HN user
Yes: no software, no ports, all media is tangible, no storage, hard to fingerprint (compared to say inkjets).
However there is one problem with typewriters people generally forget: you can extract everything that has been typed from used typewriter ribbons...
As per all magazines, it declined to 10% content and 90% advertising that you had to pay a large chunk of cash for. Good riddance to them all.
there are no facts in your comment. Please cite your sources. Not only that, my hypothesis hasn't been debunked or disproved.
My point is that it will show up in a diff of the binaries.
Actually that's a really bad comparison.
The binary size may be similar but that doesn't mean the content is. Consider the two cases below to back up my assertion:
000000WE_COME_IN_PEACE000000000000000000
000000WE_COME_IN_PEACE_SHOOT_TO_KILL0000You underestimate the problem.
Look at just chrome for example, which I've posted my rationale for here:
https://news.ycombinator.com/item?id=6035091
Not a chance in hell.
That assumption relies on two uncertainties:
1. The probability of discovering something in all that binary code, especially with the intricate and non-orthogonal nature of x86/x64 assembly and odd compiler optimisations. This isn't some 80's game.
A comparison:
28500000 = Chrome binary size [1]
750000000 = Human Genome size (converted to bytes - 1BP = 2bits so 4BP per byte) [2]
2. The probability of a vulnerability being published to Google versus selling it on the private market.
Well you hope it is.
All sorts can happen between source and binary.
Yes but some things rarely work perfectly.
It scales fine. In fact it scaled for hundreds of years before electronic calendars appeared. They mobilized two world wars with pens, paper and typewriters.
Synchronisation is not an issue if you have one single source of truth.
Congratulations on making me snort tea up my nose :)
And before someone else says this is bollocks, I've worked at a company with shared source access to Windows and there are still bits you can't get at like the CSP implementations so there may still be stuff like that in there.
No open review for any crypto functions in Windows is possible.
That's basically a fucking massive red flag.
Well considering targeted advertising is fundamentally based on surveillance, it's pretty much the same thing.
I can't really be bothered with all that after basically writing that entire solution for a private company. Another metalanguage to learn...
Plus I have to have a network connection and log into a UNIX box for it which I can't do whilst I'm on the phone as it's stuck to my head at the time.
This is where Microsoft was 6 years ago. Before that, IBM/Lotus.
Times will change again.
Evil is boolean if you ask me.
Corporations don't deserve a second chance either.
You don't always hear reminders and invitations don't always work properly.
I'd rather have neither than any that aren't trustworthy.
I spent several years using outlook, apple ical and windows live calendar. I then ended up having to write a backend for icalendar protocol and saw what a shit crock everything is. It's a miracle any of it works.
Last year I moved back to a paper diary and a mechanical pencil. It is considerably more flexible, readable and editable. I hit several brick walls with computer based calendars particularly with having to add complex events like "every third Friday but not the 25th because I'm on holiday then".
Free form is sometimes better than structure.
Its like going from a relational database to a document store.
LINQ is not a magic bullet. It has some serious pitfalls.
I'd like to aee people using the right tools for the job.
True. I only connect to two hosts and ssh on from there so non issue for me.
It's also really easy to shoot yourself with unless you understand everything and think each problem through.
Four killer issues I've seen so far:
We had a major production performance issue which turned out to be a stray ToList which was causing a massive memory ballooning. Didn't get noticed in test as the test cases passed but it hit prod and 2000 users bashed it and tried to allocate 20Mb each causing our cluster to shit a brick.
Null reference exceptions! There are so many dereferencing operations in an average LINQ expression that you really have no idea which one is blowing if it goes pop in production in release config.
People using .Single(...) and getting more or less than one result back. So frustrating.
If you push an IEnumerable<T> over an interface boundary the performance and memory semantics are not preserved and you end up with a leaky abstraction. These are shits to resolve. Example: queries executing inside the view which is outside the transaction scope.
We've had to ban it in some circumstances.
Or even better, right click putty on your taskbar in windows 7 and all your profiles are listed.
To be honest I think exposure was the software as you could reliably underexpose it and get the result you wanted.
Colours, probably dynamic range as you state. Looking at the curves for a pure white photo suggests range of blue was a little tight which is what made it look like out of date Kodak ektachrome on a good day. That may be software again bit I'm suspicious as the sensor on my D3100 doesn't exhibit that problem. Perhaps its better quality control.
Yes. They are cycled out after a week. Our data is useless after then as most of it is real time. Audit logs are kept offline in gzipped daily text files per client and these are handed over to the company and deleted.
I know the details and understand pixel oversampling etc. In the real world they produce acceptable photos for the type of device but it is nothing special and definitely over-marketed.
I used an 808 for a week when they came out. Wishy washy colours, constant over exposure and distortion due to the pissy little lens. Not what they were promoting.
Whilst I compare this to a DSLR, a cheaper compact can do a better job of being a camera.
Personally I drag a DSLR around everywhere with me but that's because I started doing that sort of stuff before we had mobile phones and have developed an appreciation of quality prints.
megapixels are pretty meaningless.
A decent bottom end DSLR (Nikon D3100 for example) will run rings around it simply because of the sensor size and optics.
Any system of significant complexity be it hardware or software cannot be trusted unless it is 100% open source both from the hardware level to the software level and all systems that are used to manufacture it are open source as well and the whole process has open oversight.
Even if you tape out a CPU and ship it to the fab, they could still add stuff before it is packaged.
In conclusion, no you can't trust anything we use today. Even Stallman's open-everything laptop is open to compromise.
Pen, paper, box of dice, OTP or accept these facts.
HP are lying. The article title is correct. You can use many well known kernel vulnerabilities to get root on the box after logging in as hpsupport. At that point you can do block reads off anything you want.
1. SSH to the box
2. Username: hpsupport
3. Password (from SHA1 lookup): badg3r5
Yes that shit.
We have some of this kit in and I've tested it and it works absolutely spot on. Fortunately it's all firewalled off but it's not the sort of crap you want on your doorstep.
Nothing to do with the NSA this - just a crappy decision somewhere which is designed to make HP support's life easier. As someone else said: this bug is as old as time.
It's perfectly possible and reasonable to delete all of your data.
It's just hard.
We have to do this with our clients if they shift off our platform and believe me, it's not much fun deleting 20-100Gb datasets from a shared database with over 2000 tables in it on production kit.
But we do it, because we are honest.
Facebook are dishonest. Simple as.