Quite often I need to save wikipedia pages to pdf as an offline reference. This is done through the clip to devonthink plugin or the create pdf bookmarklet.
However, I have to repeat this step a few times until the pdf contains the images in the wiki page.
It is reproducible in latest version of Firefox or Chrome. e.g., https://en.wikipedia.org/wiki/Factor_analysis
Is there a setting I can tweak to improve the saving behavior here? I want to do it just once.
I am clipping with the option “pdf (one page)”.
The print layout (= same as the PDF layout) is determined by stylesheets in the backend of the website. Try to save the page as a webarchive
Thanks for the input.
However, my case is that if I repeatedly do the conversion to create multiple pdf, after a few attempts, I will get the pdf with the images. So I think it’s not related to stylesheets.
I prefer to save to pdf because it is easier to annotate and able to export back to finder without additional issue.
Just tried in Safari and it behaves the same as in Firefox and Chrome.
I wonder can others reproduce the issue using the example wiki page?
My hunch is that once the link is passed to devonthink, it didn’t wait until the images are loaded to create pdf conversion…
You can try Fireshot in FF
Thanks for the suggestion. However, fireshot’s pdf is not text-selectable. it’s just an image in pdf format. Firefox can save the entire page as an image without installing plugin.
I understand that clipping will not work well across every website, especially those require login and powered by complex scripts for fancy effects.
I suppose Wikipedia is not one of them and it is a popular knowledge portal for many of the internet users that worth fixing related issues
That’s strange as I have no issue in Safari. What OS are you running?
Here is my computer settings:
Is there a DT setting related to this behavior that I can try to adjust?
Did you try to Save as PDF to DEVONthink in the Print menu?
I save webpages from Safari as PDF all the time and found Printing to PDF to be best way as it a) shows a preview and b) allows to chose between either the “normal” view (the result depends on the settings in the print CSS of the webpage) or the Safari Reader view. With the latter being most of the times the best option for me.
I have issue on Mojave and Big Sur too, the missing image issue is frequently reproducible and quite annoying.
What URL are you trying to capture?
Sorry for my late reply, it can be reproduced by the url @jakemkc provided.
It seems all the pictures are fine(mainly mathjax pictures which are small) except one slightly bigger picture.
Could it possible due to the resource dependency resolving timeout was triggered?
DT3 has had this problem from the beginning, in my experience. It is clearly a bug in the program, as it repeats exactly the same every time.
Images are included with Wikipedia clips only in the third iteration. Doing anything else with DT3 in between iterations means you have to start over with three iterations again.