Using the Safari contextual menu and its Share > Add to DEVONthink command for Rich Text does just copy only the selection Iāve made on a webpage and nothing else (which is what I want), but it only copies it as plain text (even though Iāve selected the Rich Text option) without the bold text (or other rich text) formatting and without the embedded URLs in my selection.
Again, what I want DTP to do, which Evernote does flawlessly, is to import my selectionāand only my selectionāwith whatever rich text formatting and embedded URLs that exist in the selection.
Itās the same with dragging a selection to the Sorter; the selected text does import, but only with some of the selectionās original rich text formatting and in some sort of weird faded capture (see screen capture below and compare to kornās original post).
And why should I have to use such an inefficient process as dragging the selected Safari content onto the desktop as a .webclipping and indexing or importing the clippings collected there into my DEVONthink database? And itās the same inefficiency with invoking āTake Noteā with the keyboard shortcut and dragging the selected Safari content into a noteās text field (which Iād first have to create a note to do).
I havenāt tried either of these last two methods to see if it would do what I want, but why canāt I get the selection I want just by invoking DEVONthinkās webclipper?
I prefer DTP over Evernote for a variety of reasons, but at least with my OS (Mac OS 10.11.4), Evernote cleans DTPās clock with respect to capturing just a webpage selection with original formatting and embedded links. I just donāt get why DTP canāt accomplish the same task with its webclipper.
DEVONthink (and Evernote for that matter) is at the mercy of the webpage. Iāve had rich text copy failures in DEVONthink, as described by the correspondent above, and in Evernote. Sometimes a selection from a given page fails to come across as rich text in DEVONthink and succeeds in Evernote, and other times itās the opposite. The web is a mess of bad coding ā I wouldnāt blame anyoneās clipper for failing since success depends on exogenous factors out of any developerās control.
Note: we donāt have 400+ employees available to work on the browser extension alone. So, while nothing is just āgood enoughā for us, we also have to allocate our resources far more judiciously than they do.
Also, as korm pointed out, Evernote also fails with their extension. In fact, just recently I had to use it for a Support Ticket and it failed on several pages our extension was working on. The Web is indeed the Wild West, despite whatever standards may be in place. (Itās also why I wish we could remove the word āautomaticallyā from peopleās vocabulary when referring to software and the web.) And consider any standard created now has many years, and literally billions of web pages of legacy (ie. non-standardized) code that no one would EVER backwards maintenance. Sometimes, itās amazing capturing web data works at all!
For me, the benefits are, highest priority to lowest is:
get the full-text, for searching within DT
save some visual representation of the page, so when scanning, remembering things is easy because āoh yeah, I remember itās this mostly blue page, with large white headersā
I donāt know if DT has the capability of running JS under the hood- the OSS projects I referenced further below allow for clipping seamlessly, directly into HTML files.
Saves as a standard HTML file, without the lockin of a .webarchive.
When browsing my stuff in DT, highlighting an HTML file in the main results would bring up the preview right away, vs having to drill down to the āmainā file. These are what my clipped notes from Evernote import look like:
I tried the Chrome ex, I love how it runs asyncronously, with the little notification in the bottomā¦ this lets me continue what Im doing without interruption. UX on this is really important- hopping back and forth between things quickly without losing flow.
A solid clipper would mean all in on DT and no more Evernote
better web clipper is critical for knowledge management db
saving as html or web archive aināt best because of the size and cleanliness
websites such as reddit contains much valuable info in the discussion section which can be used as future reference. however, the current clipper is weak in clip the whole reddit thread in clean format
A better web clipper would be a huge benefit. Right now Iām using the Evernote clipper, then taking the note in Evernote, printing to pdf and then importing that into DT3. Itās painfulā¦
Love DT and have been a user since 1.x. I think the web clipper, working properly, would be a huge deal.
Iāve been using DT Clip to Markdown or copying and pasting into a Formatted Note when I need fairly simple text and images.
But when I want to keep an accurate copy of a web page, I use a Safari extension called āPage Screenshotā (available in the Mac App Store) that allows me to take a āscreenshotā of either the visible page or the entire pageāas a single imageāin either PDF, JPG, and/or PNG format. That preserves exactly what Iām seeingāthen I make a PDF with text by using OCR.
The only downside is that it captures the page as a single imageāwhich makes for a very tall imageāIām not sure how one would print it to paper. But it does allow me to keep an exact, WYSIWG āarchiveā of a page.
And Iāve found that when I select āKeep full retina resolution qualityā checked, the file is very large (15+ MB) and wonāt OCR properly in DT 3āI donāt know if itās the size or dpi or what, but I keep that unchecked if Iām going to want to extract a text layer from it.
All, I believe that this is still not resolved. Is there any hope to see the save as article function? Just now I tried to capture a question and answer from Quora and just despaired. Whatever I saved was only showing the log in screen of Quora.
As this thread is long, some info is old and the word āreloadā doesnāt seem to be in here, I was wondering:
Why does the web clipper āreloadā a page before itās clipped? The problem I experience is that cookie walls and pop-overs get clipped, and I have a hard time to manually remove them from the webarchive or html.
Other clippers like the one from Evernote or Nimbus Note seem to use another capturing trick, so the loaded page gets clipped as-is. The latter even allows me to edit / modify the contents before itās stored.
Thatās just the way the browser extension works at this time.
Note: There is no āclipping standardā. These things are developed independently and with their own solutions. Though our extension works in many instances, we have it on our list to enhance in the future. Thanks for your patience and understanding. (Also, note that Evernote has 300+ employees and millions of dollars in funding. We are a small development house, completely self-contained and funded through our sales alone. And at one point, I heard rumor (though I didnāt try and substantiate it) they had at least 40 people working on the clipping extension technology.)
I appreciate the frustration. And BLUEFROGās explanation makes sense: apparently itās just not that easy to consistently clip material from a huge variety of web pages. Iāve come up with two solutions, using Safari:
āPage Screenshot for Safari on the MacĀ AppĀ Store, per my post above, then OCRing it to pdf; the advantage of this (even over Evernote) is that it preserves the exact look and layout of the page. Inline links, unfortunately, donāt work.
With some help from DT Support, I made an AppleScript shortcut that creates and opens a blank Formatted Note with the correct URL and page title. Then I simply copy and paste whatever I want from the page into that open note. If the formatting is wonky, I will usually just select āReader Viewā before copying; that at least gets the text and usually any inline images. The helpful post with the script was here: Difference between clipping Safari page to formatted note and copying/pasting into formatted note - #8 by pete31
Of course the third option is just to use Evernote, and then import into DT3 if necessary. Thatās not something I do too often, but it works for some scenarios. When repeatedly clipping simple things ā for example, I keep a list of words and definitions Iāve looked up, and DT3 is useless at clipping the dictionary I use ā I just use Evernote.
And I didnāt know about Nimbus Clipper; Iāll check it out!
Indeed! Now, that was some time ago when I ran into that. The finer point of it is, Evernote has always had a big influx of capital investment and a team that far exceeds ours in numbers. So they have the ability to create a group of developers who can concentrate on singular features or smaller sets of functionality.
Rehashing the discussion, and I think this question is related to the last comment from Jim about āreloading the pageā.
Is there a setting to change from giving permission to Devonthink to clip every page or URL every time one tries to clip a page, instead of giving the entire Chrome application permission one time?
