The to-do list problem | Hacker News Reader

The to-do list problem

funkykaraoke.com

55 pointstomazstolfa48 comments

Slightly off-topic, but never auto-play video (or audio) content on a website. If I want it to be played, I click on it, otherwise it's just annoying.

I'm glad someone beat me to this. This is the first thing I thought when I went to the post. It soured my opinion of the whole article before I even read it because of that small usability mistake.

"It soured my opinion of the whole article before I even read it"

Seriously? The fact that an embedded video started playing soured your opinion of the article? Maybe you guys should chill out a bit. Sometimes comments like this lead me to believe the author wants recognition at somebody else's expense.

I'm not saying I didn't read the article because of it but it does just start the article off with a bit of annoyance. You can't argue that anywhere you enjoy going in the online or physical world is determined by emotional responses.

Imagine when you got to a restaurant that has music annoying loud. You may still enjoy the dinner food, but you had a bad experience eating the good food because you were annoyed with the music. That scenario similar to this article. You could still enjoy the article content but you were annoyed because of the bad experience. And after so many of those bad experiences you may just think "you know what, I'd rather eat somewhere not as good just because I won't be annoyed by loud music."

If you don't think that small things like this (that can easily add up over time) do change how your viewers see your site/company (regardless of the content), then you need to think again.

However, I could be wrong. I mean, everyone did love the animated images, colors, and sounds on GeoCities which lead to their industry takeover and incredible success. ;)

"However, I could be wrong. I mean, everyone did love the animated images, colors, and sounds on GeoCities which lead to their industry takeover and incredible success."

Let's not just assume that all annoyances are created equal. One is an integrally bad user experience, the only way out of which is to stop using the service wholesale. The other amounts to mousing-over a large area of the screen and clicking a button.

Fair enough. I will give you that. It was a bit of a extreme on that example. But I imagine you do understand what I generally mean about the small things having the ability to add up.

> Seriously? The fact that an embedded video started playing soured your opinion of the article?

Sure, what's wrong with that opinion. I don't know how you browse but when I browse hacker news I'll open up 2-3 articles and their associated comments and start reading one.

When a video starts playing in the background when I'm not even on that tab, then yes, it can sour your opinion of the entire website.

Little things matter.

That's your interpretation based on your use. I do not browse that way. I pick one maybe two that look interesting and read them. I find the autoplay useful.

See that? Two different interpretations of the same feature. Just because you hate autoplay doesn't mean everyone does. Certainly one or the other does not "sour my opinion" of an article. The fact that the video starts playing when the page loads has exactly nothing to do with the content in the article. That's what's wrong with the opinion.

Browsing habits aside, I find something else annoying about it. When I click on the article, and the video sits in the middle of the page with text above it, I want to read the text until I reach the video, then watch the video, then continue with the text below it. As soon as the video starts, reading is disturbed. To me it completely kills the "flow" of the article.

And just my two cents: There is nothing wrong with any opinion. It's an opinion after all isn't it? An opinion is not a universal truth. It's something personal that people like to share for some reason. That what makes discussion so hard at times.

> See that? Two different interpretations of the same feature. Just because you hate autoplay doesn't mean everyone does.

Sure, all that makes perfect sense. None of that invalidates anything that I said though:)

This is one of the two benefits of flashblock for me. If I want to watch a video, I click on the circled F, otherwise it's just a box that takes up space.

The other benefit is that I don't give my cycles and peace to flash advertisements.

The cycles part is actually non-negligible. Those things are usually so horribly written that they'll eat up processing time in big chunks. But that's not a Flash-only problem. I've seen JS-things on webpages actually spinning up my laptops fans even though there was really nothing much on the screen, and it's not like I've got a bad laptop either. It's not high-end but an i7 with 8gb of RAM shouldn't notice the load of a pretty average webpage.

But that's just my 2 cents.

My dad didn't understand why his (admittedly quite old) laptop really struggles on some websites. For most of his usage - buying tickets, checking emails - it suffices just fine, but a badly coded flash advert can bring the entire machine to a crawl.

It was very difficult to convince him that yes, it was actually the crappy slideshow advert on one tab that was making his entire computer unusable.

Agree. Since I tend to skimread the frontpage and open all links I think is interesting in new tabs, autoplay leads to a closed tab without even looking.

I agree with you. The embed does not have a non-autoplay option.

that's a bit of a sweeping statement. I would agree with you if the autoplayed video has audio or is instructional in some way. In the article I agree it should not auto-play

However, auto-play can be nice in the right setting. Lots of car manufacturers auto-play video as part of the design. Or some product shots like the MackBook page where the laptop opens and spins is nice.

I completely disagree. I thought that was pretty cool actually. I rarely view videos embedded in articles unless they autoplay.

I hate to-do lists. The worst are the ones where others can pile up things for you.

The problem with most to-do lists is that they treat your time as infinite. You just have to take one item after the other and chunk trough it. You're just the CPU executing the tasks you've been handled trough like a puppet. Soon enough the list becomes a big graveyard of things you didn't have the time to do and it's depressing.

It's better if the system forces you to filter tasks upfront. Like a tweet, there is only so much you can fit in a day. Get over it, you won't have the time to do X, just forget about it. It's important to have quality over quantity.

Isn't that merely a case for periodic pruning and reordering of the list(s)?

I'm not very dedicated in my usage of to-do lists (probably to my detriment), but I think people who use them consistently revise their list at lunch and again at the end of the day... perhaps at breakfast too. Unless they have an assistant who does all that for them.

If there were an easy way to determine the time required (or time box allotted) for every task, then this seems feasible. Asking the user without being cumbersome may be difficult though.

Did any of the previous commenters actually watch the video? Ignore the title, it's not about to-do lists. If you ever want to be successful at a startup, watch the video. Understand what he's saying about those three things that are required to get action from someone. Then look for his other work on Captology explaining how computers can manipulate people (in both the good and bad sense).

Seriously, WATCH THE VIDEO.

YT link for easier control and downloading: http://www.youtube.com/watch?v=ZYE3Jal7Yaw

Thanks! I've replaced the video with this one to avoid the auto-play issue.

I agree that one needs motivation and ability to complete a task. I'm not sure any app or electronic service is going to supply that. If it did, you'd find a lot of people out of work.

As for this quote:

"How about having a reminder showing up when you get an email from a relevant person that can help you accomplish your task, or a reminder showing up when a relevant person is calling [through vox.io of course]."

I think you're falling into the trap that most tool-builders fall into- over-thinking the problem. Most people are pretty simple. That's why we still keep pads of paper and post-it notes next to our $1500 workstations. For a lot of tasks speed and simplicity provide far greater benefit than any expertly-engineered web service ever could. In any case, why would I want a reminder to complete a task if I am getting emailed by somebody related to that task? In that regard I hate to say it, but email will win out over an integrated task list. Why? Because the email itself is serving as your reminder. A secondary service nudging you to complete the task after the email seems superfluous.

> To build an intelligent to-do list it will take more than polished pixels, so please start putting more intelligence in your to-do tools. Your users will be thankful and will hopefully have much shorter to-do lists.

After trying many things, I have now switched to Trello. Using it I've become convinced that "intelligent to-do lists" are a fallacy, what we need are dumb todo lists. Something not much smarter than a whiteboard or a piece of paper.

For reasons I don't understand, those just work.

Totally agree. I have used index cards as a planning tool and a to do list for a long time. Trello, in all its nothingness, is the only thing that has come close to the efficiency of a pencil and stack of cards.

Agreed. I have used the same method since about 1985 and it works wonderfully. The only optimisation has been change to a mechanical pencil.

I've lost my card stack a couple of times in those years and there has been no negative effect.

I've played with and written various bits of software but it just doesn't make it any better either.

This problem is solved already.

I think the problem is workflows for todo lists are incredibly diverse and trying to incorporate support for them with a single UI without cluttering it to the point where it adds more friction than it removes is an incredibly challenge.

E.g. grouping things on paper: write down a few headlines and write your stuff underneath each one of them. Want a fresh todo list daily? Write down a few lines every morning, transferring from past list as needed. Want a historical record? Well, you just got that for free anyway. And so on.

A "dumb" todo list interface solves our workflow problems by abdicating the responsibility and making us "implement" the workflow manually and just provide support for implementing workflows. And so the need to complicate the UI in ways that will almost inevitably end up adding friction goes away.

Creating a mass market application with a sufficiently "fluid" workflow support to allow this in a "smart" todo list app without adding endless configuration screens is something that'd likely be incredibly hard.

For my part, I have my own heavily customized app that works for me. But it works for me because I can tweak the code itself to my workflow whenever I decide to change any of it, and can as a result take lots of shortcuts I wouldn't have been able to with a product intended for anyone else.

"what we need are dumb todo lists" so right. However I found Workflowy even dumber than Trello and having used them side-by-side Workflowy also did it for me usability wise. In Trello there were just too much sub categories and sub windows and too much navigating around for my liking.(note: I use it as a todo list/note taker only. I don't share anything or try to make it act like a bugtracker)

Trello is too big for "pick up bananas at the grocery store".

I just used it for that last week. A little clunky but manageable with the Android mobile app.

textphile.net may be up your alley

This is the problem that GTD tries to solve. Omnifocus is a great piece of software that's built on this methodology

http://www.amazon.com/Getting-Things-Done-Stress-Free-Produc...

http://www.omnigroup.com/products/omnifocus/

I read GTD and practiced it for a few years. I always felt on top of my tasks, but gradually seemed to fork off my own system, although, I definitely still track tasks in a list* wrt context, due date, and time estimation.

One cautionary note with GTD is that it's easy to get caught up feeling like you're cranking widgets and lose sight of the forest in favor of the trees. I welcome suggestions for this.

With respect to BJ's talk / the Fogg model:

- Triggers: The GTD system helps you capture all your task triggers in one place. This is a great idea that works well in practice.

- Ability: The most common behaviors for interacting with your GTD system are adding new tasks, deferring tasks, and reviewing. The concept of a universal inbox accomplishes task insertion very easily, but one must build the habit of processing the inbox into their own GTD system __daily__ to really trust it. That was difficult for me in practice.

- Motivation: The weekly review system provides a renewal and closure based on what you've done recently and have upcoming. The higher-level reviews "20,000 to 50,000 feet" are a harder habit to form since practiced so infrequently. I never really became great at this.

* I use one plain text file for tasks. I group with headers based on the date I plan to work on each task. Some items have hard due dates, but most don't.

  # TODAY
  x completed task
  - uncompleted task 0.5h /* estimate of how long it'll take */
  - another task (due 9/8)
  - ...

  # TOMORROW
  - ...

  # NEXT WEEK
  - ...

  # NEXT
  - ...

  # SOMEDAY / MAYBE
  - ...

And 3 separate lists for errands, groceries, and calls:

  # @GROCERY
  - ramen
  - gatorade
  - etc.

I built a very simple todo list + week planner based on a similar concept, that I use every single day now. Check it out sometime - http://weekis.com

Nifty. I'd love to see your iOS concept. The pain of editing my text file on the iPhone causes me to glance, but rarely update on mobile.

Listary ($0.99) has a nice synced list implementation that works well if you don't mind dividing the lists into separate files, but it doesn't support due dates. I didn't consider separate files a realistic option for my purpose.

The problem is that there are multiple workflows people use and dont want to change. I for instance rely on invox zero and calendar, so they function as todo list. Many people i know use other methods.

On the other hand, some people argue that if tou cant keep your todos in your head, you have a bigger problem ;)

To be effective, your to-do list should also include when and where beside the task.

The post assumes humans are bound to sense habits and gives little credit to the power to making decisions - where one can change their mind and act upon that decision.

All you need is a mental transition between the previous activity and the new activity. How? Well, the problem has existed as long as humans, so just figure out how parents convince kindergarteners to go to school, why some cultures pray before eating food (try to think beyond thankful-for-food reasons), and how famous leaders can led entire movements. The common denominator to all these will be your answer.

Primitive version of GTD. GTD solves the issues raised in this post and more. Many todo apps work with a GTD workflow. I use Toodledo (web app) and Due Today (Android), they sync perfectly. Re-reading the GTD book every 6 months or so helps to be consistent with applying it, because the devil is in thinking 'oh I'm just going to skip this or that task for now', or not being diligent in e.g. assigning contexts. For GTD (any system, really) to work, you need to hammer it into your habits until it's stuck so snugly that it feels weird to do things any other way.

I'm really glad that this works for you, but it sounds like hell.

Not half as bad as not having a system, and having to spend significant mental resources on keeping track of what to do, what to prioritize, when to do them, etc. Not having to think about that meta-information frees up resources to actually do things.

It is. But once you slog through hell long enough, it's not so bad anymore.

Just last night I was thinking that Reminders needed a "remind me at sunset" option to remind me to round up the animals. As it is I have to change the time of that reminder every month.

There are probably a ton of triggers that could be useful. Off the top of my head: time, location, weather, the people you are currently with (4sq checkins), current call, recent texts, or emails. I'm sure a bunch more.

I'm not sure that I'd call this a todo list, but it does sound like a GTD support system.

This article discusses what can be done with just time and location, if you just treat them with a bit more smarts: http://news.ycombinator.com/item?id=4489121

Also, "right before I leave the house". GPS has no idea if I am going to work until after I left.

Very closely related to Fogg's behavior model of "trigger", "execution", "motivation", Charles Duhigg in "The Power of Habit" writes of the steps of the habit loop as "cue", "routine", "reward".

Habit is coming up to be a big topic in app design and development.

"The Power of Habit" is a good read which systematically lays out some science behind what habits are, how they work, and how to leverage them.

It has never worked for me. I ended up creating a not-motivated-to-do list.

  import willpower
  # ...

Reply on news.ycombinator.com