Sunsetting Nylas Mail Development | Hacker News Reader

Sunsetting Nylas Mail Development

nylas.com

129 pointssachinag39 comments

These guys need to work on their message. Here was my read of the article:

1. Some company I've never heard of is shutting down. I'll go read about it.

2. "Exciting Year", "Incredible Journey" They're shutting down their Mail Thing. Raising prices for old users to keep the lights on until "sunset" date. Ouch.

3. I wonder what their "Mail Thing" was. I'll click the homepage.

4. Strange. Looks like the rest of the company hasn't got the memo yet. They're still talking about their mail thing on the homepage as though it still exists. Looks like some Sync API for email providers. Shame. That might have been useful.

5. (later) Read discussion here. Evidently the thing they shut down was some tiny side project and the company still exists. Did not get that at all from the shutdown notice.

It was my understanding that they made a desktop (electron based) mail client to showcase their serverside features. But now they're saying they will only sell the serverside features now and the desktop client is too expensive to maintain.

Is this correct? I haven't kept up with Nylas, all I know is they stored all your emails on their server at first, but then got rid of it because of server load and people didn't like it.

Yeah, mostly correct, but Nylas Mail did no longer store your emails on their servers, only the previous version, (Nylas N1) did.

This is correct - it wasn't just that it was too expensive to maintain the client; it split our focus between the client and the API.

So we decided to focus on the API since we have a larger business there and we wanted folks who work at Nylas to be less stressed out.

We originally piloted using Nylas's APIs in our product about a year and a half ago. In that timespan, their pricing has changed significantly about 4 times and they do not seem like they have a particularly good grasp on their business model. They also, despite advertising 10 free test accounts, arbitrarily delete said accounts and then when you ask to get them reactivated you get connected with a sales rep.

Their API sync-engine [0] either hasn't been updated since March, or hasn't been kept open source--neither of which is a great sign. We're currently using the open source Nylas sync engine in production as we were uncomfortable trusting an important feature of our product in a company that doesn't seem to have a firm idea of what they want to do. But with the seemingly abandoned open source project, we are now working on building out our own syncing applications to ditch it altogether.

I would strongly caution anyone considering using the sync APIs to think about what they are getting into and the switching costs.

[0] https://github.com/nylas/sync-engine

This sounds bad, but I'd suggest an alternate interpretation:

1. Company invests heavily in engineering and builds a sync engine. Progress is visible on GitHub and rapid.

2. Company launches and repeatedly adjusts pricing and business models. Drives everybody crazy BUT:

3. Company finds it's value prop and customer base and the changes stop. Entire company shifts to building a sales team and keeping the wheels on the bus while they make their first customers happy.

4. Engineers switch from working on the core sync codebase to surrounding services, infrastructure, and scale.*

* Last week's developer dashboard announcement: https://www.nylas.com/blog/announcing-the-nylas-dashboard-2....

And somewhere along the road: Company realizes customers that are willing to go through the hassle of provisioning, deploying, and maintaining their own version of the open source sync engine are not the customers to focus on. ;-)

Ben's got this exactly right.

Perhaps customers wouldn't be willing to go through the hassle of provisioning, deploying, and maintaining their own version of the open source sync engine if they had more confidence that Nylas is someone they would trust to do it. Doing all those things is a pain in the ass. I would LOVE for us to pay someone to do it, but we've completely lost confidence that you are those people.

Open sourcing everything? More like sunrising! Thank you for doing that.

You're very kind - and sunrising is a brilliant term.

As always with Nylas, I've got very little understanding of what's going on here. The API business must be successful, but I wonder who is using it, other than people making e-mail clients, and they all seem to have their own!

N1 was the first good desktop alternative to Outlook for Exchange in a loooong time, so I was hopeful this would be a flier. Is it going to be possible at all to use the new open source app (or any forks) with Exchange?

N1 1.5.0 was the last version that supported Exchange, and that was via the Nylas APIs. We never did ship Exchange support on Nylas Mail 2.0.x, but there's no reason someone can't add it now that we've dumped all the code.

Was client-side Exchange sync ever released?

Tried N1/Nylas Mail a few times in hope of this killer feature... But was put off first by having to give Nylas credentials, and then by the fact the new local sync engine didn't support Exchange.

Glad the Nylas team is focusing. I really wanted to like the frontend client, but I stopped using it about a year ago due to reliability issues.

Mining organization data from email seems like a great opportunity. Context.io wasn't able to make it work, but Nylas looks a lot more capable.

(disclaimer: I'm the business guy at Nylas)

Unlike Context.io, the Nylas APIs don't mine email organization data at all! We're not owned by a marketing parent company, and all our customer MSAs make it very clear that we have no rights to any of the email data passing through us.

Nylas is an infrastructure company, like Stripe or AWS, and we make money by building tools and making our customer's lives and products better, not by reselling user data.

It's excellent to hear of companies taking this position.

Right, I meant organization data meaning who people are emailing, calendars, etc for building CRMish software on top. Wasn't tryint to imply you share any of that data. Just poorly worded on my part, sorry bout that.

https://blog.context.io/ seems very much alive. Do you have any other information what they were not able to make work?

They were acquired by ReturnPath and now their unknowing "users" serve as the basis for ReturnPath's data products, feeding information back to email marketers about what typical consumers' email behavior is like on the aggregate.

Maybe I'm just missing something here, but I really can't figure out what Nylas (the api side) offers?

Am i right in understanding that its essentially a Mailserver? AD / Gmail alternative kind of thing?

or is it an API facade interface over IMAP / SMTP, that i would use if i were creating a product and wanted to access a third parties email, contact etc which they would give me access to somehow?

That sounded kind of negative. Which i didn't mean it to be, I'm just confused.

The client i really like, and I'm happy that its being open sourced. I recently tried to run it locally as a jump of point for a project i had in mind, but fell down at the sign-up screen, so I'm happy to hear that that has/will go :)

Yeah, the facade. Plus a lot of stuff on top of that: webhooks, open/reply/link tracking, metrics, other stuff.

https://www.nylas.com/customers/ lists some of our API customers; maybe that'll help!

As a user, I've been loving Nylas Mail/N1. I hope the project lives on in some usable form - the blog post mentioned Nylas Mail Lives and Mailspring, which I'll have to take a close look at.

(I came to Nylas after Dropbox finally killed off Mailbox. I seem to have a good track record of falling in love with desktop email clients which then die...)

Hey thanks! I used to work on Nylas Mail at Nylas and I'm the author of the Mailspring fork.

It's definitely sad to see this post, but Nylas has remained committed to open source and I think they've done this right. Relicensing everything under MIT ensures the product will have a bright future, even though Nylas has better opportunities it needs to pursue for it's employees and investors.

Would love for you to check out Mailspring (http://getmailspring.com) when it launches this month. We always struggled to implement performant, battery-friendly mail sync in JavaScript, and Mailspring essentially "rebases" the entire JavaScript UI on top of a new C++ sync engine built on mailcore2. You get the flexibility of JavaScript + plugins + themes with a much lighter profile.

Where can we see the source and will the client have to phone home to you?

Well, "like Nylas mail but better" is a strong value prop from my point of view; I'm signed up for your mailing list and look forward to hearing more with interest.

> Relicensing everything under MIT ensures the product will have a bright future

Absolutely. Very impressive move.

Wow! I like it.

Interesting. I used N1 for a while a really enjoyed it, but the Electron apps were killing my laptop so I ditched it. Super interested in the Mailspring fork :)

It looks like Mailspring is still going to be an Electron app, but with Mailcore2 [0] for the email behaviour rather than Nylas' sync engine (per the repo's readme [1]).

[0]: https://github.com/MailCore/mailcore2

[1]: https://github.com/Foundry376/Mailspring#high-level-goals

I don't understand, what exactly are you sunsetting? N1? Or the service that synced mailboxes?

My understanding is Nylas has two products: An email API you could use to build applications on, and a mail client they built on top of that API.

The mail cient was originally called N1 (later Nylas Mail), and is the product being sunsetted so they can focus on the API.

This is correct. The APIs have extended to include calendar and contacts in addition to the original email API, and that's where we've chosen to focus going forward.

So I won't be able to use Mailspring unless I use you as my email provider?

No, no, Mailspring and the Nylas Mail Lives forks are totally divorced from our infra.

Maybe the Thunderbird developers can take over, some of them were keen on a rewrite using "modern" technologies, this would be a good bootstrap.

We tried reaching out to them based on the blog post you reference, but got no response. :/

I used N1 during the beta, but switched to Postbox once they moved to their actual service.

For a main client, I really liked it. I'm looking forward to the forks.

For me, I find it odd how few strong email clients there are for OSX. Every app seems to have several quirks, and few provide a menubar icon with an unread count or proper theming.

Props to the team for releasing under MIT. The two forks listed look extremely promising.

Yeah, we're super excited about the forks ourselves! I'm personally using both N1 and Nylas Mail for all my work email (and paying the $12 a month subscription out of pocket) but I hope to move to one of the forks soon.

unsurprising

Reply on news.ycombinator.com