Hacker News new | comments | show | ask | jobs | submit login
Mail Pilot – The State of Email on the Mac (schneidmaster.com)
35 points by schneidmaster on Jan 22, 2014 | hide | past | web | favorite | 57 comments



>Let's face it: email on OSX sucks.

Compared to what? Since, you know, it can run most apps other platforms can (and most people use), from mutt to Thunderbird and Outlook.

>For all of Apple's skill in crafting beautiful hardware (and as much as I like OSX as an operating system), Mail.app is outdated and still has limited support for Gmail labels

Labels? Why should it support a proprietary technology -- and much more from a competitor, and much more one that might kill IMAP at any time?

Labels aside, I never had much problems with Mail.app. And, spam aside, I get around 200 mails per day.


> Labels? Why should it support a proprietary technology -- and much more from a competitor, and much more one that might kill IMAP at any time?

Because 90% of tech nerds are fucking obsessed with google and think everything that comes from them should be immediately adopted by the entire industry.

As I've said before - IMAP has a spec for "keywords". People always piss and moan about it not being supported by mail clients, and conveniently forget that Google's labels weren't supported either, when it was introduced - and yet some clients have added supported for it.

What are the chances that a) more competing email providers and b) more mail clients would support IMAP keywords today, if Google had used the existing standard instead of inventing their "labels" system?


> As I've said before - IMAP has a spec for "keywords".

That's just because IMAP has a spec for everything.

That's the huge problem with IMAP: it can do too much and it's one of the more complicated "classic" protocols (from the area where RFCs were much more concise and bare-bones).

There are so many implementation bugs in various clients and servers that IMAP feels like being kept together by duct tape.

As such, people are very reluctant to add more protocol features to their implementations and even if they did, with some likelihood, the feature would only work in very specific client/server combinations.

In case of labels, the incentive for Google was to produce the implementation that works best with the existing clients (use pretend-folders for labels), and the incentive for client is to implement labels the google way (the only widely-deployed server that currently supports them)

Email is such a mess.


>That's the huge problem with IMAP: it can do too much and it's one of the more complicated "classic" protocols (from the area where RFCs were much more concise and bare-bones). There are so many implementation bugs in various clients and servers that IMAP feels like being kept together by duct tape.

IMAP does much because we have many needs. And HTML does even more, but we still have several top notch browsers competing.

In any case, if IMAP development is so difficult, have people join ONE IMAP library project that tries to be as complete as it can, instead of each mailer creating its own IMAP client libs.

Like we have Webkit/Blink which is used across multiple browsers.


> the incentive for Google was to produce the implementation that works best with the existing clients

I doubt that very much. If Google could turn off IMAP tomorrow and not lose a huge chunk of their clients, they would.


> Compared to what? Since, you know, it can run most apps other platforms can (and most people use), from mutt to Thunderbird and Outlook.

It wasn't really a comparative claim. Email also sucks on most other platforms. The possible exception is iOS- the iOS7 Mail.app, Mailbox, and Mail Pilot are all actually good (rather than least bad) options.


Mail.app in 10.9 seems slightly less crappy than in 10.8, but the overall experience has been going steadily downhill for years now. I've been using Mac OS X + Mail.app since the public beta was released in Sep 2000.

Having tinkered around with other mail clients for the Mac, I'm starting to believe that something in the vein of mu4e[1] holds the only truly brighter future for me, but it will be such a radical break after years with Mail.app that I'm hesitant to take the plunge. On the other hand, I am "living" in multi-term[2] now rather than iTerm2, so it probably won't be all that jarring to emacs-ify another element of my daily computing experience.

[1] http://www.djcbsoftware.nl/code/mu/mu4e.html

[2] http://www.emacswiki.org/emacs/MultiTerm


I've been using Postbox[1]. It's the power of Thunderbird with the clean GUI look of a native Mac app. It has emails in a conversation-y view and includes gmail.com shortcuts. It also has powerful smart folders.

Plus, I can use Markdown-Here[2] to compose my emails in Markdown.

[1]:http://postbox-inc.com

[2]:http://markdown-here.com


I think I've been happiest with Postbox, but holy cow, that periodic re-indexing is painful. Hooking up ~500k emails means it grinds away for minutes on a HD, something like once or twice every hour.

That aside, it handled those 500k emails way better than e.g. Mail.app (whose performance routinely falls off a cliff somewhere back between 200k and 300k), synced quickly, etc and all without the normal annoyances that keep me away from Thunderbird. A very solid, cheap option.

That said, the re-indexing eventually got annoying enough that I disabled a bunch of IMAP folders in Gmail and switched back to Mail.app. I'd love to find a real alternative, but everything else I've touched has been awful in absolutely-critical ways (usually one or more of: can't handle a couple hundred thousand emails at all, can't display HTML emails, can't stand the UI, or useless/nonexistent filtering tools).


What "normal annoyances" keep you away from Thunderbird?


The UI has always rubbed me the wrong way, which I haven't really quantified. Not "wtf, no gradients? fail", but just a preponderance of small things like weird defaults, chunks of menus / screens that are inconsistent with the rest of the application (so you end up checking 10 places for a simple thing), and minor inconsistencies with other applications that I keep re-discovering after using it for a few months that I have to go back and fix.


I'll second Postbox. Search is fast, and it handles multiple accounts easily. Stable and clean.


we could all really use a better email client, but as discussed to death in the sparrow threads, the app store has reset software pricing perceptions. People will now bitch endlessly about having to pay the princely price of $10, which theoretically comes with free perpetual upgrades. This is down from the $50-$100 people used to pay for each major version of an email client in the 90s. While this may be good for pocketbooks, it has also probably destroyed desktop mail clients for non-enterprise as a market for a software company that wants to pay landlords or mortgages in something besides appreciation.


I don't think many people complain about the $10 price tag of these clients. Looking to replace Sparrow, I just purchased both Mail Pilot and Unibox (hadn't heard of either until this article).

The reviews of both apps don't contain complaints about price — unless the app is not working for the customer, which is a pretty reasonable complaint.


People don't complain ... they just don't buy the app. It is astounding the level of quality & support expected from low-cost apps.


I sell a $10 app on the App Store and have provided three years of free, major upgrades and dealt with many thousands of support emails. I never really felt like I was giving too much to people who paid $10.

Although I have never sold software at a higher price in the past, so I have nothing to compare with.


My impression is that the expectation for software quality and support is the same for a $1.99 iOS app and a $99 desktop app. Given that upgrades are always free in the App Store (unless you go the Omni route), the cost different is probably even more skewed, and if your popularity continues (which is hard in the app store), your support burden will grow without additional revenue from those users.

It sounds like you're an indie dev. You probably could charge more than $1.50/minute for your time ... so if you spend even five minutes supporting a customer, you're wiped out all your revenue (not even counting development time!).


>While this may be good for pocketbooks, it has also probably destroyed desktop mail clients for non-enterprise as a market for a software company that wants to pay landlords or mortgages in something besides appreciation.

Does not seem to have stopped other companies, from the Omni Group to Pixelmator. And didn't iA sold hundends of thousands of copies of it's editor when other editors where 1/5 to 1/10 the price? If it solves a need, people will pay for quality.

Not to mention that today's $10 correspond to 10 to 30 times more people using Macs than the times of $50-$100 email clients did.


People will now bitch endlessly about having to pay the princely price of $10, which theoretically comes with free perpetual upgrades.

I don't know. Some people still value quality and attention to detail. On my Android devices I'm using MailDroid pro, which is priced at around $20. And that's for a mobile app. But it's good enough for me to think that's worth it.

I'm paying $40 a year for Plex, and $180 for Spotify. I think that's good value too. I pay gladly.

Obviously there will be "cheapskates" always going for the cheapest option, but those people have always existed. Now they just have more & cheaper options and there's no way for a developer to escape that.

I guess they'll have to price that (smaller premium market) into the premium-price and hope it wont make it even smaller.


Out of curiosity, why did you decide on an annual Plex subscription over the lifetime option? Do you just doubt the viability of the lifetime option or is something else going on?


It's a case of try it properly before you pay the helluva-much-higher price.

Since then I've come around considering lifetime subscription, but now they've upped the price for it a little more. And right now I have enough other things sucking up my funds.

In a perfect illustration of human irrationality, I will probably get around to it later.


I gladly paid $49.99 for MailMate on my mac ( http://freron.com/ ) and am still extremely happy with the purchase. It's the best email client I've used for a long time and well worth it.


Perhaps the solution is to Kickstart the development of a desktop client?


Mail Pilot was actually the product of a Kickstarter, although it's been pretty thoroughly redesigned since the original concept: http://www.kickstarter.com/projects/1380180715/mail-pilot-em...


Thanks! I wasn't aware of that.


Mail.app is near perfect for me - however, Apple REFUSES to fix its IMAP support. If you access your email from more than one device (hey, that's why I'm using IMAP), Mail does not properly query header counts, so folders get out of sync. If you restart Mail, then it does the right thing. I've resigned myself to doing that 10 times a day. Bleah.


I loved Mail.app after switching to SSD a few years ago. Near-instant local search and a large set of smart mailboxes made for a fast and highly personalized workflow in a native retina desktop app. Email bliss.

Unfortunately, Apple began syncing Smart Mailboxes via iCloud in Mavericks, and the Smart Mailboxes saved on iCloud became corrupt. This corruption causes Mail.app to crash on launch when it tries to fetch and apply remote settings, and affects every computer that uses iCloud.

Senior support at Apple confirmed it's an iCloud issue and passed it to support engineering to resolve, which have been MIA. Issue remains open after 2 months, during which I've tried every alternative desktop email client and found them severely lacking - some are highly functional, some have pretty graphics and typography, none both. (to tie this back to the original article, I ran the Mail Pilot preview a couple months ago)

Heed my warning - iCloud will wreck the things you hold dear.

/rant.


"near perfect" and "[restarting] 10 times a day" seem like incompatible notions to me.


Near perfect = it does everything I want to they way I like it.

[restarting] 10 times a day = it has a bug that needs restrarting 10 times a day to do so.

He didn't say it was perfect as in bug free. Just as in "does what I want".


It's not incompatible, if that is the only issue your parent is facing. Personally, I just Mail.app clunky as hell.


So, the thing that scares me is this:

http://www.mailpilot.co/bestuseguide.html

The repetitive "Finding messages outside of Mail Pilot" sections make me think that it's not going to play well with other clients -- unless it means that it's storing some sort of metadata/copies in those other places, which is not the impression I get (it seems to be saying it will actually make a bunch of folders and move your mail into them, which is Not OK).


Personally I do not want my mail client to be my task manager. I use a Todo list app (Things by Cultured Code) that does a smooth job of linking a todo to an email. I simply create a task linked to an email set for 3 days from now, I delete the email from my inbox and have 1 'inbox' for my actions. I believe the various other todo apps on the mac handle this just as easily.

Seeing workflow wrapped into an email client strikes me as wrong—I want my mail client to be for communication, not for managing my next actions.


OS X and iOS qualifies as cross-platform these days?


Maybe that wasn't the most precise term, apologies. I assumed the immediate clarification of what I meant by "cross-platform" was sufficient.


I've tried every Mac mail app under the sun, and Mail Pilot is the only one I've actually enjoyed using. I've been using it as my exclusive e-mail client since the early preview and it has truly revolutionized how I handle e-mail. I've committed to their to-list idea of keeping up with e-mail, and I actually get to inbox zero almost daily (I hadn't reached true inbox zero in years). Mail Pilot still needs some more polish, but the updates are frequent and it already surpasses everything else out there in my opinion.


I purchased Mail Pilot and Unibox because of this article — with the intent for one of them to replace Sparrow (which is amazing, but there are bugs that will never be fixed).

I liked Unibox more. It has a way of minimising the complexity of my inboxes. Being able to see all the attachments to and from a particular person is very nice.

Mail Pilot feels a little less polished and less smooth. It also makes a very strong effort to educate me (popups and explanations are attached to everything, intro tutorials and videos) but I still found it a little confusing to navigate. Initially the "New Message" button simply did not work until I restarted the app.

Unibox did not make much effort to teach me. And despite completely rethinking my inbox I've found it straightforward to use. Unibox seems to place the information you are looking for right where you expect it (for example, dates fade in and attach to the scrollbar as you swipe through your emails, changing as you scroll). It's very cleverly designed to show and hide information as you navigate, without showing too much.

If you don't mind switching to a people-centric interface, I'd recommend Unibox. Mail Pilot seems aimed more at people who like to organise their email (the reminders system, completion checks, sorting). I just like to read my email and write, and Unibox does a great job of it.


I really want to love Unibox. I used it full-time for a week or so. But I just seriously cannot get over the way it deals (or rather, doesn't deal) with folders and labels. If I put something in a folder or archive it, it disappears from the people-centric UI, and there's not an easy way to drilldown through folders. Also, things like group messages are super awkward with the people-centric thing.

If you like the interface and concept, I would definitely recommend Unibox over Mail Pilot- it's more polished in general. But it's a very opinionated mail client and I really couldn't get over its opinions.


Yeah I never use folders (or labels) — I am just not organised enough to actually use them. I tried for years. So Unibox really meshes with the way I deal with email.


Mail.app, smart mailboxes for everything in the past month, past week, and past month's unread. Mark stuff as unread if you need to deal with it later. Screw inbox zero, screw a zillion folders and a zillion rules.

Though I do sort junk from social sites into their own folder and filter them out of those smart mailboxes.

No other mail client supports smart folders. They all want me to make a bazillion folders and a bazilljon rules to support them. It all seems so primitive.


I know they say this on their site, but can someone very clearly confirm that this is purely a client and uses IMAP/SMTP and doesn't send anything to Mail Pilot (the company).


CEO here, can confirm. Mail Pilot runs IMAP and SMTP directly to your email server and back; nothing in the middle. Our first client, released in 2012, did make use of a 3rd party server, but it made too many people uncomfortable, so we found ways to implement the advanced functionality of Mail Pilot without needing a 3rd party server (here's the blog post of the announcement that we were moving away from using a 3rd party server: http://mindsense.co/blog/6-major-announcements/).


Awesome, thanks for confirming this and sorry for being overly paranoid.


No problem & don't apologize - I too would be paranoid about giving a random company my email credentials. That's why we moved away from it.


Ok. So I did give it a try, and it was the worst 3 hours I could spend with a mail client before asking for a refund.

I tried it both with a GMail and a classical IMAP accounts. Both were a total disaster. Two years of development for THAT?!

Dates were messed up. Some mails would never appear, some would, but with another sender! Gmail labels were not properly managed, Lists cannot be deleted, even when emptied. And those new MailPilot.* folders are just plain wrong, it breaks your flow in ANY other mail client.

What a waste of time.


Just bought it because I heard a few recommendations, and ran into a couple of sloppy bugs. An email that arrived on January 2nd is showing up as arriving January 21st - that's a basic sorting bug. And lines that wrap on other email clients aren't wrapping, without letting me scroll - so there are some lines in my emails that I can never read.


I haven't found a mail client that integrates cleanly with Gmail yet. Mail.app never respects my settings/preferences and is still quite buggy (on OS X 10.9.1). So I'm eager to find a good mail client replacement for it. But in the meantime, I'm still using gmail.com. Fortunately with keyboard shortcuts enabled it's not that bad.


I've had pretty good luck with Airmail, which happily supports Gmail keyboard shortcuts. Might be worth a look.


I agree - Airmail is a good choice. Also, I like MailMate and it also supports Gmail keyboard shortcuts.


This looks fairly nice, but useless for me without Exchange.

This prompted me to finally just break down and spend the $2 on AirMail. What surprises am I in for, both good and bad?


Good: It's a fairly decent mail app that does all of the things you'd expect.

Bad: Still get bugs and crashes that I'd expect from a beta, not a production-ready app.


I've been running the Airmail beta version. Overall it is very stable, plus it gets bug fixes faster. One great feature is that it supports Exchange, and the most recent beta overhauled the whole Exchange engine.

One exchange oddity, the contact lookup only appears to work from a "Global address book" drop down under Window.


The exchange overhaul sounds good, since when I just tried to set it up, it wouldn't auto-discover my company's server.

I guess I'll give it a shot with gmail in the meantime.


I would really like to try Mail Pilot, but it doesn't have S/MIME-support. Some people actually use that actively :(.


Eudora still works. That's pretty good for a program that came out in the late 1980s.


For me, Unibox is incredible. Hard to believe I ever used email any other way.


I hate that Unibox grouped together all my Github emails. I want separate threads for every issue/project.


Have you found a decent way to use Unibox with folders and/or labels? Or do you just not use them?




Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: