Google Chrome - New open source web browser

**Sinkingships7** · September 4th, 2008

Originally Posted by slavik

Then there is the negative thing about it. Firefox (most likely) uses a single cache for pages, so if you close a tab and open a new one, it will use the page from the cache. If Google's browser doesn't do it, you have to load the page all over again. Once again, this is one of those tradeoff areas ...

Why couldn't Chrome use one cache as well? Why would each tab having a separate process disallow this?

**mssever** · September 4th, 2008

Originally Posted by slavik

Then there is the negative thing about it. Firefox (most likely) uses a single cache for pages, so if you close a tab and open a new one, it will use the page from the cache. If Google's browser doesn't do it, you have to load the page all over again. Once again, this is one of those tradeoff areas ...

I think that this is only relevant in the case of the memory cache. If an efficient disk cache exists, than there's no reason that couldn't be shared.

**slavik** · September 4th, 2008

@sinkingships7: read mssever's post, are implying a disk cache? a unified memory cache is not so simple since processes have their own address space (this is oneof the major differences between a process and a thread).
EDIT: making processes talk to each other while not difficult, it is not the same as having multiple threads in a single process (address space) use global variables to communicate (which is much easier).

@mssever: a unified disk cache would work, then there wouldn't be memory usage complaints as there were with firefox2. on the other hand, most if not all consumer hard drives spin down while not in use, spinning them up takes some time. this would be the time added to the latency of looking up a page in the cache.