someone passed me the link to the comic describing the new google browser (chrome) today. i looked at a couple of pages and said “was this drawn by scott mccloud?” sure enough, when i got to the end it credited him. it’s a pretty good read, and gives some good info about their new take on a web browser. it also has some pretty funny visuals when describing things, as well as funny things in the details.
there isn’t currently a linux version, so i couldn’t download and install it myself. (although they are working on one. i signed up to be notified when it is available.) also, the download was apparently disabled for awhile this afternoon. but it was available an hour or two ago, so a co-worker installed it on his windows box and we played with it for awhile.
my first — very early — impression is pretty positive.
having been the victim of firefox’s slow memory leaks, as well as feeling the pain of a runaway page or plug-in crashing everything, i’m really excited about the fact that each tab runs as a separate process. that’s something i had always wished other browsers did. (or at least a separate process for each window.)
but they went even further, and have a task manager for the browser. it shows each tab and how much memory and cpu it’s using, as well as network use. because they run plug-ins separately, they also detail those. and you can kill each one individually. all of this adds up to something beautiful in the case where you have a runaway plug-in or page. you can see what is chewing cpu or memory and kill just it. you don’t lose all of your browser tabs or windows.
i think they’ve built it using some good ideas. and since they were sort of starting from scratch they were able to design based on how browsers are used today, instead of the design of most browsers, which were built based on the type of usage and web pages in existence years ago. they also released all of it under an open source license, so any other browser (or whoever) can take and implement any of it.
it’ll be interesting to see what happens. whatever the case, i think this will probably be good for browsing in general, as well as open source projects.
Skip to content
failure is not an option. it's a requirement.
Man, that is good news. Firefox has been pissing me off lately anyway… but I still can’t bring myself to use Exploder. So, I’m not the only one that noticed the memory leak, huh? Have to kill it everyone once and awhile just to get some RAM back. separate processes, huh? Sounds brilliant (simple, but brilliant. Simple in the sense of “duh… why isn’t everyone doing that?”). You might remember the story, but there was once a dude a the place I used to work at in Austin (he didn’t last long) who said, “Threads? I don’t think that name properly describes what you are saying it does. Besides, if you did your design correctly, you wouldn’t need them!” Ok, so if anyone out there has a design for taking data off of a sampling board without threads, I’d like to see it!