Archive Page - Save Webpages on Internet Archive's Wayback Machine

[Joshua Shew]

Archive Page on GitHub

 

I just created my first workflow, largely inspired by Wayback When. You can pass in a URL to automatically snapshot it on the Wayback Machine, and the resulting link is saved in your clipboard.

 

You can use it with the keyword (currently set to "way") or as a universal action.

[FireFingers21]

Welcome to the forum @Joshua Shew!

 

Seems like a handy workflow. I'd like to contribute a few suggestions:

• Some documentation would be helpful, it's not immediately clear how to fully utilize all the features of the workflow.
• Consider adding some indicator or notification on the status of the running workflow. I tried saving a URL in the background, but was left guessing when I should check my clipboard for the output.
  • This could additionally allow for more specific error handling. For example, with https://google.com, it would be useful to know that I've reached the 5 captures/day rate limit (extrapolated from HTTP status code 429), instead of just seeing error in my clipboard.
• You make heavy use of the Arg and Vars Utility. Objects like Open URL and Copy to Clipboard can encode {var:} placeholders, so you don't need to convert your variables to {query} first.

[Joshua Shew]

@FireFingers21 thank you for looking over this and providing suggestions.

 

I will definitely take those into account as soon as I get around to it.

 

15 hours ago, FireFingers21 said:

Consider adding some indicator or notification on the status of the running workflow. I tried saving a URL in the background, but was left guessing when I should check my clipboard for the output.

 

Yeah, this is something I've wanted out of it myself. I think I'll go with notifications and see how that works out.

 

 

[Joshua Shew]

v0.1.3

 

- Updated description and bundle ID
- Simplified design of the workflow to reduce the number of instructions
- Added notifications for status updates (request sent and URL copied)
- Usage documentation has also been added to the README file since the last release

[Joshua Shew]

v0.1.4

• Added optional logging
  • Logs can be viewed in the workflow data folder
  • Logs are truncated every 10 executions
• Fixed the duplicate Universal Action

[Joshua Shew]

v0.1.5

 

• Added error detection with appropriate notifications
• Replaced custom logging with builtin debug messages
• Implemented advice from "Using {const:} to keep workflow details up to date in objects."
• URLs no longer have to be passed in when using the keyword
  • When no argument is passed, the URL of the frontmost browser tab is used instead

[Joshua Shew]

I think this is where I'll leave this workflow for now. Of course, I'm always open to new ideas/suggestions.

 

In terms of further improvements...

 

I was considering adding archive.today as a fallback service, similar to how Wayback When provides the option to query archive.today if no snapshots can be found on the Wayback Machine. Unfortunately, it seems that archive.today is not designed to handle automated requests, and adding support for that is not on their radar either (StackExchange discussion, Archive.is blog post).

 

Here are other ideas floating around in my mind. If I come back to this, I would probably start here:

• Switch to returning a URL instantly. It's annoying how it can take 10–15 seconds for the archive URL to be copied back into the clipboard. I can't do anything about this wait time because that's how long it takes for Internet Archive to do the processing on their end. However, I might be able to switch to an "instant output" structure if I return a link to a search that would reliably find the snapshot (that is yet to be created).
• Add a "wait-and-retry" fallback option. Right now, some links aren't being processed by Internet Archive, but there is a chance they may be supported in the future (e.g x.com links). Or, more simply, Internet Archive might be temporarily unavailable for other reasons. This could be fixed by saving links that fail, and try again with a sensible retry policy.

[Joshua Shew]

v0.2.0

• Changed the modifier for opening the website from "command" to "control"
• The workflow now checks for existing snapshots within a given tolerance before creating a new one
  • Using the "command" modifier skips this check
• There is now a part of the workflow that directs users to submit unknown errors to the forum
• Requests to Internet Archive now timeout after a waiting period
• Added configuration options
  • Keyword
  • "Always on" logging (off by default)
  • Timeout duration
  • Tolerance for existing snapshots

[Joshua Shew]

v0.2.1

No new features; just cleaning up code. I made an effort to try to reduce the object count this time around. I also realized I was using the Script Filter entirely wrong, so I switched it out for a Dialog Conditional

• New icon!
• Script Filter --> Dialog Conditional for the error reporting menu
• Condensed collections of objects into script
• Clarified the 429 response code notification: it is not the users fault, it is a Wayback When policy

[Joshua Shew]

v0.2.2

Minor user experience improvement and improvements to the internal design to prepare for v0.3

 

User

• Added notifications to indicate when the workflow is looking for existing snapshots

Internal

• Moved all scripts into external files
• Switched to using a SQLite database for keeping track of metadata
• Added an internal tracker for the version number

[vitor]

Did some quick tests and found a couple of bugs.

• Adding via the Universal Action fails, because the URL isn’t correctly passed (there’s a Run Script in the way sending other information).
• The Request Snapshot Availability script has a typo (resposne).

[Joshua Shew]

@vitor thanks so much for checking. I'll fix those asap.