Tornation

Threads merged

Kid 1: Would you hurry up and take the picture?  I can't hold this smile much longer!
 
Kid 2: What are the many ways I can kill my companions in their sleep?
 
Kid 3: Look, this is the best smile you're going to get out of me, so just take the picture.

Seems legit.

What the heck is with these kids 

Using that logic I would probably be a fine pulp by now.

It's a disease, not a choice. They need therapy, not punishment!

Yeah I wish that was the case for me . But i guess I'm justt going to have to live  with this for a while now while I'm learning.

It is quite likely (given post-code 62) the motherboard you have has an older bios which does not support that cpu. In that case you will need to flash a newer bios using a cpu that works with the existing bios.

Get one of these CPUs to flash the BIOs and return it or sell it, whatever works
https://pcpartpicker.com/products/cpu/#sort=a7&f=57,73

I herd cats.

Does the makey makey even come with 2.54mm pitch jumpers? I thought it only included alligator clips. Anyway You could just buy spools of wire and crimp either dupont connectors or alligator clips.

If the price is similar, go with the i5 as they are way more powerful than any M3 (Not saying that M3 CPUs are weak).
Edit: Not "Way more powerful". They are considerably more powerful though.

m3 is fanless and silent
i5 has a fan and can get a lil loud under load. but m3 is enough for most tasks imo, it's a nice processor. my mum has a m3 based machine for watching video, web, etc, and it keeps up with her old i5-U based laptop fine.

M3 will be fine, Y vs U is max like 20-30% slower but much better battery so...
Video editing will be OK at like 1080p with 720p timelines 

How does this look if I wanted to go for a 13 inch? 
https://www.amazon.ca/Dell-Inspiron-Laptop-Generation-i5378-2885GRY/dp/B01L8PDMO6/ref=cm_cr_arp_d_product_top?ie=UTF8

http://www.newegg.com/Product/Product.aspx?Item=N82E16814137048
White accents not solid color, but not bad.

http://www.newegg.com/Product/Product.aspx?Item=N82E16814137048
White accents not solid color, but not bad.

I actually don't understand feedback these days.
 
I swear people don't watch the video - Or if they do they just skip to graphs and ignore everything I say. 
 
"But this isn't the whole story and.. AMD fans.. before you head down into the comments and start screaming at me for being KIND OF a shill at this point.. we must recognize that the 1060 is 25 to 50% more expensive than the RX480 and that's where the conversation changes.. In the dollars per frame race the RX 480 wins EVERY SINGLE TIME."
 
Then I just talk about how both the price and performance deltas are significantly different 
 
I didn't say the 480 was bad
 
1) " he said the 1060 walked all over the rx 480. The highest difference was 10 frames in a dx11 game. "
 
I did.. That's because 10FPS is "walking all over something"
 
2) " When he went to give amd credit for something, his voice lowered a bit, and seemed hesitant. "
 
You're hearing things you want to hear. Sure there was a tonality change, I was changing what I was talking about. Became hesitant? Come on dude.. I scripted my own video. I decided to put this in.. That's not hesitation. 
 
3) Where's Vulkan (which is supported by Valve) or DX12, where the rx480 beats out the 1060?
 
Vulkan came out like 2 days before the video was released. Due to editing scheduels and stuff our turn around time isn't that fast. Not only was I not even in the same country as our office when Vulkan was released but I didn't have time to completely re-do the video
 
DX12 is a broken mess that is having support pulled from multiple games due to glitches and bugs. 
 
I have a video that I'm working on overiewing both of these cool but not ready technologies 
 
4)  Next up is how scarcely I see Linus criticize nVidia products anymore.  It seems strange.
 
He doesn't review GPUs anymore.. Maybe that's it?
 
5)  Nest up is things that I've heard about involving nVidia and reviewers. Here's what I've pieced together: If a reviewer talks bad or puts a review criticizing nVidia, they don't get samples from nVidia anymore, and are discouraged. If a person leaves a good review though for nVidia, and make competitors look worse, they get more free stuff. Seeing as if we go with the talking I've heard that is industry standard for nVidia , and that Linus still receives items from nVidia, it is possible to conclude that he is creating a form of bias content.
 
Where's our resident conspiracy theorist? We could use him here.
 
We have a company of 13 people and a few contractors outside of that. You think 1 free GPU is worth risking our fanbase over? That is insanity and baseless. What might make an OUNCE of sense logically is if we were getting payouts but that is ALSO not happening. We are very transparent about what we're paid for. AMD has given us a lot more money for content than Nvidia has. Simple facts.
 
6)  I've also seen a bunch of nVidia fanboys lurk around the comments section, copy pasting stuff from a site as a reply. LTT is becoming an nVidia fanboy magnet from this. I'm pretty sure nVidia fans don't like people that talk about amd beating  nVidia in dx12 or Vulkan. This could also draw the conclusion that LTT is biased as well.
 
So because our fans like nvidia we must like nvidia as well?
We don't mind disagreeing with our fans. Hell Youtube seems to be thriving off of drama and controversy lately.. maybe we should do more of it.
 
7)  My last and final piece of evidence is the channel's subscriber count. If LTT had 10k or 100k subs, nVidia wouldn't care as much about the channel. But LTT is one of the biggest channels out their, nearly at 3 million subscribers. If they criticize nVidia, nVidia would take quite the hit. In order to keep this from happening, nVidia most likely pays and gives free stuff to LTT to keep nVidia afloat.
 
This is evidence? What? Your EVIDENCE is LITERALLY just our subscriber count? Nvidia doesn't pay us for reviews in any form.
 
One of my more successful videos lately literally had a whole section in it where I went after them for mistreating their enthusiast audience. 
 
 
I'm open to criticism.. Especially when it's constructive. I'd be more than happy to address legitimate claims with legitimate evidence. This is not that. People want to be mad right now, AMD and the RX480 are the current flavor of the month and with good reason.. It's a wonderful card at a VERY appealing price point.. especially for how it performs. Be happy with your new card, if you bought one - then proceed to move on. There is no conflict here, you're just creating one.
 

Barycentric Tehnique
It's based on a simple observation :
Consider a triangle ABC and a point P. If you take the cross product of AB and AP, you will get a vector pointing outside of the triangle if P is outside the triangle and inside the triangle vice-versa. This way we can easily figure out which side of a line a points resides in. But how should we do it for a triangle? The triangle can be oriented in any way, there isn't some value to compare with, so we take C as a reference point. Thus, for a point P, if the vector AB cross AP doesn't point in the same direction as AB cross AC, then the point P can't be inside the triangle. Otherwise, we'll also need to test the other edges. So, if P is on the same side of AB as C, on the same side of BC as A and the same side of AC as B, then P is inside a triangle.
 
We'll say the triangle can define a plane in space and we'll pick a point as an origin of that plane. The others will be relative to it. So let's pick A. Let's pick two edges of the triangle which touch A. We can get to any point of the plane by starting at A at going along AB, then in the direction of AC, etc., etc.
We'll use a barycentric coordinate system to describe a point on our plane. A barycentric coordinate system for a triangle looks something like this :
Considering a triangle with the verticies A,B,C, for every point P inside the triangle there is an unique sequence of numbers x,y>=0 and <=1 such that :
P=A + x*B + y*C
A point P is not inside the triangle if x or y <0 or x or y > 1.
P,A,B and C are constants so we only need to know x and y. We can easily get a system of 2 equations by multiplying (as in dot product) both sides once by B and then by C. Solving this system of 2 equations is trivial. You could also rewrite the whole thing under a matrix form.
 
This is basically what @fizzlesticks used in his example.
 

For a counter clockwise triangle a, b, c and point p
bool pointInTriangle(vec p, vec a, vec b, vec c) { //transform so p is at origin a -= p; b -= p; c -= p; //X = cross product vec u = b X c; vec v = c X a; //* = dot product if(u * v < 0) { return false; } vec w = a X b; if(u * w < 0) { return false; } return true; } untested but something like that.
edit: just realized you're doing 2d, it would still work if you gave everything a z of 0 but there would probably be a more efficient solution.

I'd suggest against using C++ for this, you would be much better off programming this in a language like Python, JS, or PHP, heck, even Java or C# would be better suited for it. 
 
Ninjaed by Clechay!

Pick up knitting.

There are a few race conditions and deadlocking issues. The main one being that shouldStart may be in a waiting state when the second thread exits which will cause it to sit and wait forever. In the simple case of this code that can be solved by adding a call to SignalStart after exiting the loop in the second thread, if the first thread is actually doing work you'd need to add some checks to see if it should exit early. Others involve data read/write races which can be fixed by adding more locking.
 
Condition variables have a problem called spurious wakeup. This means the wait() function may return before the cv has been notified. To fix this, the wait() function always needs to be done in a loop or by using the other overload which includes a predicate to check if the cv should actually have woken up.
 
The lock.unlock()s you have aren't necessary as the mutex will be unlocked automatically when the lock is destroyed. (you may want to keep them if you're doing work after waiting for the cv and the lock is no longer needed)
 
Here's some updated code that I believe won't have any issues. It may being doing extra locking that isn't necessary but that's better than deadlocks. Lines with //f were added or changed.
Edit: Those unique_locks that aren't being used for condition_variables should be lock_guards, they're slightly faster if you don't need unlocking and relocking capabilities.

Check out http://www.piedpiper.com /s

If i remember correctly linus did general studies or some english thing and bailed hard early on
 
I did Software Engineering then moved to comp sci because I was horribly mislead and made a terrible decision(read: shoulda stayed in soft eng)
 
I then tried to do some computer systems degree crap becuase i was unhappy with compsci.. bailed on that 1 term before my degree because I wanted to pursue this job(read: yolo)