re: Do Wifi routers "wear out" over time?

quote:

---

I know, I'm just trying to get at the fact that digital data and signals still must rely on analog systems, and analog is messy.

---

I am guessing you are referring to analog as the power system. So I can't disagree there.

quote:

---

And what causes clocks to get out of sync?

---

The crystal that drives the clock has error, even if we buy the same 12Mhz crystal it is not exactly the same frequency. They have built in tolerances such +/- 50 ppm.

quote:

---

And if a noticeable amount of data must be resent, how does that impact transmission rates / performance?

---

It gets hard to compare apples and oranges. But I am sure with your experience you know how much error there is in simple transmissions using TCP/IP. There are 1000s of variables at play for all the different error factors that cause re transmission.

quote:

---

Or they can put out enough power to boot a device but not enough to operate it under heavy load. And the way a device behaves when it is starved of power depends on the design of the device.

---

Obviously I don't design every input power supply. But I would be shocked if they didn't function the way I described. If the proper power can not be supplied, the regulator will droop and probably force a brown out condition. Most modern micros have brown out protection so it will force the device into reset.

quote:

---

I don't think that's the case, but if so that's yet another way a hardware failure can degrade performance.

---

Yea.. but that is not the circuit boards fault!!

quote:

---

I've seen plenty of routers reach total failure, but I can't recall any that went directly from working perfectly to not working at all

---

Never seen one with a failed cap? Then tend to bulge, or melt. Typically causing a power to ground short and thus will never turn on.

quote:

---

Software can be pretty shitty, but it can also be pretty good at hiding and correcting hardware problems until it can't anymore.

---

I am starting to believe you write software for a living... I am guessing you know what I do at this point.

quote:

---

Are you now agreeing that hardware malfunctions can result in degraded performance and not necessarily immediate outright failure?

---

I can agree, but do not think circuit board failure is the primary reason people chunk routers so quickly these days

quote:

---

Never seen one with a failed cap? Then tend to bulge, or melt. Typically causing a power to ground short and thus will never turn on.

---

Never seen one in a router personally, but I'm sure it happens a lot. There's a cap in my soundbar that pops every so often, I've replaced it a couple times already. Hardware fails all the time. I'm just saying it's not always immediate or catastrophic, and many systems can continue to operate (possibly erratically or in a degraded way) with failed or failing components. The behavior of hardware depends on its physical environment.

quote:

---

I am starting to believe you write software for a living

---

I do write software, but if I did it for a living I'd be dead.

quote:

---

do not think circuit board failure is the primary reason people chunk routers so quickly these days

---

Well, it can't be due to software failure, because software cannot fail! Hear me out.


Software will always operate as written. It may work in a way that was not intended, but it will 100% of the time work as written. So if a given program or set of programs no longer operates a device like it used to (after a full wipe and reinstall), then there are really only two basic reasons this can happen:

1. The inputs have changed somewhat permanently (this includes things like a different set of other devices interacting with the software, different settings, cached data used as input not cleared, etc)

2. The hardware has changed, or the way the hardware functions has changed


So whenever a system behaves erratically and you can eliminate #1, that only leaves #2 as the cause.

quote:

---

You two are playing word games now. Part failure that doesn't result in complete failure of the component can be much different from performance degradation over time. So are we talking about failures or degradation? Fault trees would be different for these two issues. Failures can be the result of degradation, but there are tons of other modes of failure.

---

I'm not trying to play word games. I just had to take the scenic route to the point that failing hardware can absolutely produce symptoms as OP described (weak signal and/or connection drops). Dropped connections can be caused by software issues, but I can't imagine a scenario where a software problem would result in a weakened signal. That's got to be environmental or hardware.

quote:

---

They're running as basic Linux servers which is hardly "full linux distros." I mean I just searched for firmware for the latest ridiculously overkill Linksys and Netgear Nighthawk routers and the firmware was <50MB... They usually have like a half gig of storage

---

LEDE Wiki

LEDE is a great example of the amount of features they are now putting in these routers.

quote:

---

Why not? You wanna give me specifics about how this happens and not just generalizations like "there's a reason things happen" and "it's like Windows"?

---

Never looked at the firmware release notes for a router?

Amplifi Release Notes

RT2600 Release Notes

Obviously these guys were fixing bugs that cause various performance and connection issues caused by software

quote:

---

We're clearly talking about stock devices here, are we not? Bringing up third party firmwares packed with features is a strawman argument.

---

I would say Netgear and other manufactures are putting in similar features. Only reason I brought up LEDE is because you can see the evolution over time

quote:

---

If the only argument you have is "look at a changelog" and "look, they fixed bugs" you don't have an argument at all.

---

You obviously didn't read them. How else am I supposed to prove that software caused performance issues. Want me to write some examples of bugs???

quote:

---

You are shifting the goalposts all over the place and seem to have forgotten your original argument. Your original argument was:

---

Fair enough. Sort of got wrapped around the axle. I will simplify my arguments.

I still stand by my argument that most hardware does not degrade in performance over time. It is more realistic to see it fail suddenly and never recover.

Most performance issues are either network interference from something new added to the environment or issues with router software ( pre-existing bugs or updates introducing new issues)

I do feel like people push the degraded hardware issue to sell more routers. Most people have way more router than they need. That was my original intent. Unfortunately I made terrible arguments on the software side.