the instructions are for the task I assign to Claude, most of the times it was stopping for feedback, I prefer to check in the end rather than spoon feeding it
A project with the complexity of an email client cannot be considered stable in a month. There are bugs that will only show up after a year. It has a calendar after all.
> The UI looks pretty good too. It's really not hard to beat thunderbird.
I don't understand this mentality. It's an email client. There's not a lot needed UI-wise to make a functioning interface. Display a list of emails, show the full message when you open it, etc. I don't see anything functionally different between this email app and basically every other one I've ever used.
You've either got insane standards or are using email way different than I.
Can’t speak for others but for me Thunderbird’s UI feels distracting and almost grating in some ways. It’s technically functional but doesn’t feel right.
I know things like that don’t bother some folks and that’s fine, but others like myself can’t turn the visual parts of their brains off and so some software feels like nails on a chalkboard to use.
> It’s technically functional but doesn’t feel right.
That might be the most accurate description of Thunderbird that I've come across. It's hard to pinpoint exactly what's wrong with Thunderbird, despite me using it every day. Somehow it feels like the emails aren't the important bit of the UI, but I also couldn't tell you what I'm looking at instead.
It’s a gut feel sort of thing that’s usually grounded in some set of UI design principles that haven’t been adhered to. Defining it just pins down what the issues are rather than making the feeling go away.
Glancing at screenshots of TheBat! and Pegasus Mail, i don’t find them objectionable in the way I do Thunderbird. They’re just old school desktop clients like I used in the late 90s/early 2000s, which I generally find workable. They’re not quite my preferred style these days (which falls into the “it’s not what I’m used to” bucket you mentioned) but I could probably get used to them again if I had to.
Thunderbird on the other hand feels wrong somehow, and the recent UI “modernization” didn’t do much to change that.
> There's not a lot needed UI-wise to make a functioning interface
I remember waiting for ~15 years until Thunderbird finally implemented rich content support in the e-mail list (needed for a proper 3-column layout, as otherwise the e-mail list, being a simple table, wouldn't fit on the widest screens available).
So, yeah... Not a lot, but Thunderbird has been missing the bar for over a decade.
There is a case to be made to always start at 1.0, especially libraries with public API’s. The reason is that if you use semantic versioning then below 1.0 you can’t properly declare a new release to be a feature or bugfix release (for example, 0.4 to 0.5. Is this bugfix or feature?). Above 1.0.0 this is easy (for example 1.1.0 vs 1.0.1).
In libraries, 0.x is for when you’re still figuring the API out. Basically you’re saying that no API is stable at this point, and you’re not doing semver yet, and anything can change.
Naturally, the users should treat those libraries as beta versions and only use them in production at their own risk.
> you can’t properly declare a new release to be a feature or bugfix
In practice, 0.5 can be assumed to be breaking but then how do you specify non breaking new features? 0.4.1? Moving to 1.0.0 makes things a lot clearer. For breaking, just go to 2.0.0 etc
> but then how do you specify non breaking new features?
You don’t. As soon as you feel the API is stable enough, move to 1.0.0. 0.x is for the “yeah the API doesn’t feel right just yet but it might be useful for early adopters” stage.
I know there’s the tendency for mature projects to live in 0.x for a long while, yeah – definitely try to avoid that. But 0.x is a useful first step for libraries, especially if the API is non-trivial.
This is interesting, I am doing an experimental effort with similar goals. What is your approach to testing? The bug surface of something like this is quite large. Also, do you have a plan for sandboxing?
I use AI to aid with development, yes, thanks to this it has boosted my productivity by a lot, but to say it's completely vibecoded, not quite. but thank you :)
Because I typically see GTK as an ancient buggy toolkit with bad documentation and an ugly cross-platform experience. I would love to be proven wrong though.
There was an entire talk given about the pain of needing to switch Linus Torvalds' divelog app from GTK to Qt and they seemed to have the same opinions I do: https://www.youtube.com/watch?v=gGZyVSOnqm0
Perhaps GTK4 is different and looks better cross-platform, I don't know, but I haven't heard of any new/prominent cross-platform apps that actually use it.
And the way I have been observing several top gnome devs' attitudes and abusiveness towards their own users the last handful of years does not sit well with me.
You realize QT predates gtk right? Gtk was a reaction to qt having weird license issues, which it still does, doesn’t make it any less old though. Good luck writing anything in Rust in QT. That’s why people are either choosing gtk or a native toolkit.
Then my 'amail' client seems to fit the bill.. (OSS, rust, I use it on Linux and haven't tested elsewhere) ..not really suggesting it – it's pretty barebones and just whatever I've needed, but you're welcome to it obviously! Especially if you have issues/PRs to spur me on to features I've wanted too...
I would love an email client that automatically sorts my mails via a local LLM.
As long as it tries to be the "I" in "AI" and do something useful like automatically detect conversations with my accountant, per project and not per sender mails including somehow labeling mails referring to multiple projects with multiple labels etc.
Second that. I can’t really think of a better application for LLMs than this. In fact, this might be the only good application.
Drafting emails with LLMs is... ugh.
And assistant chat? Really?
Okay, two things might be great: translations (preferably local, e.g. using Project Bergamot) and message summary (in case somebody does send an unedited blob of LLM slop to me and social dynamics don’t let me send it to spam).
Has AI functionality available if you enable it, but doesn't come bundled with anything, you connect your local LLM or Claude if you have to, you can also use OpenRouter or whatever you want, go cloud or go local, up to you (if you're interested in AI), otherwise, disable it
Maybe I'm not the target here but an embedded chat ai window is a huge turn off to me. Having a CLI I can give my own agent to access/query the data, if I choose to do that, would have been preferred but if the target was my mother, then a chat window is probably a better option.
That's a very nice looking app - I like the look of it it gives me Mail.app vibes.
A mail client is always a tough thing to get right so I'm glad you're working on one
I love then idea that as the total cost (time/money/attention) of personalizing software continues to come down, more people will be able to encourage a diaspora of options, tailored to each niche.
Visually it resembles Mail.app, but if it’s going for a similar level of functionality I think modern GNOME’s hamburger-and-three-dots-buttons approach to menus is going to become a problem.
This is one of those types of app that really benefits from a proper menubar with several top-level menus so the user isn’t stuck spelunking submenu tunnels.
Because "Find me that mail from last year from Peter or Pat? that I never replied to" is easier for most people then setting a bunch of filters and regex.
Quick grammar correction can also be nice. Not super important, but nice.
But these MUST have a local option first and foremost, this is something that any local model can do, even the smallest ones.
But even before that there would need to be a set of APIs from the DE, imho. I think that GNOME is going into this direction with their AI center, i whish KDE would start looking into that too
I agree. The other day I reinstalled some old gaming pc of mine with the latest bazzite. Then I discovered it had lots of old photos on it, but in many many backup folders.
So I had Claude whip up a super fast image browser that had some features I really like from a Mac app I use. It looks better than 90% of Linux apps, is super fast and fits me like a glove.
Of course some of these mail apps will be abandoned, but they also provide an opportunity to make things more attractive on the Linux side and win some users.
Ohhhh man, that is a killer idea, thanks for sharing this. I use a program called "nomacs" on Arch Linux, QT bases and super lightweight but everytime I update it requires a 5 minute recompile due to all the extra features it has that I never use. I'm totally going to make a lightweight version for myself now.
But yeah, thanks for reminding me; I've always wanted to make my own file manager and image viewer, since literally every single one out there is garbage. How is nobody able to get such basic programs right is beyond me.
You need a secure, supported mail client first and foremost. Better looks definitely would help Linux adaption, but safety and functionality takes precedence in this case
This is really not about better or worse UX but preferences. Lazy loading means your client will be available from second 1 to open an emaik but would take ages every time you search for something. Building an index at first import allows almost instant search once it is complete.
I compared for instance evolution and outlook, evolution has always had much faster searches.
Having said that, having a bit of digital hygiene goes a long way.
> They will probably all be abandoned within the year.
This could be a proclamation for ALL vibe coded software. As a matter of fact, I think this is what marks software as "Vibe Coded". It's fun while the vibes last. The first few GitHub issues bring excitement because, "hey! people are actually using my project!". So the author prompts a few bugfixes or features. Then the GitHub issues become unmanageable. The author doesn't really KNOW why things aren't working. They can't find anything in the codebase without grep (this is true for the agents too). After awhile, it's not a good vibe anymore.
Correct, nothing the parent mentioned is specific to vibe-coding. At best we could say it's easier to start a new project, so there could be more abandoned projects, but maintenance is either the same or easier than before.
Also, I'd like to know what "after a while" means, concretely. Seems to be working out all right for me on the 4 new apps I've started that I wouldn't have time for otherwise, thanks to AI. I'm at 184 revisions in one of the projects and it works fine so far.
this is my daily driver, and I want to keep maintaining it for the foreseeable future, I agree there are lot's of apps that end up dying. But that's not the case for Penguin Mail. And to clarify, I have used AI to develop parts of Penguin Mail, but I've also contributed a lot of code, I've been a developer for the last 30+ years and I think AI is a tool to increase productivity, embrace it :)
> - can’t choose a sender email?! This is major dealbreaker, if you have many aliases or domains.
Ugh, this drives me nuts. It's the same reason why I stopped using my Supernote Nomad for email.
I use iCloud+ email with a custom domain. With that arrangement you authenticate to Apple's servers with your iCloud "primary address", like someone.somename@icloud.com. Then you can send and receive email with your custom domain email address, which is the one you're actually using. But not with Supernote, or apparently Penguin Mail. My Nomad will only send mail with my @icloud.com address, even though that's only used for authentication.
So yeah, I just stopped using it altogether. That's not my address. It's not how people reach me. It's literally just a username.
Thank you, I will keep checking it and if it gets most of the official fasmail client features I will make it my main one, not a big fan of the official one it’s too resource hungry.
I wouldn't say it's something that can be vibecoded in a month or two. Proper design and implementation take time if you want to handle accounts with hundreds of thousands of messages seamlessly.
These days with the convenience of having LLMs at developer’s disposal, it feels unjustified to use electron like apps.. whenever I see “one app all/multi-platform”, I feel discouraged to give it a genuine try.
I tried a bunch of email clients and then crawled back to mail.app + Calendar.app on MacOS simply for how native it feels.
For other HN folks(or bots) passing by... KMP = Kotlin Multiplatform
Apparently, unlike Electron, it lets devs write shared Kotlin code once and compile/use it on multiple platforms. KMP can compile Kotlin to native binaries for platforms such as iOS..
Interesting. TIL about this. I added myself to the waitlist.
I think this looks good, but I'm not on a Mac anymore.
For people that want a terminal client with tabs, integrated html viewer(including images), markdown compose, MCP, with unified inbox, I built notmutt.
Did you not read the part from the author saying that there is currently no linux support? And yes, it looked like a macOS app to me. [1]
But apart from given what I've seen from my experience with handling maildir, actually implementing mail sync, storage, and all the other things associated with mail AND calendar, it sounds risky to use a project this young.
Yet another GTK4/libadwaita, no thank you. I mean, if you're going to use Rust and vibe code thing, why not use a more sensible, Rust-native UI like egui?
Went to watch the Youtube intro [1] for Penguin Mail and was greeted by the most egregious scam video. Didn't click the ad, but simply went to the domain (usbest1.com) and it's just chock full of AI generated content hoping to scam you out of your money. Going through the checkout flow, it even asks you if you want to tip your driver (presumably for delivery).
Sorry for the off-topic, but Google (and all ad networks) should be legally required to vet their advertisers. That's gotta be the worst I've seen.
YouTube has some seriously scammy ads. To an extent that these ads have motivated me to buy premium for my parents as they tend to fall into the trap of fake dialog boxes shown in ads warning a PDF app needing an update.
I went through a phase of reporting them (to YouTube, there's a button while playing to do so) and did get automated reply that a lot were removed. Some took several reports, others they continued to insist didn't violate their policies, but they will remove them.
Interestingly, one of Poland’s wealthiest individuals (who has been the target of scams ranging from malicious to macabre) is now engaged in an openly declared war with Meta[1][2]. There is an entire foundation dedicated to tracking ads and identifying scams; they calculate the profits generated from these schemes and lobby for regulations that would mandate the vetting of advertisers, recently on an EU level.
Well if the rich finally decided less ads was better who are we to argue? Hopefully we can have a "I AM RICH" category on the Internet so corporations can keep targetting us with ads while sparing the poor rich people.
Well, in his case it was more of very visual fake news that he was dead, or handcuffed and arrested. Also while it started personal, at some point became an issue of “if a billionaire with positive rulings of a court and other govt institutions cannot win, who can”. Now he’s more on a crusade to stop scam on SoMe in general, not only to remove the ones about him.
Amazed anyone here isnt using an ad blocker for youtube tbh, it feel like the first thing you'd want to avoid the mess.
Ublock is still a thing. Or use Brave as thats got baked in ad blocking which can be combined with ublock to make it even more effective. Then adguard or pihole on your local network so no analytics/tracking data is still being pushed out.
I need to update that intro, that's how the app started, it's very different now.
I use Penguin as my daily driver, and I made it exactly because of what I was missing on other mail clients, but idea after idea brought me here to share it with everyone, my intentions are to continue development and not keep it stale or let it die, there are loads of things i want to add but I also don't want to lose any stability, a nice balance is what I'm looking for
Again, really sorry to derail the conversation. What you've created looks amazing! I do typical use the command line for mail (mblaze), but I'd certainly consider giving your client a spin if I wasn't on macOS.
How do you gather it's a scam? It seems like a genuine Chinese seller selling alibaba/temu/whatever nonsense? Not that I would buy any of it but doesn't look like the descriptions are scammy?
On the shop itself that you linked though I don't see any unreasonable price (or any SSD for that matter). Just normal TEMU-level stuff with appropriate prices.
How so? Any specific critique? Ive been using Thunderbird tot décades and while e-mail seems to have some inherent clunkiness I find it overall quite fine.
The thing with all these vibe-coded email clients is: how well do they sanitize the full-content view of HTML emails?
Thunderbird and the like go to great lengths to make viewing those emails safe, e.g., by preventing inline JavaScript and all kinds of tracking techniques (pixels, CSS references, etc.). Without having checked, do these vibe-coded clients just render the HTML in a WebView, or do they have similar protections in place?
It seems the <sarcasm> tag was missing, I assume,
but as shown by many companies and open source projects, agents with the right prompts are much better at finding security problems than 99% of developers (for various reasons).
Many devoted fans of AI like comparing LLMs to compilers for higher-level programming languages.
They argue that just as programmers once wrote ASM and C, the arrival of Java and Python meant you no longer had to be a neckbearded autistic kid to write software that doesn't leak memory or segfault.
Arguments like this show me you’ve never done software engineering seriously.
Despite compilers (written by some of the smartest engineers on Earth) trying hard to prevent developers from writing shitty software, those new age "engineers" still managed to ship software that crashes, leaks, and generally sucks.
I'm not saying this email client is not secure (although I'd bet 5$, it is). But it seems beyond silly to assume software is automatically safe just because a new piece of tech can potentially find vulnerabilities. Intent still matters. And the 99% would never have the intent to write secure software.
> People don't—and aren't incentivized—to give the slightest shit about quality, security, you name it.
This normally happens on two scenarios: a) they are not paid enough to care that much or b) they are paid more than enough to not have to care that much.
Either way it’s almost always down to how much the company is willing to pay for people that actually care about these things.
It's not about the pay, it's about the expectations. Turnout time and business outcomes are the expectation. Pay me an amount I'll accept (modest) and I'll give as many shits as I can, unless the expectation is that I'm doing things on a wildly compressed timeline where a checkmark is more important than data leak that might never come up. I care as long as I'm not punished for doing so.
Half the security stuff AI seems to "find" for me have been complete fabrications, with the other half that it says is "critical", but has actually been so minor that you'd basically have to have access to my own computer to "hack it". So, yes, it is much better at findings issues, too bad most of it seems to be made up.
This is actually a very tricky problem. We solve this by never rendering mails in trusted origin. It's basically sanitisation + isolated sandboxed iframe for defence in depth + standard security stuff like CSP headers. When I was looking into this, I was surprised how many popular mail clients relay primarily on sanitization and render mails within trusted origin (ex: marco, superhuman).
Other interesting problem to solve for mail clients is the editor. You should be able to forward/edit untrusted html content. Source: I develop tecotype.com for mac/linux/windows.
From the website I only see download for mac. Where are those linux/windows builds? I'm interested in testing. Could not find any repository or something resembling a more technical page.
Sorry, it isn't publicly released yet for linux/win. I'm planning on releasing it via Flatpak soon. On the technical side, you can reach out to me via email or ask here. Happy to answer any of your questions. This is a paid, closed source product.
Was looking for exactly this type of email app last week and nothing had the mix of features tecotype has and definitely not as polished. Yes, I know i could cobble all this together but you've basically answered my prayers. Eagerly awaiting the linux release!
Thank you very much for the warm words. Mail clients are surprisingly tricky problem and it took way more effort than I initially anticipated :)
For example, I'm using catch-all address with fastmail. I wasn't able to find any client with the right design that would allow me to define sender address and email account separately during composing, so I can use "alias@example.com via me@example.com" as a sender.
You can send "ping me when linux available" and I'll follow up.
I use a similar catch-all scheme and would love to define the sender and email account separately. So many companies see that my account email is: company@mydomain.com and require support requests to come from that address. So then ive got to then turn that alias into a mailbox - its quite frustrating.
If I can be a beggar and a chooser: The vim keystrokes to go to the top or bottom of the current email list and/or the conversation thread items are sorely missed.
2. If there's a way
My perfect mail client would be something that does not use a browser to show the HTML emails at all. And would need to have text-only option I can enable. And for HTML emails, I would prefer something like html to markdown, that markdown can be mapped to native UI toolkits, they all have some "rich" text support.
I did that in my IUP fork, bold, italic, headings, link, image, table, that can all be mapped to native text widget, can load and save markdown from that.Will it look as intended, no, but, do not create websites when you send emails and everything will be fine.
“Email client” is an interesting puzzle. Email involves multiple activities:
Viewing: I have a message or messages stored on disk in files or mboxes or a maildir. How do I render them? Example use case is evidence in a law suit or, more likely, emails I’ve put in a shared drive to help me and my partner argue with our landlord. Yes, I could just have a Gmail folder for these but I don’t want to have one foot in a file-filesystem and one in a mail-filesystem. I want “lease_2024.pdf” and “20240401_deposit_confirmation_002.eml” to both be files. But then of course we have…
Filesystem interface: Emails have important attachments with MIME file names. Or perhaps a more keen observation is that some named files have emails as attachments! I’d like to symlink into a virtual FS that keys files off a UUID / Message-ID, and be able to mount that virtual FS from any mail store I have available (Gmail API, JMAP, IMAP, .tar archive, mbox, maildir, or .eml files.) Most conversations I have with people don’t involve files, and most files don’t involve conversations, but when they do I want them to feel natural and not some kind of special category of filey-messagey thing that doesn’t play well with plain files or, less commonly, plain messages.
Composition and sending: Creating a new message involves making a new file and then using that file to drive a sending API. I want to compose a beautiful email using a tool that is good at that job. Maybe I’m using vim to draft it in markdown with macros and tools to machine build certain complex parts? I also need some mime aware tooling — sometimes — to help prepare a bunch of attachments that may or may not have names related to their names on disk! I want all of this to be scriptable, on the occasions when it’s complex (30 part message for a client’s loan application!) and easy when it’s ad hoc (to: mom, see cat.jpg).
Search: email is notoriously the extension of and replacement for my long term memory. I have gathered message history from many sources and accounts over the years. I need to manage these into packed archives for the old ones, editable archives for the current ones, promote the former to the latter, and arrange new messages into the system. This feels like another separate tool.
I’m sure there are other potential tools lurking here as well. Separate rendering, composition, sending, receiving, searching, and archiving, feels like the goal. Most new email apps I see try to implement all of the above in one new package when, really, it would be great to see innovation on each branch of email as separate tools.
My PDF viewer doesn’t need to innovate on file browsing! My SVG drawing tool doesn’t need to innovate on mounting remote filesystems via SSH!
Nice UI but untrusted email messages is something I'd rather open in [an uglier app with decades long reputation for being reasonably secure](https://www.thunderbird.net/en-US/)
Would not be a problem with me, download source, let the agent search for backdoors and security problems for some hours and it would be fine for me. I would need to do the same with thunderbird if I'd need to trust an UI frontend to my email.
Not sure if this is gaslighting, an absurdist joke, or we are going through a collective psychosis.
KindOfCoders, AI is a very useful tool, but nothing works like the way you're explaining. I am kind of finding it hard to articulate the problem, because the assertion is just a bit wild. So I will do my best.
Thunderbird has been in production use for over 20 years, many many smart people have put their time into it to make sure it works correctly, yes, it is a bit clunky, it has a lot of craft, but you're not going to replace that much labour with running an AI agent for a few hours; despite the AIs usefulness in finding bugs and vulnerabilities. Not yet at least.
You are correct; however, agents can and do find terrible bugs in some of the most popular software in the world made by some of the most competent engineers on the planet.
I assume what KingOfCoders suggested is to run an agent almost like an antivirus which I guess sounds somewhat interesting but you'd have to re-run it on every update and probably spend many more hours+tokens for such a system to work properly.
But perhaps once inference gets truly cheap repository maintainers should include AI security checks for new software as part of their pipeline? I think that would make sense. Especially for repos that are frequently pwned like npm!
181 comments
[ 0.23 ms ] story [ 62.8 ms ] threadI like the idea of a native Linux email client. The UI looks pretty good too. It's really not hard to beat thunderbird.
Calling it version 1.0 is bullshit though. It's only about a month old: https://github.com/c9dev/penguin-mail/graphs/contributors?al...
A project with the complexity of an email client cannot be considered stable in a month. There are bugs that will only show up after a year. It has a calendar after all.
I don't understand this mentality. It's an email client. There's not a lot needed UI-wise to make a functioning interface. Display a list of emails, show the full message when you open it, etc. I don't see anything functionally different between this email app and basically every other one I've ever used.
You've either got insane standards or are using email way different than I.
I know things like that don’t bother some folks and that’s fine, but others like myself can’t turn the visual parts of their brains off and so some software feels like nails on a chalkboard to use.
That might be the most accurate description of Thunderbird that I've come across. It's hard to pinpoint exactly what's wrong with Thunderbird, despite me using it every day. Somehow it feels like the emails aren't the important bit of the UI, but I also couldn't tell you what I'm looking at instead.
I often find myself in situations where someone wants to break the UI I like, because they're used to something else.
It seems that most of the time, "feels right" just means "it works like I'm used to in other software". So: GMail, Outlook, whatever. Which is sad.
I wonder how would you react to "TheBat!" or Pegasus Mail?
Glancing at screenshots of TheBat! and Pegasus Mail, i don’t find them objectionable in the way I do Thunderbird. They’re just old school desktop clients like I used in the late 90s/early 2000s, which I generally find workable. They’re not quite my preferred style these days (which falls into the “it’s not what I’m used to” bucket you mentioned) but I could probably get used to them again if I had to.
Thunderbird on the other hand feels wrong somehow, and the recent UI “modernization” didn’t do much to change that.
I don't understand this mentality.
Human communication is one of the most complicated activities in existence. Most (all) clients are modelling it very poorly.
I remember waiting for ~15 years until Thunderbird finally implemented rich content support in the e-mail list (needed for a proper 3-column layout, as otherwise the e-mail list, being a simple table, wouldn't fit on the widest screens available).
So, yeah... Not a lot, but Thunderbird has been missing the bar for over a decade.
There is a case to be made to always start at 1.0, especially libraries with public API’s. The reason is that if you use semantic versioning then below 1.0 you can’t properly declare a new release to be a feature or bugfix release (for example, 0.4 to 0.5. Is this bugfix or feature?). Above 1.0.0 this is easy (for example 1.1.0 vs 1.0.1).
Naturally, the users should treat those libraries as beta versions and only use them in production at their own risk.
> you can’t properly declare a new release to be a feature or bugfix
0.5.0 vs 0.4.1?
In practice, 0.5 can be assumed to be breaking but then how do you specify non breaking new features? 0.4.1? Moving to 1.0.0 makes things a lot clearer. For breaking, just go to 2.0.0 etc
You don’t. As soon as you feel the API is stable enough, move to 1.0.0. 0.x is for the “yeah the API doesn’t feel right just yet but it might be useful for early adopters” stage.
I know there’s the tendency for mature projects to live in 0.x for a long while, yeah – definitely try to avoid that. But 0.x is a useful first step for libraries, especially if the API is non-trivial.
Can this be compiled for other operating systems than Linux?
(This used to come at the cost of code quality, but increasingly does not.)
There was an entire talk given about the pain of needing to switch Linus Torvalds' divelog app from GTK to Qt and they seemed to have the same opinions I do: https://www.youtube.com/watch?v=gGZyVSOnqm0
Perhaps GTK4 is different and looks better cross-platform, I don't know, but I haven't heard of any new/prominent cross-platform apps that actually use it.
And the way I have been observing several top gnome devs' attitudes and abusiveness towards their own users the last handful of years does not sit well with me.
Yes but I mentioned that because to me gtk still looks ancient today while Qt looks native on each platform IMO.
> Good luck writing anything in Rust in QT.
There is this thing now: https://www.qt.io/blog/rust-ui-framework-via-bridging-techno...
And there was already multiple Rust bindings: https://wiki.qt.io/Language_Bindings
As long as it tries to be the "I" in "AI" and do something useful like automatically detect conversations with my accountant, per project and not per sender mails including somehow labeling mails referring to multiple projects with multiple labels etc.
Not that I think this mail client can do that...
Drafting emails with LLMs is... ugh.
And assistant chat? Really?
Okay, two things might be great: translations (preferably local, e.g. using Project Bergamot) and message summary (in case somebody does send an unedited blob of LLM slop to me and social dynamics don’t let me send it to spam).
A mail client is always a tough thing to get right so I'm glad you're working on one
I love then idea that as the total cost (time/money/attention) of personalizing software continues to come down, more people will be able to encourage a diaspora of options, tailored to each niche.
This is one of those types of app that really benefits from a proper menubar with several top-level menus so the user isn’t stuck spelunking submenu tunnels.
Quick grammar correction can also be nice. Not super important, but nice.
yeah, that too.
But these MUST have a local option first and foremost, this is something that any local model can do, even the smallest ones.
But even before that there would need to be a set of APIs from the DE, imho. I think that GNOME is going into this direction with their AI center, i whish KDE would start looking into that too
They will probably all be abandoned within the year.
Linux really need applications more polished, that’s what they are bringing to the table.
So I had Claude whip up a super fast image browser that had some features I really like from a Mac app I use. It looks better than 90% of Linux apps, is super fast and fits me like a glove.
Of course some of these mail apps will be abandoned, but they also provide an opportunity to make things more attractive on the Linux side and win some users.
Talk about a low bar ;)
But yeah, thanks for reminding me; I've always wanted to make my own file manager and image viewer, since literally every single one out there is garbage. How is nobody able to get such basic programs right is beyond me.
Linux mail clients tend to look how the desktop toolkit look.
What is wrong with how kmail, Geary, Elementary-mail or Evolution look and behave?
Mail client has been a solved problem for decades already.
None.
All of them slowed them to a very painful crawl.
I compared for instance evolution and outlook, evolution has always had much faster searches.
Having said that, having a bit of digital hygiene goes a long way.
This could be a proclamation for ALL vibe coded software. As a matter of fact, I think this is what marks software as "Vibe Coded". It's fun while the vibes last. The first few GitHub issues bring excitement because, "hey! people are actually using my project!". So the author prompts a few bugfixes or features. Then the GitHub issues become unmanageable. The author doesn't really KNOW why things aren't working. They can't find anything in the codebase without grep (this is true for the agents too). After awhile, it's not a good vibe anymore.
Also, I'd like to know what "after a while" means, concretely. Seems to be working out all right for me on the 4 new apps I've started that I wouldn't have time for otherwise, thanks to AI. I'm at 184 revisions in one of the projects and it works fine so far.
- can’t choose a sender email?! This is major dealbreaker, if you have many aliases or domains.
- no favico for emails, this is also a deal breaker for me, as I used it to quickly skim and find or check new emails.
- folders orders is missed up, trying to arrange them made it worse
- trying to color a folder and it says “server doesn’t support that”
Also the UI needs more polishing overall.
I'll stick with the web UI in that case. Are you using any other clients on Mac or Linux?
Ugh, this drives me nuts. It's the same reason why I stopped using my Supernote Nomad for email.
I use iCloud+ email with a custom domain. With that arrangement you authenticate to Apple's servers with your iCloud "primary address", like someone.somename@icloud.com. Then you can send and receive email with your custom domain email address, which is the one you're actually using. But not with Supernote, or apparently Penguin Mail. My Nomad will only send mail with my @icloud.com address, even though that's only used for authentication.
So yeah, I just stopped using it altogether. That's not my address. It's not how people reach me. It's literally just a username.
I wouldn't say it's something that can be vibecoded in a month or two. Proper design and implementation take time if you want to handle accounts with hundreds of thousands of messages seamlessly.
These days with the convenience of having LLMs at developer’s disposal, it feels unjustified to use electron like apps.. whenever I see “one app all/multi-platform”, I feel discouraged to give it a genuine try.
I tried a bunch of email clients and then crawled back to mail.app + Calendar.app on MacOS simply for how native it feels.
Interesting. TIL about this. I added myself to the waitlist.
Also it is made by pivotd ? What is it ? What is the end game ?
Genuinely curious as I am in the mail space with klar.im
For people that want a terminal client with tabs, integrated html viewer(including images), markdown compose, MCP, with unified inbox, I built notmutt.
https://fishman.github.io/notmutt/
But apart from given what I've seen from my experience with handling maildir, actually implementing mail sync, storage, and all the other things associated with mail AND calendar, it sounds risky to use a project this young.
https://news.ycombinator.com/item?id=49990805
That’s about a completely different client: https://news.ycombinator.com/item?id=49990103
Perhaps you’ve accidentally replied to the root thread instead of that comment?
Sorry for the off-topic, but Google (and all ad networks) should be legally required to vet their advertisers. That's gotta be the worst I've seen.
[1]: https://www.youtube.com/watch?v=0PyJCsw1FSE
[^1]: https://www.bloomberg.com/news/articles/2026-08-27/-locker-k... [^2]: https://www.aa.com.tr/en/europe/polish-minister-accuses-meta...
Ublock is still a thing. Or use Brave as thats got baked in ad blocking which can be combined with ublock to make it even more effective. Then adguard or pihole on your local network so no analytics/tracking data is still being pushed out.
I use Penguin as my daily driver, and I made it exactly because of what I was missing on other mail clients, but idea after idea brought me here to share it with everyone, my intentions are to continue development and not keep it stale or let it die, there are loads of things i want to add but I also don't want to lose any stability, a nice balance is what I'm looking for
https://wccftech.com/customer-buys-32tb-external-hard-drives...
On the shop itself that you linked though I don't see any unreasonable price (or any SSD for that matter). Just normal TEMU-level stuff with appropriate prices.
Unlike slapping an AI agent on top while making zero promises about its capabilities.
They argue that just as programmers once wrote ASM and C, the arrival of Java and Python meant you no longer had to be a neckbearded autistic kid to write software that doesn't leak memory or segfault.
Arguments like this show me you’ve never done software engineering seriously.
Despite compilers (written by some of the smartest engineers on Earth) trying hard to prevent developers from writing shitty software, those new age "engineers" still managed to ship software that crashes, leaks, and generally sucks.
I'm not saying this email client is not secure (although I'd bet 5$, it is). But it seems beyond silly to assume software is automatically safe just because a new piece of tech can potentially find vulnerabilities. Intent still matters. And the 99% would never have the intent to write secure software.
This normally happens on two scenarios: a) they are not paid enough to care that much or b) they are paid more than enough to not have to care that much.
Either way it’s almost always down to how much the company is willing to pay for people that actually care about these things.
… but as shown by many shovel sellers… their shovel…
Here I fixed it.
Other interesting problem to solve for mail clients is the editor. You should be able to forward/edit untrusted html content. Source: I develop tecotype.com for mac/linux/windows.
For example, I'm using catch-all address with fastmail. I wasn't able to find any client with the right design that would allow me to define sender address and email account separately during composing, so I can use "alias@example.com via me@example.com" as a sender.
You can send "ping me when linux available" and I'll follow up.
If I can be a beggar and a chooser: The vim keystrokes to go to the top or bottom of the current email list and/or the conversation thread items are sorely missed. 2. If there's a way
I did that in my IUP fork, bold, italic, headings, link, image, table, that can all be mapped to native text widget, can load and save markdown from that.Will it look as intended, no, but, do not create websites when you send emails and everything will be fine.
Should be TUI-based and for Linux + Mac.
Viewing: I have a message or messages stored on disk in files or mboxes or a maildir. How do I render them? Example use case is evidence in a law suit or, more likely, emails I’ve put in a shared drive to help me and my partner argue with our landlord. Yes, I could just have a Gmail folder for these but I don’t want to have one foot in a file-filesystem and one in a mail-filesystem. I want “lease_2024.pdf” and “20240401_deposit_confirmation_002.eml” to both be files. But then of course we have…
Filesystem interface: Emails have important attachments with MIME file names. Or perhaps a more keen observation is that some named files have emails as attachments! I’d like to symlink into a virtual FS that keys files off a UUID / Message-ID, and be able to mount that virtual FS from any mail store I have available (Gmail API, JMAP, IMAP, .tar archive, mbox, maildir, or .eml files.) Most conversations I have with people don’t involve files, and most files don’t involve conversations, but when they do I want them to feel natural and not some kind of special category of filey-messagey thing that doesn’t play well with plain files or, less commonly, plain messages.
Composition and sending: Creating a new message involves making a new file and then using that file to drive a sending API. I want to compose a beautiful email using a tool that is good at that job. Maybe I’m using vim to draft it in markdown with macros and tools to machine build certain complex parts? I also need some mime aware tooling — sometimes — to help prepare a bunch of attachments that may or may not have names related to their names on disk! I want all of this to be scriptable, on the occasions when it’s complex (30 part message for a client’s loan application!) and easy when it’s ad hoc (to: mom, see cat.jpg).
Search: email is notoriously the extension of and replacement for my long term memory. I have gathered message history from many sources and accounts over the years. I need to manage these into packed archives for the old ones, editable archives for the current ones, promote the former to the latter, and arrange new messages into the system. This feels like another separate tool.
I’m sure there are other potential tools lurking here as well. Separate rendering, composition, sending, receiving, searching, and archiving, feels like the goal. Most new email apps I see try to implement all of the above in one new package when, really, it would be great to see innovation on each branch of email as separate tools.
My PDF viewer doesn’t need to innovate on file browsing! My SVG drawing tool doesn’t need to innovate on mounting remote filesystems via SSH!
KindOfCoders, AI is a very useful tool, but nothing works like the way you're explaining. I am kind of finding it hard to articulate the problem, because the assertion is just a bit wild. So I will do my best.
Thunderbird has been in production use for over 20 years, many many smart people have put their time into it to make sure it works correctly, yes, it is a bit clunky, it has a lot of craft, but you're not going to replace that much labour with running an AI agent for a few hours; despite the AIs usefulness in finding bugs and vulnerabilities. Not yet at least.
I assume what KingOfCoders suggested is to run an agent almost like an antivirus which I guess sounds somewhat interesting but you'd have to re-run it on every update and probably spend many more hours+tokens for such a system to work properly.
But perhaps once inference gets truly cheap repository maintainers should include AI security checks for new software as part of their pipeline? I think that would make sense. Especially for repos that are frequently pwned like npm!
Like the 10 years 0-day kernel bug?
"There are no black swans because I haven't seen one"
Ugh. I thought at this point we'd be over this. Yes, claude can build any kind of application.