Hacker News new | past | comments | ask | show | jobs | submit login
Show HN: A personal Gmail
127 points by jubari on Nov 19, 2012 | hide | past | favorite | 103 comments
Hey guys,

so I've been building my own Gmail for the past few weeks and would like your feedback.

Find a preview here: http://i.imgur.com/UDz2y.png (Mouse cursor is over the "REPLY" button, to show the hover-effect)

The app you see in the screenshot is what I've been using in the past week and it starts to feel better than the gmail interface.

Just a quick explanation:

- It's conversation-based like Gmail. - Conversations with new emails are on the left. - Conversations may be labeled ... based on predefined filters or manually. - Every conversation stays in the left sidebar until marked "Done" or "Pinned". - Done emails are accessed through the "check"-button next to the "MailApp"-logo. - Pinned emails are moved to the right sidebar. This is basically a to-do list or for future reference. - Both sidebars may be filtered by attributes or labels. - The "list"-icon, next to the filter in the left sidebar, gives you the classic gmail list in the center-panel.

Some technical background:

- It's build with Rails and PostgreSQL. - E-Mail sending/receiving is based on the MailGun infrastructure. - Turbolinks (https://github.com/rails/turbolinks) and Memcache for a speedy UX. - All mail data is saved in a multi-tenant PostgreSQL db and for backup purposes in a IMAP mailbox on MailGun. - Hosted on heroku.

I'm quite happy with the app by now and only go back to Gmail for older emails.

So my BIG QUESTION right now: Would you even consider using the app, in case I'd make a commercial product out of it? Would you pay for it?

Searching HN yielded that quite a few people are looking each month to leave Gmail (for various reasons), but of course there are quite a few gotchas with my approach, especially considering trust-issues.

Let me hear what you think.




> Would you even consider using the app, in case I'd make a commercial product out of it?

No.

I use Google because of the stuff that happens in the background, that I never see. Their infrastructure, SPAM blocking, Android app, intrusion detection, seamless support for custom domains, filtering and search, etc.

GMail looks like a simple app, but it's actually a herculean effort on behalf of Google, a multi-billion dollar company. There's a reason they have very little competition in that sector.


Since upvote counts are no longer shown, I want to say that I agree with this post. I'd keep using GMail even just for the two-factor auth and security. The stellar spam filtering is another big plus.

Your app does look great, though.


not just gmail's infrastructure, but the google ecosystem.

it's like a microsoft enterprise ecosystem wet dream, but for individuals and small businesses, and without ANY cost. (iirc, it's free for companies with up to 10 people)

ps: if you don't get what i mean by ecosystem: docs/gmail/sites are the (un)holy triangle.


You say that like Gmail is the market leader here, they're not. Most of the features you mentioned do exist in other services and other services do have features that Gmail doesn't. The OP can definitely make a business out of email, it is far from a solved problem.


> You say that like Gmail is the market leader here, they're not.

They are as far as I'm concerned. I'm curious, though, who you do consider to be the market leader? I'm going to have a hard time keeping my composure if you say Hotmail (or "Live" mail), Yahoo, or something similar.


In terms of users they obviously are the market leader. I use Outlook.com personally and consider it to be overall better than Gmail. I'd gladly debate this if you'd like.

Never the less, my original point was that all of these Gmail strengths that you think are very important are obviously not very important to people using Yahoo or their ISP webmail. The idea that everyone uses Gmail and any competitor must compete feature-by-feature is just incorrect.


This sounds an awful lot like telling your customers they're wrong. And if you find yourself doing that, it's usually because you don't understand what your customer wants.


"obviously are the market leader"

Gmail only just recently (October 2012) overtook Hotmail.


Typo on my part, meant to say that Gmail isn't the market leader in terms of users, Yahoo is.


Is market leader defined in terms of quality/features? Or number of users?


Market leader would mean number of users most likely. Otherwise it's an opinion on which is a better email client and service.


Gmail may be a leader in the US, but Hotmail does have a much larger market penetration in Asian countries. And in the UK, form what I recall.


when i looked for a replacement for gmail, http://www.runbox.com seemed best, but that was a couple of years ago (instead i use procmail, mutt and mairix, but i imagine that's not for everyone...)

huh. has hn always inlined links? maybe i have never added the http: prefix before...?


Spot on. The app currently is quite egocentric, how I would like to read/organize my mails. If there is a small niche, who'd like to work the same way AND pay for it, I'd seriously consider making a business out of it.

The great thing is, to respond to the initial comment: It's build on top of MailGun, so all the nitty gritty "details" are already solved. It's a great service, really.


Email is not a solved problem? How do I use it daily without thinking about it. Solved in my mind.


That's what they said about horses. Then along came Henry Ford.


No "they" didn't, nobody ever said that horses were the ultimate form of transportation, and long before Ford came along the Railways were already whisking people along in greater safety and speed than horses.


I want to add to this, people actually hated horses with a passion, especially in cities. They were a transportation necessity, you couldn't move heavy goods without them ... but they would defecate absolutely everywhere and most cities weren't terribly organized in terms of cleaning up the poop. Which means they reeked of horseshit constantly. As noisy and polluting as cars were, people considered them a godsend.


Building email clients are big gnarly tasks...more on that in the am...


I use the Google apps version of gmail, so ymmv, but recently the spam blocking has been flagging so many legitimate emails as spam that I am actively looking at leaving the platform.


This plus trust. Not that I trust Google all that much, I just think they wouldn't risk their brand to see if I have nudie pics* in my account.

It looks really great in the screenshot, though. Pitch it to someone with an established brand?

* For the record, I do not.


Open source the app. There's a huge lacking in good web based open source mail clients. Fill it. Offer a pay-for premium hosted version. Have us pay for service, not software.

If you released this, assuming it doesn't suck balls, I'd be running it immediately. I'm currently using Rackspace for my email hosting; check them out. They're good, good be better. ;)


Haha, thanks for your frank words. I actually like your idea for 2 reasons:

1) OpenSource would definitely improve the quality of the app while still maintaining a way to profit from my work.

2) Finally giving back to OpenSource. Always felt kinda bad I haven't contributed much to a great community.


There is definatly a lot of service you could provide even if the app is open source. As GP said, hosting and infrastructure, but also better spam filtering, taking not get and remove domains from spam lists (SPF, DomainKeys, checking your domain and custom domains against spam lists and notifying the owner, &c).

Also, a nice export to S3 or a tar.gz or something would be nice if you were a private service.


I already backup all incoming mail as JSON on my Dropbox and plan to do the same periodically as MBOX. I'm not entirely sure what to do with the JSON, though. It's a "leftover" from my first prototype, where I stored ALL DATA on Dropbox.


Agreed, I'd also be running this immediately if it was open source.


completely agree. I think if you were to open source this and did it right you could get an enormous community behind it.


I've had similar thoughts to you, and it looks like a good start.

However, for me, one of the main reasons I'm dissatisfied with GMail these days is the clutter and UX. Unfortunately your app looks too similar to GMail for me to find it an improvement.

For me, the ideal GMail killer would be simpler with more clearly labelled user controls (i.e. less cryptic symbols and more standard UI buttons).

It would also have separate FORWARD and REPLY buttons. (Why on earth these most used buttons are in a drop-down is beyond me).

Don't get me wrong - I think you're onto something - I just think you need to be very careful which parts of GMail you copy, which you improve, and which you axe completely.

If you could deliver a non-bloated, snappy GMail I would consider paying for it. I would probably prefer a cloud solution though (like GMail) because I hate setting up my own email.

Good luck with it!


I had the same problems with GMail's UX until I (1) turned on button labels so that buttons displayed the action word instead of symbol and (2) learned GMail's keyboard shortcuts. The keyboard shortcuts were the biggest help because now I barely ever need to take my hands off the keyboard to perform standard actions.


Wow, very nice! Please reach out ot me: ev@mailgunhq.com I want to make sure you're not paying too much for Mailgun. We love smart and ambitious customers.


I just want to say that things like this have instantly added to the respect I have for Mailgun as a company. I know that's unrelated to the OP, and I may get down voted for this, but I feel as though for as much as a complain about an absence of good customer service - I would be remiss not to acknowledge it when I see it.

Thank you.


I'm with you.

I have had a few questions/issues during development and the live-support was simply brilliant. Meanwhile I switched 2 of my Rails apps from Sendgrid to MailGun.


Off-topic; looks like there is a mistake in the PHP example on your homepage:

    $request->setPostFields(array('from' =>'Excited User <Excited User <me@samples.mailgun.org>',


I just switched to mailgun because of your comment. Hopefully you will still be in touch with users after you get a lot bigger.


Here are the first three questions:

1. Is this something you host, or something I host? 2. How are you handling searching existing mail, tech wise? 3. Does this start to slow down after you have 10k emails in it?

If this supports custom domains and is plugged into decent spam filtering, I'd take a closer look at it.


1) Considering the hosting: It's currently basically a SaaS.

I initially had a prototype, which used Dropbox-as-a-Database. Basically I stored all emails and metadata (labels, etc) in a private dropbox in JSON-format. It actually worked, but the performance was so bad, that I ditched this approach.

2) PostgresSQL comes with a great full-text search. It's not perfect right now, but it does it job.

3) Honestly, no idea :)

Spam filtering is done by MailGun. CustomDomains are a must, of course.


It looks quite nice. I'd pay for an alternative to Gmail. Here are my suggestions / needs.

* Shortcut keys, preferably the same as Gmail but customizable

* Custom domains

* Both IMAP and ActiveSync support

* Everything stored in an encrypted state, no back doors, privacy and security as core components of your service

* The same or more storage space as Gmail provides


This app looks incredible but I have absolutely no interest in using a Gmail look alike without all the Gmail features aka android integration, amazing spam blocking, etc, etc, etc (this list could be extensive and most other people have already covered it).

You have the looks down but what you need to seriously focus on now is the back-end and how you might be able to bring all the gmail features people love along.


Absolutely. I'm quite happy with how it looks (thanks btw), but such a mission critical app requires a well though-out backend. I'm not even sure that hosting the data myself (on heroku postgresql) is even a viable option.


I'd pay in a tune of $100 for an installable version of this, but would never consider using it as a service.

PS. Though you could probably license it to smaller, well-established email service providers like pobox.com.


I would consider using the app and I would pay for it.

The primary motivator for me to change from Gmail is the privacy, security issues. So long as you make for sure everything is encrypted and I only have the key to unlocking that, I would switch in a heartbeat.


Arguably the one thing missing from Gmail right now is the ability to pick up the phone and speak to someone when something isn't working.

If you can offer that level of support at the same time, I can't see any reason for a company not to give you money.

Oh, and don't be discouraged by the people saying "no, never!" You can use third parties for handling spam, and it'll be easy enough to get mail onto peoples' phones via IMAP or Exchange. The thing to remember is that plenty of techies will say "no" because they see email as a free service. Regular consumers - and companies in particular - will be much more open to giving you money.


Thanks for the kind words.

Actually: Most of the points you raise have been solved. MailGun has (as far as i can tell after a week of usage) a great spam filter and supports IMAP.

To be honest: I wouldn't even consider building the email infrastructure part myself.


To be honest: I wouldn't even consider building the email infrastructure part myself.

Good plan, especially when there's no good reason to do so.

In all honesty, I didn't go over the technical details you provided in any detail; I just got a bit upset reading the number of "I wouldn't pay for that!"-type responses.

The thing is, I've previously seen a lot of people here claim that there's no money in time tracking tools, or any other market when, plainly, these markets do make a large amount of money. I want to emphasize that any pricing advice here should be taken with a pinch of salt. And when it comes to people who would never pay for a given service, discarded entirely ;)


To me, this looks like a reskinned Gmail. If it were that, then yes I would use it as it's an improvement over the current layout, albeit marginal. However, there are no ads which Gmail needs and imagine if Gmail didn't need adds they would use their real estate more wisely as in your design.

I use Chrome, which allows you to add styles to custom.css to override the look of any website. As such I'm able to add my own styles to customize the look of Gmail to look cleaner, hide ads and customize like in your mockup. If you package the new look as a Browser Plugin to restyle Gmail, then yes I would use it, but now I would not pay for it.

If you are attempting to recreate gmail itself, then also no I would never use it. Because once again this only offers an improvement in design. Gmail is great compared because of Google. Aka, best spam prevention, far better than Apple, Yahoo and Outlook and scaleability and up-time. There is simply no way you can compete with the resources Google has. Which is why even Apple who has incredible resources and Money and amazing design sense still can't compete with Gmail's infrastructure and servers.

There is a market for retheming however, whether in the web interface or through Apps such as with Sparrow.

So in reality stick with theming Gmail and be aware that Gmail can change their look at any time and that you won't get any direct money out of it but will get your name out there. Good luck.


I think people are leaving google because of privacy concerns. Providing a new service does not change anything from the privacy perspective.

Privacy concerned people want to host their own mail server.

Cheers, Jan


I'd consider paying for a clone of Gmail classic, but their new UI sucks IMHO. Another compelling feature I'm looking for is the ability to self-host. It's clear that many governments and cloud service providers don't consider hosted data "really all that private", despite best intentions. It only takes a single national security letter to remind us why it's nicer to have our data reposed locally.


Right, I totally understand that the people who leave Gmail for privacy concerns, most likely would host their own mail.


This looks really well done, but as a consumer I wouldn't trust a small-timer (no offense) with something as ubiquitous as my email.

As a company looking to self-host, however, your product looks very appealing, but then your battling other players than just GMail. I think in the corporate sector, it would be very hard to dislodge Outlook, which is a shame really.


This is actually my biggest concern, so no offense taken:

Despite Google's privacy policy and ad behaviour (yadda yadda), they ARE still a trustworthy company, regarding uptime , backups and everything technical. Though I remember them losing quite a few accounts about a year back. But still, your point is very valid. Why would anyone trust a small company (or even single hacker) to maintain your precious emails.


I think some of the things that could help are:

* A very straightforward exit strategy: make it easy for users to at least extract their mail from your service, if they host it on your servers, and want to have an offline backup.

* Offer an option to license for self-hosting: hosting their own email could take away their fear of uncontrollable downtime.


"I think in the corporate sector, it would be very hard to dislodge Outlook, which is a shame really."

I'm not sure how hard it would be to take some Outlook Web Access share away if targeting smaller companies since the cost could be potentially lower and licensing agreements potentially much easier to understand.


Have you seen office365.com? $6 a month per user for cloud Exchange, web e-mail, full Outlook integration if you want, plus online storage, view and editing of Word, Excel & Powerpoint. It's a lot to compete with for $6 a month.


Oh wow, ya it is a lot for a little and it is only $2 per user more if you go over 25 users. I didn't realize that you could just get online only.


I do pay $40/year for Fastmail because I like having a service I pay for that is separate from anything Google offers. I still use Gmail day to day, but the new Fastmail interface is pretty damn good.

I would definitely pay for a premium email service, but design alone wouldn't be enough for me. The key factor in my decision was trust/reliability.


The reason I'm using GMail (as opposed to FastMail) is primarily their terrific integration with Android. The excellent GMail app, the contact and calendaring syncing.

I'd really love to use Fastmail, but I see no way I could sync my contacts in a seamless way.

I can't see how your solution would solve this, so no, I would not use.


I personally would use it on top of something reliable like Fastmail. I mainly use Gmail over IMAP and it's painful slow and glitchy.

If what you're building could give me a great UI on top of Fastmail for when I want to use the web, and access to some kind of self-hosted archive, I'd probably go for it.


I pointed out further down, that my initial prototype ran completely on Dropbox. Right now, I backup all incoming mail instantly to Dropbox in JSON format for backup purposes.

I might want to look into a "on top of X"-solution. Thanks.


If you can give a trial, and nice transition option, as well as backup option to mbox format, I'd pay around the same rackspace is asking for managed email. [Added on a second thought]: I'd also consider to pay a license fee for a domain or several domains for client websites.


Very cool idea. I'd like to use the app but my main concern is spam blocking. Maybe you can partner with someone like Cloudmark? I think Cloudmark has an API or something you can use to block spam effectively.


Clickable: http://i.imgur.com/UDz2y.png

Also: Turbolinks: https://github.com/rails/turbolinks


Yes. I would surely love to use this as an email client if it works well with my Google Apps account.

And I would definitely be willing to pay a nominal amount for a year long license or a lifetime license.

What I like about your app -

* UI

* Well thought-out panels

* Utilization of white space

What you can possibly work on -

This is one feature that is not there on GMail and I really really crave for it. Sometime I feel the need to create a note for myself for a specific email I have received. But I can't do that. It would be a great addition to the Mail App.


Kudos on making a slick looking product. I'd say that if your UI was less like GMail, I'd consider using it. If it was just that, a UI for any web mail from any domain, like a Mail.app for the web, I'd definitely pay for it.

And to anyone who says they use Gmail for Security: They're an advertising company who openly mines and sells your personal data. Exactly how is that safe?

Keep on working man. Doing good.


I'd use it, and pay for it, if you gave me direct and private access to my data, hosted on a third-party. Control and portability are a huge concern of mine, and one of the reasons I'm terrified about keeping my data life with Google where I can get locked out with zero notice.

THe design looks great, and I'd love to see a host that helped me negotiate ways to protect my privacy and portability.


The UI looks good - but its not even remotely intuitive.

I had to read your entire post to figure out what the right side bar is. Its very difficult to just focus on the main content (which is what the user should actually care about). Also the title seems disjointed and un-connect from the email, I had difficult determining if they were related, or if the title was some other kind of reminder.


Are you kidding me? A "Show HN" to an imgur link of a photoshopped UI? Be honest, this is thinly (poorly) veiled ad for New Relic.


It's an actual app I've been using for a week now. Apart from the local dev database (wouldn't want to show my actual incoming emails), that's how it looks right now in my browser.

I chose the NewRelic newsletter because it somewhat fit the color scheme. So there's that...


I immediately saw and thought the same thing. I do like the UI though.


Looks good. I would love to see this as a Rails engine so I can just drop in a messaging system for users of my existing apps.


I would use it if I could host it on my own server and run it on top of Gmail using IMAP and Postgres as a cache.

I could pay 30-50$ for it, but I think this would be even more awesome as an OpenSource solution, your product rocks compared to other alternatives as SquirrelMail, and you can always offer a hosted plan as a way to make money from it.


I would definitely try it out if something can match Google's spam filters and search.

From a first look, I like the UI however, in my opinion, there's way too much white and it's not intuitive enough. I'd definitely add a little bit of colour, but it looks like a very good start.

You have nailed the visual appeal though. Great job on that!


You should disable images in mails by default (the way gmail does it) or at least add an option for disabling them. Images are usually used to track if and how many times you read the mail.

You're also missing the chat.

As for your question. I would not consider switching from gmail or paying money for an email client.


I will use it because of the UI (but for free). I love the simplicity. I'm also happy that you are using your own product. It's a good start that you satisfy needs with it.

If it has a very special feature, or will revolutionized the email, I'm sold! And I might pay for a subscription.


How did you make the layout? Graphics design is my biggest weakness, all my webapps look like the design is made in Paint. And yours is so clean, the fonts are good, everything looks so professional.


Hey thanks.

I make extensive use of http://twitter.github.com/bootstrap/ (a godsend for hackers), though I use it more as a style-guide and rebuild most parts myself.

and for icons:

http://fortawesome.github.com/Font-Awesome/

The font is "Open Sans": http://www.google.com/webfonts

The ui/ux evolved in this app, which means I started with the bare minimum requirements (it looked bad. really bad) and continually improved the "look & feel" as needed. No photoshop involved.

It's still not finished, but it has come a long way :)


Open source it. There is more value in it. You'll attract devs who want to use it and give back to it.

Is that you btw: https://github.com/jubari :)


As a quick intermission:

This is a GREAT discussion regarding the state of webmail. Although my app clearly isn't "there yet" (regarding going public), I'm having a great time reading your comments.


This interface looks really good, but as previously stated, there are a lot of behind the scenes features (security, spam filter and keyboard shortcuts) that keep me with gmail.


This is awesome! Forget the haters and keep building it. I don't think I would pay more than $5/mo but I like the direction you're going.


Hi,

I haven't seen all of your user interface to give my opinion, but I know that Sparrow (sparrowmailapp.com) had a nice one. They've even been bought by Google.

My 2 cents


I'd like it if it was turned into as a drop-in webmail solution, I haven't seen any on the market that look even remotely as good.


I have no read on market size, but it seems like there are apps that would benefit from having a fully-baked, built-in email tool.


This sounds interesting. I'd have to play with a working demo to know if its something I'd use and pay for, I'd love to see that.


I'd definitely at least try it out. I don't have a lot of bones to pick with gmail but an awesome UI can often convert me.


I'd rather like it into as a drop-in webmail solution, I haven't seen any on the market that look even remotely as good.


I would use this if you gave me a product that I can use to run my own mail (and not a service)


No, I would not pay for a different design and a few extra features of gmail...


Put an email in your profile, so that people can reach out to you.


Done. I've only been lurking & upvoting until now, so thanks for the heads up.


It is a nice attempt, but I wouldn't pay for it. Used to gmail :p


Fair enough :)


Nice design, but I don't see myself paying for it.


A lot of people here seem to like gmail.

Well gmail sucks. Its slow, buggy, now has a bad user interface and is generally not what it used to be.

It was awesome when it was first released, but there is better now,much better.

We now have outlook.com(the new hotmail) which is just awesome in every way.

And we have fastmail which I don't use, but the new ui is really groundbreaking.

I left gmail the day outlook was released; funny thing is I didnt even realize how much gmail sucked until I moved.


This reads like a bot trying to pass a Turing test? The grammatical structure is so robotic. Obviously also a total pile of stinking anecdotal nonsense.

You switched to Outlook, well, good for you. Something tells you you wouldn't recognize a good from bad ui if it bit you on your SCSI cable.


I'd of course been wanting to move to something less proprietary than Gmail for a while. The recent Gmail suckiness finally prompted me to switch to Roundcube (hosted by Gandi) a few months ago. Your new thing seems as cluttered as Gmail, so I wouldn't use it.


I like it, I think people will adopt it if it was a Chrome plugin that converted Gmail into this UI. Their infrastructure and your skin. You could charge for some features for power users that Gmail does not have.


I like your idea as it would eliminate a few of my "bottlenecks". I will check out if a chrome plugin might just do that, OR even see how http://www.streak.com/ accomplished their goals. Haven't looked into that, yet.


Awesome UI, but I wouldn't pay for it. Email is and should be free


Funny, I feel the opposite. My email is way too important to be left to a free service. I pay for Fastmail, and they do a great job. If I were to have a problem, they provide support.

I greatly prefer that model to letting Google mine my email. As others have said, you are not Google's customers with email, their advertisers are.


Actually, you're not paying for e-mail. Every common OS nowadays is equipped to be able to send mail, so paying someone else for that would indeed be not-so-great.

You are paying for the interface. He wrote software and probably devoted many hours, and that's what you are using. I can imagine you still say it's not worth paying for, especially since there are even entire operating systems that are free (of cost), but I think it's reasonable to ask. If it works better than Gmail, I can see why many would pay and switch. You may not, that's your decision of course :)


Gmail isn't free. You are paying with data about yourself, your habits, your medical conditions, your employment, etc.


How about paying a one time license fee and install the app on your own server/get your own Mailgun account etc. That way you have more control over your data.


"If you're not paying for it, you're the product" - yep.




Join us for AI Startup School this June 16-17 in San Francisco!

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

Search: