Fastmail Desktop App
Posted3 months agoActive2 months ago
fastmail.comTechstoryHigh profile
heatednegative
Debate
80/100
FastmailDesktop AppElectron
Key topics
Fastmail
Desktop App
Electron
Fastmail released a desktop app built with Electron, sparking controversy among users and HN commenters about the choice of technology and the value of the app.
Snapshot generated from the HN discussion
Discussion Activity
Very active discussionFirst comment
1h
Peak period
125
Day 1
Avg / period
20
Comment distribution160 data points
Loading chart...
Based on 160 loaded comments
Key moments
- 01Story posted
Oct 13, 2025 at 12:14 AM EDT
3 months ago
Step 01 - 02First comment
Oct 13, 2025 at 1:28 AM EDT
1h after posting
Step 02 - 03Peak activity
125 comments in Day 1
Hottest window of the conversation
Step 03 - 04Latest activity
Oct 22, 2025 at 5:52 PM EDT
2 months ago
Step 04
Generating AI Summary...
Analyzing up to 500 comments to identify key contributors and discussion patterns
ID: 45564619Type: storyLast synced: 11/20/2025, 8:28:07 PM
Want the full context?
Jump to the original sources
Read the primary article or dive into the live Hacker News thread when you're ready.
I haven’t had a chance to download this yet, but hoping that it has native keybindings. (Cmd+N) on mac for composing a new email or something similar.
I know fastmail’s built in keybindings are robust, but I can’t keep track of them all.
Just a webpage with a bundled web bowser (electron).
This is like the worst of everything. A terrible noncustomizable browser and a poor email client glued together.
Don't hate on the tech stack.
So let me understand something: you obviously have a free will to not use something, but you would prefer something TO NOT exist and NOT solve problems for users, because you hate the technology so much?
hoping servo makes it better
[1] https://get-notes.com
- Not yet, but adding support for Math blocks is planned.
- Yes, images are stored locally in the ~/user/.config/Awesomeness/attachments folder
- Yes, but only for pre-defined fonts. This is by design - I wanted the app to be opinionatingly simple. But since I got many requests for that, I'm gonna add an option in the settings for that.
- My FOSS version uses a simple plaintext editor (QTextEdit) with some Markdown highlighting. Daino Notes uses a block editor (a la Notion) I wrote from scratch[1] - this allows me to add complex blocks *in the middle of the documents* - so things like Kanban, Images - any custom component that I wish basically - so also math blocks, tables, etc in the future. It's also allows me to do nifty things like cool drag & drop[2] and the editor is even more performant than QTextEdit (especially on large texts).
[1] https://rubymamistvalove.com/block-editor
[2] https://rubymamistvalove.com/blog/drag_and_drop_internal.mp4
In order to abide by QT's license, you have to follow the appropriate set of rules, depending on how you use it. You can use it LGPL, at which point you need to release the QT source you used and dynamically link it in your program. You can use it GPL but you have to release the source to your app. Finally, you can give QT money, and use it closed source. Okay, that wasn't that complex, but those are the rules if you want to use and distribute QT legally.
https://doc.qt.io/qt-6/licensing.html
This is wrong. There's a misconception that you can't statically link your app when using the open-source LGPL version of Qt. From my reading of the LGPL license this doesn't appear to be the case[1]. The LGPL allows you to statically link your app as long as you provide the object files and allow users to relink your app with a different version of Qt.
I've observed many people spreading this misinformation about only being able to dynamically link with the LGPL version of Qt. Please stop this.
[1] https://www.gnu.org/licenses/gpl-faq.html#LGPLStaticVsDynami...
> In case of dynamic linking, it is possible, but not mandatory, to keep application source code proprietary as long as it is “work that uses the library” – typically achieved via dynamic linking of the library. In case of static linking of the library, the application itself may no longer be “work that uses the library” and thus become subject to LGPL. It is recommended to either link dynamically, or provide the application source code to the user under LGPL.
https://www.qt.io/licensing/open-source-lgpl-obligations#lgp...
That last sentence is key. It is a recommendation to dynamically link OR provide application source code. We don't want to be forced to provide application source code, so according to QT.io, dynamically linking is the way to go.
On top of that, there are GPL-only licensed QT modules that aren't suitable for use by closed source applications. Programs using QT on locked down OSes aka iOS and Android are subject to LGPL version 3's anti Tivo-ization clause. I'll revert to my original claim: Using QT is complex for commerical products.
Specific questions about dynamic linking and derivative works re: LGPLv3 have never been tested in a court of law, in any jurisdiction, so all we have to go on are tea leaves and armchair Internet lawyer-ing. The FSF and GNU are certainly entitled to their own opinion about their own license, but they also have a vested interest in casting as wide a net as possible. There are plenty of dissenting opinions out there, by actual lawyers. The important point is that GNU/FSF wouldn't be one deciding what's what in a court of law, a judge is. That judge gets to decide if the specific technical differences between static linking vs dynamic linking are even relevant in the first place. They might decide there isn't one!
Since it hasn't been tested in a court of law, different people are going to have different opinions on what's actually right. QT.io gives a specific recommendation on how to use their library, so I'll defer to what they have to say, but everyone's free to make their own decisions.
> Licensing costs
> Atrocious DX
> Still non-native UI
No problem at all.
Even discord is basically the same thing in the browser
When I want to take a simple note, I'll either open apple note, Google keep, anything that is native because it opens instantly whereas you got to wait like a second for obsidian to open.
1 second is small but also long when you just want to quickly write something and move on.
Agree.
> and a poor email client glued together.
Totally disagree. FastMail webmail is really good, probably the best webmail/mail client.
Actually, I would use this desktop app if it was compatible with IMAP.
Fastmail also have a decent API if you're keen to glue together a better client yourself.
What does their webpage wrapped in a fairly poor, memory hungry browser get me over just going to their website in my browser?
- Block sender (whole domain)
- Report phishing
- Mark as spam
Because that stuff should be done on the server, not on the client and if you mark stuff as spam in Mail, it's on the client afaik.
Also creating masked mail addresses in the first place.
Also being able to make it your default app for email (and hopefully calendar?)
Thunderbird has had a good number of QoL improvements, and the calendar plugins etc are quite niec. Just if one day search could... uhh... work, that would be nice
Fastmail does, of course, probably consider its UI chops etc to be part of their value add. Just seems like if they also want to win over people who like native clients then "here's a bunch of shit that makes native clients also work very well with our offerings" is less of a lift while being more helpful overall. Maybe.
[1] https://fmail3.appmac.fr/
Or if you gotta have a spearate app+:
Safari > File > Add to Dock
+ which I do because it's so much easier to go to my mail instead of wading through browser windows and tabs. I've been using that way for a while now
You'll still be able to do that.
I sometimes think we forget that Electron would have allowed them to ship this to customers super quickly, across all desktop platforms, and get a nice-looking application in to the hands of their customers (who probably have been requesting this for years).
Don't complain about Google taking over the Web, when everyone keeps packaging Chrome with their applications and calling it a native app.
Did that app work on all operating systems that were in use at that time? With a simple install? On DOS, Amiga, Apple, Acorn, Unix? The same app, with the same functionality?
If not: Would you rather live in a world dominated by native Windows applications, written in C with Win32?
Naturally before that there where no phones with Email to worry about.
There were games available across all those platforms, which are a specific kind of apps.
Wordperfect is good example of something available all over the place during the 1980-90's.
Thankfully C isn't the only way to use Win32, as it was already not the case with Win16.
So why didn't it? The company launched two decades ago
There is already a fastmail desktop app. It's called thunderbird. And there are many more, for all possible tastes!
https://josephg.com/blog/electron-is-flash-for-the-desktop/
It is always better to use a well-working application with your native UI-Toolkit: The “Electron” from Signal is one of the best applications using Electron. It is fat. Even in this case people resist. Signal isn’t supporting a stable API but: https://github.com/boxdot/gurk-rsTUI :)
Electron is used, when a company wants to save on developer. All users pay with suffering from bad UI and their hardware resources. In this case it is something nobody asked for?
For email, I mostly don’t care whether it takes too much RAM, if the app is usable. I work with it, then I close it. That’s my workflow, at least. I believe I’m not alone in this. My iPhone is the mini server that gets all the notifications for the emails I need. (By being connected with the default email client.) So, if I want to reply from my laptop, I’ll open my app. Otherwise it’s closed.
Flash was awesome. It was proprietary, unstable, insecure, but in other regards it was much better than competition at the time.
Electron is open-source, stable and secure (as long as you stay up to date).
You’re making a compliment.
> * Non-native UI toolkit
That’s a drawback, not a plus. You can’t easily style the app with CSS the way you want it (VSCode, Obsidian).
> * Massive waste of resources (CPU, RAM, Battery, Disk). Consumption of 500 MB RAM for idle is the norm.
Nonsense. Obsidian is sitting at 115 MB right now, zero CPU.
> * Slow.
Define slow.
> * Often issues with autonomous, local operation.
???
> * Security -> Browser Engine
The only real drawback so far. Can be mitigated by staying up to date.
> It is always better to use a well-working application with your native UI-toolkit
It is always better to use whatever works best for you and solves your problems. Not OCDing around technology and not playing identity wars with crazy people online.
> Electron is used, when a company wants to save on developer.
Also a big plus. Features are being delivered to all platforms with less money spent. Win-win.
> All users pay with suffering from bad UI and their hardware resources. In this case it is something nobody asked for?
You need to seek therapy if you’re “suffering” from application using the toolkit.
P.S. Thunderbird performs worse than ANY Electron app I’ve ever used, and I’ve been using Electron since its inception.
VSCode isn't really styled with CSS if you're a user. It's done by some json or yaml. This kind of hack is required: https://marketplace.visualstudio.com/items?itemName=be5invis..., and it's clearly not supported.
You know what does have theming? Win32, GTK and QT. GTK even uses CSS for it.
And unlike electron apps, when you change how a button looks in your GTK theme it affects all GTK apps rather than just the single electron app.
> Nonsense. Obsidian is sitting at 115 MB right now, zero CPU.
I just downloaded obsidian for the first time and launching it spawns 8(!) processes using a total of 827.3 MB of RAM. That's at the launch screen, before it's doing anything.
Of course it's using no CPU when idle. That's basically the bare minimum bar for any application. It's well known that javascript is at least an order of magnitude slower than C. That's where it's wasting CPU, not when it's idle.
> Define slow.
Well to start off with, I can count the seconds obsidian takes to start up. Most native apps I have installed (other than browsers) start faster than I can reasonably react.
Typing has noticeable lag, see: https://github.com/microsoft/vscode/issues/27378
And this is on a workstation computer, I don't want to imagine how terrible the experience is on low-end hardware.
> P.S. Thunderbird performs worse than ANY Electron app I’ve ever used, and I’ve been using Electron since its inception.
Thunderbird isn't a native app either. Just like electron it's using a browser engine to show the UI.
If your base is base-2, yes. (It is well-understood that when you don't have your thumb on the scale—e.g. selecting, whether deliberately or carelessly, poor/pathological code to benchmark—that the expected slowdown factor of executing on a mainstream JS engine instead of AOT is in the neighborhood of 2x to 4x. Of course browsers JIT instead of AOT out of necessity—a constraint that doesn't apply to programs loaded from disk.)
> I do have two questions. The app on my macOS system is using 700MB RAM; is that for Electron?
Shall we discuss any of the other claims? Your security solution are constant updates?
Invests in efficiency scales with users. Most things which suck are caused by laziness and cost saving. It would be far more efficient to help either Evolution, K-Mail and Thunderbird with access to special APIs and keep them stable.
The same is true for Firefox: Almost the entire Firefox UI is written as HTML page with JavaScript. In fact, Firefox, Thunderbird and their predecessor Mozilla Suite were the first desktop apps written as webapps, and are the direct precedessors of Electron. And they also show that it can be done well. You just have to do it well, which is not easy.
There was a reason for that decision: Netscape tried to implement all its UI natively on 3 platforms - Windows, Mac and Unix -, which was too difficult in practice to keep them in feature parity. The big new thing about Mozilla - apart from the Gecko engine - was to implement the UI of the desktop app as web app, and run the same code on all platforms. That idea is the technical foundation of Mozilla.
Electron is just using that same idea with Chrome and node.js.
The only reason why using Linux as desktop today is realistic, is that web apps put an end to the monopoly of native Windows apps - which was the ultimate goal the whole time.
/Ben Bucksch Long-time core developer of Thunderbird
[1] https://searchfox.org/comm-central/source/mailnews/local/src...
But I admit that you got me curious, do you really expect a substantial increase in power consumption due to Fastmail getting a % of their customers to run Electron instead of Thunderbird? I'm really bad at this kind of back-of-the-envelope math but I struggle to imagine that it's substantial.
Like wouldn't one kid in Nebraska playing an AAA game in anger for half an hour, on one computer, once a day, use more energy than all those electron installs combined?
I’d rather Fastmail ship their web client in Electron than waste engineering resources on reinventing the wheel for a client that’d get it’s own share of endless complaints from people wanting it to be more like Outlook or Thunderbird or Mutt.
That’s the beauty of open standards, everyone can choose their favorite tool for the job depending on their preferences and skill levels.
I don't think for instance this was keeping a lot of people from switching to Fastmail from let's say Gmail, which also doesn't offer a desktop client.
If Fastmail has an adoption problem it's still from people not wanting to pay for their email, a desktop app is not going to change that.
I could very easily imagine that if a company wants to switch over their employees from another provider where the users had an app that had email / calendar /contacts all in one app that they would like to have that same setup again on Fastmail. In the end packaging their web app in a wrapper to satisfy that need of certain customers groups doesn't seem like something that should be very controversial.
JMAP being Fastmails open source json api for email.
You paid them for services rendered. They’ve offered an additional service they didn’t previously for no extra charge. Now you’re upset, even though you’re still getting the same service you were previously at the same price?
Electron is the most cost effective, least wasteful way to produce a desktop app on all 3 OSes. They’re providing a feature to people who wanted it, while wasting less money from ungrateful customers like yourself who don’t care. It’d cost much more to build & maintain a native desktop app when electron works just fine.
Fastmail are an awesome developer, JMAP is great, their mobile apps are great and actually offer some reason to exist, and I like their ethos as far as I can tell. None of this detracts from the fact making an electron app that does nothing other than shell a website, in nearly every case, is a waste of time.
Vote with your feet. Go to mailprovider with native app (and leave sane people alone)!
However, there is little movement in the desktop email app space anymore, so aside from their web app, I see no other clients supporting this.
Evolution added some nifty features lately (Markdown integration). And allows to use Client-Side-Decoration annd classic menus - usability wise awesome. Thunderbird got this year a complete redesign.
The other part are we as users.
How so?
The one time I sent out a mass email using another server and it arrived in my spam folder, Fastmail Customer Service gave me instructions which would allow them to look at the email and help me figure out what the issue was, but I decided not to bother and didn't use that server again.
Is that one thousand plus? (Certainly not one million plus I hope.)
Either amount is pretty bad, although I suppose if you are a well-known person that people are trying to reach, that would make sense.
If not, and it's a bunch of newsletters or spam, I think you should try to make better email choices. (Be more discretionary with how you give out or publish your email address(es).)
Recently they let through a phishing attempt against their own service. It was clearly not from them (to me) and didn't have a green dot or whatever. I was surprised and a bit disappointed that I had to see it at all.
I like their service.
Not certainly, because that’s not how paying a subscription works. You would have to contact them too discuss directly paying for a specific feature. But potentially!
Lol, What do you mean your hard-earned money ? You already spent your money. It is their money now. Consumer != Investor.
Your equivalent is someone going to their local grocery store and asking them to sell stop selling Avocados because they are allergic.
And what about them making this has detracted from your experience as a Fastmail user? I've found the desktop site to be perfectly fine, and they recently addressed my primary complain with the mobile app by allowing emails to be saved offline. I haven't found anything worrying that I feel their dev time needs to urgently address that was ignored due to this. Fastmail just works.
There are so many good clients out there, and I'd rather have 1. The team focus on their core offering, and 2. the existing email client is for the same reason (limited developer time, and matureness) a much better choice for security
This is from August 2019
https://www.fastmail.com/blog/jmap-new-email-open-standard/
This announcement is a slap in our faces.
It's probably because people want easy access, and have all features supported flawlessly. Mail has come a long way, but there are always specific features not integrated well.
Also, most of those apps are more a thin wrapper around the web-interface, adding some interface-sugar for desktop-integration and serving as a playground for devs to test the web-apps offline-abilities.
> There are so many good clients out there, and I'd rather have 1.
I've yet to see any good client for me. They all are kinda flawed and limited, many suffering from age or not fitting modern demands. Thunderbird seems to be the only trustable Linux-compatible one which is still actively developed, and even this is app is lacking on many corners. Add-ons are supposed to fill and round up the corners, but without anyone developing them, what's the worth in having them?
Fastmail at least seems to work on developing the mail-standards, and having their own client is probably helping them in figuring out how well those improvements are working and where they are lacking.
I'm sure there was some deal that didn't get completed because of this.
109 more comments available on Hacker News