Memtest86 is so 1986 & outdated

[David (PassMark)]

because it was made in 1986?

1994 in fact. By an individual called Chris Brady.

86 refers to the type of CPU that the software runs on. Except just recently we also added support for ARM CPUs. So it really is a bit of a misnomer now.

I mean it can't even tell me which RAM slot is giving errors?

We would love to do this. But Intel keep the memory mapping details secret (and it is complex and varies from one CPU to the next).
For ECC RAM you can sometimes get the correct slot. More people should use ECC RAM, then MemTest86 would barely be required. But again, Intel has been stopping this happening to increase profits.

it's mind blowing to me as to how & why this program is popular

Simple. It has been around a long time and it works better than anything else available. (excluding very high end dedicated testing hardware, which might cost $5K to $20K)

regarding inconsistent results

The software is deterministic. Meaning you get the same results each time if the environment doesn't change.
Unfornately the environment does change. RAM can be timing sensitive, voltage sensitive, temperature sensitive, EMI sensitive and subject to a whole bunch of "soft" errors due to radiation and cosmic rays. If the hardware operates in a random fashion, then the software running on that hardware can also be expected to run randomly.
Yes it is massively frustrating that some errors are random, but that is the way the world is. Again, ECC RAM fixes a lot of this.

4 passes is enough, no it's not enough, 8 passes but some say nothing less than overnight & others say no no leave it for 24 hours

These are all other peoples opinions. We have no control over that.
4 is enough (in our opinion) most of the time. But if you were making medical equipment, flight control systems, mars rovers and other critical stuff, then more is better. When dealing with random intermittent faults then the longer you test the more chance there is that you'll catch the random event.

XMP, turn off XMP ...... single channel or dual channel

It makes sense to test in the setup you are planning on using long term.
Which means XMP on and dual channel for most people.

Yea sure you'll give me stories of people who successfully diagnosed their faulty RAM

There are 1000s.
But yes, The result is not always conclusive. Computers are complex, sometimes you need to swap out parts and modify configurations to get to the root cause. It is more of a technicians tool than a tool for the home user.

[banned]

Seriously we're in 2021 is this the best this crappy program can do? Is it literally called Memtest86 because it was made in 1986? I mean it can't even tell me which RAM slot is giving errors? & the inconsistencies with the results geeeze... it's mind blowing to me as to how & why this program is popular yet with a few google searches it's pretty sad to see just how much confusion & frustration it has put people through regarding inconsistent results, sometimes errors, other times no errors, the whole circle of death with your confusing FAQ. It's a memtest but HEY it's not 100% guarantee it's your RAM, it could be your CPU, it could be the weather, maybe it's your BIOS, it could be your PSU, run the test again for half a year & maybe you'll see errors, 4 passes is enough, no it's not enough, 8 passes but some say nothing less than overnight & others say no no leave it for 24 hours. You should turn on XMP, turn off XMP, try overvolting, try undervolting, try it in another PC, try another PC's RAM into your PC & it goes on & on & on. I mean just in Reddit alone (let alone the whole internet) the confusion over inconsistent results & 7000 different opinions is horrendous.
Hey are you running single channel or dual channel or triple channel or quad channel? Sometimes memtest will give errors in single channel but not dual channel but only on Tuesdays after 4pm & it has to be Winter with a hint of sunshine but only when the postman arrives with jam but not any jam, it has to be apricot. I mean the confusion & the 3 billion different recommendations to test & configurations is utterly ridiculous. Again, we're in 2021 is this really the best we've got?? Is this the best you can do? Really? What do you people do in your spare time? I'd rather pay $100 for a consistent proper reliable & informative program rather than a freeware confusing & unreliable program. Oh god the HOURS & HOURS of wasted precious time this program has cost people is truly unbelievable. Yea sure you'll give me stories of people who successfully diagnosed their faulty RAM with fairy tale endings but what about all the people that don't get fairy tale endings & are stuck with more confusion then they started & not even sure how to interpret results. If I could kick this program to the face 47 times followed by 7 headbutts then dowsing it with petrol & lighting it up I would not hesitate even for a nanosecond while stomping on it's face repeatedly.

To summarize this program as a meme it would have to be the "Goats are like mushrooms, if you shoot a duck I get scared of toasters" meme.
Please do the world a favour & just stop pretending you're still improving this program & work on something a bit more 2021 & not 1986.

I mean after all the reading I don't even care to post my results & problems because what's the point? The results can change the next day or even after a few hours. Then I'll just get generic vague replies very similar to the memtest86 FAQ. Honestly what do you people even do at your headquarters?
Please change the name from Mem-test86 to Maybe-test86.