Hacker News Reader

> Little UI design changes are NOT what you should be spending time/money on; they are a waste of effort given K9's already strong usability

I'm going to go out on a limb and disagree. I left K-9 ages ago, because unfortunately for ages it had no Material Design.

These upcoming changes are a modernisation. That's good for mass appeal. Mass users don't want an app which looks antiquated.

On the other hand, I can't imagine leaving an app for that reason.

As one who fairly loathes Material Design (especially on mobile), I was hoping cketti would fix what's wrong with Thunderbird, rather than Thunderbird ruining the best Android email client out there, which is now looking like an almost certainty.

I really hope there's a good fork.

I'm neither for or against Material Design, but when I used Android, but MD was generally a good sign that the app was well supported, and the devs cared about UX. When I'd search for an app, the high quality apps were almost always MD. Maybe k9 was an exception, but I'm sure they lost potential users.

Design for me is one of main important on mobile platforms. If the design is bad, is basically a big no for me at least. I wouldn't be leaving because I wouldn't be installing in first place.

I was using Outlook on Android, but I was finding it pretty heavy/slow. Looked for other alternatives, like K9 and Spark, and choose Spark exactly because of the UI/UX.

I get the feeling of wanting a thing you're used to keeping the same looks, but k9 definitely does not have a good user experience. I don't know if the new look will be that, but definitely needs a face lift.

For what it's worth I tried K9 in the past and ended up not using it because it was somewhat clunky and ugly.

So here's one potential user who would welcome a more modern interface.

> I left K-9 ages ago, because unfortunately for ages it had no Material Design.

Maybe you should give it another shot – they changed the UI quite recently (a year ago?)

Thanks, I'm aware they relatively recently landed UI changes. However, for now I'm happy with FairEmail, which is likewise open source.

> That's good for mass appeal. Mass users don't want an app which looks antiquated.

That's pretty spot on. Another example is Eclipse. I swear by it, and it handles everything for me. But when I show it to someone the first question is when did it last updated, because it looks very old.

Eclipse releases a new version every three months.

As a short-time K9 user I disagree. You have to pick between the display name and email of senders. There is no way to show both. This is because of problems with the UI design that risks being misleading with malicious names.

The think there are also lots of UX improvements that would be great. Right now the threading UI is awkward. I'm really glad that swipe actions and swiping between messages have landed as well.

I like K9 but I can definitely see the benefit of better UX.

> Little UI design changes are NOT what you should be spending time/money on; they are a waste of effort given K9's already strong usability

"This rotary phone in the office is perfectly fine and usable! The new-hires just need to learn how it works, and then it's the exact same thing."

Learn how it works? It's a rotary phone! Just turn the disk the appropriate amount for the number you're dialing and loose. You could train a chimpanzee to do that.

You've rather missed my point, I think: an unfamiliar or aging look over what is a common interface is a huge problem for gaining new users.

Further, I have witnessed people not understand rotary phones. Plunk one down in front of 10 teenagers and I would wager few indeed would intuit it immediately.

Ah, I see. Well, if you were to ask me whether I think a group of teenagers is smarter than a chimpanzee...

Regardless, on the topic of drawing new users, if the users truly are "new", then any look will be unfamiliar to them, but yes, luring users over from competitors may be tricky.

I don't really know about the dated look, though. For example in computer UI design we had flat, then went shapely, then went flat again. It reminds me of architecture: there is always the disconnect between people who like an older style (like when Classical Mediterranean architecture came back in vogue and we saw a bunch of public buildings being built with all pillars and triangles) and people who want some kind of new experimental style (like when Jugendstil started appearing).

There is always a balance that can be found between finding a presentation that aligns with market trends and a presentation that stands out and draws attention. It's up to the people running the show where on the axis they decide to plant themselves.

"Well, if you were to ask me whether I think a group of teenagers is smarter than a chimpanzee"

Well, your example was training a chimpanzee to do it vs. can teenagers use it without training. And I think they can and maybe even think it is cool, but unless you want to harness that coolness retro factor it makes 0 sense to use a rotary phone for actual use. It is way slower and therefore less efficient. But of course it would beat a modern phone that just looks cool, but only randomly works.

Which would be my requirement for any tool. Functionality first, looks second.

Why does an open source project need to lure new users? Especially at the expense of limited dev time when there are actual security features in the issue tracker?

"Why does an open source project need to lure new users? "

To get some traction and momentum to actually become a serious alternative. New users can also mean new funds and not just more work(like it should be), like for example the blender foundation is proofing that this is possible. Make something that really works and once some actual buisnesses are using it, there will be ways for funding it.

What's wrong with rotary phones, other than being different to the current norm?

"Intuitiveness" is AKA "resembling the current norm" when it comes to UX. Updating an aging UI to fit current patterns makes it more cohesive with the operating system, attracts more new users, and avoids confusing those new users.

Attempting to paint "strong usability" through the lens of existing users only (the guy who already knows how to use a rotary phone) is a lopsided view of the goals for an app creator.

It's slow and tedious to dial numbers with many high digits in them. And there's no quick dial function for commonly used numbers. Both of these things directly impact UX.

You can always dial zero and get an "operator".

An operator is like Siri or Alexa, but capable of passing a Turing test.

And now I want to build a rotary phone that connects to Alexa when you dial zero.

Don't be that evil, at least connect it to something more useful, like Google assistant.

Assuming we’re comparing stuff you can buy at a random shop:

- you can blindly hit the keys on a dial pad if you remember their position, on a rotary you probably need to count the clicks

- on a rotary you probably can’t look back at the number you already inputed

- traditional ones won’t have * and #

- no speakerphone feature, which many older people would benefit of

- no mute

- no phone number memory/speed dial

It’s actually fun to think about how much it has changed since the earlier phones, as the land lines are progressively getting obsoleted as well.

I could imagine a modern version of the rotary UI, as a swipe input interface for instance, but it would still be clunky IMO

I'd wager that rotary phone can dial 911 unlike a Google Pixel.

Constantly evolving UI thought experiments are largely responsible for the sorry state of software today.

Maybe the new-hires need a healthy dose of shut-up-and-listen, and do need to learn how it works.

"But we hired designers and if we don't make small design changes they'd be out of work"

Seriously that's how most of the UI trends feel like

Would JMAP not be a better focus both in general, and to attract those same users?

https://github.com/thundernest/k-9/issues/3272

There’s only one significant mail provider that supports JMAP.

...and it must also be said that the selection of client software that supports it is somewhat limited. For this discussion, Thunderbird does not. I'd also add that some fairly major providers, e.g. anything based on MS Exchange, and Google's Android client, do their own thing (e.g. ActiveSync).

For myself I'm aware of a bunch of bugs in Thunderbird affecting my experience and I'd rather these were fixed first.

And in larger scheme of things that provider is anything but significant.

> Signed: someone who still calls a tall building in Chicago as the Sears Tower

I think all locals/natives to the 312 and 773 still refer to it as such. Ain't changing!

confirmed. it's not changing. haven't lived there for a while, but it'll always be the Sears Tower if you ask me.

What flaws are there in K9 that still motivates you to stick to Outlook?

I use Outlook on my Windows PC and transfer all personal contacts, calendar entries, tasks, and notes locally. I would love to move to Thunderbird primarily because "Microsoft bad" but then I'd lose my local transfer functionality (not to mention Xobni - which still crashes every-so-often - but the functionality is worth it).

But is there a demand for desktop-based sync?

You can run your caldav/carddav service in your local network, if you are so inclined. You will get local sync, and the client authors do not have to build specific desktop-based tools.

I for one do not want to return 15 years back, to serial/irda/bluetooth sync, having to think about sync, and eventually forgetting. Running a local service is much preferable, with devices syncing themselves as they see fit.

If you have a link for running a small footprint of a caldav/carddav service on my Windows laptop, I'm all ears. I played around using WSL and setting up a NextCloud instance but why do I want to use 2GB+ just to sync?

At the end of the day, I want all my personal events, contacts, todos, and notes on my laptop and able to sync directly with my phone. I'm happy enough with my current bluetooth sync and wouldn't trade it for UI changes.

Personally, I'm using a small Synology NAS and the Contact and Calendar servers that Synology provides. It solves the problem for being a host for multiple devices, with my laptop not having to be on, or even present, at all times. Also having dns entry helps.

But for running on your laptop? Yes, nextcloud is overkill, it is not contacts/calendar server in the first place. I would look for smaller, more focused tools instead, baikal for example. For another inspiration, you can have a look at what davx5 -- the caldav/carddav provider for android -- tests against (https://www.davx5.com/tested-with).

I use davx5 with my NextCloud setup on the cloud. But that's not what I use (or want to use) for other events, contacts, to-dos, etc.

You can try DAViCal https://www.davical.org/

Tried it when I played with WSL. Like with radicale [1], I don't remember what specifically wasn't working properly but it didn't.

[1] https://news.ycombinator.com/item?id=33872707

You might be interested in https://radicale.org/v3.html. Runs on my odroid board with 26mb memory. The documentation is particularly good. I've used it as a replacement for the Synology CalDav and CardDav services.

It's very easy to install and does not require a DB. As a bonus, it stores everything as files which can be read and edited manually. It does require python.

I had played with it for 2 days before I gave up. I don't remember what specifically wasn't working properly but it didn't. Even if I had gotten it to work, syncing between a Windows PC and Android device should not require that level of effort.

I've been using Baïkal for about a decade. I have it in a docker container on my home lan. Android and iOS transparently and gracefully tolerate the server being unreachable outside the house.

The main issue is that I only touch it every five years or so, and at that point, I've forgotten what it is called or where it is running!

(I use synology's HyperBackup to E2E encrypt + backup it and a bunch of other docker containers.)

Here is a list of 11 options:

https://medevel.com/11-caldav-os-servers/

Why not just run Thunderbird on your windows pc, then? It should synchronize perfectly happily using caldav et al.

Caldav requires the cloud.

Hm? We sync our calendars using Nextcloud, but admittedly, that's not everyone's cup of tea.

To avoid syncing my calendar and contacts to the cloud, I'm using wireguard as a always on VPN on my phone, to connect to a locally running CalDav/CardDav server. Wireguard on Android can be setup to only affect specific apps, so my DavX app is the only one using the VPN.

mailbird or emclient

K9 doesn’t support modern authentication.

Have you tried FairMail?

If you mean OAuth, they've added it earlier this year.

Do they support OAuth2?

What other OAuth is relevant in the context of email?