Hacker News new | comments | show | ask | jobs | submit login
Checklistomania makes it easy to keep track of relative tasks (gsa.gov)
127 points by hugs on Apr 22, 2016 | hide | past | web | favorite | 51 comments

It isn't clear from the description, but does this allow you to have tasks become 'triggered' for lack of a better term once a task has been completed? eg I want my task list in the morning to say make eggs and bacon, then once I've done that a new task pops up for me to do the dishes.

Yup. I worked with Tony on the tool. That's exactly what we had in mind. We didn't want people going through onboarding to be overwhelmed — and we also didn't want to have to manually date each task relative to their start date. i.e. If you do this on day on (April 22), then you must do this task three days later (April 25.) This tool simply lets you enter April 22 and then figures out what the next due date should be, based on the time length you specify.

Lead dev here - Yes, it supports that. It supports any task which can be expressed in the syntax "complete this [X] days after this [list of items] is done, where X can be positive, negative, or zero, and [list of items] could include just the special first task "dayZero" which sets the base date for the checklist or any other combination of tasks.

Cool, I'm going to check it out later today.

MyLifeOrganized has that (look at 'Dependencies' - http://www.mylifeorganized.net/#comparison-table) but I think it's probably too robust (and expensive) for the average person.

I use MLO to keep track of my professional work. I agree it can be overkill, but it's perfect for managing tasks that depend on other tasks.

Some additional features I find useful are:

- the ability to add multiple tasks (line separated)

- you can automatically generate a list of tasks ordered by urgency and priority (can be customized)

- you can easily nest tasks into folders

- you can easily edit many tasks at once (priority, dates etc.)

- folders flagged as "projects" display a progress bar overlay

- you don't need to go through any menus to add/view task-related notes; notes appear in the sidebar depending on what's selected

- great for multi-tasking because it dynamically combines all your task lists into one

I was on board w/ 29.95$ for an android app, I balked at $60 for the windows version.

In case you want to quickly try out your own version. I created a fork that uses Auth0 for authentication so you can use any identity provider plus it has a deploy to heroku button. You can try it out here:


"When a new hire joins the federal government, existing employees complete over 60 tasks to bring them on board."

"This is particularly useful in government, because we need to complete quite a few training sessions and pieces of paperwork."

I can see the value in building a tool to track all of the moving pieces in what appears to be a bureaucratic, inefficient system ... but I would love to see more work put into resolving those inefficiencies.

Can any of those tasks be automated?

Is there any way to reduce the amount of paperwork required?

Can the training sessions be replaced with online tutorials?


We're automating what we can - i.e. accounts and stuff.

But a lot of the tasks are training related i.e. "You have to learn about this or that."

You can see the work we've put into onboarding here: https://18f.gsa.gov/2015/12/01/how-we-dramatically-improved-...

The biggest efficiency, I think, is creating a Slackbot that trickles out a lot of this information to new hires: https://18f.gsa.gov/2015/12/15/how-bot-named-dolores-landing...

We've also rolled out a handbook and classes — the checklist is basically to make sure people have seen what they need to and so they're not overwhelmed the first day. (Something like learning how to travel is something you need to learn, but maybe not until week two...)

You're doing government contracting but allowed to use Slack...?

We're federal government employees (and are allowed to use Slack).

You don't have data confidentiality issues? Perhaps working for the Federal government directly is quite different than the DoD.

You can see our negotiated ToS here: https://slack.com/terms-of-service/government

> Can the training sessions be replaced with online tutorials?

In many cases they have. And generally speaking, I doubt they communicate with any central checklist system, likely because one didn't exist. Until now. And pretty much every department has its own checklist. Hell, I worked for a university and I'm pretty sure the automation processes I wrote for new account creation didn't absolve departmental IT from additional work for hires specific to their department.

The other fun thing is the process that constructed the current system. Something bad happens, and another annual training is added to the list. Over time, a lot of bad stuff happens. This leads to scenarios where military officers literally have more training requirements than hours, and cut all kinds of corners and lie about it: http://www.armytimes.com/story/military/careers/army/2015/02...

The problem isn't moving paperwork or tutorials online that's already happened. The problem is one arm of the administration imposing requirements on another without regard any sort of budgetary analysis.

Also legislatures, the federal government, etc. Requirements have many sources, and only some are self-imposed, unfortunately.

This tool makes those inefficiencies more visible to everyone, so I think this looks like a good first step.

This could be applicable for user-side documentation of any form of bureaucracy as either an employee or a customer (most enterprise fiefdoms basically have to consider other employees their customers) because it tends to follow a really long Gantt chart combined with a digraph (potentially cyclical, unfortunately I found). It also happens to address the usual problems I have with most to-do systems where I can have so many unforeseen problems that it is pretty much impossible to set and achieve goals within a certain allotted timeframe (the GTD system, for example, makes zero promises that you'll be able to achieve any of your set goals even if you follow it perfectly and to your best abilities, after all). It reminds me all too much of writing a nasty mess of nested asynchronous calls where the instant anything went wrong you'd probably wind up at square one.

It's unfortunate how this kind of to-do list is something I've almost never seen considered in an enterprise workflow for UX design because it shows a great deal of integration points that are weak and could use improvement, potentially with minimal resourcing. In my limited practice, I've only ever seen the Fortune 500 basically write pie-in-the-sky Visual Studio sticks and boxes or the equivalent of a nasty nested asynchronous call stack.

After running a local version, I thought I would add some notes about my experience:

- This seemed really cool since we have various processes and this would let us both document and operationalize the processes. Things like: add/change user permissions or create a new vm/container, etc. We could make it easy to show where documentation is and what should be updated.

- Installation is simple, but it is hardcoded to authenticate against Github and look and see if you are a member of the 18F group. If not, you are refused access. Changing the group is just a simple edit to the app.js file, but changing authentication from Github to another source would require adding another passport config file into api/ (as far as I can tell), and possibly additional dev.

- The checklists are all json files, there's no UI for creating the checklist. In fact, there's no CRUD for checklists at all, they are flatfile managed with a text editor. This is not ideal, but building an editor might not be too rough. The main thing is that checklist creation is a developer operation.

- There is no UI for associating checklists with role/users or grouping the checklists. Everyone sees all of the checklists and then self-selects which checklist they want to add to their plate. It looks like the tasks are all user-specific, so you couldn't have a checklist that was divided up amongst a team, it is all single-user focused. I suppose you could create a dummy team user that multiple people had access to, but that seems icky. Similarly, while you can see other user's tasks, you can't take one of their tasks and help them out by doing it.

- There's no context for the tasks. If you had two new employees or two VMs to provision, you would have to do them one at a time. Adding a second checklist while you have a first in progress just gives you a pile of duplicated tasks that have no context as to which initial action they are related to. You can add a note to the initial checklist assignment like "SE3: Jane Doe" which will give some context for that assignment, but I don't see where that is related to the tasks. If the tasklist showed "Checklist Name - Checklist Note" in the table, then at least you could keep them separate.

Otherwise, I would love for this to work. I am in academia where we have two problems that this impacts: 1) we have lots of bureaucracy 2) we have lots of turnover in IT staff because we pay poorly. For both of these reasons, having a checklist tool would be fantastic.

Devs: please comment!

This is a great list. (I work at 18F.) Everything we do is available for anyone to fork, adapt, copy, use, etc. Someone could fork our code as a starting point and add the features you mention. I really like the idea of adding context for the tasks.

As the person who converted the tasks to JSON, I can say IT TOOK A LONG TIME. If there was a front end or an easier way to do this, it would be much easier to use. An editor would be my #1 feature request.

- We wanted everyone to see every checklist, but the idea of having checklists shared by a team would be REALLY USEFUL for us too. Think about a team who shares onboarding tasks for a user. I would love to add that functionality.

- If you feel comfortable, you could create these as issues in the repo so that we could share them and see if folks want to tackle them. We saw this as a first step to meet a need, and would love to see it improve.

- Mel

This is great feedback, thanks so much! I've created issues for your requests so we can track them: https://github.com/18F/checklistomania/issues. They all seem like they would add a lot of value. It is our hope that others in government and the public use this code, so we will do our best to deliver in a timely fashion.

After re-reading this, it might be cool to just use this to drive creation of redmine issues (or trello cards, or whatever). This tool could just be a checklist selector / issue builder, and then you could get all of the "collaborative work" benefits of the other tool.

My tax dollars paid for 94% test coverage??

Are you saying that's too high or too low?

High, by my standards. But was tongue-in-cheek regardless.

FWIW, a related product, more general but less mature (I'm the author): http://onemodel.org/ (AGPL). Highly efficient at managing lists but doesn't yet share data between accounts except by exporting the data as txt or html. I use it heavily, daily, to manage checklists/tasks and notes on many subjects.

I see it as the beginning of a platform to change how individuals (or mankind) manage knowledge overall. Future features involve exploiting the internals for collaboration (linking instances, sharing data, subscribing to each others' data, mobile, etc)

Feedback or participation are appreciated. If one has any interest at all, I suggest signing up for the (~monthly?) announcements list at least.

Is this easy to deploy in your own shared server such as GoDaddy and HostGator?

It's mongodb and nodejs, so yes.

A few years ago I worked in a research lab for the DoD. They definitely aren't joking about the amount of onboarding tasks, training, verifications, etc. It took most of my time over the first two weeks. Some of the offices are only open for very limited hours and only on certain days and so they're always backed up. This will be very useful. Even just having access to modern web apps is a welcome change.

This is fantastic, thanks for sharing! This gives me so many ideas I'd like to fork and try out. Does this support recurring tasks?

Not right now. Pull requests are welcome!

Don't forget about TopChart.io (YC) and Trello, two other very easy ways to make checklists.

Trello doesn't have checklist dependencies.

How is this different from Asana or Wrike or the glut of tasks-as-a-service products?

The reason the government builds their own version of many popular consumer products is to run on-premises behind complex firewall configuration. Because they either don't or can't trust a third party with the data.

I checked those - but really wanted the relative task functionality. That was the key feature that we needed (and that our users requested.)

I hope something like that can be used in medical applications.

"When a new hire joins the federal government, existing employees complete over 60 tasks to bring them on board"

They sound down right proud of their bureaucracy.

I would argue they sound self aware and that's better than many orgs, large and small. Many companies have no idea what the number of tasks are required to bring a person onboard and would underestimate the tasks required by 50% or more.

We actually have all of our checklists online: https://github.com/18F/onboarding-documents/blob/master/Chec...

so you can see exactly what's required. The checklist program has made it easier for new employees to keep track of the tasks (and is easier for many people to follow.)

While I don't doubt bureaucracy plays a role, I think comments in this thread insinuating that there is something wrong with having an organized and thorough onboarding process are wrong-headed.

Many, many mature organizations probably have more than 60 tasks to complete to fully onboard a new employee. If you really want to see what inefficiency looks like, put your head in the sand on those 60 items and wait to perform each of them when they become urgent.

No, they're pointing out the exact problem they're seeking to address. That's a step that many developers and founders never bother to take.

Furthermore, 18F doesn't control the onboarding process of (other) government agencies.

A worse alternative is having an inconsistent and lacking onboarding procedure, where there is no checklist in any form. I've seen that happen, very talented people coming on board and having nothing to do for weeks, not even having a computer ordered for them before they started.

Apparently HN is rife with bureaucrats.

Sixty tasks might not be that onerous, particularly if they're very finely grained. For example, 1. Fill out form to request ID card. 2. [+1 hr] Get picture taken for card. 3. [+2 day] Pick up new ID card at front desk.

1. Ask employee to bring in a blank check. 2. [+1 day] Use check to complete direct deposit form. 3. [+2 weeks] Check with employee to make sure direct deposit went through.

We're 10% of the way there and we haven't done any sort of actual orientation, set up any accounts, or done any training.

Not far different from the name of our list tracker - http://youresam.team-duck.com/listomania/

Not far different from the name of a Phoenix song https://www.youtube.com/watch?v=uF3reVVUbio


In 1839 Liszt began an extensive tour of Europe, which he continued for the next eight years. This period was Liszt's most brilliant as a concert pianist and he received many honours and much adulation during his tours. Scholars have called these years a period of "transcendental execution" for Liszt. During this period, the first reports of intense responses from Liszt's fans appeared, which became referred to as Lisztomania.

This is true.

Not far different from the name of a book, probably named after the song - http://www.amazon.com/Listomania-World-Fascinating-Graphic-D...

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact