@bm13kk

GTD was designed both as a general philosophy AND as an example of how this philosophy can be implemented on paper.

The philosophy is that you should avoid scheduling tasks in advance by date or time etc (except for appointments etc.). Instead you should decide right now what you will do right now, based on the situation. GTD talks about Context, Energy, Time (required/available) and Priority as a way to describe and match the current situation against the requirements of each task.

When done on paper, you simply split your Next list across a number of Next lists, one list for each Context (and/or for Agendas, shopping lists etc). This is practical. You cannot get it much better on paper. Like all single-choice categorizations you will always find that there are some things that would fit in more than one category, and this is something you will have to live with. Also you have to live with the fact that Energy, Time and Priority are not represented at all in this simple categorization system - it is something you have to use your brain or gut to do when reading the tasks.

The "paper" categorization principle can be used on a computer. Many apps are built like this.

But of course, in principle, on a computer you are definitely not really restricted to classifying tasks in just one way. You could classify them in many different ways, and then select whatever you want to look at. A computer can generate such a selected view in a fraction of a second. For example, one task may require you to use a phone, so you mark it Phone (a tool context), and maybe you need to compare with some excel sheets etc when you are talking, then mark it Computer also (another tool context), and maybe you will need to have some documents handy as well while you are talking, and those documents are in you office, so mark it Office also (a location context). And maybe the person you need to talk to is called Peter in the Finance Department, so mark it Peter (or Finance) also. And so on. This can be very useful, if you are reasonably strict about how you define these categories and apply these category markings consistently such that you can trust them.

In Doit you have both Contexts and Tags. The Contexts are best for the paper version of GTD, whereas Tags are more flexible for a more generalized computerization of the GTD philosophy of finding those tasks that best match the present situation.

In your example you mentioned several things that sounded a bit "funny" to me, but maybe I misunderstood you. (For example, what do you mean by &item &question &fix?)

I would recommend that you tag your tasks for any unusual REQUIREMENTS that the task has, for example, maybe some tasks need a lot of effort (time/energy) - then you can have a tag for BigEffort. Maybe you often have some tasks that require silence, then tag those with Silence. You can have people tags for the collaborators you rely on most frequently, etc. Then when you are in a given situation, usually with some restrictions, you can filter for tasks that have those particular requirements that the present situation can satisfy, or hide those tasks that the situation cannot satisfy. But I would advise you not to use tags to represent the type or purpose of the tasks, for example marketing or bookkeeping - use it for objective requirements that should be possible to match against a real situation.

More versatile quick filtering is needed to be able to use tags properly, and I have understood that this is being designed by Doit. (Especially a NOT filter is needed = be able to exclude/hide certain tags from the list).

My advice to you would be not to build "every" aspect into your tagging system. Implement only tags that you will use often, because it takes time to apply the tags, to review them, and to filter by them.

And my advice to you also would be not to wait for the new Area feature. In those apps that have this, this is just a "permanent filter" that people toggle when they go to work or go home. For most people it represents either a "location context filter" (office vs Home) or "group of AoRs" filter (Normally, in GTD, you might have 5-10 AoRs within your work sphere and equally many in your Private sphere.) With the Areas feature you will be able to hide your private stuff when at work and vice versa, but it has very little to do with your general question about tagging.

Hope that helps.

@Folke this is grate answer!

I think about it, and see, that i'm mixing tasks and information. For example, i don't need shopping list, only task 'go to shop'.

Problem is, that this info is needed. And there is no better place to store it (my brain is not enough, obviously).

I fully agree about better search options. And i think, stupid limit on 5 tags should be removed. And i need tags inheritance.

@bm13kk

I am glad that you were helped by my answer :-)

As for shopping lists and checklists and similar things, one way of doing this is to use the task notes, or use subtasks, or use a separate (external) web-based sheet or Evernote document (that you can link to from the task, if you like).

Personally, I tend to use subtasks for things that I want to be able to check off as done, and the regular notes for caveats, phone numbers etc. In your case, perhaps you could put "check amazon", "check ebay", "place order" etc as subtasks and check them off as you go, and write down the results of your research (prices, specifications etc) as regular notes (which could include links to other websites).

Sometimes - especially if I need to be able to separate the subtasks by context, tags etc - I turn the task into a project, but I often avoid this because it makes my project list too long and the task was usually already a part of another project. I wish subtasks could have tags and contexts - just like regular tasks.

Yes, I agree, I too see no reason for a limit of five tags, and yes, I too would very much like tag inheritance - very useful indeed.