Slashdot Mirror
Ask Slashdot: Options After Google Chrome Discontinues NPAPI Support?
An anonymous reader writes: I've been using Google Chrome almost exclusively for more than 3 years. I stopped using Mozilla Firefox because it was becoming bloated and slow, and I migrated all my bookmarks etc. to Chrome. Now Chrome plans to end NPAPI support — which means that I will not be able to access any sites that use Java, and I need this for work. I tried going back to Firefox for a couple of days but it still seems slow — starting it takes time, even the time taken to load a page seems more than Chrome. So what are my options now? Export all my bookmarks and go back to Mozilla Firefox and just learn to live with the performance drop? Or can I tweak Firefox performance in any way? FWIW, I am on a Windows 7 machine at work.
Have a question for Slashdot's readers? Take a look at other recent questions first to see if someone else has had a similar question. And if not, ask away! The more details and context you include, the more likely your question will be selected.
It is a mistake to discontinue the NPAPI: there are *lots* of commercial/corporate/etc. plugins using it (!)
Keep an older copy of Chrome around?
Manual installs always offer this as an option, if you have disabled the autoupdate (which sucks a ton of bandwidth anyway).
If your Firefox install and profile are reasonably old, you'll probably have a bunch of cruft. Start fresh (reinstall and start a new profile), import bookmarks, install only the addons you need. Should be plenty fast after that.
Only problem is that it seems for every new version that comes out, you have to install more and more addons just to keep the browser the same. You could always just use Firefox only when accessing a site that requires java, and use another browser for everything else.
Don't believe me?
Go to peacemaker benchmarks? Firefox uses less ram and cpu bloat. On my atom surface chrome is twice as slow and borderline unusable.
FYI you could ... gulp use IE for your work sites? You won't continue your cpu with that filth of IE 6. IE 11 is bug free and us ok. Not awesome but usable and standards compliant now.
EDGE in Windows 10 will be the best browser from what I see so far so if you're willing to upgrade next year that may solve your problem
http://saveie6.com/
There are uses for static typing and other S&M limitations on programming.
If I had a medical appliance or anything my life depends on, I'd prefer it not even do any memory management - all memory should be pre-assigned.
In academia they're emphasizing proofs of correctness too - they're all mathematicians not engineers.
And a language like Java that not only lacks dynamic types but also lacks all abstraction that could obscure what code does, such as macros or templates or overloading - it's horrible to program in, but it saves companies from the effects of having truly stupid engineers and even more incompetent managers who don't allow programmers to document their code let alone require it.
So horrible Java code has the advantage that it never does anything that can't be understood by reading the code long enough...
It's the "I can't hire competent people to save my life" department's friend. But a good programmer can accomplish a lot more in a more powerful language.
On the positive side Java and .net have better garbage collection, more scalable gc more scaleable multithread support than their competition, so there's a niche in hugeness.
You did get the part where he's talking about using Java for work, in a secure environment, yes? You aren't seriously claiming that everyone that uses SuperMicro servers doesn't care about security because their IPMI interface is a Java webstart application, are you?
I mean, for my own part, I have two choices when doing hardware tests of our appliance builds: I can drive across the Twin Cities from my home office and stand at the R&D rack in a cold and noisy staging area for several kickstart/chef bootstrap/chef converge cycles. Conversely I, as a professional, can assume the risk of using a Java IPMI interface to access a server I physically took from a box and placed in the rack of a secured staging room over a secured subnet accessed over a secured VPN connection on my development VM (with a weekly maintenance snapshot, taken every Monday morning, which I don't hesitate reverting to 'cause SystemD, but that's another story), using HTTPS with the SSL cert from that box I physically placed in the rack.
If you are somehow cracking past all those barriers into the imaging subnet of our R&D department's subnet, you've already got half a dozen usernames and passwords and have changed a cert that lives on a box whose OS has an average lifespan on the order of an hour (that is, owning that box isn't incredibly useful in and of itself). Even at that point, the new SSL cert is going to tip me off. But if somehow you managed to get past all that, with all that knowledge just to infect my desktop VM, it seems to me that you already have the keys to the kingdom, so to speak.
That is all to say, just because someone has, or even chooses, to use Java doesn't mean they don't care about security. I'm sure I don't need to explain to you of all people (I read your username and it immediately rang a bell; a quick Google search confirmed my suspicion - I run a lot of code you wrote, and most likely vice versa but to a much lesser degree)that security is about defense in layers, attack surface, vectors and risk/reward. I'm sure there are plenty of other people that use Java in their professional lives that understand and accept the risk of how and where they use it.
If I mod you up, it doesn't necessarily mean I agree with what you've said, sorry.