Nexus 4 Design Flawed

Filed in Misc Leave a comment

I finally received my nexus 4 last week and 3 days later it was already bricked. Not software bricked, hardware bricked. Here is the story of how that happened, and my interpretation of how the nexus 4 hardware design is flawed (and should be recalled or at least result in a free bumper to mitigate the flaw).

To start off, let me describe my smart phone handling history. This nexus 4 was just my fourth smart phone, i generally kept my phone for about a year and trade up, usually to a newer used phone. I usually end up selling my old phone because it’s always in good condition and I can load it up with the newest firmware so it’s “almost” as good as the newest retail phones. The system works well as it keeps offsetting the cost of a new phone, except when your old phone is completely bricked (ugh!). I am no stranger to dropping my phone, but I can confidently say that I have never fatally dropped a phone, or even aggressively dropped a phone. I could probably claim that no drop has been from more than 3 feet in the air, often just falling out of my pocket onto my desk when I’m reaching for it clumsily. That said, my phones usually have no more than a few scuffs or scratches on the bezel and almost always have a flawless screen (though most screens have been protected by a skin).

Enter the nexus 4, replacement for my galaxy nexus. I had read (and seen, on youtube) that gorilla glass 2 was extremely resilient so I was entertaining the idea of just leaving the nexus 4 naked, but I decided to leave the included plastic lining on the glass panes until I made my decision. My decision was soon completed, and I purchased a spigen skin, figuring that it would probably be good just to have a slightly less slippery surface on both sides. Three days after I received my nexus 4 is when I realized just how fragile the phone is. I had two minor drops that day, no drops what-so-ever before. The first drop was less than a foot, onto a soft-ish surface. I would be completely shocked if the first drop had any effect on the phone’s integrity so I won’t even entertain the idea. The second drop was onto a harder tile surface, from about 3 feet up. Definitely not a drop I am used to and not one I would ever like to repeat, but my galaxy nexus had experienced that type of drop before with nothing more than a small scuff mark on the bezel. I checked my phone after the drop, looked fine. Inspected it a bit further and saw a similar small scuff on the corner of the bezel. I was unsure if the drop had produced that mark, but it was likely enough. Either way, the phone looked fine so I put it in my pocket and forgot about it. I next went to check my phone 30 minutes later, when I pulled it out of my pocket the screen was shattered (pictures below). The phone literally shattered itself sitting in my pocket, I was in shock and disbelief.  To make matters works, the digitizer was also no longer functional, so the phone was had become totally unusable, essentially bricked.  The hardware buttons still functioned but they were acting strange (sometimes the power button wouldn’t register a press). So I turned the phone off and starting to imagine how I was going to deal with this catastrophe.

My first thought is that I have a defective unit, my screen exploded without instigation, there is clearly a manufacturing flaw at play. Little did I realize, I was onto something though not what I originally thought. I originally started thinking that my nexus 4 unit received some defective gorilla glass 2, and that it had just shattered due to that defect. However, after reflection on the situation, I now believe that the problem is in a design flaw with the nexus 4. Gorilla glass 2 is indeed very strong, but all of their strength tests apply impacts at vectors orthogonal to the glass plane. It is my belief that the metal bezel of the nexus 4 acts as an amplification component to impact vectors that are coplanar with the glass plane. While a plastic bezel would likely just flex, the metal bezel pushes nearly all the impact force into the glass plane, causing a shearing effect (at least when impacting on the corners, which will happen most often). This behaviour would explain why a small scuff on the corner could cause the opposite corner to fracture so violently. I seriously doubt that I can convince Google or LG to reconsider their design for this phone but perhaps I can at least start a discussion on the potential design flaw that exists in the phone.  The main reason that I think this flaw exists is how my phone was destroyed. I didn’t receive a crack, the glass has completely fractured itself to pieces. I peeled back the shipped plastic screen guard and glass shards literally started falling off (not what you would expect from the world’s strongest cell phone glass). Additionally, this type of glass destruction would explain how the digitizer could get so damaged, only a small amount of force is required to produce a lot of shearing in the glass (this is due to the fact that gorilla glass 2 is so much thinner than the original gorilla glass).

So here is my warning: because gorilla glass 2 is so thin and the bezel is metal, the phone only requires a small amount of force to shear the glass to the critical point of fracture. Corning does not perform any stress tests that are coplanar to the glass plane (as far as I know, corning or someone correct me please), so they probably have no idea that this is an issue. Other phones have used a plastic bezel which can absorb a much larger amount of force (visualized as dents) which then transfers only minimal amount of force coplanar to the glass plane. I have a feeling that the original gorilla glass was just thick enough to not be criticized for small drops (i.e. you had to really drop the phone quite hard in order to produce enough force to shear, so the cracked screen wasn’t as surprising). Furthermore, older phones did not have such large screen as we see now, so the ratio of thickness to shear vector were not quite as large as they are now. All these factors together have culminated to produce a product that suffers from a glass shearing potential from significantly less coplanar force than most of us have become accustomed to.

The solution? well if my hypothesis is correct, you just need to dampen those corners. The edges also need to be damped but it’s so unlikely that you would drop the phone perfectly on an edge, it would also hit part of a corner. So the best solution is a simple case that can give some flex on drops that land slightly off angle, and that can dent and absorb some critical force on those hard corner drops. Personally, I think I will be picking up a Rearth Ringke slim case (sadly only available via eBay in Canada). This case is very minimal in terms of size, has decent no slip properties, should provide some level of dampening, and it’s very inexpensive. I was initially interested in the official Google bumper, but after reading up on user reports, as I expected it makes the phone feel quite bulky and large (for an already large phone).

Here are some shots of the broken nexus 4:

Close-up of the crack origin
Close-up of the crack origin

Close-up of the scuff in the opposite corner
Close-up of the scuff in the opposite corner

Full damaged phone shot
Full damaged phone shot

I don’t usually link G+ photos, so I don’t know if these will actually work. If they don’t the post they exist in is available here.

I welcome comments and criticisms of my hypothesis, so feel free to light me up if you think I’m wrong, just be prepared to defend your position with a proper argument first. Also, full disclosure, I have purchased another nexus 4 even though I think it has a design flaw. This is mostly just because I can order a case to minimize the flaw (and also be ridiculously careful now that I know the level of fragile). The truth is the phone has fantastic components that are unparalleled and I would have to shell out $300 for another galaxy nexus anyways, so might as well just get another  nexus 4, which was available in stock today. So Feb 13th, I should have my replacement. I still have some hopes that I can get Google or LG to warranty replace this (or that they will recall this first batch of phones) but I won’t hold my breath for that. Pretty sure I’m just going to have to eat the dead nexus 4, wait for the screen/digitizer replacement prices to come down and repair it myself.

, , , , , , ,

Making Google Chrome Work With a SOCKS5 proxy (i.e. putty ssh tunnel)

Filed in Software Developement 6 Comments

I really like chrome, but something that is an absolute must is a SOCKS5 proxy.  This is due to my ultra restrictive corporate firewall, i need to tunnel http content through an ssh tunnel.  And ssh creates a SOCKS5 proxy when you use the -D option.  Chrome seems like it assumes that your proxy is SOCKS4.x and just fails on the ssh tunnel proxy.  But there is hope, i found a way to work around this and it isn’t even complicated!

Just a quick note, i actually use a plugin from chrome called Switchy! which helps me quickly switch to and from the ssh tunnel proxy.  It is certainly no foxyproxy, but it works well enough that i can use it to solve most proxy related problems.

Now, the secret to this solution is to use Proxy Auto Configuration scripts.  These scripts allow you to specify which version of SOCKS to use for your proxy.  So all you need to do is create a file somewhere on your computer (say called pac-ssh-tunnel.pac) and then add the following to it:

function FindProxyForURL(url, host)
   return "SOCKS5 localhost:8080";

Now, take note that i am creating the ssh tunnel proxy using the following command:

ssh -N -g -D 8080 username@remote_server

Just FYI, “-N” means don’t execute any commands on the server (i.e. strictly a tunnel connection only).  And “-g” allows the remote host to connect to locally forwarded ports (which i must admit, most people will never need this, but handy if you do complex tunneling of data).   Finally, “-D 8080″ means dynamically forward the proxy data through local port 8080.

So once you have your SSH SOCKS5 tunnel up and running, set chrome to use your proxy automatic configuration script that you created (either through chrome’s options, or through Switchy! if you prefer).  Now you can proxy traffic over your ssh tunnel.

One final note, PAC files are basically javascript files (with a few built-in functions).  You can actually create some complicated PAC files that do all your complex proxy selection for you (basically a lot of regular expression matching).  So if you get adventurous, you can just add all your proxy selection logic to that one function (or break it into multiple functions) and then you won’t even need to change your proxy around ever!

Happy surfing!

, , , , , , ,