I don't think we have any evidence of a human living longer than 500 years, predicting that most people currently alive will not reach that 500 year mark is probably a pretty good bet.
That was my idea too, it is shifted this year. Mostly last year.
Here in the Netherlands, It was the wamest june, wettest august, warmest winter and I hear it is going to be cold this week. Eventhough they didn't expect any winter-like weather anymore this "winter".
What SPDY solves is that you don't have to wait on one response for other requests/responses on the same connection to be send. This really does help a lot.
Well, a currently when you open a webpage the browser opens 6 connections per domain and the webpage author used 6 domains for "domain sharding" to make sure all his images get loaded at the same time...
Well you can imagine that it becomes a whole lot more data if you also increase the initial window size.
- Old versions of Chrome on Windows XP also had the same problem as IE (uses the same library) - all versions of Safari on Windows XP have the same problem as IE (uses the same library)
Keep alive is used everywhere, the summary of the article had it wrong. No browser opens a new connection for each and every request.
Pipelining used to be not used because there are webservers on the Internet which don't allow and which cause your webpage to not load, so the browsers that support it have a lot of workarounds. Some of which go so far as make a new connection send the request again.
SPDY does not have that problem, it is backwardcompatible.
An interresting detail is, that many mobile phone webbrowsers actually do have pipeling enabled by default, eventhough their desktop counterparts have not.
You can get SSL for free at https://www.startssl.com/ so it isn't a matter of price. An IPv4-address could be a bit more expensive at your provider though.
SPDY is mostly useful for large sites which want to speed up loading of their site because they already fixed everything else they can fix.
When it is part of Firefox, someone just needs to write an extenstion to change the UI to allow for easily generating lots of browserids (verifiable email-address like things you own: let's you have you your own domain).
Slashdot Mirror
Floods and tsnunamis have a lot less impact on people if we don't live/work near the ocean. That would be the equivalent of man made.
I don't think we have any evidence of a human living longer than 500 years, predicting that most people currently alive will not reach that 500 year mark is probably a pretty good bet.
Would you like to play a game of global thermonuclear war ?
That was my idea too, it is shifted this year. Mostly last year.
Here in the Netherlands, It was the wamest june, wettest august, warmest winter and I hear it is going to be cold this week. Eventhough they didn't expect any winter-like weather anymore this "winter".
Actually, without the synchronization I would expect it to work better on multicore.
This doesn't make the process single threaded, just the runtime within the engine.
btrffs sucks for frequent filesyetem sync calls, even the author of btrfs says that.
Firefox is known to do many filesystem sync calls.
So I'm not surprised the combination sucked.
You should try ext4, you could even choose to use ext4 without journaling.
The free cert is domain validation, like any other non-EV cert.
Most users have no idea they can even click on the HTTPS-indicator to see the details.
My guess would be:
Average Page Load Time.
What SPDY solves is that you don't have to wait on one response for other requests/responses on the same connection to be send. This really does help a lot.
Firefox 11 is the daily build (Aurora)
I don't think so, part of the code in Firefox came directly from the Chromium project.
The Chromium project uses the Firefox SSL/TLS library and as SPDY uses HTTPS it is build on top of that library.
Well, a currently when you open a webpage the browser opens 6 connections per domain and the webpage author used 6 domains for "domain sharding" to make sure all his images get loaded at the same time...
Well you can imagine that it becomes a whole lot more data if you also increase the initial window size.
SPDY does not increase the initial window, Google, Microsoft and many others increased that on their server to get the content faster to you.
Within Google a completely different department I'm sure.
SPDY uses less connections a good thing to reduce bufferbloat.
Someone above already mentioned Jim Getty mentioned SPDY will reduce the effect of bufferbloat.
Also, I forgot something else: widespread deployment of IPv6 ? Maybe that will mean no more need for SNI ?
Whichever comes first :-)
You forgot 2 minorities I know of:
- Old versions of Chrome on Windows XP also had the same problem as IE (uses the same library)
- all versions of Safari on Windows XP have the same problem as IE (uses the same library)
If only SNI would be properly supported as well, then HTTPS deployment would be even more widespread.
Keep alive is used everywhere, the summary of the article had it wrong. No browser opens a new connection for each and every request.
Pipelining used to be not used because there are webservers on the Internet which don't allow and which cause your webpage to not load, so the browsers that support it have a lot of workarounds. Some of which go so far as make a new connection send the request again.
SPDY does not have that problem, it is backwardcompatible.
An interresting detail is, that many mobile phone webbrowsers actually do have pipeling enabled by default, eventhough their desktop counterparts have not.
The number of users behing a caching proxy has gone done considerably over the years.
As a webdeveloper I can tell you, they usually caused more problems then they solved. :-(
You can get SSL for free at https://www.startssl.com/ so it isn't a matter of price. An IPv4-address could be a bit more expensive at your provider though.
SPDY is mostly useful for large sites which want to speed up loading of their site because they already fixed everything else they can fix.
Like Facebook, Yahoo and Gmail and so on.
The good thing about this is, Firefox has a 25%+ marketshare and this will be part of the browserUI to make it really easy.
Here is an old mockup:
http://www.flickr.com/photos/azaraskin/4128966575/sizes/l/
Mozilla actually wants to solve that problem I hear. They are exploring ideas and haven't created a framework for it.
When it is part of Firefox, someone just needs to write an extenstion to change the UI to allow for easily generating lots of browserids (verifiable email-address like things you own: let's you have you your own domain).
Here are some old mockups:
http://www.flickr.com/photos/azaraskin/4128966575/sizes/l/
It is just mockups ofcourse.
Then you would be wrong.
It is a system to verify your email address, but only ones.
It is a protocol, which means it can be implemented inside the browser UI, unlike OpenID (Mozilla tried that, that wouldn't work).
Browserid.org is only used because email providers and browsers don't yet support it directly.