feelix

Interesting. And also interestingly, they use the same method in their "fast_memeq" implementation:

if (* src1_as_int ++ != * src2_as_int ++) return 0; /* compare as ints */

Yes, that's true, because memcmp returns the sign based on the first byte, and this compares four bytes, on little endian architectures it may return the sign differently from the built-in memcmp(). That should be mentioned along with the caveats, that it's not a suitable drop-in replacement when people are not simply testing that the result is 0 (which I imagine to be the majority of cases).

It checks that it wont overflow beforehand by making sure the length is >= 4 (and it's still faster).

And I'm not sure why you think inlining would result in the speed increase? I believe the compiler will inline it when the standard memcmp is used too.

Things like this are why I will never trust any computer with privacy.

There are processors and memory systems everywhere. There are so many entry points for snoop attacks which can be completely invisible unless you know exactly what to look for.

If your freakin' hard drive can be running linux or a web server then how can you ever trust anything?

The first line of the article is: "The solution to all of Earth’s metal problems, or the space rock that would completely obliterate the world economy"

writing random data on it, rather than zeroing it out, will avoid block de-duplication and successfully completely fill it up.

Here is a tool written for exactly that (although not intended for securely erasing a drive, it will have that effect too): https://github.com/rentzsch/stressdrive

To someone who has not tried living overseas this may sound like a classic 'first world problem', and I suppose it is.

But what I can't stress enough is how severe it can be. I went and lived in various places around the world for a decade, from about 23 to 33. The place I settled down most was in Ecuador, where I learnt the language, met my long term partner, and made some close friends and integrated with the culture. But I spent years in Europe, the USA, and Asia too.

All of the above was achieved just by coding software and selling it off some wordpress websites I put together.

And now I've been back in New Zealand for exactly 3 years, and it's awful. Life is so repetitious and small and boring. I've never been more miserable. And it's for the same reasons in this article, and that other people are mentioning in this thread.

The funny thing is that before I left I had no particular desire to go (I just did it because it was easy and my friend wanted to), and I was happy. But now, because I have had those life changing experiences, I'm not.

I have found that the source device often reveals more than expected with public figures. Here's Tim Cook's, for another example: https://saucetweet.com/?username=tim_cook There is a lack of tweets from a Mac compared to other device types (iPhone/iPad).

In the case of Trump, he only tweets from his Android, so all of the other tweets are done by his team. If you read through them side by side the difference is night and day in their intonation too.

This is all made easy by the fact that the 'source' is a publicly available part of the Twitter API: "'source' 'String' : 'Utility used to post the Tweet, as an HTML-formatted string. Tweets from the Twitter website have a source value of web.'"

From: https://dev.twitter.com/overview/api/tweets

Any chance of sharing/providing the source code of the listener bot that auto responds?

Direct link to their press release: http://ir.kitepharma.com/releasedetail.cfm?ReleaseID=1014817

> And the gig has gotten harder for longtime drivers. In 2012, Uber Black cost riders $4.90 per mile and $1.25 per minute in San Francisco, according to an old version of Uber's website. Today, Uber charges $3.75 per mile and $0.65 per minute. Black car drivers get paid less and their business faces far more competition from other Uber services.

Considering that he did change the price from under the driver, and that is very likely what did indeed bankrupt the guy, I don't blame him from acting out in any way he can.

The fact that the CEO just screams out "BULLSHIT" when he points out the price decrease which affected him and that his only response just makes me lose respect for him completely.

It's not. I submitted it earlier and quickly deleted it.

Me personally? I never had that fear. I launched many things, and like with everybody else who launches things, most of them failed.

If you find yourself unable to launch stuff because of that specific fear, then I would say it has some extremely personal reasoning behind it, and maybe introspection would be helpful. Things that deeply ingrained are usually related to childhood and or family.

Ironically, I'd probably suggest some psychedelics as the fastest way to get to the bottom of it.

You can run Marathon 2 on the latest macOS, Sierra https://alephone.lhowon.org/

No, it's permanent. Try the app and I think you'll see a longer history.

If you're concerned about artifacts being left around, then you should also know that there is a log file of every file you've ever downloaded, including in private browsing mode. You can see it with this: https://macdaddy.io/cleandisk/

I wonder if this issue could be helped with a good rating system

It is indeed slow (of course, pretty much everything is slow on something like a Pi). But It's still fast enough for some uses. If you can even get one inference every 5 seconds that still has a lot of applications. And that was what I was saying when I said that I don't agree with the assumption that google are working on giving TensorFlow cloud support for all of their support of the Pi. Running locally could have a lot of uses too.

Besides which, they have been implementing things like 8 bit graphs for processing on low power devices. That should result in a large performance increase for these devices. I tried it on mobile and I got decent FPS (I can't remember the exact figure) by using it. https://www.tensorflow.org/versions/r1.0/how_tos/quantizatio...

I understand by your name that you may be a neural net enthusiast, but I would question how much practical experience you have.

It's a perfectly valid statement to say that a pi will run inference on 90% of the models out there, and I have experience with the same. It would be similar to claiming that (if it could), a pi could run 90% of the games out there.

Once again, I am speaking from practical experience from having implemented tensorflow neural nets on low power devices. And I sincerely get the feeling that although you're enthusiastic, you have no clue what you're talking about.

Rather than making offhand comments like facepalm, I would challenge you to either offer up some evidence to the contrary (you could start by trying to find a tensorflow model that a pi wont run), or spend more of your time doing something more practical than acting like a clueless rabid fanboy.

That is false. A raspberry pi is capable of running inference on 90% of the models out there.

Inference as in the opposite of training.

An example on inference would be feeding it an image and classifying it with a neural net.

The example I linked to above has inception running, which classifies things you point the camera into 1000 different categories.

It is very easy to set up (I have done it, and it only took a few minutes)

A raspberry pi is perfectly capable of running inference at around 10 FPS or more easily.

You obviously wont do any training on the pi, but low power devices have been used for inference for years now. For example here it is made to run on a phone: https://github.com/tensorflow/tensorflow/tree/master/tensorf...

Considering that it's costing them their business and fame, I doubt that they would see it like that.

Apologies if I'm being stupid, but I can't find the details on how to physically connect the hardware together anywhere. Is this still on the todo list? I'm interested in applying this tutorial and making an autonomous RC car.

Besides being an indicator of inefficiency, RAM uses a lot of power. In the region of 10-20% of overall system power. That's not even factoring swap space into account, which is a greater burden on power.

>so it seems to me that (&arr + 1) dereferences the pointer to the array

It doesn't. Because an array is already a pointer, in (&arr + 1) &arr is a pointer to a pointer (ie, a handle) so *(&arr) is dereferencing the handle to the pointer. So it's still one pointer level deep - it doesn't dereference it completely.

"it shall not be used as the operand of a unary * operator that is evaluated"

he doesn't use the * operator on it, he just calculates its position. If he were to access it (ie, use it with *) then that would be breaking the rule

In practice, although HDD's are meant to be able to deal with bad blocks, and remap them, I have found the opposite to be true.

(I've been working at a low level with disks for over a decade, writing data recovery apps and such)

I have found in practice that by the time a HDD develops a single bad block, 99% of the time it's on a death spiral and it's going to completely die within a month or two.

It seems like none of this bad block remapping, spare space, or anything of the sort has any effect whatsoever. The only thing that keeps them alive is being well enough engineered not to break at all.

Please correct me if my experienced misinformed me and I'm wrong.

I wonder if a good way to go would be to implement standard encryption (AES in most cases), and then take that encrypted payload and encrypt it with your custom code that you designed.

So even if it's actually true, and that you shouldn't roll your own encryption in practice, even if they do break yours they'll still have to get through the AES as well. So it's a no risk addition.

I've always wondered why they dont stick a camera on the front of it, then overlay the virtual world onto the physical world.