Comment Re:And this is why we're moving towards SSL only (Score 2) 230
That should go over really well for internet banking and other security sensitive uses.
That should go over really well for internet banking and other security sensitive uses.
The owner of the copyright on the web page isn't getting free anything from Comcast. In fact, if they're getting hit up for protection money (nice website you have there. It'd be a real shame if it took 5 minuted to load....)
But Comcast is leading the user to believe that the page looks like their modified version. If the user mods the page with plugins, they know it isn't being displayed as I intended. I don't mind the user doing that, but I do mind an intermediary doing it.
Perhaps a plugin that checks the integrity of a page against an embedded signed hash and launches a DOS against the ISP if it has been corrupted.
I would, but they don't answer any of my questions. For example, why must it be PID 1 to do those things? Why cant the service managers be in the form of optional helpers for the init scripts?
And others like Why do you keep re-inventing the wheel and then welding your 'solution' in place? Shouldn't it be up to the user if they like your solution or someone else's better?
Sure they can. They can either be in the right group or use a simple helper daemon that is not PID 1.
You plug in the USB stick, a file manager pops open. So dreadfully hard! Stick in a data cd and a file manager pops open. Plug in an audio disk and a player pops up. No systemd in sight. Next excuse please.
Actually, it is. Pid 1 shouldn't have anything for the GUI to depend on.
Systemd is a toddler with a tube of superglue and a house full of things that aren't bound together....YET
But note he indicates that he solved those needs with just a bit of customization. He then explained his objections to systemd. The rebuttal below reveals itself for what it is by claiming a laundry list of things the Swiss Army knife approach allows. All of those features already exist in systems using the old init.
Note the many many existing and well understood tools for processing and manipulating text. Wouldn't it make sense to have the 'language' they speak (text) be the standard? Every suggestion for processing those logs seems to start with use X to convert them to text...So how about just make them text?
The binary log files make as much sense as writing all English language web pages in Russian so the browser can translate them back to English for display.
Here's your car analogy, You get a new car with all the bells and whistles. It's perfect except the cup holder won't accommodate your jumbo coffee cup, so you decide to replace it with one that holds one big cup instead of a small cup and spare change. But alas, you learn that the car won't start if you swap out any of the accessories. That is systemd.
That was my first thought. Port it back to Linux so we can go back to regular old init.
Agreed 100% ESPECIALLY because computing power continues to increase.
Every single difference he mentioned is entirely addressable through different packages. In most cases they are no detriment to the other functions. How would having an excellent accelerated graphics driver harm a server? Worst case, don't load that driver. How does heavy IO capability harm a desktop machine? It's not even a case of something useless. Just try editing a video on your desktop with wimpy disk I/O.
Some days, I question if the FBI knows how to turn a computer on.
They certainly don't know how to process forensic evidence.
Meanwhile, we do know that the NSA has illegally hacked a great many routers and tapped many fiber connections. I find it more likely that they used their illegal resources to locate the server through traffic analysis. No need to invoke any superpowers.
I find it at least plausible enough to require proof from the FBI in a criminal investigation (not that it will likely be forthcoming or that the lack of it will derail the railroad).
Yes, and it too shows that the warming is real.
I don't really care to seek the needle in the rather large haystack you pointed at. I have proof of human caused global warming. I stuck it in a bottle and tossed it in the ocean. Feel free to look it up.
And yet the measurements all show an increase in temperature. Alas, the graph doesn't show what those climate models were. I could easily make climate modeling look bad if I throw in a couple nutty models to skew the results.
Always draw your curves, then plot your reading.