Feature Request: task prioritization!
 

Just got started with my sneaky peak, and OF is just plain *awesome*. However, there's one killer feature that is keeping me from moving over from iGTD permanently and taking my chances with the Alpha: prioritization of tasks! I realize that setting priorities is a bit "anti-GTD" to purists, but in my case it's absolutely necessary.

I'm a network engineer, so my @Online and @Computer contexts have somewhere in the neighborhood of 500 tasks between them. Without some kind of prioritization, I'm lost when it comes to wading through those. In KGTD, I hacked around this by creating contexts like @Online-P1, @Online-P2, etc., however I'd much rather be able to have all my online tasks be grouped under one true context and sorted by priority. Makes scanning, reviewing, and re-prioritizing *much* easier.

iGTD does this pretty well with their 1-5 prioritization column. I'm hoping Omni can come up with some kind of similar solution for OF.

Flagging could be used to distinguish one level of priority.

Nested contexts work well too, so you could set up the different Online contexts you mentioned, and group them under a general Online context. Select the group to view all, and using the filter it's easy to group by priority context.

I agree with BlueCoat. The option to prioritize tasks would be a HUGE benefit to me. This is my first posting and have to say that so far O.F. is fantastic!

If priority were a preference that I could use and/or hide, I would be all for it. One thing I didn't like about iGTD was that priority couldn't be turned off -- at least that I could tell (admittedly I didn't use it too extensively).

Given that the prioritisation of tasks is not canonical GTD, then if this feature were to be added, it would be crucial IMHO that there were a preference to turn this off altogether.

I suppose it's a fine line that the Omni crew are having to tread - whether to aim for a pure GTD implementation or open up to a wider audience practicing variations on GTD or even unrelated productivity systems.

I'm new to GTD, but one of its features that is really improving my productivity is that it doesn't rely on prioritization. So I'm with the others here who say that if prioritization were added, the user should be able to steer clear of it altogether, hiding all related UI elements with one click in preferences.

[QUOTE=HiramNetherlands]I'm new to GTD, but one of its features that is really improving my productivity is that it doesn't rely on prioritization. So I'm with the others here who say that if prioritization were added, the user should be able to steer clear of it altogether, hiding all related UI elements with one click in preferences.[/QUOTE]

yes! absolutely.

Stay GTD Canonical, at least for the first release.

Prioritization is not canonical.

I'd like to have prioritisation.

How about a solution like the tabs in Safari? Make a pref pane where you can enable it and then, once enabled set specific prefs for the feature if there are.

OmniFocus has a notion of priority already: it's the order in which you arrange your items. If you want something to have a higher priority, simply move it up in the list.

This gives you much finer-grained control than a typical priority system, which typically only has a few levels of priority: OmniFocus effectively has as many priorities as you have items.

Does that make sense?

In a previous post, Ken Case [URL="http://forums.omnigroup.com/showpost.php?p=14611&postcount=24"]mentioned meta-data support[/URL], similar to that of OmniPlan. Seems to me that all of these feature requests that fall outside "GTD Proper" (prioritization, tagging) will be handily addressed by this extensibility, without getting in the way of users who don't need them.

Bluecoat93, you really have 500 actions in your @Online/@Computer contexts? It sounds to me like you either need to be more granular (@Photoshop, @Coding) or take advantage of sequential projects. Surely most of those 500 tasks aren't actionable yet.

[QUOTE=Ken Case]OmniFocus has a notion of priority already: it's the order in which you arrange your items. If you want something to have a higher priority, simply move it up in the list.

This gives you much finer-grained control than a typical priority system, which typically only has a few levels of priority: OmniFocus effectively has as many priorities as you have items.

Does that make sense?[/QUOTE]

Yes, of course! I didn't even think of that. Duh!

[QUOTE=Ken Case]OmniFocus has a notion of priority already: it's the order in which you arrange your items. If you want something to have a higher priority, simply move it up in the list.

This gives you much finer-grained control than a typical priority system, which typically only has a few levels of priority: OmniFocus effectively has as many priorities as you have items.

Does that make sense?[/QUOTE]

And that's why you're CEO, Ken.

[QUOTE=Ken Case]OmniFocus has a notion of priority already: it's the order in which you arrange your items. If you want something to have a higher priority, simply move it up in the list.

This gives you much finer-grained control than a typical priority system, which typically only has a few levels of priority: OmniFocus effectively has as many priorities as you have items.

Does that make sense?[/QUOTE]

Just to play devil's advocate (I'm happy with OF the way it stands), wouldn't that solution allow you to prioritize [I]within[/I] a project but not [I]across[/I] projects? (Or am I missing something?)

Non-canonical?
 

I, too, don't believe that we need a 'priority' field (at least not as a high priority :) but would love to see some way of flagging projects that will make all tasks of those projects stand out in all the different views. I'm not sure how this would integrate with the planned "review" functionality, but I definitely need a way to mark projects that need an hourly review to make sure they're on track.

As for the complaint that priority is non-canonical... I don't have The Book right here, but I do remember that, when discussing the 4 factors for doing (first by context, next by time available, then by energy available and - FINALLY - priority) DA was talking about the possibility that we might at one point have technology that allows us to enter in all 4 elements for all tasks and then filter based on all 4 items. However, based on this, we would need to have an "energy available" field before a priority field.

Usually, energy available depends on the context. So a nice implementation that doesn't require entering so much info for each task might be to assign a default level of energy for each context, and then be able to make the really difficult call and the really easy blog post that you must draft stand out from the rest of calls and drafts you have to make. What do you think?

Traditional prioritization for non GTD purists
 

Although not part of the GTD purist approach many (most?) people who plan their tasks use a weekly/daily planning model and prioritize their tasks (A1..., B1...C1....). Can't this or a similar prioritization (optional) capability be added to OF to open the application to more potential users who are not bought into GTD? Being able to order all tasks doesn't cut it for me as there is no obvious focus on tasks that MUST be done first versus reminders and minor tasks in order to be effective in my work.

[QUOTE=Charliee]Although not part of the GTD purist approach many (most?) people who plan their tasks use a weekly/daily planning model and prioritize their tasks (A1..., B1...C1....). Can't this or a similar prioritization (optional) capability be added to OF to open the application to more potential users who are not bought into GTD? Being able to order all tasks doesn't cut it for me as there is no obvious focus on tasks that MUST be done first versus reminders and minor tasks in order to be effective in my work.[/QUOTE]

Charliee,

Have you tried flagging your must-do actions and projects? With the recent flag-based filtering options I'm finding flagging invaluable for the prioritization piece of choosing actions. I put flags on the projects related to my monthly goals. Now I can filter just flagged items to focus on what is important and group by due date to focus on what is urgent.

I don't mean to deny that some would find finer grained prioritization useful. Just wanted to offer a suggestion that's worked well for me. I actually found it very liberating to abandon a Covey-style prioritization system. Now my daily and weekly reviews focus on identifying next actions rather than arranging existing actions.

One of the reasons GTD abandoned prioritization is that is often doesn't work.

Say you're meeting someone in a downtown hotel for a working lunch at noon. You hit your subway connections well, your lunch partner called to say she's running late, so you have about 30 minutes waiting time in the hotel lobby. You pull out your laptop, look at your To Do list, and the top priority items are: Office Cleaning, check stationary order for typos, call printer and review design for ad.
None of these can be done while sitting in the hotel lobby.
With a GTD system, you can look at your "laptop-offline" and "phone-mobile" items, and see the tasks you can do there/then, without being reminded of the things you can't do. Maybe they're lower priority, but they still need doing (or you wouldn't bother to list them), and you CAN do them.

--Liz

Just posted a big thing about this...

[url]http://forums.omnigroup.com/showthread.php?p=17429#post17429[/url]

BZ

[QUOTE]OmniFocus can be used to implement the "Getting Things Done®" work-life management method developed by David Allen, but it's designed to be flexible enough to accommodate different styles of personal organization.[/QUOTE]

Arguments about prioritizing in GTD aside, OF is designed to be useful even if you don't use GTD. As long as it doesn't force a user to do things in a non-GTD way, I don't see what harm is done by including non-GTD concepts.

Specifically on prioritization, I don't think it's anti-GTD at all. GTD helps you define what you can do at any moment, and I think the ability to prioritize those results is useful. It would be best implemented as a way to sort the results of your filtering, and not as a filter setting itself.

If I'm making phone calls, and have 3 calls to make: 1) I can call to reschedule an August doctor appointment, and it's due by the end of July while they still have appointments available. 2) I can call Payflex about submitting hospital expenses for reimbursement, also due by the end of July. Or 3) I can call my friend Steven for a chat since we haven't talked for a while, and I should call him in July because he's a teacher and school starts up in August again.
Three calls, all 'next actions', all are 'due' by the end of the month, and I have the time, energy, and context for making all 3, so which do I make?

I should make the Payflex call first, it's hands down the most important. But if I have a long list of calls to make (and the phone call to Steven can easily be multiplied by 20, I'm notoriously bad at keeping in touch with people), I would like to be confident that I'm doing one of the most important things available for doing.

[QUOTE=GeekLady]As long as it doesn't force a user to do things in a non-GTD way, I don't see what harm is done by including non-GTD concepts.[/QUOTE]

There are at least two potential harms:[LIST][*] Including non-GTD features could delay the release of the software. (Or more selfishly for those of us in the sneaky-peek, delay the fixing of bugs.)[*] The art of elegant software design is largely about what to leave out. Including every requested feature results in products like Microsoft makes. "Is this GTD?" is one reasonable criterion among many for evaluating feature requests.[/LIST]
Arguments among alpha testers about what features to include is always a popular sport. We often see the selection of features as a zero-sum game. "If you get your priorities, I might not get my meta-data fields. Egads!" It's somewhat more complicated than that, of course.

That said, I agree that priorities are a reasonable feature request and can fit well within a GTD methodology, though they can be a planning rat-hole if one isn't careful.

[QUOTE=GeekLady]

If I'm making phone calls, and have 3 calls to make: 1) I can call to reschedule an August doctor appointment, and it's due by the end of July while they still have appointments available. 2) I can call Payflex about submitting hospital expenses for reimbursement, also due by the end of July. Or 3) I can call my friend Steven for a chat since we haven't talked for a while, and I should call him in July because he's a teacher and school starts up in August again.
Three calls, all 'next actions', all are 'due' by the end of the month, and I have the time, energy, and context for making all 3, so which do I make?

I should make the Payflex call first, it's hands down the most important. But if I have a long list of calls to make (and the phone call to Steven can easily be multiplied by 20, I'm notoriously bad at keeping in touch with people), I would like to be confident that I'm doing one of the most important things available for doing.[/QUOTE]

You set the priority and that is the point. No software is going to know which one of those calls is the most important or process that information as quickly as your brain can.

Look at a list of 5 calls and you will know instantly which one you need to do first and you will.

BZ

You're missing my point. Of [I]course[/I] I set the priority, but I set it when I process the task, just like I set the due date and time estimate. Once processed, it goes into the system. And when I look at a context, I can choose to have the results sorted in the order of the priority that I assigned to them.

It's easy to look at a list of 3 calls or 5 and know which is the most important. It's not easy to look at a list of 10 or 20 tasks associated with different lab projects and know which is the most important, especially when you aren't the person determining a task/project's priority.

[QUOTE=BwanaZulia]You set the priority and that is the point. No software is going to know which one of those calls is the most important or process that information as quickly as your brain can.

Look at a list of 5 calls and you will know instantly which one you need to do first and you will.

BZ[/QUOTE]

I personally wouldn't use a priority field, but I think it would be fine to include it as long as it can be turned off.

I try to use contexts as much as possible to help me make the decision about what action to do next, rather than other fields. I don't even use the 'time estimation' field. Instead, I have contexts that reflect the time and effort that tasks will take.

In fact, I have a 'high priority' context for tasks that I should do in the quiet couple of hours I have in the morning when I'm most focused.

For me, contexts are not just places or resources, they are 'pre-decisions' -- the most common reasons why I might choose one task over another at any one moment.

I agree with GeekLady here. I see the need of having priorities after trying OF for roughly a month now. Also, as far as I understand GTD, priorities are a part of it even if only the least of it.

I want to use task prioritization in addition to other features (some things have due dates, some are repeating, etc.) to make sure that critical things do show up on the top of the list.

If having priorities means the release of OF 1.0 would be delayed, I'd be ok with a 1.1 or so goal for the option, as long as it wouldn't make me wait for a year or so.

[QUOTE=Weasel][...]I see the need of having priorities after trying OF for roughly a month now. Also, as far as I understand GTD, priorities are a part of it even if only the least of it.[...][/QUOTE]

You can drag items up and down the list to assign them priorities. This, to my mind, is much more precise than giving them numeral values for priority. What does priority 3 mean? "Ignore for now"? "Not very urgent, but still deserving attention within the next ten days"? Priorities are, by definition, relative, so dragging them up and down makes a lot of sense. And you can have as much priority levels as you have items (as Ken Case pointed out earlier).

[QUOTE=Weasel]I agree with GeekLady here. I see the need of having priorities after trying OF for roughly a month now. Also, as far as I understand GTD, priorities are a part of it even if only the least of it.
[/QUOTE]

Priority is a part of GTD, but priorities change as circumstances change. Ignoring this reality is a certain step to failure which is why the GTD framework summarily rejects arbitrary priority numbering systems. I just see no value in trying to predict priorities ahead of time (which is all assigning a priority in OF would amount to) when I'm perfectly capable of deciding what on my current list of action items (after I've already narrowed it down by context, time available and energy available) is the highest priority right now. The whole point of GTD is to train your mind to make better choices in the moment (when most choices end up being made anyway, all planning ahead aside) so that your decisions effect your life positively rather than arbitrarily. Relying on priority assignment systems ultimately undermines that goal.

I think you're asking for OF to facilitate a broken system rather than taking the (admittedly scary) plunge into GTD the way it's meant to be used. That's fine as a personal choice. I'm by no means a pedant when it comes to pure GTD implementation, but I don't expect non-GTD-like features to show up in GTD-specific applications.

If it can be done in such a way that not only is it optional, but it can be completely hidden, then I see little harm in adding this feature to OF. I think it could be the start of feature creep. I think it could encourage new GTDers to use the priority field as a crutch that ultimately stands in the way of them really understanding the usefulness of the next action list. I think it could slow down the release of OF 1.0. And I think it's borderline pointless if you can order your tasks in the list anyway to mimic a priority feature.

But on a personal level, as long as I can turn it off, it won't interfere with me.

MEP hits it out of the park!

BZ

1. You can only do this ordering in Project, not in Context, so it's not useful for actually doing things. And even if can eventually order tasks in context, having to drag things around constantly is annoying.

2. Numerical systems aren't easy to read, but "High" "Medium" "Low" and "None" are reasonably understandable, even if they're not the most granular.

3. Using Flags doesn't work well, because you can only see flagged tasks or all tasks, what prioritizers want is to Sort tasks, not Filter them.

[QUOTE=HiramNetherlands]You can drag items up and down the list to assign them priorities. This, to my mind, is much more precise than giving them numeral values for priority. What does priority 3 mean? "Ignore for now"? "Not very urgent, but still deserving attention within the next ten days"? Priorities are, by definition, relative, so dragging them up and down makes a lot of sense. And you can have as much priority levels as you have items (as Ken Case pointed out earlier).[/QUOTE]

To Mr. Curt:

My quote was merely to illustrate that OF is already described not as a GTD-only piece of software, but as a more flexible productivity application. So, I my opinion, arguments about including or excluding features based solely on GTD canonicity is kinda silly. Feature-creep is another argument altogether, but task prioritization is pretty basic.

Ok, I know a lot of people here is very orthodox in regards to GTD and I'm aware that most of you don't want to see anything outside the 'holy book' of GTD in OF. But there is a whole wide world of people and potential users of OF that might want some things on top of it.

As for my own case, I've been using LB for years and I was all the time looking for another solution since LB doesn't seem to be evolving and besides many good things it has a horrible interface, it is a pain to add projects/tasks and a few more issues many other users have brought forward too.

I have tried several other GTD related apps, having tried iGTD for x times (I download it about every 2-3 weeks start working with it and throw it out again after 2 days - I can't say what it is but it doesn't feel right). Some time ago I tried again kGTD and got somehow started with it. About a month ago I got the invite to OF alpha and since then I have been using it in parallel with LB (safety net).

I have during these four weeks tried many different ways of arranging my data, I have also read DA'a book and I have read almost every single post on this forum.

The best system I have come up with so far is:
on the project side:
[CODE]STUDY [I](most important thing in my life)[/I]
- Administrative
- University
-- Course 1
-- Course 2
-- .....
- Other university
-- Class 1
-- Class 2
- Single Tasks
- Books to read
WORK [I](not as important as study)[/I]
- organizing [I](all GTD and tagging related stuff incl. sorting desk, etc.)[/I]
- website 1
- website 2 [I](there are about ±10 sites I'm hosting/maintaining)[/I]
- Computers
-- [I](7 macs listed by their name for different maintenance tasks)[/I]
EVERYTHING ELSE [I](exactly what it says)[/I]
- Single tasks [I](everything that doesn't fit anywhere else)[/I]
- Personal [I](from hairdresser to doctors appointments)[/I]
- Dogs
- Chores [I](anything in the house like A/C maintenance etc, mostly repeating)[/I]
[/CODE]
on the context side I have actually only two main contexts
[CODE]HOME (I'm a distance student and working from home)
DOWNTOWN (has sub-contexts by mall, store etc.)
[/CODE]
When I tried to make more contexts the thing got totally out of control. I'm also not trusting myself that I would always think about checking various configs of filters etc. allthough I do use perspectives to focus.

The point where I see the need for priorities is that I had e.g. a task to do related to one website, critical, due date set and it got just lost in a pretty list of black and purple. I also need to have e.g. the Frontline and Heartgard repeating tasks for the dogs to show up top of the list when it is due.

[B]How would I use the priorities:[/B]
If there are 5 levels of priorities, be it numbers or words, I would make 'Study' by default 2 (but books to read would be 4 or 5), Work by default 3. Then the dog meds would have 1 as well as other really important things too.

I guess there would be a way to achieve this with tons of contexts and filters, but then I will need another list (OO checklist?) to make sure I checked all possible configurations.

If everyone keeps on talking about the fail-safe system, for me a system like that allows me to [B]not forget (overlook) important stuff without making me spend excessive time on checking the various possible list configuration[/B].

[QUOTE=Weasel]
I guess there would be a way to achieve this with tons of contexts and filters, but then I will need another list (OO checklist?) to make sure I checked all possible configurations.

If everyone keeps on talking about the fail-safe system, for me a system like that allows me to [B]not forget important stuff without making me spend excessive time on checking the various possible list configuration[/B].[/QUOTE]

I actually do use OO checklists for Morning Review, Evening Review, Weekly Review, Monthly Review, and Bi-annual Review. Periodic reviews to keep your system working are a key tenet of the GTD approach. You can get some benefit of GTD without doing them, but the benefits multiply when you have all the pieces.

The key is the definition of "excessive time". I probably spend about 30 minutes a day on daily reviews (10 in the evening and 20 in the morning). My weekly review probably takes an average of an hour. I don't consider that time excessive at all; in fact, it is essential.

[QUOTE=curt.clifton]The key is the definition of "excessive time". I probably spend about 30 minutes a day on daily reviews (10 in the evening and 20 in the morning). My weekly review probably takes an average of an hour. I don't consider that time excessive at all; in fact, it is essential.[/QUOTE]

Very good point. It's taken me a long time to learn that I need to plan around half an hour at the end of my day to review what I've done--not just in terms of what remains to be done, but also how well I've done it and anything I could do better the next time (or need to fix this time). OmniFocus could use better tools for this, though Perspectives, once they are more robust, could help a lot.

[QUOTE=BwanaZulia]Look at a list of 5 calls and you will know instantly which one you need to do first and you will.
[/QUOTE]

This is my thought — if I can tell which activity is highest priority today, I should still be able to pick it out of a list 3 days from now.

Priorities ARE important — but do they need to be written down? I'm not so sure.

GeekLady, using your 3 phone calls example, you know the call to Steve is lowest. But of the other two, one takes 3 minutes and the other takes maybe 20. If you have 5 minutes for phone calls right now, you have time to set up the appointment but not deal with voice mail h*ll at the insurance company.

I'm beginning to realize that I need enough contexts -- and supplementary info such as duration, dates, etc. to be able to view my stuff in little tiny groups. Then I can focus in on the tiny group that fits my needs right now, and do the task that jumps out at me in that group, without being nagged by others. If I have enough supplementary data, I don't need priority.

Which is not to say you don't need it.

--Liz

Indeed, but I did specify in my example that I had the time to make any of those 3 calls.

Perhaps a more complicated example is in order. I'm a research technician. At the lab, I have, at any one moment, a minimum of 20 active projects. I also do not have the priviledge of setting my own priorities - my PI sets my priorities (and these can change hourly depending on his mood anyway). But my PI's priorities as a researcher differ from my own priorities as his lab manager. Part of my job is to make sure the lab will keep running, that the computers all continue to work, that we have enough supplies for experiments, etc. But this part of my job is essentially invisible to my PI, he notices if it isn't done, but only because it interferes in what [I]his[/I] priorities for me are - that I assist the post docs and students in their research projects.

It is easy to remember and focus on what my priorities are. It's harder to focus on other people's priorities (especially when they change frequently), and this is where the ability to prioritize tasks and projects would be extremely helpful for me.

[QUOTE=LizPf]
GeekLady, using your 3 phone calls example, you know the call to Steve is lowest. But of the other two, one takes 3 minutes and the other takes maybe 20. If you have 5 minutes for phone calls right now, you have time to set up the appointment but not deal with voice mail h*ll at the insurance company.[/QUOTE]

[QUOTE=GeekLady]Indeed, but I did specify in my example that I had the time to make any of those 3 calls.

Perhaps a more complicated example is in order. I'm a research technician. At the lab, I have, at any one moment, a minimum of 20 active projects. I also do not have the priviledge of setting my own priorities - my PI sets my priorities (and these can change hourly depending on his mood anyway). But my PI's priorities as a researcher differ from my own priorities as his lab manager. Part of my job is to make sure the lab will keep running, that the computers all continue to work, that we have enough supplies for experiments, etc. But this part of my job is essentially invisible to my PI, he notices if it isn't done, but only because it interferes in what [I]his[/I] priorities for me are - that I assist the post docs and students in their research projects.

It is easy to remember and focus on what my priorities are. It's harder to focus on other people's priorities (especially when they change frequently), and this is where the ability to prioritize tasks and projects would be extremely helpful for me.[/QUOTE]

I think you just gave a perfect example why actually recording and managing priorities is a mistake. They are constantly changing, are too numerous and possibly blinding because of their complexity.

Your mind has the great ability to scan a small list (context) and pick from there what the biggest priority is.

BZ

You're making the assumption that 1, priorities are so dynamic that by the time I've recorded a change, they've changed again; 2, that I [I]can[/I] pick out the tasks most important to someone else from a LONG, not short, list of things that need to be done. I have no short lists in the lab.

I just don't always remember which things are super important to my boss. That's why I find priorities so useful - otherwise, I [B]will[/B] forget something. It'll be in my list, but I won't get to it as quickly as the boss expects, because I don't recognize it as a priority.

If I'm going to be able to trust my system, it has to hold on to the things that I will otherwise forget, and one of the things that I forget, that I know I forget, is how important different tasks are to my boss. It imposes a certain amount of hassle on me, but much, much less hassle than forgetting to email some results to the postdoc that left the lab 2 years ago and is just now getting around to publishing his paper on my sweat blood and tears.

[QUOTE=BwanaZulia]I think you just gave a perfect example why actually recording and managing priorities is a mistake. They are constantly changing, are too numerous and possibly blinding because of their complexity.

Your mind has the great ability to scan a small list (context) and pick from there what the biggest priority is.

BZ[/QUOTE]

We can go around like this all day. If we base these things purely on personal preference, my use for priorities is just as valid a reason for including them in OF as your non-use is for excluding them.

Given flags, due dates and drag and drop ordering, OmniFocus has more prioritization than a project/context system (aka GTD) needs. Using those three items, the arrange function allows you to sort and prioritize on the fly, allowing your productivity to flow as rapidly as needed.

Just my 2¢.

Jim

[QUOTE=Lecter]Given flags, due dates and drag and drop ordering, OmniFocus has more prioritization than a project/context system (aka GTD) needs. Using those three items, the arrange function allows you to sort and prioritize on the fly, allowing your productivity to flow as rapidly as needed.

Just my 2¢.

Jim[/QUOTE]

1. You cannot drag and drop order in context.
2. You can't sort by flags, only choose to see just flagged items or all items.

I was using flags for priority markers until they moved it to it's own filter.

I'm already on record as unopposed to adding priorities to OF, but want to better understand what you're saying here.

[QUOTE=GeekLady]
1. You cannot drag and drop order in context.
[/QUOTE]

You can, however, reorder whole projects in project view. I organize my projects by life roles, but one could create folders in project view based on relative priority. For example, all the highest priority projects could go under the Highest Priority folder, which you would put first in the project listing. Then if you view contexts as Unsorted, the actions from the highest priority projects will come first.

[QUOTE=GeekLady]
2. You can't sort by flags, only choose to see just flagged items or all items.

I was using flags for priority markers until they moved it to it's own filter.[/QUOTE]

I don't understand what you're saying here.

Why would moving flags to their own filter cause you to stop using them as priority markers? It is true that you can't sort by flags. (I've submitted a feature request for that). But moving flags to their own filter makes them more powerful than when they shared a filter with action availability.

[QUOTE=GeekLady]You cannot drag and drop order in context.
[/QUOTE]

I also am not opposed to adding priorities to OF.

However, I didn't realize until reading Curt's post that we can sort in context view (or even see/set the View Bar in context view for that matter), but since we can sort by name in the context view, why couldn't you reorder things in context view simply by adding a priority number to the beginning of the action names and than sorting by name? That seems simple enough. Anything PI wants gets "1" and would sort to the top of the list.

[QUOTE=GeekLady]
It is easy to remember and focus on what my priorities are. It's harder to focus on other people's priorities (especially when they change frequently), and this is where the ability to prioritize tasks and projects would be extremely helpful for me.[/QUOTE]

Ah, now it makes sense.

It's not your own priorities that need recording, but someone else's. I understand how easy it is to forget what is important to someone else -- if you asked my husband this is m worst failing. <grin>

Could you add a code word in the Task name, such as HighPri AE, for the task of yours that is highly important to Albert Einstein? Then you could search on the priority code or the PI's initials.

--Liz

Exiting lurk mode for a moment. I've been following this thread closely, and I'm really sympathetic to the request for separate priorities.

I'm just also trying to avoid adding yet another built-in dimension to actions (on top of project, context, flagged, next action, start and due dates, time estimate, etc.), since the more options we add the easier it is to get lost spending time thinking about the system rather than about the actions themselves. (But if you're used to setting priorities and your new system doesn't support it, then you're probably going to spend a lot of time thinking about the system...)

[QUOTE=GeekLady]I was using flags for priority markers until they moved it to it's own filter.[/QUOTE]

Oh! Would it help reduce the need for a separate priority attribute if we restored the ability to sort actions by whether they're flagged?

(Alternatively, what did you think of the prefix hack suggested above where you include the priority as part of the action's name?)

Personally, I don't think that prioritisation fits well with GTD for the reasons I've given [URL="http://forums.omnigroup.com/showpost.php?p=17660&postcount=23"]here[/URL], so I'd support no prioritisation beyond flagging and reordering of lists (including sorting by flagging if people feel that would help).

[QUOTE=Ken Case]...since the more options we add the easier it is to get lost spending time thinking about the system rather than about the actions themselves.[/QUOTE]I wholeheartedly concur!

[QUOTE=Ken Case]Oh! Would it help reduce the need for a separate priority attribute if we restored the ability to sort actions by whether they're flagged?[/QUOTE]I think so. Plus, I can't think of a good reason why one shouldn't be able to sort by any of the columns. If a column is worth putting in (and worth being used) it should be worth sorting by.

[QUOTE=Ken Case](Alternatively, what did you think of the prefix hack suggested above where you include the priority as part of the action's name?)[/QUOTE]A adequate workaround.

Jim

[QUOTE=Ken Case]
Oh! Would it help reduce the need for a separate priority attribute if we restored the ability to sort actions by whether they're flagged?

(Alternatively, what did you think of the prefix hack suggested above where you include the priority as part of the action's name?)[/QUOTE]

It would eliminate [I]my[/I] need for a separate priority attribute if we can sort by flagged, that's how I was using flags before, but I only really need 2 levels of priority - boss placating and normal. I'm sure there are other people that would use a more advanced priority system.

However please don't add flagged back to Sort at the expense of removing the separate flagged filter, Curt might hurt me!

I don't like the idea of adding some sort of code to the front of the action's name for 2 reasons:
1 - Since I need to change them occassionally, I don't want to accidentally delete or alter my action's name trying to change it's priority.
2 - It's just so damn inelegant.

OF does already have notions of priority - in Project view. This is cool but not very useful. Priority is one of GTD's things to evaluate when choosing a task and so, in my opinion, OF needs some way to evaluate priority in Context view. Being able to drag tasks to order them (in Context) would be the most granular, but also the most cumbersome.

I have a hard time thinking of a way to implement priority when all I want is a way for things that I forget are important to keep floating at the top of my lists.

Priority IS Canonical GTD
 

OK. Let's go back to the source. I found this on page 49 of [I]Getting Things Done[/I]. Here's a paraphrased summary:

[B]The Four-Criteria Model for Choosing Actions in the Moment[/B]

[I]For me, this is what I think of when I'm in the Context view of OF[/I]

[B]Context[/B] Most actions require a specific location or productivity tool. That limits your choices on what you [I]can[/I] do in the moment.

[I]OF supports this, obviously.[/I]

[B]Time Available[/B] Having a meeting in five minutes prevents you from doing many actions.

[I]OF supports this as well.[/I]

[B]Energy Available[/B] Some actions require a reservoir of fresh creative energy, others need physical horsepower.

[I]Doesn't look like OF supports this directly.[/I]

[B]Priority[/B] Give the context, time, and energy, what action will give you the highest payoff? (His examples in the book are [I]cross project[/I] examples)

[I]OF doesn't support this either[/I]

=====

To me, the first three are filters, and the last one is a sort of remaining actions with NO grouping.

[QUOTE=timshadel][B]Energy Available[/B] Some actions require a reservoir of fresh creative energy, others need physical horsepower.

[I]Doesn't look like OF supports this directly.[/I][/QUOTE]I think this would be an "of the moment call" of the individual doing the task. To integrate a energy available feature to OmniFocus would be the worst form of micromanagement IMHO. :)

[QUOTE=timshadel][B]Priority[/B] Give the context, time, and energy, what action will give you the highest payoff? (His examples in the book are [I]cross project[/I] examples)

[I]OF doesn't support this either[/I]
[/QUOTE]If you can sort by flags, then they could be used exactly for this type of delineation. Is this a priority or not? Finer grained priorities can be bad for you. A reference follows:

[QUOTE=page 8 from the Book Of David :)]And daily to-do lists and simplified priority coding have proven inadequate to deal with the volume and variable nature of the average professional's workload. More and more people's jobs are made up of dozens or even hundreds of e-mails a day, with no latitude left to ignore a single request, complaint, or order. There are few people who can (or even should) expect to code everything an "A," a "B," or a "C" priority, or who can maintain some predetermined list of to-dos that the first telephone call or interruption from their boss won't totally [I]undo.[/I][/QUOTE]

[QUOTE=timshadel][B]Energy Available[/B] Some actions require a reservoir of fresh creative energy, others need physical horsepower.

[I]Doesn't look like OF supports this directly.[/I]
[/QUOTE]

If OF can tell how energetic I'm feeling, it's cleverer than I thought. :)

I think one of the main points of GTD is neatly summarised in Lecter's quote: you can't set priorities rigidly in advance, because the landscape of the typical working day can change so rapidly. That renders any sort of "high, medium, low" or whatever system redundant as far as I'm concerned: you'd spend longer readjusting your priority settings than you would actually doing the tasks (OK, I'm exaggerating there, but you see my point). One of the beauties of GTD for me is that it cuts out just that sort of fiddling.

I think that this constant harping on how priorities change too quickly to be worth keeping track of is really falling into one of the traps GTD is designed to help us avoid: [B]don't stop to immediately process a new task or piece of information, unless it's absolutely necessary.[/B]

My workplace is one constant interruption (as a technician, I am keeper of the lab's tribal wisdom!), and pre-GTD I was notorious for forgetting to do things I committed to doing. The one great thing it taught me that I didn't have to try to remember, and I didn't have to do it right away, but I could put it in an inbox and deal with it when I wasn't working on something.

If a priority changes, if I need to put one project on my back burner and work on something else, I don't run back to my computer right away to update my priorities. If I chip a tooth and have to call my dentist right this very instant, I don't go and upgrade my "call the dentist" task first.

If something needs immediate action, I do it. I don't fuss about whether my system says it's important - this is what the evening review is for. If something gets a priority change, but doesn't need immediate action, it goes to my inbox - a 2 minute task that probably will never see the inside of OF's database.

And in the evening when I review my day, if a lower priority task became immediate and got done, I just check it off, because hey it got done, and now it doesn't matter whether it was urgent or not.

It's useful information if you don't obsess about it. Not everything needs a start date or a due date, and not everything needs a priority, but if you don't recognize a task as important when it really is, it's useful information to have.

[QUOTE=al_f]That renders any sort of "high, medium, low" or whatever system redundant as far as I'm concerned: you'd spend longer readjusting your priority settings than you would actually doing the tasks (OK, I'm exaggerating there, but you see my point).[/QUOTE]

Here's a crazy idea for consideration. (Please excuse my rough sketches.)

What if priority was just a little icon like the current flag and note icons? Something like this:
[ATTACH]258[/ATTACH]

Clicking the priority icon would pop up a little slider control something like this:
[ATTACH]259[/ATTACH]

Rather than being a very specific value in an A-B-C, 1-2-3 priority system, priority just becomes a point on a relative scale. (I think Life Balance does something like this, but it's been awhile.)

Now here's the crazy bit...

Suppose you're in context view and sorted by priority. Dragging a task to a new position in the list would set its priority slider to the average of its two new neighbors. Thus, rearrangements in context view would "stick". You could change the view configuration then change back to priority-based sorting without losing the ordering information. And like dates and flags, the priority information wouldn't be "in your face". It would just be lurking there if you need it.

This is probably too baroque for an Omni app. In particular, what should happen if you try to drag items in context view when sorted by something other than priority. But it seems like a fun idea to play with.

Without rehashing the debate about the GTD-osity of priorities, I'd love to hear what others think of this idea.

I love the idea, it is kind of similar to LB's scheme of importance.

[IMG]http://forums.omnigroup.com/attachment.php?attachmentid=260&stc=1&d=1184885516[/IMG]

Having experienced this feature for the last few years, I know I'd like to have a more graduated scale. I had asked for adding just 10 little marks below the slider. Some kind of visual clue...

On the same note I prefer the little squares on the screen when adjusting the volume on a mac over the drop down slider from the menubar.

I love it.

Beautiful.
Elegant.

Probably too much for 1.0, but I would cope with flagging if I knew something like this was coming in 2.0
[QUOTE=curt.clifton]Here's a crazy idea for consideration ... Without rehashing the debate about the GTD-osity of priorities, I'd love to hear what others think of this idea.[/QUOTE]

[QUOTE=curt.clifton]Dragging a task to a new position in the list would set its priority slider to the average of its two new neighbors. Thus, rearrangements in context view would "stick".[/QUOTE]

It turns out that this is exactly how we order actions now: each action has a "rank" which is assigned based on its position relative to other actions in the list.

It sounds like the proposed change, then, would be to allow actions to have individual ranks in context view rather than forcing them to only have child ranks within their own project. This would let you interleave actions from different projects in context view, effectively assigning individual priorities to each action (based on their position).

[QUOTE=Ken Case]It sounds like the proposed change, then, would be to allow actions to have individual ranks in context view rather than forcing them to only have child ranks within their own project.[/QUOTE]

I should mention that we've considered this approach somewhat already, and the big sticking point has been that we're worried about performance when you have to rerank everything to make room: right now you only have to rerank the other actions in your group (hopefully not more than a dozen or two—which also makes it unlikely that you'll ever have to rerank at all), but in this other system you'd have to rerank every action in your database (including the completed ones, since you want to preserve their order as well) whenever you run out of ranks between two items (which would happen much more often).

Of course, that's not nearly as big a problem if you have a separate priority field rather than using it as the only indication of order. (Maybe an easier path would be to just give people more flagging options, so you can flag something as "high" or "low" priority rather than just "a" priority?)

[QUOTE=Ken Case]right now you only have to rerank the other actions in your group (hopefully not more than a dozen or two...[/QUOTE]

Pardon me while I reply to myself yet again, but I thought I'd mention that this is not to imply that it's unreasonable to have hundreds or thousands of actions in the same group, just that that's a lot less common than having that many actions in the entire database.

Ken,

What would it do to the performance if the priority was stored as a floating point number instead of an integer? For all practical purposes there are an infinite number of floats between any other two floats. (Of course that isn't technically true, but I suspect disk capacity issues would arise before anyone exhausted the precision of 64-bit floats.)

[QUOTE=Ken Case](Maybe an easier path would be to just give people more flagging options, so you can flag something as "high" or "low" priority rather than just "a" priority?)[/QUOTE]

Multiple levels of flags would be fine and quite help to achieve my goal if the sorting by flag status is possible.

Another thought, if you add multiple levels of flags, would this not be after all a priority system which some of the forum members so vehemently fight against?

And as I understand many of the anti-priorities faction use one level flags and they would probably not be happy with multi-level flags?

Would it not be possible, and maybe more considerate to all sides, to add the priorities system as a separate feature and have an option in preferences to enable/disable it?

[QUOTE=GeekLady]I think that this constant harping on how priorities change too quickly to be worth keeping track of is really falling into one of the traps GTD is designed to help us avoid: [B]don't stop to immediately process a new task or piece of information, unless it's absolutely necessary.[/B]
[/QUOTE]

I don't think that really has much to do with whether pre-assigning priorities to tasks is canonical GTD or not. I'm pretty sure it's not, but I think we'll have to agree to differ on this one as I don't want this to turn into a "priority vs non-priority" holy war. :)

Suffice it to say that I don't mind if there is a priority system in OF as long as it's hidden until you choose to use it and it doesn't degrade perfomance. I personally wouldn't use it, but if it works for you that's great.

I've said previously that I couldn't see myself using a priority feature (although i don't mind if it is there for others)...however, i probably would use a dragable priority system like Curt suggested

[QUOTE=al_f]I don't think that really has much to do with whether pre-assigning priorities to tasks is canonical GTD or not. I'm pretty sure it's not, but I think we'll have to agree to differ on this one as I don't want this to turn into a "priority vs non-priority" holy war. :)[/QUOTE]


Too late. :)

BZ

ho ho. It's a very well-mannered war though. :)

I can sort by flagged again! Thank you!

[Edit] As holy wars go, it's been a very polite one. Someone break out the cupcakes.

[QUOTE=GeekLady]Someone break out the cupcakes.[/QUOTE]The Soup Nazi says, "No cupcakes for you, priority zealot!"

<said with tongue [I]firmly[/I] in cheek>

[QUOTE=curt.clifton]Here's a crazy idea for consideration. (Please excuse my rough sketches.)

What if priority was just a little icon like the current flag and note icons? Something like this:
[ATTACH]258[/ATTACH]

Clicking the priority icon would pop up a little slider control something like this:
[ATTACH]259[/ATTACH]

Rather than being a very specific value in an A-B-C, 1-2-3 priority system, priority just becomes a point on a relative scale. (I think Life Balance does something like this, but it's been awhile.)

Now here's the crazy bit...

Suppose you're in context view and sorted by priority. Dragging a task to a new position in the list would set its priority slider to the average of its two new neighbors. Thus, rearrangements in context view would "stick". You could change the view configuration then change back to priority-based sorting without losing the ordering information. And like dates and flags, the priority information wouldn't be "in your face". It would just be lurking there if you need it.

This is probably too baroque for an Omni app. In particular, what should happen if you try to drag items in context view when sorted by something other than priority. But it seems like a fun idea to play with.

Without rehashing the debate about the GTD-osity of priorities, I'd love to hear what others think of this idea.[/QUOTE]

Outstanding!

[QUOTE=BwanaZulia]
Your mind has the great ability to scan a small list (context) and pick from there what the biggest priority is.

BZ[/QUOTE]

The mind should not be used for scanning, ordering or reordering of any type. Software does a better job.

3 calls is nothing... until you get to 100 calls. Trying to scan those is ridiculous.

It's like processing the inbox. We don't scan the inbox. We take the top one and process.

How do we get OF to show us the most important thing to do at the top of the list so we don't have to scan.

[QUOTE=SpiralOcean]The mind should not be used for scanning, ordering or reordering of any type. Software does a better job.

3 calls is nothing... until you get to 100 calls. Trying to scan those is ridiculous.

It's like processing the inbox. We don't scan the inbox. We take the top one and process.

How do we get OF to show us the most important thing to do at the top of the list so we don't have to scan.[/QUOTE]

This might be where the Focus command helps. You might have 100+ items in your Phone context, but if you focus on a particular project/folder/whatever and then look at the context, it should be a lot easier. When I'm wearing my graduate program director hat, for instance, and I'm focused on that folder, I'll only see the actions that are relevant to that. It almost makes me wonder whether OmniFocus needs another way to focus, so that you don't have to switch back to the Project view to do so. But that might be too baroque; the app needs to remain reasonably simple.

Come to think of it, a more robust Perspectives would go a long way toward solving this problem.

I've been thinking this over once more, and if we can, at some point, reorder tasks in context view by dragging and dropping them, that will solve the issue for me. Adding yet another attribute to a task (priority, however prettily represented) will not.

[QUOTE=GeekLady]Arguments about prioritizing in GTD aside, OF is designed to be useful even if you don't use GTD. As long as it doesn't force a user to do things in a non-GTD way, I don't see what harm is done by including non-GTD concepts.[/QUOTE]

I would wholeheartedly agree with you there. OmniFocus does just about everything I need it to do for me right now; if the extra bells and whistles can be turned off, I'll be a happy camper.

[QUOTE=GeekLady][...] I should make the Payflex call first, it's hands down the most important. But if I have a long list of calls to make (and the phone call to Steven can easily be multiplied by 20, I'm notoriously bad at keeping in touch with people), I would like to be confident that I'm doing one of the most important things available for doing.[/QUOTE]

I realize I'm verging into "what does GTD tell us?" mode, but I think the key concept David tries to get across in the book is that [i]it's all equally important in one way or another[/i] in terms of the mental resources it takes to keep track of stuff. You may need to call Payflex, but that might not be the call that you can knock off 5 minutes before you go to lunch, so it's not the highest priority thing you can do [i]right then[/i].

Still, there's a subtext that runs through the GTD book I try to keep in mind; you need to use a system that you will trust will give you the information you need when you need it. If you need task priorities to work, more power to you. If the Omni Group wants to include that feature in OmniFocus, I only ask that I'm able to turn it off (or not have it enabled by default).

add me in the "if it's a preference you can hide and never see again, I don't mind" camp. For work related stuff, I rely on due date and time it takes to complete, the rest of the thinking is done by the GTD system itself. I don't have a need for priorities...

getting priority out of my head
 

I'm jumping in late, but....

I guess my concern is this: If the beauty of GTD is getting stuff *out* of my head so that I can focus on what I need to do, then "how important these various things are" is something that I shouldn't hold in my head.

Otherwise, what happens is this: I look at a list of things I have to do, and I think, "Oh, I have to figure out what to do now, and I'm not sure, and I can't decide, and I kind of am not sure I want to do that one, but I really ought to, and maybe this other one is more important, and... oh, look, a bird!"

Deferring the decision as to what to do is, in some ways, deferring a very important and potentially anxiety-provoking decision to a time when it is maximally likely to derail me.

I would like a little more granularity than "flagged or not," and a little less than "an arbitrary number of decimal places." 1-5 is plenty. Just trying to get a sense of the difference between "this would be kind of nice sometime" and "this needs to be done ASAP or something awful is going to happen."

-- Aimee

I believe the phrase you are looking for is "ooooh SHINY!"

[QUOTE=aimee;18540]"oh, look, a bird!"[/QUOTE]

[QUOTE=aimee;18540]I'm jumping in late, but....

I guess my concern is this: If the beauty of GTD is getting stuff *out* of my head so that I can focus on what I need to do, then "how important these various things are" is something that I shouldn't hold in my head.

Otherwise, what happens is this: I look at a list of things I have to do, and I think, "Oh, I have to figure out what to do now, and I'm not sure, and I can't decide, and I kind of am not sure I want to do that one, but I really ought to, and maybe this other one is more important, and... oh, look, a bird!"

Deferring the decision as to what to do is, in some ways, deferring a very important and potentially anxiety-provoking decision to a time when it is maximally likely to derail me.

I would like a little more granularity than "flagged or not," and a little less than "an arbitrary number of decimal places." 1-5 is plenty. Just trying to get a sense of the difference between "this would be kind of nice sometime" and "this needs to be done ASAP or something awful is going to happen."

-- Aimee[/QUOTE]

Good points, especially the part about getting priority out of your head and into your system.

I think the main point of GTD regarding priorities though is that they change based on context, current time and daily changes in your responsibilities that are beyond your control. I can't take any of my tasks and give them a 1-5 priority that applies to all situational contexts.

For instance, let's say I have two clients I am working for and each of them have big projects currently in the works. Client A is a long-time client with a very lucrative contract and they've brought plenty of business my way through referrals in the past. I like Client A and generally consider any project for them a high priority.

Client B is... well not to be mean, but they're a pain in my ass. I bend over backwards for them and it's never enough. They push me around, demand meetings on short notice for stupid piddly crap that could wait until next week and every other contractor that works with them (and by association, occasionally with me) has to put up with the same stupid crap. I get Client B's work done, and I do it as best as I can for the sake of my portfolio, but honestly, they're not my highest priority.

I have a meeting with Client B tomorrow and there's work that needs to be done before that meeting. The next milestone deadline for Client A is still three days off and I'm ahead of schedule.

Now when I look at my current list of tasks, it's pretty clear to me what's higher priority right now. Client B's work, as much as I hate it, has to get done today. But in the overall scheme of things Client A's work is much higher priority for me and much more valuable and important.

So the question is this: Last week, long before I knew what situation I was going to be in today, when I was doing my weekly review, which priority do I assign to these tasks, and how do I put that information to use right now today?

Do I put high priority on Client A's work and low priority on Client B's? If so, then when I look at my action list today, Client A's work will be at the top of the list even though right now it is not the highest priority.

Client B's work is the highest priority right now. If I'm sorting by priority, Client B has to be at the top of the list, but the only way I can do that is give Client B's tasks a higher priority in my system, which is unlikely to happen because Client B really isn't that important to me. They can jump in a lake for all I care in the grand scheme of things (which is why I should probably dump them as a client, but I should've made that choice before committing to the project that currently should be my highest priority).

Should I try to predict what I will and won't get done between my review and every day following the review so I can assign priorities to my actions that will work on... oh say next Tuesday? Should I just assign priorities appropriately in the grand scheme of things during my review and then look through my entire action list every day to determine if those priorities are actually valid right now in this very moment (which they aren't, and frequently won't be)?

I could assign a due date to Client B's work, in fact, I have. But that due date tells me everything I need to know about that task's priority. "This project needs to be done by this date. That is all." I don't really need to know any more about it's priority than that.

I think priority isn't something you carry around with you in your head. Priority is something you decide at the moment, and it changes from moment to moment. Trying to assign priority to tasks during the weekly review is like trying to predict the future for every future moment you will look at your task list and then choosing the one numerical value that you think will be accurate in the most situations (but unlikely to be accurate for [b]all[/b] situations even if you do predict them all accurately). I don't remember the priority of my actions so that I can recall it when I look at my action list. [i]I choose the priority of my actions every time I look at my action list.[/i]

That's the difference between GTD and other systems. That's the goal that David Allen is trying to get us to work toward.

[QUOTE=MEP]I think priority isn't something you carry around with you in your head. Priority is something you decide at the moment, and it changes from moment to moment....That's the difference between GTD and other systems. That's the goal that David Allen is trying to get us to work toward.[/QUOTE]

An excellent point.

I have, however, found priority very useful in my next-action review process. As I look through a long list of avaiable tasks (or next actions on many active projects), I may want to move things to the top of the list, or the bottom of the list. Dragging and dropping items on the list works to some extent, but I find it requires too much thought during review.

In iGTD, I like having a 1-5 priority selector, because I can work my way through the list once, tag everything with its priority at this moment, then sort by priority and start working down the list.

Dragging and dropping, on the other hand, causes the list to change while I'm reviewing it, which makes it a little harder to focus on a quick pass through the list -- I'll worry that I missed something as I shuffle through the ever-changing list. Additionally, when deciding where to drop something, I have to re-evaluate the tasks around its destination, even if unconsciously. I find this added mental load distracts me from quickly passing through the list.

Do you have any other ideas how one might prioritize a long list without lots of dragging & dropping?

And, unfortunately, OF doesn't seem to let me manually order my tasks in context view (which is where I normally work). It's always sorting by some criterion. Using the built-in notion of priority that Ken Case mentioned wouldn't really work here, because that appears to be priority-within-a-project; so these internal priorities can't be compared between projects.

This all suggests to me that there's a place for explicit priority (even if some people abuse it in non-GTD ways).

During the review, I think dragging and dropping to change the order of tasks should work really well. Since each project or sub-project is its own little list, prioritizing in each one shouldn't be too hard. If you have a really long list of tasks in a single project, that really sounds like it could be broken down into smaller sub-projects. I have some truly daunting projects on my task list, but after reducing them down into their component sub-projects, I don't have a single sub-project that is really more than 6-7 tasks long and most of those are sequential so priority doesn't even enter into it.

It sounds like a lot of refactoring, and it can be if you let it, but ultimately I think breaking things down into smaller chunks helps a lot when dealing with long task lists. Any way you slice it, prioritizing or even contextualizing a large amount of tasks is going to take some amount of work. By breaking it up, you can prioritize all of your tasks within the smaller and easier to conceptualize context of a sub-project, and then each of those sub-projects can be prioritized in relation to each other rather than in relation to every single task.

So there's a lot of dragging and dropping the first time you set it up, but after that, it's just large chunks of data moving as single units all at once as you drag projects around and you only have to prioritize tasks themselves very rarely (if you grouped everything right which is the trick). It's hard to say without really seeing what kind of task list you're looking at, but that's what I would suggest.

As for priority in context view, that's a hard one. I really like scanning my entire visible action list in context view (after it's been narrowed down by context) so I can actively decide what's most important to me right now so I'm not a big fan of putting the most important things at the top so I can develop a habit of ignoring things at the bottom.

I think the most important thing here is to really use contexts to reduce your visible action list as much as you possibly can first. If location contexts (@Home or @Computer) don't narrow your list down enough, use sub-contexts (@Home:daytime for yardwork or @Computer:Web for researching things online) to really get that list as small as you can when you're looking at it right now.

If the list is small enough, scanning it visually and deciding what you have the time and energy to do and what is most important right now shouldn't be too hard. If the list is still too large I think you need to rework your contexts to make it smaller -- which kind of sounds like a cop out, but I don't know what else to suggest. This took me a long time to do myself. I think it was about two years of trying GTD before I finally got a list of contexts that I was satisfied with for more than a month at a time (my current context list has remained almost unchanged except for occasional additions as I've taken on new responsibilities for about a year). It's one of the hardest parts of setting up your next action list and I think most problems people have with their next action list can be remedied with rethinking their contexts (at least that's what I've found for myself).

I would suggest really examining where and when you assign high or low priorities in your task list and then ask yourself, "Is this really just another context?"

Could set your contexts up as, for example:

@Home - High
@Home - Medium
@Home - Low

Or am I missing something here?

[QUOTE=strongmace;18573]Could set your contexts up as, for example:

@Home - High
@Home - Medium
@Home - Low

Or am I missing something here?[/QUOTE]

Not quite what I meant. Let's say I examine all of my high priority tasks and discover that most of them are related to a certain kind of work. Well, then I've found a subcontext of @Work. I can stop using the priority for that and just use the subcontext instead to narrow down my task list when I'm working on those kinds of projects.

A personal example that won't apply to everyone: I'm studying industrial design right now and I have a context set up called @Campus. When I'm on campus, this is the list of things I need to be concerned with. This list started to get really large and unwieldy so I took a good look at it and realized that a large number of those tasks all had one thing in common, I needed to be in a lab or shop of some kind to complete them. So I created a sub-context called @Campus:labtime where I put all of those tasks.

I have, in my calendar, the lab schedule for department so I know when the wood shop is open or when I can get in a welding booth. When I'm on campus during those times, I look at the "labtime" list rather than the campus list.

It's possible, that when looking at your high and low priority tasks, you'll discover that the high priority tasks have something in common with each other beyond just priority. That something is your new context.

[QUOTE=MEP;18569]If you have a really long list of tasks in a single project, that really sounds like it could be broken down into smaller sub-projects.[/QUOTE]

My problem is that, in my line of work, I can have many dozen available projects at a given moment. (They're all small, but can be put on hold, prioritized, etc.)

Even when I'm in a specific context, I may still have many dozen next actions available. (Some of the projects won't appear because their next tasks aren't in the current context.) And it's the long list of available actions in my context that I need to, once or twice a day, review top to bottom to prioritize them.

Your point about getting into a bad habit of not looking at the end of a prioritized list is well-taken. But I can't really afford the overhead of going throught a LONG list (that really can't be whittled down) after every action.

I realize my situation may be a little unusual, having this many active tasks in a current context. In fact, many of the tasks will never get completed, simply because there isn't time -- but I still need to record them when they're first proposed so that I can consciously not do them if other, more important tasks displace them.

[QUOTE]If the list is still too large I think you need to rework your contexts to make it smaller -- which kind of sounds like a cop out, but I don't know what else to suggest.[/QUOTE]

How about priorities? :) I do appreciate the value of this approach, but unfortunately, it doesn't work in all cases -- sometimes you simply can't reduce the list of tasks, but you can prioritize them with varying levels of permanence.

And I really do need the lists to be this long, so that I have a single place to look for tasks and projects, other than in my head. Ah, the joys of having all possible irons in the fire simultaneously.

[QUOTE]I would suggest really examining where and when you assign high or low priorities in your task list and then ask yourself, "Is this really just another context?"[/QUOTE]

The examples you gave of subdividing like this all had to do with situational or geographical contexts, which is indeed how I use contexts. The only way I could subdivide further is to mix the notion of context with priority (as strongmace suggested), which I'd really rather not do.

[QUOTE=pjc;18583]
The examples you gave of subdividing like this all had to do with situational or geographical contexts, which is indeed how I use contexts. The only way I could subdivide further is to mix the notion of context with priority (as strongmace suggested), which I'd really rather not do.[/QUOTE]

My examples were not the best and that's because I haven't used priority in years, so I can't really remember the last time I dumped a priority in favor of a context. If priority works for you right now, then use it.

I'm really only saying that when I finally ditched priority completely, it made a huge impact on the overall effectiveness of my entire GTD system. And I had similar problems letting go of priority because I couldn't figure out how to change my contexts to eliminate its necessity -- I hadn't really analyzed why I was assigning certain priorities and found the underlying semantic connections that allowed me to tune my contexts. I personally think that anyone's task list can be contextualized in such a way as to eliminate the need for priority altogether, but I also know it took me a long time (years) to accomplish that and there's no way my contexts will work for you.

[quote]As I look through a long list of avaiable task[/quote]

There is your GTD problem, the lists are too long.

DA doesn't talk about it (or I missed it) but a good trick to GTD is keeping your lists short enough to scan through them pretty quickly. Since the human mind can only handle 8 items, I like to keep my lists in any one context (where you should be scanning) down under 20. If they get bigger than 20, create more specifc contexts.

With a list under 20 you should be able to quickly scan that context and pull priority by looking at the list.

Again, the point is prioritizing things in the future is pretty much a waste of time as it changes too much to put a system to handle it and your brain is much more adept as figuring out priority based on time/energy and current situation.

Just my $5

BZ

[QUOTE=BwanaZulia;18595]There is your GTD problem, the lists are too long...If they get bigger than 20, create more specifc contexts.[/QUOTE]

I agree that having short lists is an excellent aspect of GTD, however it doesn't work in all circumstances. In an environment where tasks are coming in rapid-fire, I would have to spend a lot of effort coming up with artificial context designations to get the list this short.

Fortunately, I don't feel bad about accomplishing only a few tasks on an extremely large list -- not everything is possible to get done. But I still need to record the full list, at full speed.

[QUOTE]Again, the point is prioritizing things in the future is pretty much a waste of time...[/QUOTE]

Agreed. I just need some way to pass through the list on a daily basis to figure out which few tasks I'm going to try to complete. If I finish them all, I can move on to others.

It occurred to me this afternoon that I can actually use flagging for my needs. I can pass through the list in the morning and flag the tasks that look like they're the most pressing for the day (and unflag the ones that no longer seem important). Then I can sort by flag status until I'm done with flagged items.

So now I'm back in the priority-neutral camp. :)

[QUOTE=pjc;18600]I agree that having short lists is an excellent aspect of GTD, however it doesn't work in all circumstances. In an environment where tasks are coming in rapid-fire, I would have to spend a lot of effort coming up with artificial context designations to get the list this short.

Agreed. I just need some way to pass through the list on a daily basis to figure out which few tasks I'm going to try to complete. If I finish them all, I can move on to others.

It occurred to me this afternoon that I can actually use flagging for my needs. I can pass through the list in the morning and flag the tasks that look like they're the most pressing for the day (and unflag the ones that no longer seem important). Then I can sort by flag status until I'm done with flagged items.

So now I'm back in the priority-neutral camp. :)[/QUOTE]

Effectively what you're doing with flagging is a weekly review -- a daily review if you will -- by flagging a next action you make it real, otherwise it is a "someday maybe" task that you don't want to see until your next (daily) review.

I don't have access to the beta yet, but I'm interested to see what its support for weekly review, "Someday Maybe", and "Waiting For" is. Hopefully doing the review and processing those those lists is as fluid and easy as people describe entering and viewing lists of actionable items is.

[QUOTE]With a list under 20 you should be able to quickly scan that context and pull priority by looking at the list.[/QUOTE]

I agree, but I'd still like to have some kind of sorting in context view, whether by priorities, or by drag and drop. Actually, I'd prefer drag and drop, ideally a "manual sort" in the view bar that remembers the relative positions from session to session. (Or maybe there'd need to be some kind of reset, or option to have the manual sort start from the current view. I guess that could be accomplished just by allowing manual sorts to be started from any view, and have it switch automatically if you drag a change. Then that arrangement gets memorized till the next time you start a new one. Tricky.)

In general, in the morning, I like to look through my contexts and my hard landscape, and plan my day. Obviously, things can change, but that just means changes should be easy to make. OF isn't smart enough to put them in the right order for me, especially since things change, and dragging around actions and projects, then going back to context view is unweildy. Plus, I prefer to keep my projects ordered by life area.

I've used Curt's OF > OO script to enable this kind of planning, but then I have my list in two places, which is less than ideal.

[QUOTE=Ken Case;17719]It turns out that this is exactly how we order actions now: each action has a "rank" which is assigned based on its position relative to other actions in the list.

It sounds like the proposed change, then, would be to allow actions to have individual ranks in context view rather than forcing them to only have child ranks within their own project. This would let you interleave actions from different projects in context view, effectively assigning individual priorities to each action (based on their position).[/QUOTE]

Ken,

I have hundreds of projects, with many hundreds of actions, but almost no deadlines or calendar appointments. My priorities are constantly shifting, based on the demands of my organization. Except when conducting a Review, I want to know absolutely nothing about most of these projects, and I do not ever want to be unnecessarily reminded of them. And I certainly don't want to waste my time assigning due dates to projects (since I don't have any ... ).

I like flagging in OF -- it allows me to display just the projects and actions that I want and to control when and if I look at my larger lists of projects. What I want though is for the flagging concept to be expanded in a way that the user can implement flexibly -- not simply a 'hi' / 'middle'/'low' flag concept.

In particular what I would like to see is a full implementation of HIERARCHICAL flags. Hierarchical flags would allow me to create the following nested grouping of flags, for example: Current / Month / Week / Today . (I'd really want different categories than these -- that's why I'd like these to be user defined rather than pre-defined. ... similar to the Hierarchical contexts you have already created -- and therefore quite in keeping with the program logic).

I should then be able to have the option to "Filter By Flag" and to choose which flag I filter for. Today Projects would be the only ones to show up if I filtered for 'Today' but if I filter for This Week, I would get both this Week projects and Today projects.

Thanks for your continued good work.

- Art

[QUOTE=otter;18660]In particular what I would like to see is a full implementation of HIERARCHICAL flags. Hierarchical flags would allow me to create the following nested grouping of flags, for example: Current / Month / Week / Today . (I'd really want different categories than these -- that's why I'd like these to be user defined rather than pre-defined. ... similar to the Hierarchical contexts you have already created -- and therefore quite in keeping with the program logic).[/QUOTE]

I wonder if a hierarchy is adding too much complexity. It sounds to me like what you're really looking for is arbitrarily named "tags", and the ability to filter on a selected set of tags. This frees you from having to create or manage a hierarchy.

I do see how a hierarchy might simplify the selection process, but I think it imposes too much structure that would limit the flexibility.

[QUOTE=rjfay;15371]If priority were a preference that I could use and/or hide, I would be all for it. One thing I didn't like about iGTD was that priority couldn't be turned off -- at least that I could tell (admittedly I didn't use it too extensively).[/QUOTE]

You can show/hide the Priority column with the little icon top-right in the main window. In iGTD I tend to ignore both Priority and Effort columns.

[IMG]http://www.olpin.net/blogpics/iGTDdisp.png[/IMG]

[I]*sits patiently watching his inbox for a sneaky peak invite so he can switch to OmniFocus and restore the cosmic balance*[/I]

[QUOTE=pjc;18752]I wonder if a hierarchy is adding too much complexity. It sounds to me like what you're really looking for is arbitrarily named "tags", and the ability to filter on a selected set of tags. This frees you from having to create or manage a hierarchy.

I do see how a hierarchy might simplify the selection process, but I think it imposes too much structure that would limit the flexibility.[/QUOTE]

Filtering on tags is fine -- but being limited to filtering on only one tag at a time would be a big loss.

So for example, if I had Tag1, Tag2, and Tag3, I will need the ability to easily have a filter that has the capacity to display everything that is tagged either "Tag1" or "Tag2"; "Tag1" or "Tag3," etc. Without at least a basic capacity to 'union' individual tags into compound tags, tags/flags will be much less useful.

In the example from my earlier post, if I tag some actions as 'Today' and some actions as 'This Week' it will be basically useless if I can't have a filter that displays the 'Today' items in the same list as the 'This Week' items. Of course I also want to be able to display the 'this Week' items without the today items as well. This capacity could be implemented either through hierarchical flags/tags, or with tags plus logic. Either would work.

-Art

I urge Omni to incorporate Prioritisation into OmniFocus. Moving items up and down in a list is OK if you only have a few items, but this method is unsuitable if you have hundreds of items. It concerns me that Omni is limiting the functionality of an application because of the narrow adherence to one man's philosophy. I take on board lots of David Allen's ideas but I disagree with his attitude to prioritisation. I don't doubt that there is an evangelist for a particular outlining methodology but I doubt that Omni limits the development of OmniOutliner to be a slave to his/her methods. If you want to keep the purists happy then implement prioritisation and other requested features as options that can be switched on and off via preferences. I'd be interested to know what others think about this.

And I urge OmniGroup to stay true to their design rather than diluting it.

Watchit, I notice that you didn't even mention the flag functionality that is already in place. Have you used it? If so, why is it not working for you?

I would argue that the numerous websites, blogs and dedicated followers define a [I]wide[/I] acceptance and acknowledgment of one man's philosophies. I would posit that adding generic prioritization features is narrow adherence to priority fanatics' desires. To dismiss the design of OmniFocus (which draws fervently from David Allan's system/philosophies) is to go against the core audience that will drive the success of OmniFocus.

Implementing optional features should be low to nonexistent on the initial milestone list. Lets get a stable 1.0 out that is rock solid and true to its core design, and then consider optional features for a 1.5 update, if they are justified.

[QUOTE=watchit;19880]It concerns me that Omni is limiting the functionality of an application because of the narrow adherence to one man's philosophy. I take on board lots of David Allen's ideas but I disagree with his attitude to prioritisation.[/QUOTE]

Although Omni obviously want to make it saleable to the widest audience they can, I think from what's been said by them so far OF is at heart a GTD app. Hence I agree with Lecter that features that are the direct opposite of what is suggested in GTD (i.e. pre-assigned task priorities) shouldn't be in there, at least at 1.0 (and if they're there after that they should be able to be turned off!).

If an app that implements a pretty canonical version of GTD doesn't fit your workflow you might be better off with another app: as an analogy, although you could use Keynote if you worked in an organisation that required all presentations to be given in Powerpoint format you'd probably find life easier using Powerpoint.

Lecter… where is the dilution of design in adding OPTIONAL features that can be turned off or on by the user. The whole point of using a Mac rather than Windows is the flexibility and customisation that allows users to work the way they want. I do not see how adding OPTIONAL features will drive away dedicated David Allen devotees but it will enable Omni to cast a wider net to reach people who like some of the guys ideas but are not fanatics and would like to use a flexible application. As to the use of Flags, yes they are great, but they are either off or on. Prioritisation is a matter of degree, not just yes or no.

al_f… yes, there is a more flexible app which uses the basic GTD philosophy, iGTD. It is quite good, but it is being developed by one bloke in a garret. Who wants to enter a huge amount of data in app that might suddenly disappear? Omni is an established developer making some of the best apps on the Mac. They seem to be very responsive to incorporating user requested features into their other apps without being slaves to a particular ideology.

But this debate is a bit like prohibition. If you don't like being drunk, then don't drink, but don't try to stop the those of us who do! :)

I won't stop you, but I don't see why you expect non-GTD features to show up in GTD-specific apps when there are dozens if not hundreds of other task management apps out there that don't share OF's GTD-centric strategy.

If it's added as an option, I won't object, but everything else should be done first.

Watchit, by definition, an optional feature is a dilution. If it was a part of the core design, it would not be optional (it would be integral).

Adding optional features drains development resources that could be applied to core features. As I mentioned in my previous post, I would prefer that optional features be added in [B]future[/B] versions, [I]after[/I] version 1.0 is rock solid stable.

I do realize that prioritization is a religious debate. From my perspective, something either is a priority or it isn't (which fits the flag system perfectly). In my experience, granularity of priorities is busy work that prevents the actual work from progressing.

I don't want to get off topic here, but I have to point out that flexibility and customization are (IMHO) some of the lowest points for using a Mac. Stability, interface and reliability are much stronger arguments for me.

And as far as prohibition is concerned, no one is stopping you from drinking your prioritization koolaid. We just prefer our brand of D. Allan's Ojai Blended GTD Whiskey. :)

Earth to Omni Forum...
 

I have to throw my 2 cents in here.

I've just waded through 10 pages of people debating whether to include priorities in OmniFocus. To me, this only proves two things: (1) a lot of people use priorities and (2) a lot of other people do not. Trying to force one group of users to conform to another group's workflow is only going to end up alienating a large portion of your market share (who may end up turning to other solutions).

I would suggest that priorities be included for those who use them. But there should also be a Preferences option to turn these off, for those who don't.

There is nothing wrong with adding features that some users find superfluous. In fact, this fits really well with the whole "shrink to fit" concept that OmniFocus is based on.

The other argument for including priorities is that both iCal, OmniPlan, and the Palm OS already have this technology installed. If you include priorities it means you can access a 3rd dimention (if you so choose) for organizing your lists! If you don't include them, all you've done is hamstring the other components of your system.

I would also say the question of whether priorities fit with canonical GTD is largely irrelevant. David Allen may have argued against using priorities, but he also says that we shouldn't be outlining our task lists (something that, apparently, we find really useful anyway).

[B]Implementation:[/B]

You need to pick a standard and stick with it. iCal defines priorities as: "Very Important", "Important", "Not Important". Similarly, BlackBerry uses "high", "normal", "low" (as does Mail). Yet, MailTags uses "Urgent", "High", "Normal", "Low", "Very Low". And both OmniPlan and Palm OS defines these as: 1 through 5.

All this means is that every time you sync between non-conforming applications, something is going to get corrupted. And if you now throw floating-point numbers into the mix (Curt's slider suggestion), syncing isn't going to have a chance!

Interesting that you mentioned that about " the David " but I've heard the same thing from others on his staff in various arcticles etc .. " nested actions " etc is a no no . They state that what is hidden may come back to bite you .

but I find the outline format very useful. Sooooo I guess omni focus isn't really canonical GTD after all ; )

[QUOTE=Adam Sneller;22873]I would suggest that priorities be included for those who use them. But there should also be a Preferences option to turn these off, for those who don't.[/QUOTE]

I'm opposed to priorities--but what if they could be renamed? Then those who want some way to tag tasks with the amount of energy or concentration required could use them for that, and everyone would be happy. I'd be delighted if I could have OmniFocus show me tasks of X minutes or less that I can do when I'm brain dead, for days when I've had a bout of insomnia.

[QUOTE=brianogilvie;22995]I'm opposed to priorities--but what if they could be renamed? Then those who want some way to tag tasks with the amount of energy or concentration required could use them for that, and everyone would be happy. I'd be delighted if I could have OmniFocus show me tasks of X minutes or less that I can do when I'm brain dead, for days when I've had a bout of insomnia.[/QUOTE]

Brian,

Why don't you use the estimated time field?

Optional "priorities" feature
 

I think the way you make everybody happy is to include Priorities, but make the feature optional. Exactly the way that the "start date" and "due date" columns are optional in the current release. If you don't want them, don't turn them "ON".

The fact is that people are diverse. And one person's "perfect solution" is going to be absolutely useless to person sitting next to them. So you have to build a solution with flexibility - enough to account for different user "Preferences" (hence the term).

I honestly don't see how anybody [I]can[/I] object to this - especially when it has no effect what-so-ever on the people who do not wish to use it!?

[QUOTE=dhm2006;23000]Brian,

Why don't you use the estimated time field?[/QUOTE]

Because I use it for estimated time. :-)

That is, some tasks take a while but don't need much brain power. Others might be quick but I want to be at the top of my game for them. I can figure that out on the fly but I wouldn't object to having a way to note estimated energy required. I think estimated time is more important to my workflow, though.

Estimating time
 

Time estimation sounds useful. Of course, if you take this one step further, you will be re-inventing the Gantt chart... and then you might as well be using OmniPlan.

Which is actually interesting, because I remember reading that Omni is planning to integrate the two apps in the future...

[QUOTE=Adam Sneller;23012]Time estimation sounds useful. Of course, if you take this one step further, you will be re-inventing the Gantt chart... and then you might as well be using OmniPlan.[/QUOTE]

Yes, but one doesn't have to go that far. One important question to answer in deciding what to do next is "How much time do I have available?" If I have a meeting in 20 minutes, OmniFocus lets me filter tasks that I think will take 15 minutes or less. That way I can get a quick email or phone call out of the way and not have to choose between it and something that I've already guessed will take me half an hour.

More abstractly, time estimated for an individual task can be quite useful even if you're not calculating milestones and dependencies. David Allen writes in [I]Getting Things Done[/I] that available time is one of the four factors to consider in deciding what to do next (the others being the context you're in, the task's priority at the moment, and your energy level).