Blog

Moving Help Center

Last night we migrated the help center and blog to a new server. One issue: if you created a non-urgent support ticket between 4PM PST and 11AM PST, and you received an error page after submitting it, please re-sumbit it as we did not receive your request.

Evolving Passpack's Privacy Model

Passpack’s job is to let you store and share private data (passwords) in way that not even we can see them on our server. While it sounds very straightforward, that definition has been in the making since Dec. ’06.

When we first launched Passpack, we equally stressed both data privacy & security and personal privacy & anonymity. It’s anonymity that is up for discussion today.

Currently Passpack’s architecture has some throw-back structures to these early beginnings. Moving forward, these residual elements make it increasingly complex to solve what should be simple problems for you.

We’re planning some significant changes to the architecture, which will allow us to get a move on with some of the important features you’ve been requesting. But before we do so, we wanted to let you know about what those changes are, and give you a chance to voice your opinions.

Shared Tags, Global Groups & More

First – I’m going to details the pros and cons here, and at the end of the post there’s a survey for you to express your thoughts.

Share / Transfer Tags

This is the feature request that first made us wake up and rethink our architecture. It should be an easy enough thing to do to allow you to tag an entry, share it, and have the tags shared along with the entry itself. And technically we could do it with the existing architecture, but to do so would take an enormous amount of time and effort… and really would just be just postponing the underlying issue: are we making our users’ lives harder than need be by obsessively over-encrypting?

Current state - All tags are HPH encrypted in a single monoblock. On the server, Passpack doesn’t know the names of your tags, nor which ones are used on which entries. All of that matching happens on-the-fly on your computer after you’ve logged in.

The change – We’d like to remove the encryption from tags. This will mean that we may know what tags you are using. But unless you’re doing something really odd like using your password as a tag, this shouldn’t actually compromise your password privacy.

Global Groups & Group Admins

Global groups would allow folks to know when they’ve been assigned to a particular group and who else belongs to it. Building on that would be the ability for an account owner to designate someone inside the group as an administrator with permission to manage sharing privileges on the owner’s behalf. Both are pretty basic in most collaboration systems. Both are not possible with Passpack’s existing structure.

Current state - Passpack’s server knows the sharing nicknames of the people you’re connected with. But it does not know the names of your groups, if you’ve renamed any of your people, or which people belong to which groups. Like with tags, all of that matching happens in real-time directly on your computer. Groups exist locally, in your account, as an organizational tool only. The server knows nothing of them.

The change - We must first and foremost make the server aware of the groups you have created, and which people are assigned to them. This will reveal something about your organizational structure to Passpack, but will not compromise the privacy of your passwords.

More Flexibility in Entries

There are a bunch of suggestions in the forum which revolve around the concept of making entries more flexible. Things like customizing entries or storing things other than passwords would work perfectly fine with the system as-is. However some things, like setting reminders and tracking usage are inherently non anonymous, thus in conflict with the existing architecture.

Current state - The name, user id, password, notes, link and email of each and every entry are all HPH encrypted, in a single monoblock. If the entry is shared, Passpack knows the sharing nickname of all the folks involved, but that’s about it.

Also, the activity logs you see in your home page are not persistent. In other words, we can’t actually keep a running log.  Similar to the current approach with tags and groups, the changes in your account are deconstructed by the interface and  shown on screen as a notification on-the-fly. Once you log out, and log back in, that information is gone.

The change - We essentially want to split up the data between “needs to be encrypted” (password and notes) and “doesn’t need to be encrypted” (link and entry name).  While we could make some enhancements to entries now, it would be largely more scalable once we make this split. Also some items, like notifications, simply can’t be fixed unless we make this change.

Bonus: Customer Support

We often have folks writing in saying things like “hey I need help with the 5 entries that I shared with the group ABC.” or “the 10 entries for my tag XYZ” or simply “my Amazon entry”.  Alas, as of now, none of that information can help us locate your encrypted entries in our database. Usually, a lot of back and forth with the customer ensues to try and figure out which entries are the problem ones. It’s frustrating, or sometimes impossible and we’re literally unable to help.

The History of Passpack’s Architecture

For those who don’t already know, Passpack is Host-Proof Hosting (HPH). That means that not only is all your data encrypted, but it’s encrypted on your computer, before being sent to Passpack’s server for storage. The key to decrypt and read that data is your Packing Key, which never gets sent to the server at all. The net result is that Passpack only stores pre-encrypted data to which it does not have the key (aka: we can’t read your stuff).

That’s not changing. We are, and will always be a HPH company (heck, we’re HPH pioneers!). What is changing though is which data HPH gets applied to it, and why.

Striving for both data privacy and anonymity required us to apply HPH encryption to as much account data as possible. We jumped through amazing hoops to avoid accidentally finding out who our users were, or anything about them for that matter. But even back then, we knew a line needed to be drawn between reality and theory. We just chose to draw it in a different place than where it needs to be now.

Anonymity was a big request from our early adopter. Many early architectural decisions were based on their desire to keep their identity and activity on the web completely hidden. This was much more important to than, say, ease of use or convenience.

That was many years ago. Since then, Passpack has evolved into a collaboration tool used mostly by work groups and businesses. The large majority of these folks have a much bigger need for easy, convenient solutions, than they do for anonymity.

So… time to make some changes.

Tell Us Your Opinion

Ok, so I’ve laid it all out for you, Now tell me what you’re thinking with this super quick survey below.

Some News From The Hills

A couple of quick updates. Passpack Mobile got a mention in Inc magazine yesterday (wow!) and BizTechDay just published the video demo I did there a couple of weekends ago.

There were some interesting technologies demoing – check them out below:
(or go direct to Passpack ‘s 5 minutes of fame here)

On a side note, BizTechDay was fantastic! Edith does a great job of mixing quality speakers with an intimate sized event so that you get real face time with some pretty incredible people.

Hilights: Joan Barnes knocked it out of the park for me, with her rocky tale of a CEO going IPO, loosing herself, then finding herself again. And who can resist watching Dave McClure warn the entrepreneurs in the audience “Don’t bum rush me with your pitches when I get off this stage, or I’ll refuse to fund you just for not listening.”  Classic.

If you’re running your own company, I highly suggest you try and make the next event.

http://fora.tv/2010/10/23/BizTechDay_SF_2010_Demos#chapter_15

Making Credit Card Payments

Passpack uses PayPal to process our credit card payments. It came to our attention that many of you who’d like to pay by credit card, get set back by the login screen that appears during the payment process.

Have no fear, you do not need a PayPal login to upgrade your Passpack account. You may use any credit card.

Admittedly, PayPal does a terrible job of making this clear. In an attempt to sign up as many new users to PayPal as possible, they all but hide the link to simply make a credit card payment – without logging in.

Today we’ve added a few extra instructions to our purchasing process. Hopefully this will make paying easier for you. So if you’ve put off upgrading your Passpack account in the past, now is a good time to give it another try.

To upgrade, go to Account > Pricing and Packages inside your Passpack account.

Mobile Secure Messages: Check!

Continuing in our promise to keep delivering on Passpack Mobile, today we’ve rounded out the final features for Secure Messages.

You now have complete control of your secure messages. Beyond the sending and receiving that was already available, you now also have the ability to reply-to a message, add it to your favorites, or delete it entirely.

For tighter screen optimization, the message composing box appears only after you’ve actually selected a recipient from the send to selector. And finally, we’ve made reading the messages a little faster by removing the show/hide option on mobile (who needs the extra tap?). We’ll likely be propagating these last two tweaks to the standard web interface as well.

So now you’re fully equipped to replace texts with secure messages whenever you’re sending really sensitive information. Enjoy the privacy!

Passpack on iPhone, Android & Opera Mini. It’s Official.

After long ado, and over 1700 votes in the suggestions forum, Passpack Mobile is now official. Access your account and use your Passpack It! button to auto-login on your smart phone.

Go get it!
https://m.passpack.com

The Features Rundown

  • iPhone, iPod touch and Android in their native browsers
    (hint on your iPad, use the standard version – it’s bigger!)
    .
  • Also…  Java phones, Windows Mobile, S60, Symbian and Blackberry running Opera Mini 5.x or Skyfire 2.x
    .
  • Manage your passwords, messages and shared users
    .
  • Supports Third Party Logins (ex. OpenID, Google, Facebook)
    .
  • One-time device authentication for Two Factor Authentication
    (don’t know where to stick that Yubikey? Instructions here)
    .
  • Set mobile favorites for speedy access
    .
  • Passpack It! button for mobile - Stop tapping long passwords!

We’ve got plenty more enhancements in the pipeline, so keep the suggestions coming. And yes, further increasing speed is on our radar.

Also let us know if ti’s working on other browser/OS combinations that are working for you that we missed.

Web App or Native App?

Wondering if Passpack Mobile is an iPhone or Android native app? It’s not. No need to look for it in your marketplace, just fire up your browser and go to https://m.passpack.com.

Passpack is a web-app. We pride ourselves on finding web-based solutions that don’t require you to install anything: no plugins, no software, no apps. It’s all 24/7 access via the web, across as many browsers and devices as possible.

If you’re a developer, and interested in building out your own apps, plugins and whatnot – then vote up the API development here and don’t forget to leave a comment to tell us what you want to build!

Completing the Must Haves for Mobile

Thanks to all for taking the mobile readiness survey (it’s now closed). By combining your feedback, with our technical considerations, we’re coming up up with a working plan for the remaining mobile feature set.

What we heard from you as being most urgent was “make sure we can login with our favorite methods!” That includes both 3rd party login options, as well as being able to handle two factor authentication methods (with a Yubikey work around).

Getting You Logged In

First up, we’re in currently completing access via 3rd party login services.


(this isn’t online yet at the time of posting)

Also in the works is being able to handle your Two factor Authentication. We’re building this out to allow you to authenticate with your second factor, and save that device as “authorized” so you only have to go through the process once. This is especially handy because of the winding work around needed to get your Yubikey or email token pasted through to your USB-less phone. I’ll post instructions to the help center on how to do that.

In both cases, access is already working and we’re finishing up the relative settings pages. We’ll get these completed for you before the official launch. No release date on that, but things are progressing well and it should be a fairly quick (days not weeks) turn around.

Non-Mobile Development

After the mobile launch, we’ll be switching back to work on the standard, non-mobile application. So for those of you waiting on improvements, rest assured, they’ve not been forgotten. We’ll have another survey coming up soon.

Happy Monday everyone!

Have We Told You Lately That We Love You?

This morning, going through the replies to the mobile readiness survey I came across this very disheartening comment:

Did we forget something?
Your users who don’t use the mobile or desktop version…

Reading that felt like a punch in the gut. To the question “what’d we forget”, the answer was “your users” [cringe]. This a much bigger problem than the recent emphasis on mobile and Desktop development. It sparked me to do a little soul searching. I’d like to share some of that with you in this post.

The Fall From Grace

When we first started Passpack, we had your support, because we supported you. We chatted with you in email, over twitter, and here in the blog. We were Tara and Francescotwo founders against all odds, working from the living room and building an amazing product for you. It was awesome.

Then something changed. I think it all started with getting funded… the money, the budgets, the board members… it all distracted us. We made rookie mistakes: the blog became an oddball “industry news” thing and eventually just dwindled down to the occasional product announcement it is now; we recruited folks to help out with chatting with you out there in the wild, which backfired and came off as spammy; we hired so many new programmers that development came to a near-complete halt as we tried to train them.

We basically made every mistake in the book. We messed up.

The Incredible Come Back

In December 2008, barely a year after funding, and with a big recession looming, we were close to having burned through nearly all the cash we’d raised. So we did the first smart (though hard) thing in a while: we let the entire staff go.

Product development resumed, even though Francesco and I were frankly exhausted. We’d put ourselves, the company, and you folks through the ringer.

But we were also really determined to turn it around. And we did.

Some of our investors stepped up to the plate with follow-on cash. We wanted it to last as long as possible, so I stopped pulling a salary, Francesco slashed his, and Passpack moved back into the living room.

It worked. In May 2009 we released Passpack 7 with secure sharing. It was a triumph of blood, sweat and tears. And we did it on our own.

Today’s Reality Check

And we did it on our own.” That’s today’s reality check. It means we stopped paying enough attention to you: our users, our customers, our super-awesome beta testers and (sometimes even) our friends.

That’s bad news. As of today, I’m vowing to make it right.

The Naked Truth

Amazingly, Passpack is now in pretty good shape. Yes, we’re down to two full-timers, but we’re also nearing the much-sought-after break even point.

What does that mean for you?

For one, it means you guys are much faster at suggesting improvements, than we are rolling them out. We have to put new features in a very tight pipeline, doing just one thing at a time (the desktop/mobile juggle was an example of this). Sometimes it means we have to postpone really big projects until revenues grow further and we can invest in them properly (I believe in doing it right, or not doing it at all).

So are we as fast as lightening? No. But the good news is that we’re no longer distracted. We’ve learned our lessons. And we’re fully dedicated to you.

What Next?

I’ve been setting up some surveys here and there, and have been closely following progress on the suggestions forum. I’ve added a known issues page so you can keep track of what we’re up to as far as bug fixing. And while I’ve slowed down on Twitter personally, I’ve also set up a @passpackhelp account that you can ping for… you know… help.

But I want to do more. I’d like to talk to some of you, perhaps set up a call or grab a coffee. I’d like to hear what you have to say about Passpack – not just features, but what challenges you face with it, or how it’s succeeding in helping you.

Are you willing to chat with me? Get in touch.

And thank you. Really, really thank you.

Passpack Mobile: Express Your Opinion

Yesterday we hit a milestone with the release of auto-login for mobile (horray!). In our opinion, that was the last big hump to get over before releasing Passpack Mobile officially to the general public. But before we move forward, we’d like your opinion – is it ready for the big time yet?

Three Disclaimers:

(1) yes, we will continue development after the public release, so feel free to tell us “add it later”.

(2) No, we don’t plan on building native apps, this is a web-mobile application, and such it will remain.

And (3) make sure to check out the latest version, we just added remember me set up and fixed some bugs for Android 2.1.

So without further ado…
Update: The survey is now closed. Thanks for your help!