Category: Data Custody

Everyone’s email workflow is unique.

Most new-age email apps (such as Hey, which we examined a few days ago) impose the founder’s idea of how workflows are created on the idea of processing one’s email.

But for many, their workflow isn’t really explicit, but an outcome of some habits formed early on, and those are hard to change. For even more, their workflow is limited/influenced by the nature of their job.

This is why most innovation with email clients fails, and why there is such a large graveyard of excellent-looking email desktop and mobile apps.
I am reminded of this today by this review of the subscription email app Superhuman. The writer admits it is quite pretty, but points out some glaring limitations:

My attempts to filter out PR blasts into a single split was foiled since I couldn’t come up with a search term to encompass everything at once.

and

But the Important filter, while useful, isn’t perfect — despite Superhuman’s “A.I. Triage” — leaving me with plenty of dross in my “Important” inbox and a few key messages, including a team-wide email from our editor (which skipped the Team inbox since it was sent to the staff list, a feature you can’t disable) ended up in Other instead…

And of course every discussion of the Superhuman app must also acknowledge its USD 30 monthly fee. The writer recognises that the fee is not worth it. Though he does not say it as such, it’s pretty clear that the fee is excessive not just because of the app’s limitations but also because fundamentally it doesn’t improve the email management experience by that much.

My primary email app for years has been Thunderbird, which I have used on Linux (most of the 2000s), Windows (very briefly 2009-11) and then OS X (2011- now).

It is hardly the most good-looking email app in the market – it’s less ugly than just plain anachronistic – but it is very customisable out of the box, and then extendable further through add-ons. In terms of email management, which is what we are discussing here, I have a range of filters that have been built over the years to route and label email that makes sense for me, including backing up my email to local (on-disk) mailboxes. Then, I have replicated Gmail’s well-thought-through keyboard shortcuts via another extension. Finally, all these are in the context of the Inbox-less workflow we discussed earlier.

We have made this point before but it bears repeating: open-source software may look un-sexy and may have a slightly steeper learning curve, but you will very likely be able to adapt it exactly the way you like. And it will last you years.

📱

In yesterday’s post about the first day of my Twitter-Reddit isolation, I mentioned that I spent Sunday morning coffee writing down some thoughts that I had connected. What follows isn’t about the content of what I wrote, it’s where I wrote it – in a document that I am now syncing between my iPad and iPhone via Resilio Sync, the BitTorrent-based P2P filesharing service

I have been a user for years but have been unable to make the mental move from a centralised cloud provider, first Dropbox and then iCloud Drive, to a pure P2P one. This is ironic since my first use of Resilio Sync, then BitTorrent Sync, predated my move to Dropbox. I have wanted a multi-device information sync infrastructure that was cloud-less, so I wouldn’t be locked-in, but also because I didn’t want to have to pay for cloud storage:

It seems strange to me that I’m paying to rent a few dozen GB on some company’s servers far away when I have already paid for hundreds of GB of high-performance storage on all my devices: my iPhone has 128GB, my iPad 256GB, my Macbook 250GB – all solid-state (and they’re all either on the same Wifi network or are connected via 4G on a plan that gives me hundreds of GB of bandwidth every month that carries over).

But iCloud Drive and Dropbox were both well-supported as external file locations by apps, which meant that even in the sandboxed environment of iOS, there was a very good chance your favourite apps would be able to access your Dropbox/iCloud Drive files. But having your files inside Resilio Sync meant that there was no easy way to edit them in place, or for media files, no easy way of playing them from the players you liked. They were locked-in, which limited its utility.

Things changed significantly with the introduction of the concept of external file providers in iOS 12, which Resilio Sync implemented. This meant that files in Resilio Sync could be made available to any app just like the two mainstream cloud storage providers. In fact, many apps have now made their storage available. This is the ‘Document Browser” on my iPhone:

Now I can store my text and all other sorts of files and have them available to any app that opens files via the iOS Document Browser – which now the vast majority of apps do.

What’s also helpful about this implementation is that if you enable an authentication lock on your storage app, the Document Browser will also ask for the same auth before listing and making documents available:

So. The plan now is to create all new personal documents in Resilio Sync, mimicking the organisational structure that I had set up in iCloud Drive, and make the move to Sync organically instead of forcing it. To see if a combination of Sync’s File Provider and the gradual approach works.

(Part 1)

The most consequential change I had made was to change the default folder from my Inbox to one I called Desk (if I remember correctly).

The Manually Sort Folders extension made this possible:

The thinking was that we copied the idea of the inbox and outbox for email, but not how we actually work with it on our (paper-based) desks. The inbox was just a tray on one corner of your desk. it was not the centrepiece. the centrepiece was where you sorted, read, thought, composed.

I had set up a whole set of filters so that email from a whitelist would go directly into the Desk folder. This was my company management, my Product Management team, and key people from projects and partners I worked with at that time. Automated alerts, newsletters had their own folders. And finally, email from people I did not know would stay in my inbox.

A few times I day I would scan my inbox and rapidly triage email: I had a keyboard shortcut that would move email into Desk, and the Gmail shortcuts extension had one-key shortcuts to archive or delete email.

That meant I was free to spend most of my day in Desk, a folder with email that I knew was important, and into which no new email would flow unless I explicitly moved it there or it was from someone I had allowed in. Email could pile up in my actual Thunderbird Inbox or other folders, and I would not care. This most closely corresponds to the ‘Read Together’ feature of Hey, and meant that I could achieve flow several times a day.

Endnote:

By and large, open standards, open formats and open source software is tremendously more flexible and customizable. You can build a highly individualised experience that makes you highly productive over time. And/or you can build an experience that compensates for your unique weaknesses or disabilities in ways commercials software can’t or won’t.

The tradeoff is that because few open source projects have dedicated marketing and branding teams, the packaging will often be rather unsexy. Because the developer optimises heavily for the steady-state experience, setup will often be wizard-less (although this is a plus if it’s part of package managers like HomeBrew) and “hard”. The out-of-the-box interface will be optimised for functionality over looks, though it’ll be easy enough to tweak that with themes (and you can tweak those themes).

But if you understand that open-source software front-loads most of the work for increasingly less marginal work later, you’ll be able to look past superficial shortcomings. You’ll invest the time and effort it takes to set up your tool the way you want it. And you’ll be set for years – you won’t be hostage to software being sunsetted (Google Wave), business models changing (Dropbox, 1Password), the developer being bought by a larger company and either taking a different path or being shut down (Whatsapp, Tweetie, Dark Sky).

(ends)

“HEY is not about workarounds. it’s about workflows,”

– Jason Fried, in the demo video for his new email service/client HEY

It’s a brave take on email. The last time anyone took on email with real vigour was eleven years ago. It was Google, with Google Wave. Tim O’Reilly termed it “email if it were invented today” instead of back in the 1970s. Google itself described it as bridging the “divides between different types of communication — email versus chat, or conversations versus documents” which really was even more ambitious.

Wave lasted a little over a year. Since then there’s been a lot of innovation in the design of email clients – desktop and mobile, especially after iOS’ popularity, but not in the basics of how you handle email. Even Google’s own Inbox, shut down but with many features merged with the mainstream Gmail, was tame compared to the audacity of Wave.

HEY is the first take I have seen that is in the vein of Wave.

But – as one comment online said – it’s an opinionated take on what email should be. Specifically, the opinion of Jason and DHH who have built it. Their workflows.

Contrast this with an open-platform email client like Thunderbird, which can be modified, via extensions, to suit essentially any workflow, any layout, any shortcuts. It’s a shame it’s not as popular as it ought to be.

During a phase in 2012-13 whenI dealt with many external parties at work and therefore several different individuals across different team, I used the outrageous SEEK extension that displayed my primary Thunderbird interface like this (screenshot from the SEEK homepage):

It resembled the iTunes Column Browser mode and was invaluable in quickly toggling between different ways of filtering my email. I was/am a fan of Gmail’s keyboard shortcuts, so I used an extension to add those very shortcuts to my Thunderbird. There used to be a whole clutch of Thunderbird extensions I would use to create an experience that was uniquely mine.

Now you could make the case that these are workarounds to the fundamental problem of email, not workflows – the very thing that HEY avoids. But I’d argue that the sum of these customisations resulted in a Thunderbird experience that matched my communications workflow then.

(Part 2: The single change I made to Thunderbird to create a whole new workflow)