Ended up being really useful whenever I wanted others to see what I had experienced on personalized websites (e.g. ranked feeds that would look different on different accounts, or change if reloaded e.g. https://ctxt.io/2/AAAAiPRWFg), interview feedback, or even paywalled content if I just wanted to share a little snippet for fair-use and keep the HTML styles, https://ctxt.io/2/AAAAACI6FQ. Scroll to the bottom of the page for these examples.
The main technical difficulty here was in trying to get all the CSS styles to be as accurate as possible, and the browser extension solves many of these, including proxying requests whenever necessary to bypass some restrictions. The browser extension doesn't request any permissions, and won't inject any content scripts on your page, so as someone who cares about security, is what I wouldn't mind installing personally.
One cool thing I've discovered after building it is that it's super useful for web developers to modify the DOM of pages and redo mockups and then use Context to re-share them with others for them to also modify: e.g. https://ctxt.io/2/AAAA0KGcFA
Everything is free for now (don't worry about the $ signs), mainly interested in feedback. Thanks in advance!
How do you proxy the requests?
I think it's a great idea and incredibly useful.
[0]: uploaded it to https://github.com/samuelantonioli/richie (added "expires", good idea)
p.s.: I've solved the problem with fixed headers using a fixed header with a high z-index (unfortunately this hides the fixed header of the copied website).
p.p.s.: Maybe it's possible to build it as a small community-maintained project, that would be great.
Happy to think of ways to collaborate.
You should say this on your site. Found it rather confusing until I read your reply here.
I think "only works if you're copying rich text to begin with" is the kinda the point.
(I'm not affiliated with ctxt.io at all, just trying to understand it like you are.)
Example page: https://www.golem.de/ticker/ - I tried to copy some listed links but they just look like standard Times New Roman after contextification: https://ctxt.io/2/AAAAAPxhEA
Anyway, nice idea. I have ever used screenshots so far.
Firefox extension or the bookmarklets on https://ctxt.io/faq would solve this and the former should be out soon: https://addons.mozilla.org/en-US/firefox/addon/context-ctxt-...
Although, at this moment, my Firefox (52.0.1 on GNU/Linux) says the extension cannot be installed because it is faulty what ever that might mean.
Do you have any plans to release the source code?
Is this because you would rather verify things are what I say they are (no content stored elsewhere) and use this service, or to run your own, or just learn?
Right now, code quality isn't in a state to open-source, but I'm thinking about it.
I see lots of people mentioning this, but imo, it doesn't matter much. You can definitely upload your work and later change it as you work.
Isn't that how it should be done.
Like others here, at first I thought it was just another pastebin. I didn't realize what it could do until I read your comment here in this thread. Seems dumb, but I didn't notice I could scroll down for examples, and I also would have had no idea about the browser extension. I would have dismissed this almost immediately had I just visited the page. Perhaps you can make this things more obvious?
Also, any plans for a Firefox extension?
Note: "Wonderstood" was a typo, but I'm leaving it because I like it :)
Yes Firefox shouldn't be too hard, I don't use anything too Chrome specific, and I have built extensions for FF and Safari prior (wink wink).
It was harder that I thought, lots of quirks and behavioral differences even though the APIs used were relatively simple. Needed for example `e.preventDefault()`, requiring `<all_urls>` (unfortunately! - but that might allow me to allow copying styles from selection) and a bunch of style fixes just to get to parity.
Kidding aside, this is a cool idea
But this has a different primary use-case, more on transient context sharing, and caring about accuracy in the short term. And not about publishing per se.
http://stackoverflow.com/a/29094545
Rich text format supports font colors, images etc
Pasting this in a browser is explained well here
You aren't serving the appropriate intermediates. It's only serving the leaf certificate, but should also serve intermediates that get us to a trusted root.
Because browsers cache intermediates, this obscures the problem much of the time.
This is pretty amazing though. I copied some Go code from Gogland and got it in a browser, syntax highlighting and all! https://ctxt.io/2/AAAAwO9VEA
I think I followed instructions from here to get the certs: https://github.com/arkarkark/letsencrypt-nosudo but I can try all this again.
edit: Looks like this is the solution, thanks for pointing this problem out: http://stackoverflow.com/questions/17746643/how-to-set-the-i...
Otherwise, I like it for its simplicity.
The people that professionally create a hassle of security will beg to differ.
Edit: So just ignore them when they do that.
ITYM "within seconds, share with others the context of whatever you see". Yes?
