This depends on the view. Each column of the column view can use a different sorting whereas the left pane of 3-pane or split views is always using one sorting. But the right/upper pane of 3-pane or split views is using a sorting specific for the selected group.
Ok thank you. You are right. I only get the different sort order when i open the group up in its own window.
One more dumb question: how to I sort tags by most to least please? I am picking up a zillion tags and when I file things, so I have something like 300 tags with (0) in them and a bunch with (1), how to I sort so they appear at the top or bottom so I can skim them or delete them?
I would also agree with this comment and Mikeh’s original question about sort order. I’ve been slightly annoyed by the sort order being an all or nothing affair in the 3 pane left-most view in the past but it mattered less prior to the arrival of tags. Now that I’m starting to use the tags, I too like my groups unsorted but would greatly appreciate the sort order of a given group displaying correctly when in 3 pane view and the group is expanded.
In particular this is starting to make a difference with tags, because that’s the one group I would like to sort alphabetically or also from least used to most used, so I could easily get rid of the tags which are at 0 and perhaps consolidate the tags which have only a few entries. Right now the latter is impossible and for the former I open the tags in their own window as a workaround, to have a kind of floating tag palette window.
I understand these are small details and the focus is on finally finishing 2.0 for release with many problems having precedence, but when there is a little time for interface niceties I would also add my voice to the suggestions and questions here.
Same issue here. Really seems that if you’re dealing with a complex hierarchy of disparate elements, there will be times when there is a fundamental need for a different sort order in different groups. Not necessarily just based on level in the hierarchy.
Sometimes we need to use unsorted so that we can make the group order match what it is elsewhere in a project. But at the same time we may have another folder at the same level of the hierarchy that is a large group of folders by user name, and thus asc by name makes more sense.
It would also seem to be somewhat more intuitive to set the default sort order by folder (children inherit from parent groups until changed).
I use DT both to keep project information and documents as well as “journal” type use where i record individual meetings/emails etc. I want groups containing journal entries sorted by time/date, while other groups ordered by Kind or Name.
It’s actually surprising to me that DT cannot set individual per-group sort order … it really slows down my workflow when working in a project, as I have to continually switch back and forth between sort orders. Very irritating.
When I use the Data -> New from templates -> Registers -> Months, it is quite unintuitive to to have them alphabetically sorted from April to September.
The most natural way would be to have them sorted from January to December.
Currently, it is a pain to find the right month when doing it a lot
KLUDGE: An adaptation of one or more functions or actions in order to achieve a desired result. A cogent summary of kludge philosophy was stated by the late Senator Russel Long: “There are more ways to kill a cat than by stuffing it with butter.”
In this case the desired result would be to view a list of groups or documents in a different order than would be available otherwise in the Three Panes view of a database, e.g., a set of 12 groups bearing the names of the months of a year, in calendar order rather than, e.g., in alphabetical order by Name.
I can think of at least four ways to do that, and would probably use either #3 or #4, below.
Switch to the Columns view, which allows individual groups to independently sort their contents and maintain the chosen sort order, regardless of changes of sort order at the top level of the database, or within other groups. Choose Unsorted for the contents of the group that holds the groups that were assigned the names of the months of the year, then drag the groups into their proper calendar order.
Use the workaround noted by korm.
Modify the group Names by prepending a number, e.g., 01 through 12, so that the groups would now be named ‘01 January’ through ‘12 December’, and sort by Name. I would prefer Names such as ‘201101’ (instead of January), as that adds more potential for searches by month or year.
Modify the creation dates of the groups that are named by calendar month, using the Set Date script. For example, one could set the creation date for the January group as 01/01/2011 and so on, ending with the creation date of the group named December as 12/01/2011. Then sort the contents of the set of month groups by Creation Date.
Note: I find it convenient to add columns in a view for the sorts that will be used most frequently (View > Columns > …). The kludge for achieving an equivalent to Unsorted, manually ordered contents by modifying creation date could handle hundreds of items, but I usually leave ‘space’ in the series to insert new items by skipping three days in the sequence of creation dates, so that the series might become ‘01/04/2011’…12/28/2011’. I might use this kludge to arrange the order of the sections and subsections of drafts for a writing project. It allows for insertion of new sections or changes in sort order of one or more items with little difficulty.
Speaking of dates (and especially ranges of dates), many users will be delighted by Christian’s feature enhancements in an upcoming update.
Needless to say (but I’m saying it), those enhancements also extend what could be done with kludges based on creation date.
Thank you for having taken the time for this huge response.
For me, #1 is a workaround and I unsuccessfully tried to apply #2 (but I am going to try again!).
I quite like number #3 and I think that ‘01 January’ is enough because these folders can be in a folder ‘2011’. Such a year folder, considered as a tag, is quite convenient when searching and the result is more readable - in my opinion - than a ‘2011 01 January’ (and the information that it is the year 2011 is unique instead of being repeated twelve times).
Finally, the #4 can also be quite interesting but it makes needed to use the Column view or to use she Date sorting for the whole database. So I think we can consider it is a workaround too.
The more convenient would be an option to sort by except if manually re-sorted. But I understand that it is not so easy (it would already have been implemented if it was).
