Thunderbird Making Progress With Adopting Rust Code

[elbar]

they went for Android with TB, but think that it is delicate to deal with these kind of cards and market went away. But with their budget they are limited and Mozilla Foundation throw TB away a little time ago with expected sink but their PM(that is not prime minister maybe want developers that may think that a little big adventure is making space progress and fight with their enemies might be a lost thing before their begin their bird journey... 🗽 with something from next Olympics place...

[DumbFsck]

When you're unable to tie your shoelaces yet, Rust won't help, it will hinder you, because it's difficult itself.

Yeah, no.

If your architecture is completely fucked up and to wrangle the spaghetti is taking much more effort than you feel is justified for the feature you're implementing, and already in the pipeline there is the realisation that a major code rewrite will take place, rendering your efforts obsolete and in tandem making it easier to implement the feature after the rewrite, I don't think anyone ever would make the call of implementing said feature.

The Thunderbird crew says the new architecture debuted with "supernova" is meant to make it easier to add, change and modify features. That before this major rewrite it was difficult even to maintain thunderbird let alone meaningfully improving it.


Also, "learning" rust is not exclusionary to improving thunderbird. One of the speakers in this talk, Ikey Doherty (in my opinion a great programmer, also a great snake oil peddler) learned rust (is learning rust) after writing a lot of tooling for Linux distributions in Dlang and finding it insufficiently mature. Note that all of this tooling he is writing is done outside of his employment working hours, so it does not take away from adding stuff to Firefox.

One reason for him to change from D to Rust was also that other team members were already familiar with the language, and soon after changing it they already went out and made huge improvements to performance right after arrival.

Maybe there was something similar in his thunderbird team?

In any case, this sounds a lot like when gamers complain if stuff like "hurr they are still releasing skins and cosmetics instead of fixing the terrible networking", because obviously the artist making horse armor is the network engineer.

What you said isn't as drastic, but is in the same vein.



Btw, I like how you said easy to master for sarcasm. I agree whole heartedly that Rust is easy to learn and a good programmer should be able to hit the ground running real quick, especially with how much help the compiler gives you. So for irony only "master" works, really.

[Daktyl198]

Thunderbird hasn't been with Mozilla in quite a while.

Edit: Seems that was wrong, it's still developed by some Mozilla subsidiary.

It was dropped by Mozilla for a few years, then picked back up once the community was showing promise. Now the Mozilla foundation "manages" it, but provides no funding, no developers, nothing. They just like having their name associated with an active project.

As for Thunderbird itself, it'd be my go-to mail client if only it could minimize/close to the system tray.

[Saverios]

As for Thunderbird itself, it'd be my go-to mail client if only it could minimize/close to the system tray.

Have you tried SysTray-x? It's been working very well for me for the past few years and pretty much does what the original Thunderbird tray indicator did, only with more options. Quite stable, too.

[Quackdoc]

I still find it hard to use, kmail has a lot of annoying things with it (If you read an email under one category (IE. "Important"), it won't market it as read under "Inbox") But even then I much prefer it to thunderbird.

[mb_q]

Or maybe their true Netscape burden is the idea that Thunderbird must be founded on this firefoxy-electron abomination. For Atom/VSCode it makes some sense to allow for huge plug-ins that completely rewire most functionality, but TB is just an E-mail client that doesn't even have a web-only edition to share its code with. Also, Electron is a platform, which gives it some additional flux of attention, maybe TB should work on decoupling something similar instead of fighting with Outlook in the Microsoft-powered corporate spaces?

[jacob]

Imagine if Wine project was rewritten to Rust.
Not blaming Wine developers, if a project have lived long enough there is bound to be forgotten code paths.

Still.. that would be an impressive feat and live up to the meme..

Wine must support Microsoft's C ABI, so not sure if a RIIR would be feasible or, even if it was, whether it would bring some actual benefits.

[darkonix]

Wine must support Microsoft's C ABI, so not sure if a RIIR would be feasible or, even if it was, whether it would bring some actual benefits.

What would be the problem? Do you see a technical limitation?

[Sonadow]

I still have some gripes (relative text size settings vs. absolute numbers...)

The single biggest reason not to use Thunderbird or Gmail ever.

[jacob]

What would be the problem? Do you see a technical limitation?

If the whole code was something like:

unsafe {
  call_c_function()
}

then there is little point in using Rust for this.