I wouldn’t advise to do that. Back in 2022 I made the investment of throwing a MacMini in a corner of my appartment for the ONLY purpose of hosting DevonThink Server 24/7. Because I’m in the same situation as OP.
Do I get along with that? The web application up to this day offers a really, really bad user experience if you intend to use it for anything beyond finding your documents (which you’d like to do very likely if you’re a Windows User :)). I have to help my self out on a regular basis by accessing the Mac directly/via VNC and/or using the devon iOS App. Does it overly complicate things and blow up any satisfaction in neatly organizing your documents? Absolutely. Does it feel like I’m wasting valuable time and efforts to maintain a sketchy setup having all the negative impacts of physically hosting a server (hardware, regular electricity costs, $200 additional license costs, waste of physical space, risk of failure, damage and increased risk of fire, …)? Absolutely…
Is it worth it? No! After taking on all negative impacts and costs, the bad webapp is really the dealbreaker. It’s just pure disappointment. Some might disaggree, but I’m just not getting enough benefit out of it for the immense overhead. (And up to v3.9.4 Devon didn’t lift a finger to improve the web interface. I personally think that’s the least they could do for those who make this investment).
Technically correct. And it would be a different discussion if the solution came for free. But people are charged twice the amount of a pro license to get it. Here you would expect that the company puts in a minimum of effort to make it at least somewhat appealing to work with, right?
We have many people, including businesses and academics, that happily use the websharing interface. In fact, I can think of one such business with a skeleton crew in house and 30 agents in other locations using it.
To share things? Certainly. But that’s just one of many aspects where a DMS has to deliver. The user experience of managing data and the data-structure isn’t even mediocre, let’s be honest… Is Devon enclined to improve this? I haven’t seen any real usability-improvements since I aquired the solution back in 2022!
I’ve been using same set of UR files (or databases) for over 20 years now on a daily basis. It can act both as a document manager (like DT) and an outliner (topic based). It does both excellently.
UR served me very well on MS Windows. I see myself using it for years to come.
In addition, I’d love to be able to use MacOS and DT for my personal knowledge management, but the lack of custom sorting in DT and the fact that manual sorting does not sync across devices is keeping me away for now. I purchase DTP license 3 years ago, but so far, it is a no go due to the above reason.
When you haven’t looked at Windows software for years that interface is somewhat jarring…
I’m happy it works for you, but watching the rather clunky procedure in the videos makes me appreciate just how refined DEVONthink is, and OS X in general.
The Renumber script is a reasonable workaround for preserving a custom sort order across sync. I can’t remember where I found it. If you search this forum there are several suggestions.
Set the file order you want, highlight the files, and run the renumber script.
Here’s the one I use - apologies to the author. I don’t remember where I got this.
(*
Keep cur_level around in case you want to let it know its depth later.
*)
property add_nums : 0
property start_num : 0
property trailer_max : 0
property after_num_string : " - "
tell application id "DNtp"
set r to display dialog "Add or remove numbers?" buttons {"Add", "Remove", "Cancel"} default button "Add"
set pick to button returned of r
if pick is "Add" then
set add_nums to true
(*
set r to display dialog "Start with: " default answer "1"
set r_text to text returned of r
set start_num to (r_text - 1)
log start_num
*)
set r to display dialog "Number of columns:" default answer "3"
set pick to text returned of r
set trailer_max to pick * -1
else if pick is "Remove" then
set add_nums to false
end if
set cur_level to 0
my looper(cur_level)
end tell
on looper(cur_level)
set counter to 0
tell application id "DNtp"
set cur_level to cur_level + 1
-- CSH: Have to do this shit separately.
set the_selection to selection
-- set the_selection to children of cur_group
repeat with next_item in the_selection
set counter to counter + 10
if add_nums is true then
my add_num(next_item, counter)
else if add_nums is false then
my remove_num(next_item, counter)
end if
end repeat
set counter to 0
if cur_level > 1 then
set cur_level to cur_level - 1
end if
end tell
end looper
on add_num(the_item, counter)
set old_name to name of the_item
set clean_name to my clean_me(old_name)
set counter to counter + start_num
set good_num to text trailer_max thru -1 of ("000000" & counter)
set new_name to good_num & after_num_string & clean_name
set name of the_item to new_name
end add_num
on remove_num(the_item, counter)
set do_it to true
set old_name to name of the_item
set AppleScript's text item delimiters to after_num_string
if (count of text items of old_name) > 1 then
try
set prefix to text item 1 of old_name
set testing to prefix as number
on error number -1700
set do_it to false
end try
if do_it is true then
set new_name to text items 2 thru -1 of old_name as string
set name of the_item to new_name
end if
end if
end remove_num
on clean_me(old_name)
if (count of text items of old_name) < 2 then
return old_name
else
set AppleScript's text item delimiters to after_num_string
try
set prefix to text item 1 of old_name
set test_convert to prefix as number
on error number -1700
return old_name
end try
set new_name to text items 2 thru end of old_name
return new_name
end if
end clean_me
Is Devon enclined to improve this? I haven’t seen any real usability-improvements since I aquired the solution back in 2022!
