Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Proposal to start a new implementation of Thunderbird based on web technologies

mail.mozilla.org

Post reply on HN
Print view View original

71–80 of 243 posts

Re: Proposal to start a new implementation of Thunderbird based on web technologies

#71

Post by jcranmer »

js8 wrote: post #32
cakeface wrote: post #11

Does anyone actually have any problems with Thunderbird? I feel like it is "done" software. It works, perfectly, for all my needs. It's fast. It has great UI. It lets me read and write email.

I haven't used Thunderbird for a long long time, but the feature I was missing the most when I moved to Gmail ages ago was some sort of automatic categorization (machine learning) based on content of the emails. One application would be spam filtering, but I would actually appreciate if it could be applied more broadly to classify other emails. Is something like that now available in Thunderbird?

Spam filtering has been in Thunderbird for a decade now (based on a Bayesian classifier, although it's not quite naïve Bayes). There is a capability to extend the classifier to problems beyond spam/ham classification for extensions.

Re: Proposal to start a new implementation of Thunderbird based on web technologies

#72

Post by pmontra »

Kenji wrote: post #29

I wish I could freeze the program Thunderbird in time to keep it precisely how it is right now for the next couple of decades. I literally want not change. I know, I know. It's a pipe dream, the surrounding environment changes, breaking the program as the OS and libraries evolve. But this constant resolving of perfectly solved problems annoys me.

It's maintenance. Sometimes it has to be somewhat radical.

I hope for two things.

1) Drag attachments to folders finally works on Linux.

2) Message filters keep working exactly as they are now or it's going to be a huge pain to migrate them. And a showstopper with no filters.

Re: Proposal to start a new implementation of Thunderbird based on web technologies

#73

Post by jcranmer »

tmzt wrote: post #63
Ono-Sendai wrote: post #56

Sounds like a terrible idea, please no. I for one probably will stop using thunderbird if it gets rewritten in JS.

What is it written in now? If I remember correctly it's mostly JS with a few native extensions in the chrome.

The backend is mostly in pure C++ (~500KLOC or so). The frontend is in JS (~200KLOC or so).

Re: Proposal to start a new implementation of Thunderbird based on web technologies

#74

Post by janci »

cakeface wrote: post #11

Does anyone actually have any problems with Thunderbird? I feel like it is "done" software. It works, perfectly, for all my needs. It's fast. It has great UI. It lets me read and write email.

The filter is slow in regular-size mailbox and almost unusable on my bloated gmail. I don't mind getting results progressively as the messages are scanned, but it runs on the UI thread and UI is horribly slow while the filter is working.

The search is just broken. It gives messages that don't have anything of search string in them and misses the right ones.

Re: Proposal to start a new implementation of Thunderbird based on web technologies

#75

Post by randomdata »

Certhas wrote: post #26
cakeface wrote: post #11

Does anyone actually have any problems with Thunderbird? I feel like it is "done" software. It works, perfectly, for all my needs. It's fast. It has great UI. It lets me read and write email.

Working with big folders tends to be a bit error prone to me. I would like to synchronize my filters/settings across different devices. On MacOs it randomly hangs for tens of seconds and eats batteries for some people (me included). So yeah, it's mostly done, but better virtual folder support, and a streamlined UI would both be hugely appreciated by me. But, because it's mostly done, I also will just happily keep usi…

Is a rewrite necessary to improve on those aspects over modifying the existing codebase?

Re: Proposal to start a new implementation of Thunderbird based on web technologies

#76

Post by khedoros1 »

gibbitz wrote: post #57
cakeface wrote: post #11

Does anyone actually have any problems with Thunderbird? I feel like it is "done" software. It works, perfectly, for all my needs. It's fast. It has great UI. It lets me read and write email.

I think the point is that to keep the project relevant the underling aging Gecko underpinnings may be going away. This would cause the maintainers to have to take on maintaining their dependencies as well as the core project. I can totally understand this as a motivating factor here, however, I use Thunderbird as a local offline mail client. Though I see these are out-of-fashion these days, it doesn't make the need f…

> I use Thunderbird as a local offline mail client.

I think that the intention is to keep it that way. They're proposing mostly a reimplementation of the UI, from Gecko+XUL to HTML+JS, but also a reworking of the interface between frontend and backend code (I guess between the model/controller and view parts of the code).

Re: Proposal to start a new implementation of Thunderbird based on web technologies

#77

Post by nkkollaw »

thesmallestcat wrote: post #62
nkkollaw wrote: post #60

They've already neglected Thunderbird for years, and now they want to start from scratch? Doesn't sound like a good idea. Sounds more like another project they will shut down after a while. I used to use both Thunderbird and Firefox a few years ago as my main mail client and browser. Since then, Mozilla started spreading between too many projects, chasing the failed phone thing, and both Firefox and even worse Thunde…

Thunderbird, like Firefox, can be themed, and many are available. Well, until they finish killing XUL... what an awful decision. Point being that it's a mail client for power users, and generally requires a little customization.

It's deeper than the theme.

It's the paradigm, the absence of automatic sorting of emails (priority/promo, etc.), failure of prioritizing common actions in the UI (everything is the same, the spam folder and the inbox folder are given the same importance, "compose" is not prominent, etc.). I could go on forever.

IMO, it's kind of how good software was made 10 years ago, but now we're used to much better.

Re: Proposal to start a new implementation of Thunderbird based on web technologies

#78

Post by hashhar »

pmontra wrote: post #8

The only point I could argue with is the choice of JavaScript, all the others seem inevitable. As a Thunderbird user I'm happy to know that I might be using it for another 20 years. Even JavaScript is not that bad. I not familiar with Electron: do those apps run from a directory with the source code or are a binary file? Running from source code would make it easy to hack with the code, which is good.

Thunderbird already uses a lot of JS. The only difference I think is that Thunderbird works on the XUL instead of the HTML DOM.

Re: Proposal to start a new implementation of Thunderbird based on web technologies

#79

Post by Certhas »

randomdata wrote: post #75
Certhas wrote: post #26

Earlier quoted context omitted.

Working with big folders tends to be a bit error prone to me. I would like to synchronize my filters/settings across different devices. On MacOs it randomly hangs for tens of seconds and eats batteries for some people (me included). So yeah, it's mostly done, but better virtual folder support, and a streamlined UI would both be hugely appreciated by me. But, because it's mostly done, I also will just happily keep usi…

Is a rewrite necessary to improve on those aspects over modifying the existing codebase?

Maybe not, but from the sound of it, a rewrite is necessary just to keep the lights on and to keep Thunderbird compiling in the long run.

Re: Proposal to start a new implementation of Thunderbird based on web technologies

#80

Post by thesmallestcat »

nkkollaw wrote: post #77
thesmallestcat wrote: post #62

Earlier quoted context omitted.

Thunderbird, like Firefox, can be themed, and many are available. Well, until they finish killing XUL... what an awful decision. Point being that it's a mail client for power users, and generally requires a little customization.

It's deeper than the theme. It's the paradigm, the absence of automatic sorting of emails (priority/promo, etc.), failure of prioritizing common actions in the UI (everything is the same, the spam folder and the inbox folder are given the same importance, "compose" is not prominent, etc.). I could go on forever. IMO, it's kind of how good software was made 10 years ago, but now we're used to much better.

Compose is C-n, and the "Write" button in the default toolbar.

Not sure what priority/promo is, but Thunderbird certainly can sort by any field by default. It sounds like you want a client that is tightly coupled with Gmail's features, and you're not going to get that with a non-Google client.

Note that some customizations, like Firefox, require accessing about:config. Also, as with Firefox, most Thunderbird users will benefit from perusing extensions. I wouldn't be surprised if there are some extensions that offer the more Gmail-centric experience you want.

Post reply on HN