I have the following problem with webarchives (i use macos 10.3.9 and devonthink 1.9.3):
When i open a web-link in DevonThink and choose “seite erfassen” (sorry it’s the german version) everything works fine and also when i’m not connected to the internet i can see the website with pictures and everything.
Bevore DenonThink could handle webarchives i saved webpages with firefox: firefox made a html-file + a folder of the same name, where all the pictures and stuff where in - both were saved in a folder in the documents-folder. I imported the saved html-files in DevonThink and DevonThink took the pictures from the folder saved else-where.
Now i thought i could use “seite erfassen” with the already imported web-sites saved whith firefox and than delete the saved html-file + folder, cause DevonThink would now store everything inside its own database in a webarchive. But after i deleted html-file + folder from the documents-folder the saved webarchiv in DevonThink showed nothing any more (just a blank page).
Louise:
It would seem that DT had only a link to the Firefox-saved files. Deleting those files would leave blank screens in DEVONthink.
Try your procedure again – saving page in Firefox, then importing them into DEVONthink. Now, when you see the page in DEVONthink, select Capture Page. Now examine the DT file. Is it a Web archive or not? Is this the procedure you used?
Thats exactly what i did. I tried it again - same result.
The file i got through “seite erfassen” (capture page?) is a webarchive… (but it seems nevertheless be connected somehow to the firefox-saved-file - a path to this file on my harddisk is stored as “url” when i look at the “info” of the webarchive.
Louise:
Could you please check a few things.
[1] Is there a word count substantially greater than zero for the DT document that’s linked to the Firefox page folder? Try some searches on the DT document for terms contained in the Firefox folder. Can they really be found in the DT document?
[2] How does the file size of the DT document compare to the Firefox folder’s size?
[3] If possible, go offline and view both the DT document and the Firefox captured HTML. Does the complete page with images appear in both?
I can’t check this out on my computer at the moment, and would like to find out what went wrong. 
So i hope i can answer this. My English is not the best 
in the example i used now:
(1) the word count is 291 in the DEVONthink-html-file imported from firefox,
and 292 in the DEVONthink-webarchive capured from this file.
searches work with both, the imported firefox html and the webarchive.
(2) htmlfile in finder: 20kb, folder with images and stylesheets for this
htmlfile in finder: 36kb gives 56kb for both.
imported htmlfile in DEVONthink: 39kb
captured webarchive from this file in DEVONthink: 83kb
(3) the complete page with images appears in both cases.
i hope there is no misunderstanding:
capturing webpages in devonthink following a link from a bookmark for example works fine.
importing html-files that had been previously saved to the harddisk with firefox (what gives a html-file and a folder of the same name where for example pictures are stored) works fine (i kept the html-file + folder on the harddisk, cause devonthink loaded the pictures from there).
now with the possibility to save webarchives in devonthink i wanted capture the imported html-files as a webarchive, deleted the imported html-files from devonthink and the saved html-file + folder from the harddisk… and after doing this the webarchive shows just a blank page.
what i found out right now: after the files are deleted from the harddisk and the webarchives are no longer displayed, this webarchives can still be searched and are shown as search results…!
when i use the info-button in devonthink there appears a information “ort” in german, must be “location” or something like that:
files captured directly from the web show here: “DEVONthink > Web-Sites on DEVONthink”
files captured from the imported html-files show here for example: “Web-Sites > Wikipedia” - the folder i saved the html-files on my harddisk.
Thanks. It seems that although the DEVONthink ‘copies’ show file type as WebArchive, they don’t include much of the information from the Firefox folders.
I understand why you did what you did. As long as you retain the Firefox folders, DT will display your text and images.
But for ‘safer’ capturing of Web material, it would seem preferable to copy and paste the URL into DEVONthink as a ‘bookmark’, then select it and do a page capture. I’ve got a sizeable Bookmarks group in my database containing clickable links to the journals and news sources that I frequently visit.
I use DEVONagent as my default browser, because it gives me the option of page or note capture to DEVONthink.
You probably already knew - what i wrote about the “location” in the info palette was nonsense: it only refers to the groups in DEVONthink, also in my case I just mixed this up, cause i have stored data on the harddisk in folders with the same name.