I am very surprised a "one man show" was the author's final choice here: What if the one man gets hit by a bus? FastMail is probably worth the increased cost to avoid that concern alone.
But in general, I am excited to see anyone moving to their own email domain, decentralized is how email was always supposed to be.
EDIT: A huge terms issue with PurelyMail is "The Company may, at its sole discretion, terminate service without cause or notice." FastMail can terminate for violations of the terms or non-payment, PurelyMail could terminate you because Scott just doesn't like you anymore.
Hi! I am the "Scott" from Purelymail in question. In the hopefully unlikely event I'm hit by a bus, I do have friends who could step in to keep it running for a while. One of my infrastructural goals is also to have it run itself without manual intervention, where feasible. It's not 100% there yet, though. (Hence the beta.)
As to the terms of service, as far as I know the clause you quoted is fairly standard cover-your-ass. I've probably seen it in a few other service terms. Presumably, Fastmail words it as they do because they've covered all the reasons they might want to do so in their terms already, and have decades of legal experience.
Hey! I feel like ensuring you clarify you'll only shut down an account if they break the rules you set out is a really good practice and there are examples, like FastMail's, you could borrow from. I know the standard boilerplate is to give a service the right to do whatever it wants, but as a small indie service operator you can be better than the fairly standard! ;)
Your service running itself is a good idea for avoiding you doing a lot of work, but if you got hit by a bus it is likely someone would shut off your credit cards that pay for services your services depends on and even if it's a box in your house it will end up unplugged. I'd strongly advise some business continuity involving a very trusted friend or family member who can be your second in managing the service at minimum, who has access to everything and knows what to do already.
I hope you don't take my comments as too much of a knock, what you're building is hard and worthwhile, and I wish you success with it. I'm on FastMail but that doesn't mean I always will be, and I like to have good options by good people.
No offense taken! I think you're broadly right that I could be doing a better job on reassuring messaging, and I'll put that on my work queue.
I also hope to solve any bus factor problems before leaving beta. Right now if I were struck down, I think I'd still have about 9 months worth of free AWS credits on the account running the service, plus any profits. (It's not quite a self sustaining business yet, but it's not that far from it either.)
Your responsiveness on these questions is seriously making me consider picking up an account. I, too, am looking for an excuse to migrate off GMail (It's a significant project once you have 30+ accounts) and your product might just be the straw needed to get me moving.
Could you also commit to notifying users when the privacy policy or terms of service change? I suspect you already do this, but it would be good to have it in writing.
@Felz, hey mate, I love what you are doing. I will get Purelymail featured on SaaSHub for a few months. Hopefully that may bring you a few new sign ups. Please just get it verified there and improve the listing.
Hey Scott, Purelymail looks like a really nice service! Out of curiosity: how are you handling deliverability/avoiding spam filters? From my experience, hosting my personal email, any messages I send almost always go to spam (unless it's a direct reply to an email I've received). (I've set up SPF, DKIM, DMARC, a PTR record, etc) Is it just a matter of building up enough email volume?
I don't think it matters even with all that stuff. The first time you email anyone, it's going to go to spam. Once you have established a mail flow, then it will continue. The best setup is to have them email you first, then your replies will already be well scored.
Seeing this thread prompted me to sign up - I've been procrastinating over ditching gmail and consolidating on one IMAP provider for ages.
I shelved my one main doubt (I don't think much of Roundcube) and started the process today. Seems like you've done a really good job. Recipients have been getting all my emails, and I have found the integrated migration tool invaluable in getting set up.
Zoho Mail only costs $2 more per year vs PurelyMail (it's shown on PurelyMail's website along with other competitors) and also avoids the "one man show" problem.
I'm not sure saving $2 a year is worth using a self-proclaimed (according to its website) beta service. Then again Gmail was in beta for many years.
Indeed. I honestly would never go cheap on email again. People have traditionally expected it to be free, but it's the most crucial point of my online presence: I need to be able to rely on it and it needs to be good.
In my opinion, the fact that people would pay more for Netflix than their email provider is kinda crazy, when you think about their relative importance.
Email is obviously really important, but the cost of running Netflix is extremely different to running an email service and I would take that into account with the pricing.
But I think Netflix is extremely cheap/good value if you enjoy the content.
Last I checked, Zoho was per address; This guy specifically looked for domain support (meaning, potentially hundreds of addresses).
PurelyMail seems to charge by storage, not mailbox address; and AFAIK so does Migadu (which, up until reading this post, I thought was unique in that).
I will admit it's unfortunate to have to pay per inbox on FastMail. I can't even pay less for my second box to be smaller and more limited, it must be at the same rate as my main box.
They do allow infinite aliases which covers most of my needs, but when I wanted to use a FastMail box for my home automation system's service account it didn't make sense to do so.
Zoho is per user/inbox but a single user/inbox can receive emails for multiple domains and email ids. You can also set up individual sending identities. I am currently receiving email for 7 domains while paying for one.
There is an undocumented limit though. You can generally only have upto 30 aliases.
Fair comment, but all of those concerns are mitigated if you:
1) use your own domain(s)
2) take frequent backups
so, even if the purely mail dies, you have access to historical emails and since you own the domains you can migrate to another provider pretty quickly and again, since you own the domain any accounts that are connected to that email address/domain combo are not impacted.
How many people are actively backing up their email on their own? Backing up email seems much simpler than running your own email server, but it's still going to require some technical know how.
To me, the comment that you can just manage your own backups suggests that the service isn't right for anyone nontechnical or technical people who are too busy to backup a hosted service.
I would strongly recommend against using POP3 for such purposes. It doesn’t contain the entire history, for two reasons:
• POP3 doesn’t have the concept of mailboxes. All you will get is the messages that exist at the time of request, not any folders you’ve put them into or labels applied.
• If you delete a message on the server, it won’t be deleted locally, which means that your backup does not represent the current state of affairs.
Also POP3 is definitely designed for the “download and delete on server” approach, and various tooling may have limits on it because of that. For example, if you get Gmail to fetch from some other server over POP3, it stops working after there are 50,000 messages. This undocumented limit bit me some years back when I used Gmail in this way, and it didn’t even notify me that fetching had stopped working! It was two weeks before I noticed that no new messages had been coming in from that source.
> POP3 by design creates a continuous local "backup" that contains the entire history. IMAP doesn't.
I refute this. IMAP is a synchronisation protocol. Clients can choose to operate fully online (performing every operation on the server), fully offline (downloading everything from the server and operating locally) or a hybrid (e.g. keep only the last 30 days of messages locally). The protocol, and most clients, are fully capable of creating a continuous local backup containing the entire current state of the server. Is this “the entire history”? Depends on your definition. I’d argue it’s more true of IMAP than it is of POP3. But it’s not like a Git repository showing what happened and when. I know of no email protocol that provides that. But you can make it so yourself, e.g. sync IMAP into a maildir that happens to be a Git repository and commit after every sync. That would have the entire history, at the resolution of IMAP fetch.
The bus factor is a potential issue. Legal terms & conditions probably aren't as the corrupt civil law systems of nearly all polities make legal redress unavailable or hopelessly impractical to all but the wealthy (who probably aren't looking at $10 pa mail). Ordinary people don't read legals, partly out of laziness, but partly from a (correct) judgement that they are irrelevant to them.
I think it's very important for customers to be able to trust the business though. This means the business relationship should have clearly defined terms and both parties should be expected to follow them. Unfortunately, the trend is for services to have take it or leave it terms and to essentially be written to hold them blameless while letting them do whatever they want.
A legally sound strategy, but not one that embodies trust. And I'd argue being able to trust your email provider is very important.
I'm more worried that PM is still fairly new and in beta. These hobby businesses come and go, and changing one's email is a huge pain. What's saying that PM will be around for 10+ more years, like Gmail has?
This is mostly because changing your email address is a huge pain. Once you're using your own domain, changing email providers is like changing web hosts or domain registrars: Some work but not a big deal.
Or, as an easier solution, a dedicated email forwarding service. I've been using one for about 15 years now and changed the actual email service provider a couple of times just by updating the forwarding address.
Of course then the problem is to identify a reliable email forwarding service that won't go anywhere.
There is definitely still an upside from Gmail: The author can redirect his email address to any other service on a whim and all his inbound email comes with. So as long as his archive is backed up he does have a good strategy to restore access.
But in general, I am excited to see anyone moving to their own email domain, decentralized is how email was always supposed to be.
EDIT: A huge terms issue with PurelyMail is "The Company may, at its sole discretion, terminate service without cause or notice." FastMail can terminate for violations of the terms or non-payment, PurelyMail could terminate you because Scott just doesn't like you anymore.