I’d support several improvements mentioned by others:
Most important for me would be custom “metadata” - not just the difficult-to-access file metadata that varies by file type (and which apparently can’t be edited by Applescript either). Spotlight Comments is a start, but it’s not nearly enough. Such fields would make DT more flexible (like the fields in a database record), which is really important if you want to filter and sort subsets. Historians, for example, aren’t only looking for all the documents that talk about topic X as determined by the AI. Sometimes we want to construct a narrative/explore with the help of metadata, e.g. to see only the letters on topic X written by author A (author A may or may not discuss topic X differently from other authors), or only those written after date Q. Or to see a list of all the letters between author Y and recipients R, S as well as letters with no recipient info (author and recipient as separate fields), sorted first by where the letter was written (another field), and then sorted by chronology (another field). Which means you should be able to sort by multiple columns. These fields should also be batch editable, i.e. change multiple records at a single go (like you can with Spotlight Comments).
Tag search on main Search window.
Option so tags aren’t always sorted alphabetically.
I have no problem with the UI. But then I hate MS Office’s ribbons.
#1 in particular would be a game changer for historians.
I think that there are search engines that do this, at least askSam, the database program that I used for many years had this great feature: in the search result window, in the list of found files, beneath the file name and path was a third line with the highlighted search term (hit) and a number of words before and after the hit. One could specify how many words before and after the hit should be displayed.
The advantage of this is that when you have a lot of hits to go through, you can see the context of your search term instantly, you don’t need to click on each one and shift with your eyes from the list to the preview and back, which takes time, and also some files don’t open instantly.
Of course there can be more than one hit in a file. In that case DT could either just show the first one or there would have to be more than three lines per list item.
The search result list for a search for the term “constrict” could look like this (4 words around hits):
…convulsive paroxysms, with convulsive constrictions of the jaws (Allen…
Rosaceae - RS.rtf
…was the Sensation? Sudden constriction, sudden choking, pressure and …
…to the Boa family, constrict their hapless victims to…
Gets my vote. I love the way this is simply achieved in Brett Terpstra’s Marked app. He was acutely aware when designing the app of the pitfalls of floating window workflows so he made it simple to pin/unpin Marked as the frontmost window.
I also quite like the idea of “hazel-fication” proposed earlier. This to apply to a form of automation when documents come into a database, are processed by OCR, have changed in some way. But additionally, I would like some means of extracting information from a document (dates, keywords anything) also along the lines of what hazel does, and use this information to automatically tag, set color, etc. I use DTPO quite often for maintaining patent portfolios and being able to extract e.g. a patent’s priority date automatically, tag the document with this and then subsequently be able to use the tagging function to quickly see a “patent cloud” related to a particular priority date would be … golden.
Syncing could use some improvements. Setting it up, though very flexible, is not always clear and merge conflicts are not resolved well/easily/clearly. I use a macbook and iMac and sync between those often with edits and changes coming from both the macbook as well as iMac. I also keep a store on my NAS. I travel often and don’t have the macs in one place all the time. I do the syncing manually because the automatic method gave me problems time and again.
OCR is way to slow, can that be sped up somehow?
On the UI, I don’t hate it the way it is now. But, I really like the new iWorks/Omni UI. Making it simple (but not stupid) is quite important and drives focus when using the application. Perhaps DTPO does not fit that new UI very well, in that case keep what you’ve got and don’t force it. However, for DevonAgent I do think the new UI would be much better, DA is messy with its variety of windows and could use a clean up.
DA generates this keyword map at the end of a search query. I would love to point at a bunch of documents in DTPO and let it generate a map for me. Again, very handy for patent portfolios.
All of the above are obviously interesting and would be good, but I wonder if I’m the only one who doesn’t dislike the User Interface?
I actually like it. I use DTPO for words and nothing but words, and it just looks right on the Mac and on the iPad.
If it looks a bit old-fashioned, that’s fine. So does War And Peace when you lift it off the library shelf. I love the way DTPO looks, and would be quite happy if it looks the same forever. Reliable and does the job I bought it for.
I’ve looked at pale imitators and they gave me a headache, so much distraction from the work I got DTPO for. If you update it to be trendy, please let me keep this one!
Must agree with you - I have no problems with the interface either. I guess, were I to be choosy, I would maybe like more customisation options, like in Scrivener, where one can select your own background colours for separate sections of the UI, to suit one’s needs… But that’s really just my looking for something to add, for the sake of it - would really not be a deal-breaker in any shape or form for me, at least…
I’m still trying to get used to the new 2013(?) MS Office look - all this pale/flat stuff, almost for the sake of change, has me realising I’m obviously getting old… Having said that - I quickly adjusted to iOS7 - but iOS6 and older were beginning to look a bit old in the tooth by then, imo… Not there yet with DTPO, surely?
Recognizing no one at DEVONtech will reveal plans … yet
In my opinion the Apple WWDC announcements this week about iOS 8 and, especially, Yosemite, suggest a tighter integration across platforms and within the OS on each platform than what Apple has delivered in the past. This may be vaporware, but taking the announcements at face value the possibilities are very attractive for what I need to be doing in my work and research. (I can’t comment on anyone else’s use cases.)
Just to say it – if the OS provides tighter data and feature integration – then the future value for end users of so-called “everything” solutions such as DEVONthink and other might diminish significantly. Apple hasn’t announced anything that goes up against DEVONthink’s search and AI and robust support for customization (hello, Swift, have you met DEVONthink?). These are still valuable and needed.
But the rest of the DEVONthink features? Maybe they fall away. Time will tell. We’re headed deeper into a universe where we craft our own workspace with numerous smaller apps tightly focused on simple feature sets. If the OS provides rich capabilities to integrate data, documents, and function on the backend, across platforms, then that universe is pretty cool, IMO.
I’m very, very excited about the future with OSX Yosemite, iOS 8, and DEVONthink 3, however it still remains to be seen if/how DEVONthink will work out on iOS. Objects in the rear view mirror may appear closer than they are.