Homebrew Website Club

Tantek Çelik:

Welcome to Homebrew Website Club SF - we'll talk about what we've done on our own sites

Indiewebcamp Dusseldorf happened last weekend https://indiewebcamp.com/2016/D%C3%BCsseldorf

Homebrew Website Club is based on the Homebrew Computer Club that happened in the 1970s and seeded the computer revolution

a lot of people just use social media, but there are some of us who hack on our websites, so we do it for the web

Indiewebcamp is more based on Barcamp - a structured conference over 2 days where the agenda is made by who arrives

We start indiewebcamp with demos and discussions. The 2nd day is hack day, and we end with demos

Indieweb summit is coming up in Portland http://indiewebcamp.com/2016 followed by the w3c SocialWG meeting

Webmention is going to w3c Candidate recommendation

Kathy Ems:

My app is at http://tranquil-caverns25242.herokuapp.com determining your outfit for your next adventure

you enter your temperature and activity it filters your activities by temp and what you wore before

currently the data is in a static variable, but I'll add it to a database soon

Tantek Çelik:

can you save the outfit as a mini blog post - like instagram's #OOTD outfit post?

Kathy Ems:

I do take photos of my outfits so recording that too would make sense


I rewrote my site backend again http://notes.whatthefuck.computer/ it exposes my weird quantified self data to the world

it generates static html on mysite from scratch so it take longer and longer each time

I have a lot of differnt posts types, waling reading and more - if I want to syndicate it uses brid.gy

it gets sent out to twitter and facebook, and I pull likes back too

so now I cna go and backfill the indiewebcamp pages to say thsi supports them

It's all static so I could theoretically host it on S3 or github, though it is on a page on my own site right now

I like low contrast type in my code as I work on it late at night - my primary OS is this crazy LISP environment

it's called Arcology- it's on my own git repository, not github

I'm probably going to have to stop usingg emacs lisp as I have over 100M of files to translate into this

Tantek Çelik:

Static sites are great and more and more common, but you can't set response codes for redirects or 410

one thing we came up with for webmentions is to delete a comment using an http 410 Gone

and then send a webmention to the original place that you sent the webmention so it can delete the comment

to enable 410 for static sites, we could use a Meta http-equiv for the status header http://indiewebcamp.com/meta_http-equiv_status

I wrote this up as a spec on the indiewebcamp Wiki - you publish in the head

if you are serving the site, you should return this code. Similarly, if you receive this tag you should treat it as a 410

I tweeted it at 5am this morning, and within an hour @tommorris implemented it for Django

Never underestimate the power of going to the beach.

Ari Bader-Natal:

In s3 you can set headers for a page

Tantek Çelik:

those look like the ones that meta http-equiv as specified at the moment, so we should push for adding new ones

Ari Bader-Natal:

on my website https://aribadernatal.com I have been trying to collect various things: stars, likes and so on

I use IFTTTT to take posts frome the various silos and make it a post on my wordpress site

pocket, github, youtube, kickstarter and son on get fetched and posted ot my own site so I have a copy

I can search across all of the things I have collected; I have pages and feeds for each topic

Tantek Çelik:

it looks like a personal Pinterest

Ari Bader-Natal:

after I built all this I showed my wife it and she said "isn't that Pinterest?"

I use Amber, a wordpress plugin that makes a snapshot of sites you link to on your own site

I wrote a pst about using Amber to save links: https://aribadernatal.com/2016/02/15/saving-favorites-amber-edition/

amber has a dashboard of every page it tries to index - whenever someone links form twitter it is a t.co link

twitter robots.txt on t.co/robots.txt has User-agent: * Disallow: / so you can't spider links in tweets

most sites I link to don't actually work to be archived for various different reasons

Tantek Çelik:

personal archives is broadly seen as a legitimate reason to ignore robots.txt—you aren't spidering, you're fetching as a user

time for a photo http://known.kevinmarks.com/2016/homebrew-website-club-sf-2016-05-18

See IndieNews