Flow 1.4: Heart Surgery for the Future

We’re thrilled to announce that Flow 1.4 is available for download

Here’s what’s new:

  • [New] Secrets you’ll find out about if you keep reading…that result in 2x to 3x faster transfers, in some cases.
  • [New] You can now stop transfers.
  • [New] You can now set permissions on folders recursively.
  • [New] Redesigned transcript view, which makes it far easier for you to help us figure out what’s going on internally.
  • [New] Redesigned transfers view.
  • [Fixed] Issue where some transfers could be “Pending” forever.
  • [Fixed] Issue where connections would “reset” to the root of the directory hierarchy after a period of idle time.
  • [Fixed] Many, many improvements to the editor.
  • [Changed] Various user-interface and experience improvements.
  • [Changed] Flow now requires Mac OS X 10.6 Snow Leopard in preparation for the App Store.
  • [Changed] Support for “Local” connections removed.
  • [Changed] Support for SFTP, Amazon S3, WebDAV, and MobileMe iDisk connections temporarily removed.

Yep! That’s all folks, thanks for your patience, we really think you’re gonna lo–– Okay. Fine; let’s talk about that last change. 

The features we’ve added are cool, yeah. But still, you might be wondering why it took us more than 6 months to remove 4 out of the 5 supported connection protocols from Flow.

We have been working on a brand new product that also shipped todayand on the process of rebooting this entire company. And that explains why it’s taken us so long to get this far. But…it doesn’t explain why we’ve, in some senses, moved backwards.

The truth is, this story isn’t bullshit. Our standards are higher than ever before, and when we looked at Flow –– as it stood previously –– we only had one thing to say: "It’s not good enough."

Was it the interface? Uh, no. Flow has the simplest, most intuitive, most gorgeous UI of any file-transfer client on the planet. It won a design award, for science’s sake. It wasn’t the interface.

Was it the features? You mean like our URL integration, which was not only invented by us, but still blows all other clients out of the water? Or how about our droplets, which are so small, so svelte, that they don’t even open Flow to zip your files across the ‘net? No, it wasn’t the features.

It was Flow’s heart. The beast within Flow that controlled and executed the connections themselves. The connection engine. The problem? It wasn’t stable enough, and it wasn’t built for the future. 

So we fixed it. And by “fixed,” I mean rewrote. From scratch. Entirely. 

It’s called ConnectionEngine, and it’s beautiful. It’s unimaginably elegant. It’s so meticulously designed that we’re almost overcome with guilt about not open-sourcing it.1 It’s like what Jony Ive said about the MacBook Pro’s unibody design

In many ways it’s more beautiful internally than it is externally. I think that testifies to just how much we care.

Of course, as a brand new (and crucial) component of Flow, it won’t be perfect from the start. So that’s why we’re starting out with FTP-only support. We’re going to make each protocol available –– one by one –– only when we’re satisfied with the existing protocols. 

So that’s the story. Flow has a new heart, and will grow far stronger than ever before with it. Take it for a test drive if you use FTP, and please give us feedback.

Almost. Sorry, we’re still ragingly proud capitalists.