The end of e-mail

e-mail is a great form of communication.

For starters, it is asynchronous, which means that one can choose to deal with it when desired and not to be victim of unwanted interruptions (unlike phone calls).

e-mails can be of any length, from just one word in the subject line and an empty body (unlike a letter), up to several pages of arguments and ideas. They can be accompanied by attached documents and richly formatted¹ (unlike the limits of SMS or some «social media» platforms).

e-mails have some kind of permanence, since they can be stored by the sender or the receiver as they want (unlike most modern «social media»).

e-mail is decentralized, in the sense that anybody can have their own server and send and receive e-mails without going through a central server or authority (unlike most of the means of electronic communications we use today, be it SMS, WhatsApp, etc.). e-mail can be considered as the first service in the Fediverse.

e-mail also allows to personalize how one wants to appear on the internet, since an e-mail address can show the belonging to a group, a company, just one's own name or even a chosen nickname.

Of course, most people have learned to hate e-mail with an attitude which has become rather hipster.

• We get lots of e-mail (but much less than what we get on Twitter, Instagram, WhatsApp, etc.)
• Lots of e-mail is spam (not unlike the ads one gets on all other media)
• Long e-mail threads with many people in CC are close to spam (not unlike Twitter threads)

All these issues can be solved by using a good e-mail client or MUA which allows us to filter and archive messages to suit our preferences (unlike the algorithms implemented by Twitter, Facebook and even Gmail, which choose what one can see). We can also have a good electronic hygiene by unsubscribing from useless newsletters, implementing «Inbox zero» (if that works for us), shutting off e-mail notifications and using different e-mail addresses for different purposes².

Many have predicted, wanted and even worked towards the end of e-mail. You may remember Wave, a Google project presented as the perfect replacement for e-mail. It was discontinued after a very short life, then the Apache Foundation tried to resuscitate it, but it was finally abandoned.

Google was criticized for trying to implement this e-mail substitute as a way to capture all those electronic communications which where happening between people not using Google services. Indeed, despite the popularity of Gmail, both as an e-mail provider (with lots of storage for free) and as an e-mail client (with a very good anti-spam filter, and other useful features to deal with lots of e-mail), this service uses the standard e-mail protocol and therefore, it is difficult to have a monopoly on it.

Why would Google want to have all the e-mail on the internet go through their servers? Because being able to parse and inspect all these messages allows to extract lots of information which is useful to train AI models which in turn can be used to understand what people think, want and need³. Then one can select which ads to show them and how to keep them «engaged» in their platform.

I am picking on Google because it was the first player on this game, but Microsoft is doing the same with the Office365 platform, where Outlook is the Gmail equivalent.

Just providing a complete platform with an online office suite, e-mail (server, storage and client) does not seem enough to force all users to stay permanently on the platform. What should Google and Microsoft do to get people to do all their electronic communications on their platforms?

Until now, they have succeeded in capturing lots of users: nearly everybody has a Google account because of Android and YouTube; and an increasing number of people, at least in the enterprise and academic sectors, are Office365 users because of the «need» of collaborative distributed editing (think SharePoint and the move of Microsoft Office itself to the cloud).

If a large number of business and universities have moved to cloud services like G-suite and Office365, this means that many e-mail messages have a receiver on these platforms. What can these platforms do to get more users? The easiest thing is to make communications with users outside these platforms difficult. For instance, redirect to spam all e-mails coming from domains which are not linked to the platforms. Google will accept messages from Microsoft servers and vice-versa, but will likely flag as spam messages coming from other sources.

Fewer and fewer people maintain their own e-mail servers. It is not a matter of lack of technical skills. Easy solutions like Mail-in-a-Box exist, so nearly anybody can install their own server. The problem is that these servers are often blacklisted and the messages they sent are trashed by the receivers⁴. This makes many sysadmins choose e-mail services hosted by cloud providers, like Gandi, OVH, or Ionos in Europe.

Unfortunately, some of those, like OVH, have chosen to replace the technology they used which was based on Free Software (postfix or sendmail, etc.) by Exchange. Are they loosing their technical skills and prefer to use a commercial product with commercial support? Or are they choosing a solution which is less likely to be flagged as spam? After all, a Russian spammer will use free software instead of buying a Microsoft license.

Regardless, these providers will have a hard time competing with complete integrated solutions as those from Google or Microsoft⁵.

Now that everybody uses Google or Microsoft servers, there is still one thing that bothers these companies. As e-mail uses a set of open protocols (SMTP for sending and IMAP or POP for getting the messages from the server), there is still the possibility to access Google and Microsoft servers with clients running locally on the users computers. It is therefore impossible for the provider to display ads or monitor how the user interacts with e-mail. This is a loss of potential revenue.

Both Google and Microsoft have announced that they will be shutting down the standard user authentication for SMTP and IMAP. The only way to send and receive e-mail will be to implement an authentication which needs to register the client application with a secret token which may change periodically. That means that only applications for which Google and Microsoft will have given their blessing will be able to communicate with Gmail and Outlook accounts.

Of course, getting a Free Software e-mail client registered with Google or Microsoft should be possible, but their terms of service forbid making public the registration token, which means that it can not be embedded in the free software. Thunderbird and KMail seem to have gotten an exception, but for how long?

So that is a nifty theory about evil corporations working to destroy our dear internet. I am probably wrong and even a little paranoid here, but the fact is that it is increasingly difficult to self-host an e-mail server and use local e-mail clients.

If you have a different point of view or any ideas on how to preserve e-mail as a decentralized and open means of electronic communication, please get in touch. Contact information is available at the bottom of the page.

The sad thing is that most people don't care about these issues. Most people think that the internet is the web. From those, the majority things that the internet is Google or Facebook.

Even most people who are into politics don't care or don't understand. Left-wing anti-capitalists are on Facebook and use Gmail and then they complain when they are censored. Conservative patriotic nationalists in Europe live in Microsoft environments to write and discuss about sovereignty. What a joke!