2013-12-03 UTC
snarfed and bnvk joined the channel
# 00:15 Loqi bnvk: tantek left you a message 1 hour, 16 minutes ago: but these criticisms are coming from those who were in the community, that's the difference, that's what's new.
scor joined the channel
# 00:15 bnvk man, Loqi message later feature is awesome!
KartikPrabhu joined the channel
# 01:04 KartikPrabhu anyone know how to make form fields always show right aligned text? For example: I am using an input element to display permalinks to my posts. But the input field is small and so the url always appears with the site name, with the post name hidden to the right. UX-wise, it might be better if the end of the url i.e. the post name was visible with the site name hidden on the left. Any suggestions? See example here at the bo
# 01:05 tantek KartikPrabhu your message was cut-off after "here at the bo"
# 01:07 KartikPrabhu aaronpk: it can be done with text-align, but as soon as an input field loses focus it is back to left align :(
# 01:08 etymancer could you maybe do something funky with right-to-left text?
# 01:09 etymancer not if you reverse the text :)
# 01:09 snarfed KartikPrabhu: huh, just fyi, i don't see that losing focus bug in firefox
# 01:10 aaronpk KartikPrabhu: yeah add "direction: rtl" on your input box and it totally works
# 01:10 etymancer nice. don't even need to monkey with the text
# 01:10 snarfed i don't see the bug in chrome either. 31 on mac os x, and ff 25. just fyi
# 01:11 etymancer I like this channel.
benwerd joined the channel
# 01:19 KartikPrabhu aaronpk: direction:rtl works! except on Opera Mini for Android, losing focus reverts to ltr
npdoty joined the channel
scor and KevinMarks_ joined the channel
# 01:38 snarfed e.g. if you managed to blow away a bunch of your tweets, you could probably recover them the same way
# 01:39 tantek KevinMarks - hilarious. I think we've seen examples of "recovery from Google search cache backup" before. Not always entirely successful.
# 01:40 tantek that's because we were more comprehensive about specific blogs (at least their contents)
# 01:40 tantek when we should have been archiving their entire HTML permalinks
# 01:44 tantek KevinMarks, the irony is that it would have been *simpler* to just archive the curled permalink pages
# 01:45 tantek and would have enabled reparsing them later too (when parser/spider improvements were made)
# 01:45 tantek why we chose to store the parsed text and not the raw HTML I'll never know. Some attempt at premature optimization
# 01:45 tantek maybe by the same person who insisted we should not index Twitter
# 01:46 KevinMarks_ 'cos we were already db-focused by the time I joined (and dave wasn't even keeping posts, just links and 255 chars around them)
tilgovi joined the channel
# 01:55 tantek hard to break that db-focused culture isn't it?
# 02:00 KevinMarks_ yes, that was the bit I found odd about the "oh crap I used mongo" post
skinny joined the channel
# 02:24 tantek anybody else get home screen icons working on their personal site?
KevinMarks and skinny joined the channel
# 02:52 bret Does it have to be a picture of my face?
npdoty joined the channel
# 03:09 tantek bret - I hadn't seen it and their validator is wrong ;)
KartikPrabhu joined the channel
# 03:11 tantek bret - so no, during my research I didn't find it and it's a horribly misleading resource :(
tpinto joined the channel
# 03:44 bret benwerd: having issues sendin webmentions to your post
# 03:45 bret tantek: didn't try the validator, that sucks
jschweinsberg and lukebrooker joined the channel
# 04:55 tantek pdurbin - there's something missing from that page - a personal domain / URL for Daniel Siders
# 05:18 tantek oh Tumblr, you were so close to doing a decent default job with personal icons
# 05:28 tantek it has a iOS UA sniffer at least - and only sends the markup for iOS when it sees it (noted in prev section)
# 05:28 tantek part of the problem of doing UA specific coding
# 05:30 tantek I wouldn't hold my breath - since they've been acquired by Yahoo.
# 05:31 KartikPrabhu maybe the indieweb way will catch on. in any case Tumblr is doing much better than other silos, i guess
# 05:32 tantek it is. Tumblr has *many* redeeming design aspects.
# 05:32 tantek in fact there is much we can learn from Tumblr's UI design.
abrereton joined the channel
# 05:32 tantek everything from on-boarding to setting up your domain to editing your templates
# 05:35 KartikPrabhu was looking at POSSE. Is there a good way to syndicate art-directed posts? most syndications would destroy any custom styles
# 05:37 tantek KartikPrabhu - all POSSE is silo destination dependent
# 05:38 tantek and thus there's quite a bit of room for silo-specific "art-direction" ;)
# 05:38 tantek but inevitably, expect that the silo POSSE copy will be lower fidelity than what you can do on your own site
# 05:38 tantek however that's a *feature* - it encourages readers on that silo to click through and view the high fidelity original on your own site
# 05:40 KartikPrabhu yes. but I suspect it would be a problem with feed syndication too. sort of throws a wrench in the whole 'readers should read whereever they want' philosophy
# 05:48 tantek and thus if they're your friends, it's only polite to accommodate them, rather than make them read some other way
# 05:50 KartikPrabhu but most people would prefer to get all their internet at one place, be it a feed reader or Google+ or such. Is it a good idea to direct them back to to your site for 'the best experience'?
# 05:51 tantek rather it's good for you to express yourself as richly as you wish to, and not limit yourself by what silos allow
# 05:51 tantek POSSE is the compromise to at least give your friends something where they happen to read things.
# 05:52 tantek and yes - I personally think it's only ok to direct them back to your site when there's more content there
# 05:53 tantek KartikPrabhu - it is you that is asking the good questions ;)
# 05:54 KartikPrabhu tantek: I don't get the difference between permashortlinks and permashortids. Maybe a page on the wiki for that?
# 05:55 tantek KartikPrabhu - did you search for them on the wiki?;)
# 05:57 tantek ah - I meant permashortcitations - will add AKA
rtaibah joined the channel
KartikPrabhu joined the channel
cweiske joined the channel
melvster and LauraJ joined the channel
andreypopp joined the channel
# 07:01 tantek RSS was/is so poor a standard that it motivated dozens of people to develop the alternative of Atom
# 07:01 tantek similarly, any apologisms for Media RSS are misplace - also a poor standard from the start, and poor stewardship as well
# 07:02 cweiske I tend to believe the internet tv players vs. rss theory
# 07:02 tantek ultimately - all of these problems are just facets of the weakness that is combination of DRY violation and invisible side files that are always eventually neglected relative to HTML
# 07:03 tantek RSS was weak to begin with. It didn't take much for attack it.
# 07:04 cweiske tantek, how do you expect full text feeds to work in hatom?
# 07:04 cweiske have index pages with the full text of 10 articles?
# 07:05 tantek why ask how I expect them to work? just look at my home page for existence proof.
# 07:05 cweiske or do you expect the clients to parse the index page and follow each entry?
# 07:05 tantek my full text "feed" or h-feed of my index.html file takes *less space in bytes* than the equivalent in Atom XML
# 07:06 tantek all the work on XML sidefiles duplicates of what works *just fine* in HTML has wasted 10+ years of web dev time
# 07:07 cweiske I need a full library just to figure out the title
# 07:07 tantek then you're ignoring Amdahl's law for your own productivity, as is *anyone* who complains about parsing
# 07:07 tantek building good UX is 10x more important and 10x harder than parsing any of this stuff
# 07:08 tantek anyone who's wasting time arguing about easier/harder to parser is basically ignorant of that
# 07:08 tantek plus, yes, you need a whole library to parse RSS soup as well.
# 07:08 tantek the whole XML makes the world valid pipe-dream was always wishful thinking
# 07:09 tantek wishful thinking by programmers that don't understand the messiness of successful systems
# 07:10 cweiske then why webmention? pingback works, you only need a library to use it
# 07:10 tantek also, nevermind UX, all such parsing discussions are trumped by higher fidelity content you get by not violating DRY
# 07:11 tantek higher fidelity content > perception of easier parsing
# 07:12 cweiske does "DRY" mean that the content is only available in one format, or only in one location?
# 07:17 tantek.com edited /webmention (+834) "/* FAQ */ Why webmention instead of pingback (good to answer this up front for folks that are already familiar with pingback)" (
view diff )
# 07:18 tantek cweiske - both kinds of DRY violation are possible
# 07:19 cweiske tantek, "Less work to implement" - I can say the same against hfeed
# 07:20 tantek hahaha that's funny. you're either new at this, or trolling. optimizing for publishers over consumers has been a long-standing obvious design principle.
# 07:21 tantek (more optimal in the aggregate, number of people etc.)
# 07:21 tantek because that benefits more people, short and long term
# 07:21 cweiske there are much more consumers than publishers, so it should be the other way round
# 07:22 tantek there are far fewer programmers that write consumers rather than publishers
# 07:22 tantek "consumer" in this context refers to the code for consuming
# 07:24 cweiske moving your idea to the real world, newspapers aren't necessary if i can look at the raw data
# 07:25 tantek nope, newspapers are better UX than raw data and if you read up, that was the first thing I pointed out - UX matters orders of magnitude more than all the nits you're trying to pick
# 07:25 tantek and thus you're violating amdahl's law by focusing so much time on optimizing such a minor % of the work of making something work
# 07:28 cweiske the work the parsers have to do is per-user, not per-parser. thus it is important that it's easy to consume
# 07:29 tantek parser code re-use is far higher than publisher code re-use
# 07:29 tantek people keep writing new publishing code all the time
# 07:29 tantek then you're again focusing on the wrong thing
# 07:29 tantek human code writing time is far more a big cost / obstacle than CPU time
# 07:33 tantek Amdahl's law is about time spent optimizing something vs time that thing spends as a % of the total
# 07:33 tantek "maximum expected improvement to an overall system when only part of the system is improved"
# 07:34 tantek or rather parallelization is just one optimization strategy
# 07:35 tantek parallelization is a red herring in discussions of Amdahl's law
wyomingplease, friedcell and Jihaisse joined the channel
# 07:48 KartikPrabhu interesting discussion/debate on RSS. The publisher over parser philosophy seems to be the common theme through the indieweb ideas
srushe joined the channel
# 08:29 KevinMarks I'd argue that there is an indieweb aspect to the proliferation of readers and publishers of rss that that article documents to some extent, but that it was always hugely imperfect
LauraJ joined the channel
tpinto, glennjones, andreypopp_, rtaibah, Acidnerd1, adactio, CheckDavid, friedcell, scor, skinny, andreypopp, LauraJ, skinny_ and barnabywalters joined the channel
# 16:01 aaronpk LOL at the first comment on that comment cweiske linked to: "Would you please post this an complete page of its own on a blog somewhere. It's an excellent analysis/commentary on the situation."
bnvk and snarfed joined the channel
# 16:15 snarfed aaronpk: yup, gotta love generic comment spam. endlessly entertaining
# 16:17 aaronpk I thought it made an excellent point about owning your comments!
# 16:18 barnabywalters most comment spam is “yes i really like your blog it has good informations please continue to post”
# 16:19 snarfed my spam detector has overfitted on generic "this is a good post" comments
# 16:22 aaronpk oh good lord. there was a bug with my dropbox->flickr sync, and now the same photo has been added to my flickr account hundreds of times
# 16:24 aaronpk i'm gonna have to write a script to delete them, I can't do that bby hand!
tilgovi, dietrich, npdoty and KevinMarks joined the channel
# 17:04 XgF aaronpk: I once posted a Facebook album to a group. Facebook decided to post each image in the album individually...
# 17:14 XgF Everybody woke up the next morning with 120 Facebook notifications...
HereticLocke, KevinMarks, KartikPrabhu, jschweinsberg, fritzy, ryana, bnvk and melvster joined the channel
andreypopp, _6a68, dietrich and realzies joined the channel
KevinMarks and LauraJ joined the channel
# 19:13 KartikPrabhu I have a screenshot testing homescreen icons on android with FF and Chrome. How do I upload it to the wiki so people?
tantek joined the channel
# 19:23 tantek cool - just noticed the icon for indiewebcamp.com :)
# 19:26 KartikPrabhu tantek: FF on Android uses a low-res version of icons for some reason
bnvk joined the channel
# 19:30 tantek KartikPrabhu - just saw that! Very cool test shot
tpinto joined the channel
KartikPrabhu joined the channel
# 19:32 bnvk Hrm, weird! They each pick different files?
# 19:33 bnvk srsly, with standards likes these, who needs standards!
# 19:33 tantek KevinMarks - which OS / device / browsers were you using?
# 19:35 tantek KartikPrabhu in your home screen icons image - which browser versions were you using?
# 19:36 tantek so strange because those are the same versions I think Kevin Marks used
# 19:37 KartikPrabhu yes. and he seem to have gotten some frame for FF. But I get that on some, like adactio but not on others
# 19:37 bnvk I don't think I will buy another Apple smart device ever again- their mobile carrier lock in nonsense is absurd and has consumed 10+ hrs of my life trying to get my ipad working on Verizon
# 19:38 tantek bnvk - I thought you were very happy with your ipad mini mobile data setup - thought it was very portable and stuff?!?
# 19:38 KartikPrabhu aaronpk: The indiewebcamp.com icon does not show up on FF on Android. Adds a generic star icon instead. Could be because of .ico file?
# 19:38 bnvk I was, in Europe- apparently my iPad (bought in Germany) was locked to Dutch Vodafone
# 19:39 tantek.com edited /icon (+308) "/* Firefox for Android */ add Android comparisons section with screen shots comparing Firefox vs Chrome home screen icon adding" (
view diff )
# 19:39 bnvk but that "supposedly" was just the CDMA baseband, GSM should have been fine, but trying to sign up at ATT was also impossible!
# 19:39 bnvk Trying one last thing- fingers crossed
# 19:40 tantek Good luck - I really hope you figure it out and can document it for the rest of us.
# 19:41 tantek I was this || close to getting an iPad mini with celldata hardware because of it (likely tmobile if they have it).
# 19:41 KartikPrabhu tantek: cool. I'll try to keep it upto date if things change/improve with some icons
# 19:41 bnvk tantek: thanks, I'll do that right now. Is there a page started on IWC for this sort of process?
# 19:41 bnvk supposedly an American one should be able to go to EU and use GSM no prob- so it might still be a viable option
# 19:42 tantek bnvk - go for it and start indiewebcamp.com/ipad :)
# 19:42 bnvk maybe indiewebcamp.com/mobile-data-options ?
KartikPrabhu joined the channel
# 19:48 bnvk tantek: regarding logo icons, I notice you don't do the multiple sizes for Apple, like aaronpk did- has Apple improved the parsing so as not need the sizes="" attr?
# 19:54 bnvk Loqi: is tantek: on an asymptotical curve in respect to streamlining his textual output to equal a "pure creation / documentation" state?
etymancer, scor, glennjones and KevinMarks joined the channel
_6a68 and dentonjacobs joined the channel
thatryana joined the channel
# 20:33 bnvk tantek: I was thinking about adding a blurb re: "Updating (of logo)" but was unsure if that was something more better suited for hCard ?
LauraJ and bnvk joined the channel
KartikPrabhu and melvster joined the channel
# 20:47 tantek bnvk - yes ideally these devices would get your logo from the h-card that's already on your page
# 20:48 bnvk ah, gotcha! And hence "updating" would be the onus of the browser / OS to handle using said standard?
tpinto and KevinMarks joined the channel
# 21:00 tantek bnvk - I've thought about simply having Falcon read my h-card from my index.html home page template, get the "logo" from there, and then automatically generate the rel="shortcut icon" markup etc.
# 21:01 tantek so that I as the author/user/publisher only need to update the h-card on my home page index template to have it all work.
# 21:01 tantek reducing DRY violations on the authoring side at least
glennjones, kylewm, friedcell, eschnou, _6a68, abrereton and KartikPrabhu joined the channel
# 21:43 brennannovak.com created /mobile_data_solutions (+499) "Created page with "This page about '''mobile data providers''' that outlines alternative methods of getting on the web that meet the following criteria: * Does not require a physical wired connect..."" (
view diff )
KevinMarks, bnvk, _6a68, acegiak and lukebrooker joined the channel
# 23:00 jonnybarnes Trying to set up Indieauth and I get the following error for twitter: There was an error verifying this provider. Confirm you have a rel="me" link on this site pointing to your website.
bnvk and glennjones joined the channel
# 23:09 aaronpk jonnybarnes: for some reason we're having issues with twitter right now (their SSL cert is erroring out on the server)
_6a68 joined the channel
# 23:15 bnvk tantek: ah, yah that makes sense as a solution! Both of mine are dynamically generated from the same setting / file!
skinny joined the channel
# 23:16 jonnybarnes github is working, need to edit Google+ the page that goes back to indieauth.com is slightly different to the page I link to
skinny and dentonjacobs joined the channel
# 23:34 jonnybarnes I may have found a bug with Indieauth and app.net, my domain is verified but it doesnt work
# 23:35 aaronpk and you're linking to your https URL from your app.net profile?
# 23:36 aaronpk I thought app.net domain verification worked with a rel-me link from your site to your app.net profile
# 23:36 jonnybarnes the way Ive verified my domain with app.net is to add a txt entry to my dns
# 23:43 aaronpk well I'm not actually sure if any of the people who run https-only sites have done app.net verification