Forum Posts Feed
TreeProjects 2.5 adds database encryption
< Next Topic | Back to topic list | Previous Topic >
Posted by Ian Goldsmid
May 4, 2012 at 09:15 AM
Yaroslav Pidstryhach wrote:
custom attributes for items are on the road map.
Great news!
And column based views with filters?
Posted by Yaroslav Pidstryhach
May 4, 2012 at 09:43 AM
Ian Goldsmid wrote:
>And column based views with filters?
Yes - item attributes will be database table columns, so arbitrary filtering/selection will be possible via SQL queries. Of course, users won’t type SQL - there will be GUI to do that.
TreeProjects already has a tree-column view, albeit only modification date and item type are currently shown. It can be turned on in Settings - Display.
Posted by Alexander Deliyannis
May 10, 2012 at 08:58 PM
Given that the infrastructure underlying many contemporary outliners (mostly 2-pane) is some kind of SQL or other database with tables and the like, it is rather surprising that so few offer custom columns / metadata.
An explanation may be in their roots organising textual information, to be used alongside word processors and the like. Also, I understand that most 1-pane outliners didn’t have a database infrastructure to begin with. So even though the technology allows it, the accepted paradigm didn’t foresee it.
Posted by Alexander Deliyannis
May 10, 2012 at 09:02 PM
P.S. I believe Brilliant Database is a tool that has taken the opposite route: it is a fully faetured database management system (and therefore allows unlimited customisation of tables and queries) but also allows organisation of information items in hierarchical folders, as far as I understand.
Posted by Yaroslav Pidstryhach
May 14, 2012 at 10:00 AM
Alexander Deliyannis wrote:
>Given that the infrastructure underlying many contemporary outliners (mostly
>2-pane) is some kind of SQL or other database with tables and the like, it is rather
>surprising that so few offer custom columns / metadata.
Even with an underlying SQL engine, there are many corner cases to think about. Consider, for example, a database where custom attributes (columns) A, B and C are defined by the user. Now you want to copy and paste an item from the database to a database with attributes X, Y and Z. Should the destination database be amended with attributes A, B and C immediately? Or should these attributes be discarded? Or should attributes be strictly per-item (which doesn’t play well with simple relational models)? Or should the user be asked what to do in each case? When there’s a solution, it must also be presented to the end user in a simple, non-technical, straightforward manner. So this feature might not be as simple as we’d like it to be…