Readopting Thunderbird

Thunderbird On July 2nd 2012 an email came in for all Mozilla contributors aka Mozillians with a clear message:

“…continued innovation on Thunderbird is not the best use of our resources given our ambitious organizational goals…”.

It was a more political correct answer to say “we don’t care about it anymore“, I believe. After the announcement, the Thunderbird project had no employees assigned any longer and volunteers were asked to chip in. Mitchell Baker, the Chairwoman of the Mozilla Corporation also announced it on her blog.

As we speak, Mozilla employs over 1000 people (probably around 400 back in 2012), yet Thunderbird is so unimportant that not even a single employee can be assigned to it? There are dozens of volunteers working a full time job and contribute to Thunderbird in their free time. Half of them wouldn’t hesitate if they would get the chance to work for Mozilla, assigned to the Thunderbird project. At the end of the day it’s probably the only viable open source alternative of its proprietary counterpart, Microsoft Outlook.

I know Mozilla’s finances allow them to make and support some quite interesting projects, so why are we still hearing this ‘resources’ corporate talk? Firefox OS requires at least 50x the ‘resources‘ needed for Thunderbird, as an example. Thunderbird on the other hand is trusted by millions of users (yes, millions) for over a decade (yet we cannot say the same thing about Firefox OS). The trust alone Mozilla would get back by supporting Thunderbird is literally priceless. However, I’m deeply disappointed Mozilla still hasn’t changed their stance regarding this.

Disclaimer: Mozilla released security updates and provided the infrastructure for Thunderbird development, QA, support and documentation continuously after 2012. That is however not anymore the case (this is done 100% by volunteers right now).

Snowbird & Thunderden

It’s quite funny that Mozilla ‘let go‘ of Thunderbird in 2012 and not in 2013. It probably wouldn’t have happened in late 2013 after the Snowden leaks (at least I like to think so, no one knows). One would expect Mozilla to rethink the decision regarding Thunderbird, but almost 2 years have passed since the Snowden revelations, and there is still no sign of Mozilla changing their stance. A missed opportunity? I definitely believe so.

Thunderbird is one of the friendliest email clients for PGP encryption and with a similar Add-On ecosystem as Firefox, it offers endless customization possibilities. It is also the de facto email client for most Linux distros. Mozilla could occupy the whole email client space on all Linux operating systems with less ‘resources‘ (I love this word) than needed for most side projects Mozilla takes on.

The thing is, abandoning the Thunderbird project makes no sense neither community-wise,  neither marketing or PR-wise. How can you keep an open web when you are not taking email into account, which is a critical part of internet traffic?

Well, let’s have a look at the Mozilla Manifesto.

2 4

Are we helping cater these 2 principles (among others) by ignoring Thunderbird? It seems not. This is not any longer about “picking our battles“, this is necessary.
Also, on a market share and user base perspective, it’s true that for the average Joe, web interfaces are simply too convenient to exchange with an email client, but when it comes to business or professionals who are serious about email, they want to have full control over latter, which only email clients offer to a full extent.

David Harrison wrote a great piece about reviving Thunderbird. Quoting him here about the importance of email clients like Thunderbird

“The cloudification of mail and calendar is just so useful and popular — how could anything turn it back?!

But then I remember what it’s like when an online service goes offline and you can’t get your provider to help out. I remember the NSA, and the risks associated with having my mail and calendar lying around on someone else’s servers.

I remember promises of backups being made, only to find that after a disaster, they turned out to not work”

I must repeat I really can’t get behind why Mozilla is neglecting Thunderbird. It makes no sense in whatsoever way. We need to reevaluate this.

What does the Bird say?

There has been some good progress from the community front of Thunderbird lately though. 22 active Thunderbird contributors gathered in November 2014 at the Mozilla Offices in Toronto for the Thunderbird Summit, discussing plans and the future of Thunderbird. Reading about this made me actually very happy and hopeful for Thunderbird’s future.

Thunderbird Group Photo Toronto Summit 2014

Thunderbird Group Photo Toronto Summit 2014 CC-BY-SA Kent James

I mean look at this (one of the key points discussed at the Thunderbird Summit):

The Thunderbird active contributors are proud to be part of Mozilla, expect to remain part of Mozilla for the foreseeable future, and believe we have an important role to play in fulfilling the goals of the Mozilla Manifesto.

Reading this made me have mixed feelings. Firstly it made me proud of being a Mozillian and being on the same boat as these contributors who not only understand the Mozilla Manifesto, but breath and live it everyday (a lot of other Mozillians are in this category too of course).

Secondly, it made me incredibly sad that these efforts are not as appreciated as they should. It makes me  slowly lose faith in how well Mozilla can keep its promise for an open web (in practice, not in theory). Yes, I said it, I’m deeply concerned about this. I am also convinced Thunderbird is essential in helping our mission to keep the web open.

Learning a lesson

To summarize, there are various reasons for Mozilla to readopt Thunderbird:

  • It’s a “thank you” to all the supporters of the open web. It’s a sign that we still care. I know Mozilla has ambitious plans, but I don’t think we will be successful converting new people to open web supporters before taking care of our existing long time supporters. The thing is, Thunderbird is the best present you could give them.

  • With a lot of controversies happening in the last year (starting with Brendan Eich leaving), Mozilla has went through quite some battles. I’m afraid we have lost some credibility by making decisions which aren’t ideal for an open web, decisions one would expect from commercial companies like Google, Microsoft or Apple. Readopting Thunderbird would boost our credibility and show that we are not only about the browsers wars, but about the open web in general.

  • Marketing and PR-wise this is a much wiser decision than pouring money into marketing efforts. Just imagine how many bloggers would talk about the news. This would be huge engagement for us. It also shows that we are reflective of our past and analyze each of our steps carefully (we are already doing this with Firefox OS and its new strategy; Thunderbird deserves this too).

  • We would be able to be the default email client for the whole Linux environment. We kind of already are, but there is so much space we can improve here if we do a tidier job.

  • We created Firefox with the intention to break Internet Explorer’s browser monopoly; why won’t we do it with Outlook? Thunderbird is a very strong contendant and with some improvements it could easily replace Microsoft Outlook and other email clients. There is a lot of ground to gain here.

  • It’s a necessary tool to advocate the mainstream use of privacy protective measurements and encryption in a post-Snowden era.

  • Without Thunderbird, we are unable to fulfill our Manifesto to its full extent. It’s not an optionality.

Caring again

It’s no secret that the Thunderbird project is desperately looking for at least one employee to help Thunderbird going into the right direction. I am very positive that Mozilla is able to cater these needs, for its own best, by readopting Thunderbird. But employees (or the lack thereof) isn”t the only problem. A lot of Thunderbird contributors have not received support (or very little) from Mozilla in the past. We need to reactivate the community and make Thunderbird again inclusive of the Mozilla family. It would enforce efforts exponentially because contributors will actually notice that someone out there cares.

As of now, I’ve been traveling the Mozilla journey for 2.5 years and don’t plan to end it any time soon. Among others, I have been involved in various local communities, advocated open design within the Mozilla Corporation and helped the Mozilla Reps program to cultivate more transparency. Also, lately, I was asked to design a new Thunderbird Logo for the 38 major release. However it seems it won’t see the light of the day, due to doubts about the branding efforts. It was still a great experience designing it.

I believe that if there is a problem, there also needs to be a solution. While I have identified a problem, I didn’t give a solution yet (only hinting at it).

While this is a problem a lot of us make when we want to address something, I wish to avoid this and offer my full support and commitment to help Thunderbird be as much as part of Mozilla as Firefox is. Please reach out to me so we can start talking about this in the open. I created an etherpad for this where we can gather thoughts. Every input is appreciated here.

Thunderbird needs Mozilla, but also vice-versa. We just don’t know it yet.

  • To be clear it was the Thunderbird Team (original staff) that told Mozilla leadership they felt the product was feature complete. Most of those folks now work on Postbox.

    “Mozilla is releasing security updates and providing the infrastructure for Thunderbird development, QA, support and documentation continuously. However that is still very little compared to what Thunderbird is about.”

    Not accurate Mozilla made these commitments and did do some of it for awhile but Mozilla currently does not have anyone as their role at Mozilla tasked with Thunderbird QA, Support, Documentation or even Security.

    • Elio Qoshi

      This is not about adding new features, but about keeping the momentum of Thunderbird alive based on the values of our manifesto.

      Thanks for clarifying this, I cited that from the blogpost in 2012, so I thought that promise would be kept. Will update it accordingly.

      • I do not think this is going to change there have been conversations with community at Mitchell and Mark even asking to fund a single developer or support Thunderbird in other ways but as far as I know leadership is not interested in supporting Thunderbird at any level than the current level where infrastructure is paid for that allows the community to continue to work on it.

        At the same time I do not think Thunderbird can sustain with no paid staff for the long term as it is right now nobody is working on it full time. The team that is working on it is doing a good job of maintaining it and getting some fixes out but improving it and not letting it get passed up by Geary and others? I don’t know.

    • Andrew Sutherland

      I think there may be some confusion around the feature-complete/Postbox point. There were two Mozilla Corporation-employed developers working on Thunderbird at the time of the Mozilla Messaging decision spin-off circa 2007/2008. They and some other Mozillians formed Postbox. One of those two developers ended up joined Mozilla Messaging around the time of its formation. After having worked on Thunderbird for a long time, they left the Mozilla Corporation (which Mozilla Messaging was effectively reabsorbed into in 2011) to work at Google. All MoMo/MoCo developers still work for MoCo as far as I know. (It’s possible there are developers from the Thunderbird community who work for Postbox in some fashion, however. It’s a big community! 🙂

      • Hmm? Scott MacGregor was the lead Thunderbird Developer and founded Postbox.

        “Both McGregor and Bienvenu publicly supported Mozilla’s decision to focus on its highly-profitable Firefox web browser.”

        • Andrew Sutherland

          My confusion is that this post seems to be talking about the 2012 announcement of the planned discontinuation of Mozilla Corporation resources supporting the development of Thunderbird. But you’re citing the feelings of the 2007 MoCo Thunderbird team and events that ensued, which is the time period where Mozilla invested significant resources for several years in improving the state of Thunderbird and the messaging domain.

          When I read that line, it seemed like someone might infer that the 2012 team was so distraught by the discontinuation of Thunderbird resources that they immediately left and founded Postbox or went to join Postbox. Which is not what happened, and seemed like a non sequitur, so I figured it was worth clarifying for any other readers and you, in the event that you were missing some context.

          Aside (with apologies for the nit): I think it’d be handy if you can paste the URL of the thing you’re quoting (re: “…publicly supported…”). It’s not clear to me which of the links in the blog post you’re excerpting, or if the quotes are from somewhere else entirely. From posts to tb-planning etc. I’m under the impression you’ve done a lot of great information archaeology on the subject and it’d be great if you can more directly share the fruits of that labor.

  • Andrew Sutherland

    In regards to Mozilla Corporation providing one or more full-time employees to work on Thunderbird, I think it’s important to recognize that Thunderbird has a tremendous amount of technical debt. I also think it’s important to disentangle the benefits of a mail user-agent working on behalf of the user versus Thunderbird-the-code-base.

    I worked on Thunderbird at Mozilla Messaging for several years on the global database, the faceted search UI, the 3-pane quick-filter refactoring (with mozmill tests), and various under-the-hood things that involved touching a lot of the backend. I now work on the Firefox OS email app and its backend. The app and backend are entirely built on JS. (Notably, most of the lower level libraries are reused open source libraries authored by the https://whiteout.io/ PGP mail app team, and which we actively contribute to.) (Important disclaimer: I speak only as a Mozillian contributor and individual human software developer here, I do not manage anyone or any budget and do not set policy for anything outside my role as sub-module co-owner of the Firefox OS email app.)

    It’s my belief that the best game-plan and place for the Mozilla Corporation to allocate resources to is the Firefox OS Gaia email app and its back-end over Thunderbird. While nothing in the email problem domain is ever trivial and developing for the thus-far-shipped low-resource FxOS devices has at times slowed overall backend development, its backend and tests are in a much more promising state than Thunderbird’s.

    Having said this, I recognize, and it’s important for everyone to recognize, that Thunderbird-the-product is something that existing users know and love. And any changes are quite likely to be jarring. Joshua Cranmer has the plan of replacing the C++ back-end with JS pieces, and has been making progress on this front. But it is a mammoth undertaking with many Gordian knots to be cut. (And there are things the *multiple* full-time employees of Mozilla Messaging attempted to address.) Even with the back-end replaced, there are still issues in the front-end like XBL’s frustrating life-cycle semantics making things like the folder location combobox/move to/copy to picker be a nightmare to work on.

    Contrast with the screenshot of a prototype conversation-centric desktop mail client using the FxOS email app backend I’ve been working on at https://clicky.visophyte.org/files/screenshots/20150529-022214.png using react.js which is able to provide variable-height conversation summary display with efficient random access seeking. I’m of the mind that it’s easier to create a UI that will feel enough like Thunderbird than it is to overhaul the entire guts of Thunderbird (while keeping up with the many internal changes to Gecko.) Note that the UI in the screenshot is not one I would think Thunderbird users or humans in general would be okay with. It exists to drive development of the back-end and provide a basis for a usable UI to be derived from.

    Which is not to say that anyone should give up on Thunderbird. It’s a great product that works; my prototype is not something anyone can use for anything at this time. And Joshua Cranmer and R. Kent James, both back-end wizards, are doing great work moving Thunderbird’s back-end forward. (And R. Kent James deserves substantial kudos for his tireless efforts to advance Thunderbird in a frequently uphill journey.) But at the same time it’s entirely rational for Mozilla Corporation to focus its efforts on mail user-agents that will bear greater fruit because they are not saddled with massive technical debt and can run as apps on mobile and in Firefox desktop, etc.

    • Noel Grandin

      Is Gaia running on Windows and OSX and Linux (various) ?

      Does it have a well-developed theming and extension/plugin system?

      Does it support POP3 and IMAP and the literary hundreds of variations of these protocols out there?

      Does it have integration with native platform-specific stuff on various platforms?

      I think you are dramatically underestimating the value of the existing codebase and falling into the classic re-write trap, ie.

      http://www.joelonsoftware.com/articles/fog0000000069.html

      • Andrew Sutherland

        The Gaia email app can run anywhere Firefox runs. It supports POP3, IMAP, and ActiveSync (which Thunderbird does not support). Firefox OS does have a theming mechanism and a way for extending applications in a user-scripts sort of way, but both are in their infancy.

        There is definitely value in the Thunderbird code-base. And there is definitely wisdom in Joel’s post. Programmers love rewrites, and are equally notoriously bad at estimating the amount of effort something will take. But a central assumption underlying Joel’s post is that you have a product version N and you want to produce version N+1 that is fundamentally the same product aimed at the same use pattern on the same platform.

        The Firefox OS email app is a different product on a different platform with an inherently different mode of usage. It’s built on top of the web technology stack at a content level using HTML and JS. Thunderbird is intermingled with the web technology stack; it’s a C/C++ mail client that directly accesses the guts of the browser engine and uses a XUL UI. It wasn’t an option to use Thunderbird on Firefox OS. (Emscripten could have theoretically been used, but given how quickly things end up passing through XPCOM, it would have been far from trivial and probably would not have met memory footprint limitations.)

        The thrust of my argument against putting engineering resources on Thunderbird’s current guts is that I believe it’s possible to build a new UI that satisfies a similar product space as Thunderbird significantly faster than it’s possible to refactor Thunderbird’s innards. I’m not saying building a new UI is easy, I’m saying refactoring Thunderbird is very, very had. (And I say this as someone who worked hard on refactoring Thunderbird.)

        Things get tricky product-wise because Mozilla does have ideological goals about the web as a platform and using standards, and in many ways Thunderbird is not aligned with this. Thunderbird is very much a native app with an embedded web view, much like the apps you see in app stores. It uses a tremendous number of Gecko-internal APIs that will never be web standards. Its UI is built on XUL and XBL which helped pave the way for the flex-box and Custom Elements web standards and the future Web Components standards, but which are not web standards. As your questions point out, these are some of the things many people enjoy about Thunderbird: it’s theme-able, it supports extremely powerful extensions, it has native integration, and frequently feels very native. And even if we ignore the ideology, these are also all technologies on their way out and that do cause a non-trivial amount of bit-rot that Thunderbird does have to keep up with if it wants to keep up with Gecko. (Which is notably the opposite of the otherwise correct assumption that code doesn’t rust.)

  • People want to watch DRM-encumbered Netflix movies, not read emails with privacy. And everyone is on Facebook anyways, why use email at all? Don’t take the manifesto too seriously, we can live fine with an almost-open web :]

  • I like your post, I really think Mozilla should start working on Thunderbird again now that there is a strong focus/interest in privacy.

  • Zara

    @changetip:disqus, a rad to @fczuardi:disqus

    • A bitcoin tip worth a rad (35 bits/$0.01) from @bitsrenegadecheermaster:disqus has been collected by @fczuardi:disqus via ChangeTip.