Why do my do my webpage links keep evaporating?

It happened again. Third time.

As the title of this kvetch describes, my Scrivener research links keep, well… evaporating. Allow me to describe:

I’ve compiled a small mountain of notes in Evernote notebooks. I publish these notebooks, copy the URL, and create a webpage link in Scrivener that, indeed, contains the Public Link URL to said Notebook. Presto! Click on the webpage and the Evernotes in said notebook are visible in a column, and to the right the selected Evernote appears embedded in my Editor. Nirvana ensues. I close Scrivener.

I reopen Scrivener. The webpages in Scrivener indeed, resolve once more to the public shared URLs. Nirvana persists.

A couple of days later… I open Scrivener. No notebooks resolve. The webpage icons remain in in my Binder, sitting in the research folder, but click on them and they do nothing, say nothing, amount to nothing.

Suggestions? Theories? Keys to the liquor cabinet?

Cabinate is open to you sir…

Can you elaborate on what you are doing between nirvana and hell? iOS use? Windows use? export to folder? throw laptop out window? drive over laptop with lawn mower? dowse laptop with kerosene then light with match? consume large quantities of absinthe? Anything you can think of.

While you are thinking of things that you might do between “I love my life!” and “why does the universe hate me so much!?!?” have you verified that your editor window is not in a different view? Ex: it works in compose but not in outline. Just thinking of the obvious (to me).

In reviewing what the delta is between disappearances, I can only come up with the common axis of time. Launches? Restarts? Hard to tell, as I don’t reference these webpages every Scrivener session.

Having a research folder full of essentially null webpage pointers to my Evernotebooks, I recreated one sole instance and am observing its behavior across launches and days.

We’ll see what watching the pot yields.

To preface, there isn’t much that Scrivener is doing here that are not core features of the OS. Here is what I would do:

  • Test 1 (macOS & Safari only):
    [*] Open Safari and load your public Evernote URL.
  • Use File/Save As… and save a WebArchive file somewhere for safe-keeping.
  • In Finder, select the file and hit the Spacebar to view its Quick Look content and ensure it is okay.

[/*:m]

  • Test 2 (macOS -> Scrivener):
    [*] Drag that very same file into Scrivener’s binder, and mark it as having been imported from Finder.

[/*:m]

  • Test 3 (Scrivener only):
    [*] Using the same URL as in the first test, use File/Import/Web Page… in Scrivener to import a second copy.
  • Mark it as such and place it next to the one saved with Safari and imported from Finder.

[/*:m]

  • Test 4 (no fancy stuff):
    [*] Select either #2 or #3 in the binder and hit Cmd-D to duplicate it.
  • Use the Documents/Convert/Convert Web Pages to Text command.
    [/*:m]
    Shake well, and let ferment for a few days, relaunches, reboots, whatever.

The last one is less of a test and more of a proof of concept. That strips the text out of the page and turns it into an ordinarily editable text chunk in Scrivener. It’s never going anywhere other than where you tell it to—and if that’s good enough as a method you might not need to bother with trying to get Evernote playing nice with the WebArchive format. It doesn’t always work; the fancier a web page tries to be, the more like an “app” i tries to be instead of a document, the less likely it is for there to be useful information to extract or convert.

We would expect all three tests to fail together. Scrivener isn’t do anything Safari isn’t, and its web importer is merely a convenience for the steps taken in Step 1 and 2 together, along with some additional smarts to display a WebArchive’s origination URL in the footer bar. If they do all fail, you would probably need to contact Evernote and see if they have any ideas why the archived version goes stale after a few days. They might have a suggestion for a more static way of downloading the content.

Do you have to be logged on to Evernote to see these published notebooks? I’m thinking it’s either that something has expired (I’m thinking of Scrivener as its own web browser here). That “something” would probably be a cookie, or some other thing that creates a persistent logon status as you browse evernote.com.

Try accessing the same link you’ve added to Scrivener in a browser you don’t normally use, or even make a new user account on your mac and use the link there (without first logging on to Evernote).

Indeed, this results in a webArchive on my desktop. Hitting Spacebar results in a display of the notebook contents: a column of notes, with the contents of the selected displayed beside. In other words:

Behavior as expected.

So done. Clicking on Scrivener webpage icon results in a display of the notebook contents within Scrivener editor: a column of notes, with the contents of the selected displayed beside. In other words:

Behavior as expected.

So done. Clicking on Scrivener webpage copy icon results in a display of the duplicated notebook contents within Scrivener editor: a column of notes, with the contents of the selected displayed beside, identical to the above. In other words:

Behavior as expected.

So done. Clicking on Converted Scrivener WebPage icon results in a display of… blank page. In other words:

Failure.

Gone are the sixteen notes, their contents… only void visible (to paraphrase Milton).

Interesting. That it won’t convert to text means the WebArchive doesn’t have any text in it. Whatever you see is being dynamically loaded off of the Internet when you view it, and that can become volatile over time—meaning it isn’t an archive at all. I’d contact Evernote and let them know their pages aren’t archiving nicely with Apple given how they coded them. Meanwhile you might try PDF or some other alternative out of the browser that is safer.