Plugins are Dead. Time to Migrate to WebRTC

What were you thinking with your plugins for Chrome?
Want to run WebRTC on anything? Check out my free WebRTC Device Cheat Sheet.
3 years now we have WebRTC. We are at a point where it is stable and ubiquitous enough. Chrome 42 has just been released. Besides improvements in WebRTC, it comes with a "minor" modification long announced - NPAPI plugins.

- NPAPI is old and outdated (think Netscape old)
- It has been announced to be removed from Chrome a long time ago
- Google is now phasing it out - requiring an option change that needs to be specifically opted in for it
- Come September this year, that option will be gone as well
- There's an alternative - Google's NaCl
- Drop support altogether. Unlikely. Those that do send a strong signal to their market about their irrelevancy: "Don't buy from us" is the message
- Switch to a full fledged app. Very likely. But reduces convenience for their end users. Not a good thing to do in 2015
- Replace NPAPI with NaCl. Very likely. Not sure how much effort this takes. Those that do send the message of clinging to the horns of the altar
- Switch to WebRTC
Need to know where WebRTC is available? Download this free WebRTC Device Cheat Sheet.


Tsahi! Great article as always.
We couldn't agree more with you. WebRTC is here and companies of all sizes will gradually move over and away from the plugin.
However, what other services are going to spawn from this? Are we going to become browser dependant with all of our communication and activities going through the browser?
Laurence,
I am not really sure, but what is it that you do today out of the browser? Office has moved to the browser (or at least starting to move there). All that's left now is "Skype".
Tsahi,
I have been looking your posts since last autumn as I am intersted in WebRTC. I see the drive of WebRTC comes from two aspects: from server side, WebRTC is a simple stack/solution to deploy video service; from client side, no need to install client or plugin.
However, the drive from client side seems not strong enough, as it is just one time effort for end user to install plugin, isn't it?
I think the drive from client side is not strong enough so that we don't see the highly and quickly adoption of WebRTC in more large scale.
Roger
Roger,
There's cost associated with any action you ask a user to do. That's why Whatsapp has been so successful in passing companies like Skype - it didn't require people to build a buddy list in order to use it - it depended on the address book you already had on your phone.
The same thinking goes for many use cases. A store trying to capture a customer on its website won't ask him to install a plugin - it will nudge him for a call directly. The less the potential customer has to think, the better it is for the seller.
To top it off, your thinking is deeply rooted in looking at communication as a service. It is not. It is a feature within another service. Embedding it into a workflow and a service on the web is rather impossible without something like WebRTC embedded within the browser.