Taskserver ...!

Summary
Putting this out here for discussion. I’ve been personally poking around at Taskwarrior (and Tasksever) on one of my Freedombox installs, as an isolated service, and I think it deserves consideration for future inclusion as an app.

Lemme explain.

Problem
The main takeaway is that Taskwarrior is a fully focused, data-model–driven task-management/to-do system, which happens to support multiple front-end options.

Freedombox doesn’t currently ship a native to-do-list solution.

I know from searching that Taskwarrior was proposed at one point before on the mailing list and was summarily dismissed because it’s “not a web app”; I think that’s wrong and it also doesn’t accurately reflect the reality of several other criitical apps. For example, mail service and calendaring support are also not web-apps.

In fact, more pointedly, Radicale itself does not have a web-app front-end in Freedombox; I’ve asked for one to be included in a separate thread, but natively it’s only usable via third-party tools that are not included in Freedombox. Same would be true for Taskwarrior/taskserver (except that taskwarrior does include a decent CLI utility, which Radicale doesn’t…).

The main benefit is that Taskwarrior is data-driven, so there is a host of desktop, web, and mobile front-ends to choose from, and which work.

Technically, there are some other apps that have some sort of nominal to-do functionality, such as Radicale which (on paper) provides this via VTODO. But VTODO isn’t widely supported by client apps, nor is it usable, nor is it really all that great on its own merits, if we’re being honest.

Solution
The proposal is to include Taskwarrior + Taskserver, as a bare minimum, with taskserver configurable via Plinth to be usable via the user’s choice of client apps.

My strong preference would be also include one of the client-apps, namely one of the several web front-ends that are actively maintained by the Taskwarrior community. (There’s a massive list here: Taskwarrior - What's next? … which is searchable for “web” or “android” etc.).

I’m trying not to advocate for any one specific web-front-end between the various options on that front; there are maintenance and integration question to be rigorously considered there which I am not qualified to opine about. But there are options (and I even have some preferences).

Tasks

  1. Vigorous, healthy discussion!
  2. Adding taskwarrior/taskserver!
  3. Celebration!
2 Likes

The server is available in Debian and already included in our whishlist. A prioritized list of features and a design proposal would be nice next steps.

The most popular web interface once had a RFP. I’ve included it in our whishlist. This needs a Debian maintainer to package (and maintain) it.

Glad to hear I’m not the sole voice interested in this, of course! I am already aware that the server package is available, and I use it.

(Side note: it’s actually not entirely clear that the page you’ve linked there is a “wishlist”… It also didn’t turn up when I searched for prior discussions of taskserver, whereas the thread that alluded to, rejecting it on the grounds of not-being-a-web-app, turns up. This could turn bikesheddy real quick, but I wonder if there’s some way to better boost the wishlist-facet of that page, in its role as a resource for people thinking about what they want from Freedombox. I dunno; food for thought.)

The state of any web front-end is, indeed, important. On that front, I never touched the one you linked to (“taskwarrior-web”) because it seemed abandoned since 2019, and was also in Ruby, which I don’t use. The one I had been poking at was “taskwarrior-webui” (small namespace problem…), which is still actively being developed.

That being said, is it strictly mandatory to have a web front-end?

I’m of the opinion that it’s not functionally mandatory, and the success of other Freedombox app packages that similarly don’t bundle in their own front-end demonstrates that.

So I would not put “include a front end” on the list of required features, and not at the top of the prioritized list either (although it belongs on there somewhere below because, like Radicale, it’s more usable to offer one to users).