6a68 joined the channel
# 18:59 aaronpk cause if there are a bunch of copies of this floating around we need to know quickly if one is out of date
paulcp and KartikPrabhu joined the channel
# 19:10 kbs aaronpk: can't hurt I guess. (btw, I've settled back to wget - httrack is too convoluted to use, and I hope to do a full pull every so often anyway.)
KartikPrabhu joined the channel
KevinMarks joined the channel
# 19:29 kbs aaronpk: how are you managing passwords via the rsync mechanism? [should I use that rather than http, if you're planning to keep it around?]
# 19:30 aaronpk kbs: I figured I'd just create an "indieweb" user and put the password on the wiki
# 19:31 kbs oh, okay :) i'll check there when you've updated it
# 19:32 aaronpk that script was potentially useful, maybe link to a gist of it?
# 19:33 kbs oh :) sure, I'll put it in a gist. I changed it to start using wget
# 19:33 kbs but if you're going to enable rsync, I'd rather switch to it as well
# 19:34 aaronpk would like to know how wget works out for you and if you see other advantages to rsync
# 19:34 kbs gotcha. I think the main thing is that wget doesn't have a good way to disambiguate between wifi login endpoints and other urls it's getting
# 19:35 aaronpk at least then you'd get a cert warning if you're MITM'd
# 19:35 kbs oh right - I should have the right certs enabled
# 19:35 kbs thanks - yes, that should take care of it nicely
# 19:36 kbs gotto go here, will let it run over a few days and keep you posted if there are any issues. nice job on backups aaronpk :) hopefully enough copies around, much more safe!
# 19:36 aaronpk *headdesk* I just wrote a tiny mediawiki skin to return the contents of a wiki page without any wiki sidebar stuff, then realized I could have just run the page through the mf2 parser
indie-visitor joined the channel
# 20:03 Loqi Welcome, indie-visitor! Set your nickname by typing /nick yourname
# 20:08 tbrun I am new to indieweb and am setting up to both own my content and looking forward to installing the components that make this a success. I have my own domain name, I've set up wordpress on a hosted site and have gotten indieauth installed and woking. I trying to figure out how to initiate POSSE for twitter and facebook. So far I haven't found any place that can help me set it up. Any help is appreciated
# 20:10 GWG tbrun: I use Wordpress. I'll be happy to help.
gRegor` joined the channel
# 20:15 tbrun I installed the jetpack plugin and my second question was going to be: What is the difference btw POSSE and Jetpack? Now I don't need to advertise what a noobe I really am. Thanks
# 20:16 GWG tbrun: Publicize is a horrible plugin.
# 20:16 GWG tbrun: There is no configurability in it.
# 20:16 GWG tbrun: Although it is the only thing that I know that does G+
# 20:18 tbrun jWell now I have two opinions, what to do? what to do? This might be a stupid question, but what kind of configurability to you want/need and is there a better solution in your mind
ttepasse joined the channel
# 20:22 KartikPrabhu tbrun: I think the Jetpack plugin is a decent start to do POSSE. I don't know how configurable it is, but it is used by a lot of people here
# 20:25 GWG tbrun: I use SNAP - Social Network Auto Poster. People also like Social by Mailchimp.
# 20:25 GWG KartikPrabhu: You can't configure the output, mostly.
# 20:26 GWG tbrun: I'm also working on something with Bridgy. Which will work with Wordpress if you have Webmention support
# 20:27 KartikPrabhu tbrun: yes that is what i suspected, but for people doing it first time Jetpack should be decent
eschnou joined the channel
# 20:31 GWG I got rid of Jetpack because of how much it makes you dependent on Wordpress.com services.
# 20:32 GWG Most of the plugins inside Jetpack should not need an account.
krendil joined the channel
# 20:33 GWG But it was easier to go for Google Analytics
# 20:35 GWG KartikPrabhu: Did you decide if you were coming to NYC?
# 20:36 tbrun GWG: Is SNAP the NextScripts plugin?
# 20:38 GWG Disclaimer: The Opinions Expressed by GWG are his own, and not necessarily those of the Indieweb movement.
# 20:39 GWG I would have moved to Bridgy if it hadn't had Facebook issues snarfed hasn't been able to isolate
# 20:39 GWG tbrun: What is the site, by the way?
# 20:40 tbrun GWG: I thought it was just one person's opinion, but thanks for the disclaimer. I'll muck around and see how it goes.
# 20:40 tbrun GWG: The site is www.tombruning.com. I am just starting out with this so be kind :)
# 20:41 GWG tbrun: I'm rather opinionated at times.
# 20:41 GWG tbrun: I'm always interested in other Wordpress users getting into the Indieweb
# 20:42 tbrun GWG: I'm open to criticism, but obviously the more constructive the better.
# 20:43 GWG Have you installed the webmention and semantic linkback plugins?
# 20:43 GWG And signed up for Bridgy to send comments/likes/etc back to your site?
# 20:45 tbrun GWG: Not yet, but I guess those are the next activities to do. So webmention, semantic linkback and sign up for Bridgy. Is there anything I need to do on my site for bridgy, or is just signing up what I should do?
# 20:46 GWG Nothing on the site side, other than add webmention support
# 20:46 GWG Bridgy just takes information from various sites and converts it to webmentions.
# 20:53 tbrun GWG: I'm off to work on this stuff, Thanks for all your help, I need to deal with Mother's Day now
# 21:09 GWG Still not sure this was the best way to use spam
jedahan, paulcp and pasevin joined the channel
iangreenleaf, jeremyzilar, erikmaarten and barnabywalters joined the channel
# 22:25 Loqi barnabywalters: ben_thatmust left you a message 7 hours, 22 minutes ago: great, i searched for your site once and now every note pops up in my Google Now cards
# 22:27 aaronpk is trying to come up with a weekly summary of wiki changes
# 22:33 GWG barnabywalters: How do you do that?
# 22:33 GWG barnabywalters: I want to invade Google Now
# 22:37 GWG barnabywalters: I forgot...it's limited geographically.
# 22:49 aaronpk planning to generate HTML files that can be pushed out to Kindle for example
caseorganic joined the channel
# 22:50 aaronpk cause this would give me a great excuse to use my kindle more :)
# 22:53 GWG I keep thinking of upgrading from the other one
# 22:53 ben_thatmust GWG, barnabywalters, I'm guessing it works via atom feed which both you and tantek use and I have to have searched for your site in google
# 22:53 aaronpk it's just going to be basic html, i think it'll work on any
# 22:54 GWG ben_thatmust: Atom feed or just search index/sitemap?
# 22:56 ben_thatmust i doubt its just a sitemap, it seems to be for only pages i visit not whole sites
# 22:56 ben_thatmust i could try an make a list of sites i get updated about and we'll see if we can figure it out
# 22:56 GWG ben_thatmust: The sitemap tells Google about changes to the site.
# 22:57 ben_thatmust tantek and barnabywalters site's, xkcd.com, and thedailywtf.com all send me updates
# 22:58 GWG So, you search on Google, Google notes your search history, and when one of the sites pings Google to tell it that it has been updated, using the XML sitemap, it indexes the story and sends it to you.
# 22:58 GWG I don't think Google is using the RSS
jedahan joined the channel
# 23:12 aaronpk so I was thinking that the edit comments may actually be useful
# 23:14 aaronpk just downloaded a wiki article to my kindle for a test, it looks great!
snarfed joined the channel
# 23:22 aaronpk the awesome thing about kindle is if you send an html attachment to your kindle's email address it turns up looking like a real book
barnabywalters joined the channel
# 23:26 barnabywalters aaronpk: well, maybe this is a good place to actually set guidelines for what “minor” edits are
# 23:26 barnabywalters major edits are ones which show up in the roundup, minor ones are ones which get left out
# 23:27 aaronpk also this should encourage people to include better edit comments
barnabywalters joined the channel
# 23:34 aaronpk making the html version of caseorganic's book was a nice throwback to 1990s web development
# 23:34 Loqi aaronpk meant to say: making the kindle version of caseorganic's book was a nice throwback to 1990s web development
# 23:41 aaronpk question: is the time part of the timestamp relevant? i'm thinking just the date may be