4 ms·

Chromium Blog: Release Early, Release Often

34 points by pavs 16y ago

adg 16y ago
I think this is probably an excellent example of what can be done with a comprehensive set of automatic regression and performance tests. I'm guessing most bugs are caught immediately after each build, and they can be fairly confident releasing stuff without doing much manual testing.
- pavs 16y ago
  Also, I think there are quite a few people using beta and dev channel and chrome developers are continuously getting feedback on features being added on those channel. I have been exclusively using dev channel for 6 months, never had a problem and it updates every week.
modeless 16y ago
Chrome's pace of major version releases is already faster than any other desktop software I'm aware of, but I'm glad they're not stopping there. Other projects should learn from Chrome: remove all user interaction from your automatic update process and you can release new versions as often as you want.
- iamjustlooking 16y ago
  I keep getting burned by chrome's automatic updates. They have twice killed xmltree extension I use and I have to wait for the extension owner to update it.
  - abraham 16y ago
    Are you using the dev channel?
  - cryptoz 16y ago
    You can block their updates if you like (by editing your HOSTS file I imagine), but then you'd be running out-of-date software. I don't think that you can blame Chrome's updates for breaking that extension. It's much more likely that the extension is broken.
    - aboodman 16y ago
      In this case it was our (Chrome team's) fault. We broke an API, which we try to never do. The bug is: http://crbug.com/47726 http://crbug.com/47726. I actually thought this had already been rolled to stable, but it hasn't yet. I will check if we can get it done more quickly.
  - pavs 16y ago
    Are you referring to this one? https://chrome.google.com/extensions/detail/gbammbheopgpmaagmckhpjbfgdfkpadb https://chrome.google.com/extensions/detail/gbammbheopgpmaag... It looks like it was updated yesterday with a bug fix.
  - w1ntermute 16y ago
    I've never had an extension killed by an update and I run several, so I'm guessing it's that extension's fault.
slig 16y ago
Nice view of Chrome's marketshare after each major update: http://getclicky.com/marketshare/global/web-browsers/google-chrome/ http://getclicky.com/marketshare/global/web-browsers/google-...
heresy 16y ago
Outside of my code editor, Chrome is the piece of software I spend the majority of my time using. I've been running the bleeding edge versions since the first public releases, pretty much, and last saw the sad face in 2.x series, I think. It's such a great example of how to execute, reminds me of the IE team in the early days when they still cared. I've converted most of the office to using it, even die-hard Firefox and IE (yes, IE) users. My nagging gets them to try it, the performance keeps them :) I hope they never give up or lose inertia like IE did.

34 points by pavs 16y ago

adg 16y ago

I think this is probably an excellent example of what can be done with a comprehensive set of automatic regression and performance tests. I'm guessing most bugs are caught immediately after each build, and they can be fairly confident releasing stuff without doing much manual testing.

pavs 16y ago
Also, I think there are quite a few people using beta and dev channel and chrome developers are continuously getting feedback on features being added on those channel. I have been exclusively using dev channel for 6 months, never had a problem and it updates every week.

modeless 16y ago

Chrome's pace of major version releases is already faster than any other desktop software I'm aware of, but I'm glad they're not stopping there. Other projects should learn from Chrome: remove all user interaction from your automatic update process and you can release new versions as often as you want.

iamjustlooking 16y ago
I keep getting burned by chrome's automatic updates. They have twice killed xmltree extension I use and I have to wait for the extension owner to update it.
- abraham 16y ago
  Are you using the dev channel?
- cryptoz 16y ago
  You can block their updates if you like (by editing your HOSTS file I imagine), but then you'd be running out-of-date software. I don't think that you can blame Chrome's updates for breaking that extension. It's much more likely that the extension is broken.
  - aboodman 16y ago
    In this case it was our (Chrome team's) fault. We broke an API, which we try to never do. The bug is: http://crbug.com/47726 http://crbug.com/47726. I actually thought this had already been rolled to stable, but it hasn't yet. I will check if we can get it done more quickly.
- pavs 16y ago
  Are you referring to this one? https://chrome.google.com/extensions/detail/gbammbheopgpmaagmckhpjbfgdfkpadb https://chrome.google.com/extensions/detail/gbammbheopgpmaag... It looks like it was updated yesterday with a bug fix.
- w1ntermute 16y ago
  I've never had an extension killed by an update and I run several, so I'm guessing it's that extension's fault.

slig 16y ago

Nice view of Chrome's marketshare after each major update: http://getclicky.com/marketshare/global/web-browsers/google-chrome/ http://getclicky.com/marketshare/global/web-browsers/google-...

heresy 16y ago

Outside of my code editor, Chrome is the piece of software I spend the majority of my time using. I've been running the bleeding edge versions since the first public releases, pretty much, and last saw the sad face in 2.x series, I think. It's such a great example of how to execute, reminds me of the IE team in the early days when they still cared. I've converted most of the office to using it, even die-hard Firefox and IE (yes, IE) users. My nagging gets them to try it, the performance keeps them :) I hope they never give up or lose inertia like IE did.