Hacker News new | past | comments | ask | show | jobs | submit login
Thunderbird’s New Home (thunderbird.net)
298 points by sashk on Jan 28, 2020 | hide | past | favorite | 158 comments



I admit that software needs some sort of upkeep to not fall to bitrot, but I simply don't understand how we got to the near-falldown of Thunderbird a few years ago. You don't have to add new features, keeping it alive and simply fixing critical bugs is fine. Basically everyone I know would prefer a stable mail client not getting any new features over a few new features and addons breaking.

I'm really happy Thunderbird hasn't gone the way of the dodo.


From what I understand the challenge with TB is that it's tied into Firefox' engine and that is also in parts used for the GUI (XUL). Every time there's larger refactoring work in FF that affects Thunderbird. And FF wants to get rid of XUL eventually.


> Every time there's larger refactoring work in FF that affects Thunderbird.

Okay, that's scary. It's like a one half of a conjoined twin trying to remove the other half from their lives. Mozilla started transitioning Thunderbird out of their umbrella while TB is still highly dependent on Firefox code. Yeah, scary.

Maybe we need a successor to Thunderbird? Something that's not tied to Firefox so much? Because it seems like we can't trust Mozilla and however many unpaid/underpaid volunteers to maintain it well enough.


> Mozilla started transitioning Thunderbird out of their umbrella while TB is still highly dependent on Firefox code.

If you are referring to Mozilla Messaging and the period to 2012, no the only transitioning was in terms of product functionality. The community organized in 2014, unaffiliated with Mozilla, and it's not until about 2018 when staff were hired that community leadership were able to seriously reduce technical debt. But seriously reducing dependency on Firefox code is some distance away.

> Maybe we need a successor to Thunderbird? Something that's not tied to Firefox so much?

You are looking a millions of dollars paid developer time.

> Because it seems like we can't trust Mozilla and however many unpaid/underpaid volunteers to maintain it well enough.

Mozilla are not running the show here, the community is.

Perhaps it's difficult to keep up with the history, but https://blog.thunderbird.net/ has past relevant postings, plus developer and other mailing lists. To cite a couple, big changes in encryption https://blog.thunderbird.net/2019/10/thunderbird-enigmail-an... https://mail.mozilla.org/pipermail/tb-planning/2019-December... and development priorities http://lists.thunderbird.net/pipermail/maildev_lists.thunder... which includes a new address book.

There is a significant professional development staff which is nicely paid, and the number of developers will soon double. https://www.thunderbird.net/en-US/careers/


I really hope that my extensions are not broken. It would kill Thunderbird. Nobody is going to come along and replace the library of amazing extensions that have been built up over the years.

Unlike a web browser, email has not changed much in the last 15 years, so I don't see why development needs to advance at a high speed.

I doubt that any UX fashion statements are going to be enough to make up for the huge loss of functionality.


I thought XUL was gone with the recent change to the way addons work in Firefox?


XUL is no longer exposed to plugins (in non-ESR FF), but is still used internally in some places - where it is slowly being replaced.


XUL was used all over the place, so it takes a while to fully remove it. Thunderbird updates the underlying platform only once per year, catching up with Firefox changes in the yearly major release, while Firefox does so every few weeks.


That's not quite accurate. Thunderbird develops its software with the tip-of-trunk Firefox code at all times. It only makes releases much less frequently than Firefox.


It was XBL (XML Binding Language)[0] that was removed, not XUL - which is being phased out and replaced with HTML where appropriate.

0: https://briangrinstead.com/blog/xbl-in-firefox/


Anecdotally, I dropped it years ago specifically for the lack of new features. Namely, the lack of any sane calendar integration, since Lightning could integrate with Thunderbird, but couldn't integrate with any other calendars, and I wanted a shared calendar.


I'm not sure what you mean. I use Lightning with CalDAV, and it works fine.


Yup we here. Shared family calendar across Thunderbird and all family phones with CalDAV hosted on a radicale server set up like this:

https://partofthething.com/thoughts/host-your-own-contacts-a...


I'd be happy for non extension calendars and better exchange/o365 compatabilty But otherwise I'm very happy with TB as my main client.


I've found that davmail works flawlessly for connecting thunderbird to an exchange server.

However I have not found a way to connect it to O365. If anyone has figured it out, I'd love to know how.


The UX is that of the 90s. FF continues to see progress beyond the web browsers of the 90s, why shouldn’t Thunderbird?


Maybe, but the UX works. I'd be very happy if they didn't start tinkering just for the sake of not being "stale". (See: Gnome Shell.)

As a daily user for 10+ years I can say there is certainly room to incrementally improve TB in many areas, like calendar and contacts, but don't mess with UX success.


I used thunderbird for years, gave up, and then tried it again a couple years ago. I think the UI/UX is really bad and was the reason I had forgotten why I left, and it was the reason I gave up the second time.

I unfortunately can't remember my complaints to make them more specific for a lively discussion, but I guess I am an anecdotal counter-point for thinking the UX was actually not a success.


I'd say a bad UI that is stable and familiar to the majority of users is actually a better UX than some new thing. See the MS Office "ribbon" interface for a good example.

Car manufacturers don't try to reinvent the steering wheel & pedals UI every couple years.


That is basically the motto of the Bloomberg Terminal, so I would agree with that to some extent.


Search is really bad and every time I use it I feel something is now shown (and what IS shown is totally irrelevant and old).

Also, I see that a dark mode would be a requirement today as everything else either complies with OS's dark mode setting (Windows 10, OS X) or has it's own dark theme.


People say here that the UI is bad but nobody says what exactly is supposed to be bad about it. Or is it just enough that it doesn't look like the blown up UIs that are in fashion today?


The icons are ugly, there are too many boxy sections on any given view, the top left pane in the calendar provides no useful information and is ugly. There is too much noise and clutter, it could be simpler and cleaner. I use it every day, but I wish it was less ugly.


I gave up because the search was inexplicable Reddit level bad.


Personally, I prefer the 90s UX over the stuff we see these days.

The modernization of the FF UI, combined with the reduction in the ability to customize it enough to make it the way I want, is one of the reasons I stopped using Firefox (not the main reason, though!)

Whatever TB does, I hope they at least allow people the ability to modify the UI enough to make both of us happy!


> Personally, I prefer the 90s UX over the stuff we see these days.

Seriously, can I just have frames back instead of the "our whole site is one HUGE page" trend.


It’s hell adding recurring events in Thunderbird that only happen at non-standard intervals.

I work a rotating six week schedule with different workdays week to week. In iCal/Apple Calendar, I can intuitively copy/paste events to a different day of the week if the rotation changes. Thunderbird, with its 90s UI, doesn’t provide you with this option.

I’ve found refuge in eM Client on Windows, but it took a while before I came across the app.


Well comparing Apple Calendar with Thunderbird is not really fair. One is dedicated calendar app the other is email app.

I use thunderbird for email and calendar for calendar. When i compared Apple Mail with Thunderbird i found quirks small tech issues with Apple Mail especially when using something else than gmail.


I dunno, I see a borderless window, tabs, a settings page with big colorful icons representing each section, smooth fonts with DPI scaling, even a hamburger menu! Oh and it seems to happily support resizing the window larger than 800x600. Don't remember any of that in the 90s.


Don't fix what aint broken.


I don't mean to pile on, but could you please explain this a bit more?

Why should the baby be thrown out with the bathwater, just to chase cosmetic trends?

Can you give a compelling technical reason why this would be good for users, without using superficial UX fashionista platitudes?

Just because something isn't "on trend" doesn't mean it is bad. Some of the most important and beloved software has user interfaces that have not changed in decades.

I don't care if a UX designer would be embarrassed to use it in front of other mail clients. Trying to chase trends will usually just leave your software looking dated, anyway.

Unless you're willing to roll up your sleeves and keep existing functionality with your new trendy UX, then don't bother, because deprecating real functionality at the cost of a superficial fashion statement is really egregious and it needs to stop.

Nobody wants their workboots replaced with tennis shoes. And especially not because a UX designer wanting to stroke their ego and make a name for themselves.


Well firstly, having a great UX doesn’t mean throwing away features. It means improving the product you have so it’s more user friendly, increased readability, and more powerful. It can even mean increased features like better thread management, snooze support, better people-centric management, additional workflows explicitly supported (like zero inbox), etc.

Most of the sibling comments here are a “get off my lawn” type variety that IMHO are not fully informed by what good UX means. Great design isn’t superficial, in the least. A great user experience is a great user experience... it doesn’t mean compromising because something was hard, and it doesn’t throw away how the product is used or has evolved.

We should be aiming to improve and innovate to be helpful, not stagnate. And improvement means people’s existing needs are considered, not disregarded as many here seem to think.


> Great design isn’t superficial, in the least. A great user experience is a great user experience... it doesn’t mean compromising because something was hard, and it doesn’t throw away how the product is used or has evolved.

I don't think anyone disagrees with that. The complaint I am making, is that a lot of UX design is superficial, and done for the wrong reasons. Most of what Mozilla has done is terrible.

They've essentially gutted their browser because they want to be cool like Google.


But if we’re talking about ideals, then why wouldn’t you advocate for change towards great design versus stagnation? Now if you don’t trust Mozilla to pull it off, then that’s a different argument than so far has been made.

While I don’t personally know the designers at Mozilla, I’m doubtful that they simply “wanted to be cool like Google” and instead did their own research. I could be wrong but usually work isn’t so arbitrarily done for large products and teams.


I mentioned this in another thread... but Mozilla/Thunderbird (not MZLA) should really concentrate more on feature-rich support. Specifically better calendar and provider integration.

Second to this, they should probably work on their own first class mail and calendar server as at least open core, and probably as SaaS as well. As much as I like some of the mail hosts out there, for some, the ability to self-host is big. The various pieces and software you have to cobble together are cumbersome, and even then, you don't get anything close to what Exchange+Outlook gives you.

It would be nice to see Thunderbird fill this space, it used to be what I would consider a best of breed email & newsgroup client, now, I'm not sure I would say anything close to that.


> Second to this, they should probably work on their own first class mail and calendar server as at least open core, and probably as SaaS as well.

I quite disagree here. Servers and clients are very different in their architecture and implementation, and you are going to have very little opportunity for code reuse or even knowledge sharing between the people implementing the client and those implementing the server [1]. I don't think you can turn email servers into a profitable revenue stream unless you do full SaaS, and I don't think Mozilla is really capable of putting in the resources to make a Mozilla Email SaaS feasible.

A better bet would be to try to put together something to more effectively cobble together the existing open-source email server components and do the installation and support as a service, but I again don't think that Mozilla is the organization best-placed to provide that service.

[1] The big exception here is LDAP. But Thunderbird doesn't even really maintain its own LDAP library--it should move to WinLdap/OpenLDAP depending on OS.


> Servers and clients are very different in their architecture and implementation

True, but also the only way to really know how to develop a good one is to have experience with building the other.


A near 100% drop-in replacement for Exchange server is kind of the holy grail of messaging/collaboration. I was looking at something that claimed to be that (or close) many, many years ago when I was doing consulting, so I could try to save my clients some money. But nothing had all of the features they wanted or was as seamless as they wanted it to be. I don't remember what the software was called, but it pre-dated Zimbra (which I think I've installed once to dink around with). I think this is more difficult now, with the move to cloud services like O365 and Google Suite. But having a self-hosted, Outlook compatible, Exchange substitute would be great.


That has been my experience as well... these days Office365 is often the best option, with Google Docs/Gmail, Fastmail and a handful of others coming close. I'd generally favor o365 currently, but would love to see Moz deliver a great alternative. They definitely have most of the expertise to do so (even if UI/UX could use some improvements).


> I was looking at something that claimed to be that (or close) many, many years ago when I was doing consulting, so I could try to save my clients some money. But nothing had all of the features they wanted or was as seamless as they wanted it to be. I don't remember what the software was called, but it pre-dated Zimbra

Ximian's Evolution?


Ran Zimbra for a number of years and it was fine as an exchange replacement. Only reason we moved is we wanted SaaS and didn't want to maintain it anymore.


I think Thunderbird resources are best spent being the best email client it can. It currently lacks mail merge support, and mapi support is bad/not-working. There is an add-on for mail merge, but as they go to webextensions its future (and that of many other add-ons) are uncertain. This affects other add-ons such as enigmail too.


I would like if they focused on performance and stability of current features, e.g. by improving their IMAP code and moving to SQLite storage to improve stability and fix corruption of mailboxes.


I think that would be great too. For now they seem set on maildir as the future of mail storage.


Absolutely, I agree 100%. Instead of trying to be all things to all people, I think they should stay focused on being a great email client. Based on the comments on this page it sounds like there’s certainly enough work to go around just with that goal.


Second to this, they should probably work on their own first class mail and calendar server as at least open core, and probably as SaaS as well. As much as I like some of the mail hosts out there, for some, the ability to self-host is big. The various pieces and software you have to cobble together are cumbersome, and even then, you don't get anything close to what Exchange+Outlook gives you.

I would love for someone to do this, so we could have a nice, self-hosted store for our vital data but still access that data from all our devices, over VPN, etc.

The problem, as always with these things, is who that someone would be. While many HN readers might appreciate the potential advantages, I fear the target market for such a product is significantly smaller than for Thunderbird, which for better or worse is itself aiming at a relatively small minority of users in the era of Google Mail and friends.

For anything it's worth, I'd contribute financially to such a project personally and so would my businesses, since it would be a significant benefit in each case.


I mostly agree... I do think it's more of an "if you build it" scenario. Thunderbird isn't big as a mail client, because there isn't the level of integration with either existing providers and/or a self-hosted option.

o365/exchange, gmail/gdocs/gcalendar and fastmail are probably the big three providers that should have seamless integration at this point. A SaaS option from Moz would be nice, as would a self-hosted option. I bring these up in line with Moz's efforts to get VPN subscribers, they could likewise do groupware.


> MZLA

Why's Mozilla got a stock ticker if it isn't publicly traded?


It doesn't, that's the name of the corporation taking over maintenance of Thunderbird listed in the article.


Oh right - but I looked up https://stocktwits.com/symbol/MZLA and it's from 2010 so I don't know if that's the new company but it says Mozilla.


Given that this is Stocktwits, I assume the page you linked was auto-generated after enough people used the cashtag ($-followed by Symbol) on Twitter.


It's the name of the new company meant to steward Thunderbird according to the post.



I have in the past... now how/where do I sync this calendar, and does it allow me to share read-only data with other people?


I really like that second idea, although it might be more productive as a partnership with an existing solution like Open-Xchange, Zimbra, or Citadel.


Does this mean I need to shift my donation?

I've been giving specifically to support Thunderbird development, not other Mozilla work [0] for the past few years because I think it's vitally important that email remains an open protocol, supported by mainstream desktop clients. Without Thunderbird, I fear email will be lost to the Gmail & Apple walled garden.

[0] https://give.thunderbird.net/en-US/


Based on this Twitter discussion, it seems necessary to update the donation. Possibly because the recipient is no longer a nonprofit (previously Mozilla Foundation), but now a corp (MZLA). https://twitter.com/mozthunderbird/status/122222824579547545...


Thunderbird community manager here. This is correct. Moving from the Mozilla Foundation (non-profit) meant we had to cancel recurring donations.

We hope everyone will resume via https://give.thunderbird.net


There's no continuing option to give to the Foundation and direct support to Thunderbird? Since the new subsidiary is for-profit, direct donations will no longer be tax-deductible.

Feels like an odd choice (and perhaps, an important omission?) when the announcement calls out the importance of donations to Thunderbird.


I doubt it's legal for money to flow that way?

You'd be donating to a nonprofit and getting the tax benefit, then the nonprofit would be passing it onto a for-profit corporation. End result: tax-deductible donation to a for-profit corp. That doesn't sound like something the IRS would like.

It sounds like the ability to directly generate revenue was seen as outweighing the benefit of allowing tax-deductible donations. (This is my own speculation.)


Consider it this way: if a non-profit pays a for-profit company for services rendered (pure expense on the balance sheet), are they "passing on" your donation to the for-profit? I would think not. Non-profits need to be able to pay for things. At the most basic office-management level, they would need to rent office space, buy paper and coffee filters, contract a janitorial service, etc.

Would there be a limit? I don't think so. If I was running, say, a non-profit website host, 100% of my raised donations would be going, as expenses, to a for-profit company (some commercial cloud-hosting provider.) That's still an entirely-legitimate model for a non-profit to operate under.


I'm still looking for help upgrading/maintaining the Lookout add-on. This add-on allows opening the proprietary Microsoft Winmail.dat files

I don't have as much free time as I used to, that and Thunderbird is removing legacy add-on support faster than I would like.

https://github.com/TB-throwback/LookOut-fix-version


I switched back to mutt recently. I've tried to use and like Thunderbird, but I just can't make it work consistently across platforms. I really think mutt is the ideal mail user agent. If TB would just become a GUI for mutt, it may be useful to me, but otherwise, it's just not stable nor consistent.


I found mutt's ui difficult to use, but the main problem was that a significant part of mail I receive is meant to be rendered by a browser, for better or worse (worse). Can you say a bit more about your case?


Yeah, I'd almost give a kidney for a client with mutt's power and Thunderbird's rendering.


And Thunderbird's configuration system. I tried getting back to Mutt a couple years ago, but gave up configuring it after a few hours. I believe the biggest pain points were multi-server configuration and enabling as much caching as possible.


I, too want this. My ideal is actually mu4e, rather than mutt, but the idea is the same.

I think what is truly optimal is to keep the CLI nature of the MUA, but have a slave browser window that displays the HTML portion of a message. Then you can tile your terminal to one side of the screen and the browser window to the other. Use the already excellent UI to browse mail and compose mail. When you encounter a message with an HTML part, render that in the slave browser window and render whatever text part exists in the terminal.


Maybe MailMate?


I think it’s good to put a bit of separation between Thunderbird and the rest of Mozilla. Thunderbird may have a modest but sustainable future of its own, and ideally it shouldn’t be held hostage to the broader challenges and shifting priorities at the parent company.


Will this affect development any? Thunderbird recently broke my "Virtual Identity" extension, which I've been relying on for years. Not sure what to do now.


Current thunderbird lets you customize your from address without the use of extensions.


It doesn’t work well because you have to remember and do it manually every, single, time for every message to each recipient. V.I. would keep track of that for you.


> It doesn’t work well because you have to remember and do it manually every, single, time for every message to each recipient.

I'm having the very same problem, but never knew there actually was an extension for that - though it's a bit disheartening to learn that at the same time it no longer works in the latest Thunderbird version. Hm... (There's also https://bugzilla.mozilla.org/show_bug.cgi?id=1518025 to port that functionality into Thunderbird itself, but it seems to have possibly stalled a little?)


Ultimately, Thunderbird needs to stay up to date with Gecko for security reasons, or else drop support for third party HTML content, so even though it is nowadays officially independent from Firefox, changes for Firefox will impact Thunderbird for the foreseeable future


Yeah, Thunderbird broke a few things that were critical to me, too. As a result, I'm staying put on version 52.9.1. If that becomes unsustainable, then I'll have to figure out a different solution.


Same here, I'm locked at 60.9.1 because the new Extension redesign broke the only extension that's a must-have for work, ExchangeCalendar: https://github.com/ExchangeCalendar/exchangecalendar/issues/...


> I'm staying put on version 52.9.1

You are 18 months behind on security updates for an internet-facing program. You are vulnerable to RCE, as CVE-2019-17026 is actively being exploited (among others, as the most recent example).

Is your email address mentioned anywhere online? Public Git commits? Do you ever see spam in your client?

In 2020 it is unfortunately simply not valid behaviour to use old versions of software,


Yes, I'm aware of that.

Given the way I use Thunderbird, none of the disclosed vulnerabilities affect me, so I'm not terribly worried at this time. Should one appear that does affect me, I'll reevaluate what I'm doing.

> In 2020 it is unfortunately simply not valid behaviour to use old versions of software

How true that is depends on the software and your use of it. Should the old version of TB pose an unacceptable risk, then I'll move on to something else.


A bit overstated. It’s not a server and that flaw is firefox’s, likely won’t affect TB due to html sanitizing.


I don't think it matters whether your PC is a server, attacks using publicly known exploits are generally indiscriminate.

Re: CVE-2019-17026 - Mozilla themselves state that it affects Thunderbird, but you're right, these JS issues might not be possible to exploit with just a malicious email as the attack vector.

So how about CVE-2019-11703, CVE-2019-11704, CVE-2019-11705 (aka MFSA2019-17)? It's a serious issue that gives RCE in Thunderbird 52.9.1. Is this issue reachable from an *.ics attachment, or will something happen without interaction if you have Lightning installed or receive a malicious calendar invite? This is still quite new and the bugzilla is still private.

CVE-2019-11713 gives RCE if any embedded images are reached over HTTP/2, that seems very plausible. At least you have the defense that images aren't loaded by default. Do you ever load images in emails?

CVE-2018-18500 is an older RCE but unpatched as well. This one is really interesting - I don't know if custom HTML elements are rendered in emails, but this issue happens in the parser before rendering so it's probably exploitable, the HTML is always parsed at least.

Less interesting but still plausible are the ones like CVE-2019-11740 and CVE-2019-11709 and their ilk, there's a new one of these with every Firefox platform update where nobody has (publicly) researched the details, but the patch is there for reversing so given the economic calculus I'm sure someone out there has put the effort in.

And finishing off with CVE-2019-11707 and other Javascript things like that, if you have any extensions installed that can be convinced to call the affected function in an exploitable way. If you're stuck on 52.x it's likely for extension reasons, and it's also likely your extensions themselves are no longer getting updates and have their own attack surface.

Having looked into the list of CVEs more closely, this is an impressively short list for a 18-month old version of any program (skipping over all the XSS issues and sandbox escapes). But I would give up extensions in a heartbeat versus trusting my personal email account to any one of these.


I do not allow HTML to be rendered, nor do I allow Javascript. Email needs neither of those things. Nor do I ever allow images to be rendered in email, external links to be followed, and I don't click any links in an email.


> explore offering our users products and services that were not possible under the Mozilla Foundation

Can anyone offer insight on what this means? My assumption might be some sort of for-profit endeavor?

> Ultimately, this move to MZLA Technologies Corporation allows the Thunderbird project to hire more easily, act more swiftly, and pursue ideas that were previously not possible.

Not sure how a different legal structure would solve what sounds like an organizational problem.


MZLA Technologies is a subsidiary of Mozilla.

https://en.wikipedia.org/wiki/Mozilla_Foundation#MZLA_Techno...

I'm interested in more clarification on what this change "actually means" as well.


A corp has more flexibility in what it's allowed to do, compared to a US nonprofit. Maybe similar reasons why Mozilla has the Mozilla Corporation in addition to the nonprofit Mozilla Foundation.


As the Community Manager for Thunderbird, I can say that this comment is correct.

The non-profit Mozilla Foundation had many rules about what we could and could not do. The new Corp gives us more freedom, legally.


Freedom to what though? I'm at a loss why a structure that's OK for FF, is not OK for TB.

...unless there are plans for profit and charging a subscription or some such.


Firefox is under Mozilla Corporation, which is owned by Mozilla Foundation.

Thunderbird will now be under MZLA Corporation, which is owned by Mozilla Foundation.

Why not put it pack in Mozilla Corporation? Because Mozilla Corporation has tried to jettison it a few times for not aligning with their goals, I guess?


Rules derived from where? Could you give us an example of what the new structure allows that couldn't be done before?


> pursue ideas that were previously not possible

That's uncomfortably vague.


Why not a Thunderbird Corporation or similar, instead of another confusing 4-letters corp?


I'm assuming other projects will eventually move to MZLA


What problem is solved by creating a second corporate subsidiary? This post doesn't really explain.


To keep money streams and decision making for Firefox and Thunderbird separate?


So it can be spun off more easily?


Presumably MZLA is not a 501(c)(3) organization (a Delaware C Corp perhaps?) and can do things that a 501(c)(3) can't, even if the corporation (MZLA) is owned by the 501(c)(3) (Mozilla foundation, in this case).

Hopefully this doesn't mean we end up with pocket on our thunderbird home screens.


I came here to ask the same thing. What were the disadvantages of using the Mozilla Corporation? Will Thunderbird be the only offering in MZLA Technologies Corporation?


God I hope this goes better than Mozilla Messaging. I nearly joined MoMess, and was very sad to see where it went.


Yeah, it really feels like history repeating. For those who don't remember:

"Mozilla Messaging (abbreviated MoMo) was a wholly owned, for-profit subsidiary of the non-profit Mozilla Foundation. ... Its main focus was developing Mozilla Thunderbird, the e-mail client developed by the Mozilla Foundation. It was spun off from the Mozilla project in 2007; on 4 April 2011, it was merged into the Mozilla Labs group of the Mozilla Corporation." (https://en.wikipedia.org/wiki/Mozilla_Messaging)


> MoMess

Nomen est omen?


Indeed.


An email client from the 1990's moves across two subsidiaries within the same company. Tada


Not Sure if Brendan Eich is Reading, wondering if he has anything to share.

You need to setup an Entire Separate For Profit Company for a different product?


Brendan Eich left Mozilla in 2014.


Thunderbird is a great product, love it. Is there a way to follow Twitter users in "Blogs & News Feeds" ?


Nitter


Thanks I have started using it.


I like the ENGMAIL plugin for GNUPG support. If I want to send encrypted messages to my friends I can do so with it.


> More information about the future direction of Thunderbird will be shared in the coming months.

Why ? Months is an eternity on the Internet.

edit: aye, aye for the downvotes, that's what karma is for, but would someone please explain why more information about the future direction of Thunderbird can't be shared now ? Does it mean there are no plans at all at the moment ?


Anyone know if/when Thunderbird will start providing native support for Gmail labels (rather than using the less than satisfactory IMAP folder system that is currently used)?


Ideally, never. Tying an open platform into features that exist only in one SaaS offering from an advertising company isn't a good way to keep it open.


I understand your point, but gmail labels are such a good idea that Thunderbird should implement them as a native feature (even perhaps proposing it as extension of IMAP standard).

Also think Thunderbird should implement JMAP (Fastmail standard) and/or Google RESTful gmail API because these are open standards worth supporting.

Another feature that I would like to see more fully supported in Thunderbird (through better synchronization with mail provider server back-end) is editing emails, to include most importantly the ability to remove bulky attachments without losing email text or touching important meta-data. Unfortunately gmail treats each email as an immutable object, and the hacks that have been suggested to date to work around this limitation have all been clunky at best.


There's no need for an IMAP extension; there's already tags, but Thunderbird is pretty much the only mail agent that has some kind of support for IMAP tags. Google uses the folder hack in part because support for tags is so bad. However it would be nice if Google supported both folders and tags.


It seems like it's other way around. There is IMAP spec for keywords https://tools.ietf.org/html/rfc5788 gmail just never implemented it (for whatever reason the spec is probably bad).

I am afraid when we start adding gmail specific features at certain point google makes compatibility breaking change and becomes email. The people not using gmail will have to start because of network effect and email died.


Supporting the email provider of hundreds of millions of people with a feature that should be common for all is somehow bad for the future of Thunderbird?

Not giving people what they want will drive them away, which is the worst way to keep something open. It’s not like you need proprietary binary blobs to support Gmail...


If you put gmail features into TB, it will become a gmail client, not an email client. The reason why all email servers and clients can talk to each other at the moment is because it's been well standardized, as opposed to, say, instant messaging. I'm all for improving email, but if done irresponsibly it will introduce division due to incompatibility and damage the ecosystem.


This is also the reason we can't have nice things using open standards, like labels. Once everything is tied to an ossified but open standard that's the end of the road as far as innovation is concerned.


It's not about openness, but about federation.


Google could have tried to standardize IMAP2, but they didn't.

Now only JMAP exists as a more modern mail standard.


My guess is never, because Gmail only exposes the labels as folders via the IMAP API. As far as I know Gmail doesn't have any other API.


They are currently blocking all these APIs for everyone who does not do a 30k-100k security audit


Are you referring to this API?

https://developers.google.com/gmail/api



They do have another API that is REST. At least one gmail backup tool is already using it ...

https://github.com/jay0lee/got-your-back/wiki


this should be upvoted more


Frankly, Mozilla should be rewriting Thunderbird with a completely clean slate. Focus on privacy (ie detection of email trackers) and integration with the big clouds (GSuite and O365), perhaps adding in some security features that are today only available in really pricey business offerings. I’d love to see this happen and for the next good email client to remain open and free rather than being acquired and strip-mined.


Why would current Thunderbird users be interested in this? Thunderbird is already pretty good at blocking email trackers and a rewrite would just cause development to stall again.

I would rather that they worked on the stability of their IMAP code (maybe rewriting that part of the codebase necessary) and fixing performance issues in general.



Of course I have read Joel's post and it's accurate. In the late-1990s, Netscape didn't need a full rewrite. But this is not the late-1990s. Thunderbird's current architecture makes it exceptionally difficult to develop on and it's probably time for a reboot of the platform.


> integration with the big clouds

Integration with the big <whatever>s instead of supporting and insisting on open standards is the antitheses to the open web.


Mozilla ignores the people at their peril. Many, if not most, small business type customers have already moved to Microsoft Office365 or Google Suite or have a plan to do so. By default, they use the web interface provided by the giants, but both providers also provide excellent APIs along with IMAP support. Mozilla can help the internet stay open by providing an excellent free and open source email client that can run alongside the proprietary ones that are constantly being developed and re-developed. I think an open platform for email client innovation is badly needed.


Or you know, giving millions of potential users an option.

I would love to use providers that support open standards. I don't have that luxury at work and suffer everyday when it comes to emails and calendaring.


I think that's a false dichotomy. Both can happen.


Supporting proprietary APIs means promoting them, which harms open alternatives.


I absolutely agree. Recently switched back to Apple Mail because the Thunderbird UI is just a fricking mess. Settings are so bloated and unordered, it's a pain to configure even simple things. The UI itself is cluttered and looks more like 2000 than a modern mail client. Also, I had several issues with mail formatting jumping back and forth. I gave Thunderbird several chances and nothing happened. Now I will not go back anymore until they completely overhauled the interface and user experience.


I'm baffled Thunderbird has to essentially be crowdfunded while Mozilla pays their execs millions and experiments with mostly useless things in Firefox. The only thing I remember worth anything was Container Tabs, and that's mostly abandoned with little tooling and no sync support.


They've rebuilt several core parts of their browser in recent years, as far as I know, now being in a place again where I'm glad again I'm using fx instead of chrome. I for one am glad they focus on their core product :)


Absolutely, the release of Quantum too changed my view on FX completely.


On the other hand they alienated their core user base with the extensions mess which is still lacking to this day.


While Mozilla wastes a lot of money (for example on their high exec salaries), they also have done a lot to improve Firefox the last few years. It is much faster and more stable now.


Well, my point was that they spend a lot on things that aren't their core product (Pocket, Lockwise, Monitor, Send, Firefox OS) while being unwilling to fund their long time project with an actual audience.


Pocket was already an external product (readitlater?). Send solves a real problem people have, and it was probably easy to develop it separately. Similarly to Monitor, you don't have to touch the core product to make it. Lockwise was probably easy to make, if you already have Firefox Sync, it's just a viewer app.


> Pocket was already an external product

OP's point still stands though. Mozilla spent millions acquiring Pocket. They also made it so users can't disable it as a plugin like they used to be able to.


And this a semi-annual reminder that it has been X years since Mozilla Corp spent $29m to acquire Pocket and said it will "become part of the Mozilla open source project", and Pocket is as open source today as it was on the day before the acquisition, which is to say, it's not.

X here is 3.


Pocket played into an overarching internal strategy that I believe Mozilla has shifted away from since (but it's still useful for populating the default New Tab page).

If you want to make an ex-Mozillian shudder, sneak up behind them and whisper "Context Graph" in their ear.


You just made a current Mozillian shudder!


Mozilla abandoned Firefox OS several years ago. Lockwise replaced the old password manager. Monitor and Send complement and advertise the core product. Pocket has revenue.


Container tabs is the feature that brought me back to Firefox after a decade of Chrome.


Same here. Being able to use the AWS console in multiple accounts is a lifesaver, as is the privacy of my Google cookie not being sent to every site.


Can you cite these millions paid to execs at Mozilla?


Mitchell Baker was paid $2,294,667 in 2017.

https://assets.mozilla.net/annualreport/2017/mozilla-2017-fo.... See page 7.

Mitchell Baker was paid $2,458,350 in 2018.

https://assets.mozilla.net/annualreport/2018/mozilla-2018-fo.... See page 7.

Mozilla has not released their annual report for 2019, yet.


To be fair it's not that much for a CEO of a SV corporation, however the fact that she went from $700,000 to more than $2,000,000 in the space of 5 years and her financial accomplishments don't reflet any of it is quite surprising.

Almost like vultures who are skinning a company while they still can, emptying the bank account knowing that eventually they'll get fired for poor performance.


Yes, I thought the whole point of Mozilla, being a non-profit and all, is to be different from "SV corporations".

But apparently it's run like all the VC-backed companies.


> emptying the bank account knowing that eventually they'll get fired for poor performance.

CEOs don’t set their own comp.


Baker is Executive Chairwoman and heads the Board of Directors. But she's also been around since Netscape, and was instrumental in starting Mozilla post-AOL. Not exactly a corporate raider.


In theory. In practice, they often negotiate it with their board.


Yes, but in practice then all employees set their own comp by negotiating with their manager.


It's (semi-)public info; same as it ever was: Moz is laying off employees and paying managers waaaay too much.

https://twitter.com/christi3k/status/1217542576141524992


Reading through that thread, the main thing that came to mind was the adage "Never meet your heroes."




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: