Hacker News new | ask | show | jobs
A new way to approach Wayback Machine (github.com)
41 points by wabarc 1856 days ago
5 comments

$ curl -sf https://gobinaries.com/wabarc/wayback/cmd/wayback | sh

  uname_os_check 'CYGWIN_NT-10.0' got converted to 'cygwin_nt-10.0' which is not a GOOS value. Please file bug at https://github.com/client9/shlib
Hmm... let's add "cygwin*) os="windows" ;;" to the shell file.

skraeling@moria ~ $ curl -sf https://gobinaries.com/wabarc/wayback/cmd/wayback > wayback

skraeling@moria ~ $ nano ./wayback

skraeling@moria ~ $ chmod 777 ./wayback

skraeling@moria ~ $ ./wayback

  ==> Downloading github.com/wabarc/wayback/cmd/wayback@master
  ==> Resolved version master to v0.12.1
  ==> Downloading binary for windows amd64

  Error downloading, got 500 response from server
Am I doing something wrong here?
I recommend the Wayback Machine add-on. Let's you quickly archive pages and automatically serves an archived page for non-working sites.

https://addons.mozilla.org/en-CA/firefox/addon/wayback-machi...

That extension asks for permissions for all tabs, and for all data on all websites.

Try https://addons.mozilla.org/en-US/firefox/addon/one-click-way... which asks for no permissions.

A bookmarklet could work, too.

I've been using a bookmarklet in Mobile Safari. Looking for its origin just now, it looks like archive.org uses a Wikipedia page as a reference: https://en.wikipedia.org/wiki/Help:Using_the_Wayback_Machine...
Last updated 3 years ago (Apr 18, 2018)
So?

The extension is still being developed[1] but they haven't uploaded newer versions for Firefox recently. Check the Github Issues ([2],[3])

[1] https://github.com/internetarchive/wayback-machine-webextens...

[2] https://github.com/internetarchive/wayback-machine-webextens...

[3] https://github.com/internetarchive/wayback-machine-webextens...

> still being developed

This does not help anyone since you can’t sideload an unreleased extension in Firefox anymore (except by manually installing it every time you start Firefox, via a “Debug Add-ons” menu).

Wow did not know it’s that hard to sideload. Essentially not possible practically speaking if you want to do it all the time. That sucks.
That’s not important. What’s important is, does it still work in the most recent version of Firefox? (I’m on mobile rn so can’t check for myself.)
I was curious about the IPFS archiving and it appears that it just inlines everything. This is a bit of a shame because with IPFS's content-addressing and deduplication it would be really cool if the assets were out-of-line so that they could be duplicated between webpages.

Example of my blog. You can see that the images, CSS and JS are inlined: https://ipfs.io/ipfs/QmcEKGbtFsYrUoDLFMLv45yrZ1hP7ZyyX79x9Ts...

That being said there are problems with inlining IPFS URLs currently. Because few people support ipfs: URLs you essentially need to pick a gateway and embed that into the archived document which is unfortunate. That being said most people who run their own gateway intercept ipfs.io automatically so it is a sort of defacto IPFS scheme.

You could probably post-process this yourself if you wanted to to "pop out" the inline blobs.

This sounds like a great idea about archiving assets out-of-line, and it will be one of the goals Wayback Archiver is pursuing.
I sometimes find that when I submit a link manually to Wayback or to archive.is, they may temporarily fail to save the link and I need to do it again. Other times they fail in a different way. In particular Wayback may say that a live version is not available even though it is.

Does this tool handle resubmitting the link for the kinds of failures that are temporary? And what do you do about permanent failures?

archive.is may have enforced a strictly CAPTCHA policy, causing an exception to the request, and it is hard to resubmit currently.