The Omni Group
These forums are now read-only. Please visit our new forums to participate in discussion. A new account will be required to post in the new forums. For more info on the switch, see this post. Thank you!

Go Back   The Omni Group Forums > OmniFocus > Applying OmniFocus
FAQ Members List Calendar Search Today's Posts Mark Forums Read

 
Why OmniFocus v1 didn't support multiple contexts per action Thread Tools Search this Thread Display Modes
Regular visitors to these forums know that various flavors of "when will OmniFocus support multiple contexts (or tagging)?" is a fairly common question here. Over the years, we've answered that question in piecemeal fashion, so it's understandable that new folks wouldn't know where we stand on this. The information has changed over time and is posted in several different threads. It's confusing.

So, we're rebooting the conversation. I'll start this thread off with an attempt to clearly and comprehensively present our thinking on this subject. Once that's posted, I'll lock the previous threads - they'll still be available for reference purposes. Future threads will be merged into this one, so the conversation stays in one place and is easier to track.
 
In our experience, when an action looks like it requires multiple contexts, it almost always means that the action hasn't really been broken down into concrete, discrete steps.

We're not saying that you should go too far down the rabbit hole of breaking tasks down into sub-tasks (“Well, in order to send that email I’ll have to fill in the ‘To’ field, then the ‘Subject’ field, then compose each paragraph…”), but in our experience, several small actions are quicker to accomplish than one big, complicated compound action.

This approach also lets you independently accomplish each part of the task and mark it as complete, rather than needing to leave the compound action sitting open on your lists until you've done all the things it actually represents.

Another downside to tagging is that it tends to alter the meaning of your context lists. Rather than a short list of actions that you know you can do, you end up with longer lists that are full of actions you may be able to do.

Those longer lists require you to spend time searching through your context list, deciding whether you can actually do each item. Time you spend doing that is time you aren't spending getting things done.

We understand that a lot of folks are comfortable with a "fuzzier" approach to their contexts/workflow, but when there's a conflict between what's comfortable and what's productive, we really want our software to err on the side of productivity.

Obviously, what's "productive" is a subjective judgement, and we're inevitably losing out on sales when we do this - there are other products that take a different approach to this issue. Having lots of options is actually a good thing, though! It's more important that folks be able to use a tool that works the way that they do than it is that the tool come from us.

Myself and the other support folks will be happy to help someone resolve any problems they're having applying a single-context approach to their workflow. Let us know which situations you're struggling with and we'll be happy to help. And if the lack of a multiple-contexts option means that OmniFocus isn't the right tool for you, we'll figure out another solution. One of our core values is that we shouldn't profit from selling tools people can't use or don't need.

Last edited by Brian; 2013-02-08 at 02:00 PM.. Reason: clarify the final 2 paragraphs
 
Okay, so, if we're not going to support multiple contexts, what should people used to those kinds of systems do while working with OmniFocus? The following examples are far from comprehensive, but they cover some of the most common use cases we've seen presented.

While I'm at the office, I need to ask my boss a question. It's a high priority task. I need three tags to capture that.

Create an "Office" context, and put the context for your boss inside that one. Tasks that only your boss can help with go in their context. Tasks that anyone in the office can help with go in that one. (Plus, you can assign a location to the Office context and use the Map view in the iOS apps!)

The highest priority tasks get flagged; everything that's high-but-not-highest priority gets moved to the top of the project, so they're first to get knocked off once the must-do actions are done. Everything else gets reviewed and re-prioritized during your next review.

I need to buy something. I might get it online, or order it over the phone, or buy it at a store in my neighborhood.

This is a case where breaking the action up has real benefits. You'll often save money by breaking this up into at least four steps. You'll generally want three "research <option>" steps, and one "purchase the best option" step. Skipping any of those steps risks paying more than you should, or not buying the best item.

Of course, sometimes there really are multiple options where you could get the same item with little to no difference in price. (Say, an AppleTV or other unmodified product from them.) In that case, just pick whichever applicable context you look at most frequently. In the majority of cases, that'll get the action done soonest, and you'll save time overall.

I need to ask Anna or Bob a question. Either of them could answer it. I might call them, or email them. I need four tags.

Make contexts for Anna and Bob. Assign the action to whomever you're in touch with most often, and decide when you capture the action whether you'll call or email them. Put that info in the action title, so future-you doesn't have to think about it.

I need to show Cindy a new machine that's in a particular location. If I'm not in that place, or if Cindy's not there, I can't do it.

Assign the action to the context for the location. Put Cindy's name in the title. Check the context for that location when you're there; if Cindy's there, do the action. In the meantime, if you're meeting with Cindy at some other location, you're not distracted by an action you can't do.

Last edited by Brian; 2013-02-08 at 02:36 PM.. Reason: add third and fourth items.
 
While some things are multiple contexts some can just be handled with tags. There are many things that can be completed without other's involvement, say expense report, but if I have a tag for Susan the company accountant and I am sitting with her, I could filter and pull up related tasks (the two items I am not sure will be approved in expenses makes a this a much quicker task and done in one pass).

I also have contextual conundrums as I have a task for a face-to-face meeting with a friend, who lives in San Francisco. But, I also run into him at some conferences as well as when he floats through the town where I live. If I am planning a trip to SF the note to email my friend to set up a face-to-face and put it on the schedule is needed. For a while I had this in SF trips context, but not in conference related tasks where he was.

While OmniFocus has location context having another "person" context would help as well. Tying this to their entry in address book would be quite nice (the simple mapping names to address cards like VooDoo Pad does is really nice). When adding a person and seeing latest interactions with them (similar to the review with calendar view slated for OF2 would be quite helpful).

I really like the main multiple context filters outlined in the "what should be do..." in Brian's post above.
 
Quote:
Originally Posted by vanderwal View Post
There are many things that can be completed without other's involvement, say expense report, but if I have a tag for Susan the company accountant and I am sitting with her, I could filter and pull up related tasks (the two items I am not sure will be approved in expenses makes a this a much quicker task and done in one pass).
This sounds to me like one "complete expense report" action in whatever context is best for your workflow, and two "ask about <topic>" questions that go in a context created for Susan.

Potentially in a sequential project, so the "complete report" action doesn't activate until you have those questions answered...
 
I must say, the readings here and on the linked articles sometimes leave me smiling. One could be inclined to think of three approaches to those who (continue now to) ask for [certain ways of having ... see below] tags or multi-context views in OF.

* You are a heathen to the GTD mantra ... wise up!

... or ...

* Here's a work-around/kludge ... be happy!

... or ...

* What you really meant to do is this ... fix it!

Not that polite discourse is not common to this thread -- it is actually very much so. Just my observations, perhaps to ground the answers in a realization that ... certain types of tags or multi-context views being requested for OF are clearly not gonna happen (in your lifetime).

--
JJW
 
Quote:
Originally Posted by Brian View Post
This sounds to me like one "complete expense report" action in whatever context is best for your workflow, and two "ask about <topic>" questions that go in a context created for Susan.

Potentially in a sequential project, so the "complete report" action doesn't activate until you have those questions answered...
The action in the complete expense report has questions for Susan along with other actions. But, when I am sitting with Susan to review the finances and get a signature on an upcoming project I can also ask her about the expense report. But, I can also email her, or include the questions in the expense report, if I haven't heard back from her by the time I need to file the expense report. There may also be somebody else that could answer the same question, but Susan has the ultimate authority. I need to get the questions answered before submitting the report, or highlight in the submission so they could be approved or declined as line items and not have the whole report rejected and have to wait for the next submission window.

Similar this is in my client project projects with actions. I have a need to get a question answered or an approval by somebody in a role, with authority to approve, but one person whom I occasionally have meetings with can handle this as well as the other three actions that are tied to our meetings.

The "I really needed to also have conversation "X" with this person" problem is one I continually have. When I explain it to others I find it is a big pain point for others as well.
 
Quote:
Originally Posted by vanderwal View Post
I also have contextual conundrums as I have a task for a face-to-face meeting with a friend, who lives in San Francisco. But, I also run into him at some conferences as well as when he floats through the town where I live. If I am planning a trip to SF the note to email my friend to set up a face-to-face and put it on the schedule is needed. For a while I had this in SF trips context, but not in conference related tasks where he was.
(Separating the examples out to try and keep the conversation/reply threads distinct, btw.)

This sounds like an "email <name of friend> to schedule trip" in some sort of Online or Internet context, and I'd put that action in a "Plan SF trip" project. Can you explain how you use the "SF Trips" context? That may help me understand your workflow and make better suggestions.
 
Quote:
Originally Posted by Brian View Post
(Separating the examples out to try and keep the conversation/reply threads distinct, btw.)

This sounds like an "email <name of friend> to schedule trip" in some sort of Online or Internet context, and I'd put that action in a "Plan SF trip" project. Can you explain how you use the "SF Trips" context? That may help me understand your workflow and make better suggestions.
[thanks for breaking out distinct points in my post]

The F2F with my friend is a action that is initially tied with San Francisco (that is where he lives and where I was traveling regularly), but the main hook for it is my friend and our face to face. Years back in OmniFocus I used tools or rough location for context, which has broken a bit, but face-to-face meeting could be a Context. Meeting, particularly called out as face-to-face, has a secondary dependency that is a person or people. This issue arose this year when I missed meeting with my friend when were were both in NYC as well as he had a trip near DC where I am also near.

I have a script that roughly works to track Foursquare, Lanyrd, and Google checkins for upcoming location nearness (worked much better when it was Dopplr and Upcoming focussed) and having context of a person and F2F I can put a filter that is tied into that script (I've thought about tying a positive hit on that script to an OmniFocus flag).

But, the key is there are more than one potential context. Location is a flexible variable, but the person is a key variable. Nearness in location (which may not be fixed to a place, but proximity) is another key.
 
Quote:
Originally Posted by vanderwal View Post
While OmniFocus has location context having another "person" context would help as well. Tying this to their entry in address book would be quite nice (the simple mapping names to address cards like VooDoo Pad does is really nice).
Getting OmniFocus to make better use of Address Book data is something we're considering for the future - I'll add this post to that feature request. Thanks!
 
 


Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes


Similar Threads
Thread Thread Starter Forum Replies Last Post
OmniFocus 2 – Multiple Contexts or Tags…? [A: Not in the foreseeable future.] effective OmniFocus 1 for Mac 17 2013-02-08 12:56 PM
One action, multiple contexts dp1 Applying OmniFocus 23 2010-07-03 05:09 AM
multiple contexts and multiple projects mind full of water OmniFocus 1 for Mac 7 2008-06-23 09:31 AM
Multiple Activites for Multiple Contexts Journey OmniFocus 1 for Mac 12 2007-12-27 01:03 AM
Multiple contexts per action? pasentcom OmniFocus 1 for Mac 15 2007-11-28 06:37 AM


All times are GMT -8. The time now is 08:08 PM.


Powered by vBulletin® Version 3.8.7
Copyright ©2000 - 2019, vBulletin Solutions, Inc.