Clip to DEVONthink

[BLUEFROG]

[tudoreynon]

Yes, I only use Web clipping if something is temporary. Good advice. I think it might depend on your usage and style, I certainly think for academic work or scholarship that is essential. Thanks for explaining the technical reason for that roughly. Very useful to know.

You need PDF and I usually use a Bookends citation too, if I am quite sure I need to find the stuff again. Recently I have had a spate of broken links to web sites, not from DEVONthink 3.

[chrk]

@nomasprime

The main reason is the extension doesn’t see the page as Safari does, so you get things like cookie dialogues blocking the entire page.

‘Export as PDF’ renders the best, although it would be nice to specify the width and the URL metadata is lost.

@BLUEFROG
Same reason for me. 7 out of 10 sites clipped as PDF are unusable because of popups, grey overlays, etc. The print to PDF and Safari’s Export as PDF features are the only PDF options that reliably get good results

[BLUEFROG]

Stop going to those 7 sites.

Thanks for the clarification.

[cgrunenberg]

@chrk

Same reason for me. 7 out of 10 sites clipped as PDF are unusable because of popups, grey overlays, etc.

In the past or using the current version?

[chrk]

That’s 1 solution

The problem is that you never know if it is a usable result or not until looking at it in DT. That’s why stopped using the clipper.

What is the reasoning behind the 1000 point width you mentioned above?

[chrk]

In the past.

In case this has improved, I still think that a local PDF clipper option (like the print to PDF service or Safari’s export feature) would be better because you mostly get what you see in the browser.
This is especially useful when using adblock extensions.

A quick clipper test (PDF paginated) with the current version shows the following:

Those ad areas only show up when using the server side clipper. With local clipping, my adblock extensions cosmetically remove all this. The clipper also has issues with images, they are often lost.

Here is the example URL: With crime up and ridership down, Metro struggles to move homeless people off trains

[pete31]

@chrk

homeless people

Photos of the Most Egregious ‘Anti-Homeless’ Architecture

[chrk]

I didn’t read any of the articles by the way, it was just a random click on the website from a bigger publication. Same thing with NYTimes articles, to show the issues with the clipper.

[pete31]

I also didn’t read the article that you’ve linked. The one I linked immediately came to mind, it of course has nothing to do with your post. Probably confusing and inappropriate to reply merely with a link, sorry for that.

[chrk]

No problem. In the meantime, I took a look at the Vice article and see why it might come to mind. The LATimes headline has a slightly derogatory tone to it.

[BLUEFROG]

That’s the width in points of a captured PDF unless there are site-specific settings that override this. It was added in 3.8.3.

[chrk]

Thank you. Does this help with in-page popups and clutter in your experience?

[BLUEFROG]

Unfortunately, no. That would likely require things like disabling JavaScripts on the page.

[cgrunenberg]

@chrk

Here is the example URL: With crime up and ridership down, Metro struggles to move homeless people off trains

Thanks! The next release will be able to handle this too.

[cgrunenberg]

@pete31

Photos of the Most Egregious ‘Anti-Homeless’ Architecture

And this one will work too. Here are some examples clipped by DEVONthink or exported by Safari, in all cases without scrolling through the web page manually first:

DT1.pdf (4.8 MB)
DT2.pdf (11.4 MB)
Safari1.pdf (8.2 MB)
Safari2.pdf (7.9 MB)

Additional example URLs which should be checked are of course welcome.

[chrk]

The DT examples look good.

Just in comparison, my own Safari clips of those pages look better than the attached Safari examples because I use an ad-blocking extension, which you don’t seem to use for the Safari result.

Overall, the full-page PDF clipping seems pretty close.

Some details I noticed:

• Safari1 and Safari2 have no working links, but when I export those URLs as PDFs from Safari myself, links are working in both of my exported files.

• In DT1 the highlight color of the links is missing.
  When clipping this URL with DT as a full-page PDF, DT preserves the (blue) highlighted links for me.

The bigger difference between Safari and DT is in the paginated PDF clippings.
There, the result is usually much better when printing to DT via the service. It’s the paginated clipper that often results in pop-ups and obstructing layers in the saved files.

Are you guys using blocklists to improve the clipped results?
In such a case, I wanted to recommend the following lists I have been using without many false positives the last few years with NextDNS:

• oisd
• Lightswitch05 - Ads & Tracking
• 1Hosts (Lite) (not the Pro version though)

Thanks for working on this.

[cgrunenberg]

@chrk

The bigger difference between Safari and DT is in the paginated PDF clippings. There, the result is usually much better when printing to DT via the service. It’s the paginated clipper that often results in pop-ups and obstructing layers in the saved files.

The improvements of the next release are generic and not limited to a certain format. See these additional examples. Of course additional URLs currently causing issues are still welcome.

DT1 Paginated.pdf (11.4 MB)
DT2 Paginated.pdf (4.1 MB)
Safari1 Printed.pdf (7.9 MB)
Safari2 Printed.pdf (8.2 MB)

@chrk

Are you guys using blocklists to improve the clipped results?

No.

[chrk]

Thanks, I’ll restart using the clipper and will post example URLs here when I come across issues.

[nomasprime]

Interesting, is this for when clipping with DEVONthink?

Haven’t had a chance to implement it yet, but my current idea is to use AppleScript to resize Safari, export a single page PDF to Inbox, and then restore Safari to previous dimensions.