The AI-Classify is very interesting and can be time-saving. But when AI doesn’t find the good group for classify in the “classify”-Tab, it is absolutely not ergonomic to design quickly another group.
This is also the case if the suitable group is a subgroup of the AI-found group. You cannot switch quickly to a subfolder (or a superior group) by doubleclicking on it (if you double click on it, it will be immedately classified in this wrong group)
All this is once again the problem that IMHO, DTPO should propose “breadcrumb” navigation : a full “path” of every document or group, with clickable groups for every component of the path:
Interesting. So, what would the useful “quick ergonomic design” do? I select a document, I do “See Also”, there’s not a group proposed that I think fits … and then I do … what? I cannot visualize how beadcrumbs showing the current document’s location could help with creating a new location in a different hierarchy.
Korm, what do you do when the choice of groups doesn’t contain the group that you are looking for?
-Open a new window, put it next to your actual window, navigate to the group?
-Rightclik on the doc, “Move to…”?
In both cases , this is slow.
edit : especially when the folder-levels are very deep"
All i would need to be happy is exactly the same view that the MacOS-Finder gives us in the column-view. Every group is always visible.
In DTPO, after a double-click, you don 't see the higher groups any more…
I do the same thing that Greg suggested. Takes a second.
It is already, isn’t it? The List, Columns, Split, and Three Panes views all display the entire list of groups in the current database. WRT the current database, Groups and Tabs doesn’t know about or display groups that the group panel doesn’t.
when you double click on one group in the group pane in order to open a new window, the new windows has no groups (folder-structure) in the groups tab any more!
Perhaps i should learn to not double-click, but i often have an open window and don’t know where i am (i already opened the breadcrumb-thread for this)
If the “suitable group” is a subgroup of the AI-found group, why do you need the Classify? I’m confused. Anyways… Why wouldn’t you just right-click on the “wrong group” in the Classify drawer and choose “Reveal” or Open?
Or if it’s identified as the “wrong group” but you known the right group is a subGroup, why not then right-click and select Move > … ?
Of course, you can "right-click, “Reveal”), or “Open”, and then push the new window (the 10th new windows in the session…) a little further because it hides the window you are working in, then drag and drop.
But for a real fast workflow, this is slow.
Why not have the same behavior as the Finder does? I think we all love Mac beaucause 90% of Finder’s Behavior is really quick, fast, neat…
Why should i have ten windows open (or have to click once again to close it, now we have at least 5 clicks for classify one document that didn’t fit with AI), instead of working in the central pane with a column-view similar to Finder’s view?
Actually :
When AI finds the good group: one click, it’s classified. You cannot do better, great.
When AI doesn’t find : 4 clicks or five clicks instead of having an open "Groups and Tags"Pane to do as fast as possible everything that you need to do (create new group, while you continuesly see all the other existing groups…)
uh…oh…no. ever had a look at Path Finder?
if not, have a look and have your mind blown
ctrl+c
cmd+m
kitch
not saying you are wrong. but groups and tags without a way to filter is unusable for me and i doubt i am the only one (currently short of 1k groups but 1k+ tags).
i agree about the path selection in the classify list though. that being said, i only have documents in leafs not nodes. using another filing scheme (like you do) seems to feed into this problem.
I did, but i left Path Finder for performance reasons on my MBP. And now with Mavericks Tabs (not upgraded yet) i think Finder will offer one of the biggest advantages of Path Finder, too.
[quote]
When AI finds the good group: one click, it's classified. You cannot do better, great.
[/quote]
ctrl+c
Yes, I know, but I fear those keyboard-shortcuts. But of course, this is still alittle faster than a mouse click.
More i work with DTPO, more i think “downgrade” to a simple nested folder organization .
as David Spark does (Great iBook : “Paperless”)
He also says that Devon Think is a very good program, but he prefers working with nested folder in Finder.
Of course , i will loose
AI for classify
replicata (they are bette than “Aliases” in Finder
integrated OCR from ABBYY
But i win:
immediate Dropbox Sync of the whole Database
immediate Sync Over the air with the iPhone and the iPad
not all the data in one package, but in folders easily visible from outside for other programs (also from syncing apps on iPhone). And also visible for me on the Web-GUI of the Dropbox-Website…
-Some ergonomics
DTPO would be the perfect solution if :
it could work in Dropbox folders (or have a similar Sync working)
allow Indexing rather then copying all in it’s own database (right now, you cannot index folders on the Hard Disk, if you want to sync the database : DTPO will Sync all index folders contents in the clients DTPO-Database instead of syncing links.
So you could use the AI and other features of DTPO as a big help for organizing your folders.
These are some thoughts, will continue working on DTPO and Dropbox-Folders in the same time for some weeks or months…
This isn’t criticism @saschabur, but DEVONthink doesn’t work for everyone. It seems to be a struggle for you – when DEVONthink is working smoothly, it should disappear into the background and make it easier to work with your documents and stuff, not harder. Maybe a nice folder structure in the file system, augmented by Mavericks tagging and perhaps DEVONsphere Express or some other Spotlight augmentation?
roll your eyes as much as you want but its still tedious. sascha is right. how long does it take you to file a document that way? 5 seconds? 10 seconds? 15 seconds? more?
anything above one or two seconds is not acceptable imo
I offered a suggestion that is all. I am not going to argue about some personally imposed time limits on file-handling. If sascha or you or anyone finds it too slow, then don’t do it. That doesn’t make it an invalid method, it just makes for one you wouldn’t use.
PS: I rarely confine myself to hierarchical systems anymore and when I do, I let the AI sort it out for me, so… yeah, way faster than 2 seconds.
