Originally posted by: MrMark0673
First, EEPROMs =/= EPROMs. Not too concerned about UV influencing my EEPROMs...
I'm not sure what protos I sold to you, I've sold hundreds over the years. Were the EPROM windows uncovered? If not, the volatility of the EPROM is your concern, and the UV exposure is a moot point. Also, if Lincoln dumped it, it was overwhelmingly likely that it was a proto that had issues to begin with as I only ever sent Lincoln things that I couldn't successfully dump it myself. What caused the issues is anyone's guess, they had a long life before ending up in my hands.
When I dump a retail proto, I immediately compare it to retail. In doing such, all of the "final" versions of games I've dumped have been byte for byte identical to the retail counterpart. Pretty easy way to identify whether a byte has been changed. With hundreds of examples to my credit, and with so many examples of byte for byte identical code, I personally am not concerned about the volatility of these EPROMs.
You can certainly call it confirmation bias if you'd like, but I've had far more experience than most here in proto preservation and I have no concerns at all about taking a quick picture with the EPROMs uncovered. Far too many sideline experts demand people cover the EPROMs while doing little personally to help preserve these pieces. It just comes off as silly.
Ichinisan has been dumping unreleased ROMs since the SNES was still current so he's probably not the sideline expert others might assume him to be. Heck, it even extends beyond mask ROMs: he's recovered Nintendo confidential utilities deleted from a serviced Wii by way of a raw NAND dump and he worked with someone else to dump and extract Master Quest from the GC disc before it was released. This was also before any GC rip method/utility was made public... though that was mostly thanks to an IRC EFNET friend who happened to be local and was privy to a release group's inside utility (we already had the PSO exploit running, at least).
My phone's auto-generated dictionary frequently "corrects" EPROM" to "EEPROM" just because I routinely use that (correctly) in other discussions. It should be pretty clear now that we know the difference so I apologize for the distraction.
Other than being long-time preservationists involved with obtaining and dumping early copies and protos, Ichinisan and I have even more in common with you and BeaglePuss: we are also twins.
I'm not blaming you for anything. I was simply responding to your anecdote intended to show how hardy they are with one that shows the opposite while hopefully demonstrating how easy it is to ignore contrary evidence (especially where we can't even look/test). I only mentioned that it came from you so it would be clear how close to home this particular anecdote it is.
No, the EPROMs weren't uncovered, though they use paper and not the reflective stickers made to block UV (likely the proto assembler's choice). No one is saying you should have peeled off the dev's original sticker (GOD NO!
) or are at fault for not protecting it. I wasn't implying that it had been mishandled in any way, only that the data can be more fragile than you and others might realize. The issue that required Lincoln was actually that it used a pinout more like a maskROM than a typically EPROM (Toshiba TC571001AD). He encountered the problematic byte after he got it to read/dump.
Yes, the volatility of my EPROMs are my concern/responsibility, but my concern extends to other prototype owners who may put too much confidence in the assurances that they will be OK. I can similarly ask how much time the arcade board was outside of the cabinet and how long it was exposed to UV light. How much UV exposure was it cumulatively exposed to? Just because it didn't have an immediate effect, how can you know if it will have an effect over the long-term (hasten bit rot; shave off 5 years of the lifetime, etc)?
The retail code is obviously preserved in the mass-produced copies so comparing a retail proto to a dump of the final retail game is of limited usefulness from a preservationist perspective but it's good for determining if your dump truly is final or not and it is good for seeing if there was any immediate corruption. For an undumped prototype of an unreleased game with unknown prior exposure (as is the case with some here) it is much more important to limit exposure, at least until it is dumped.
Now, I understand that you mention comparing dumps of hundreds of bit-for-bit identical retail protos with exposed EPROMs without finding any changes as evidence that it doesn't have measurable effect regardless of how final it is. Unfortunately, no one can say how it extrapolates to the data's integrity over time. That kind of testing take more time than any of us have (decades and decades).
Even though every single bit retained the correct value, how much time did momentary exposure shave off of their life before the cells would have lost their charge without exposure? Could be none, could be years. We can't quantify that and should probably hesitate to suggest exposure for EPROMs that have not been dumped. It may be an abundance of caution, but it just makes sense when it's so easy to cover the windows.
Think about it: if they reach that point on their own even without exposure ("bit rot") then even momentary exposure could push it closer to the tipping point (even if it doesn't immediately tip it past or corrupt it any time soon). It may be exponentially more sensitive as the years pass so advice that was perfectly fine for a fresh EPROM could be increasingly risky. All Ichinisan is saying is that it's better to be safe than sorry. At the very least, make sure it is backed up first before exposing them for something as trivial as a new photograph for this thread.
Now, it's important to distinguish perspective. I can read the datasheets and learn about the inner workings of EPROMs interacting with UV and I come to the exact same conclusion as you and the experts: momentary UV exposure is generally safe. You can generally expect to pop your uncovered chip out of the programmer and into a shell and expect it to last for decades... but a 30 year old undumped prototype with an exposed window and an unknown exposure history is not a "general" scenario. It's pretty specific.
To take that general advice and say that it's OK to expose 30+ year old chips that have undumped data for something as trivial as a photograph is a completely different beast. Lacking contrary evidence and seemingly supported by confirmation bias, this general advice has trickled down even into prototype/preservationist circles. The power of suggestion trumps even personal experience, which is why so many people now blindly believe that blowing in NES games did nothing after a MentalFloss article popularized the idea. It mindlessly repeated as fact on science channels and quizzes as an example of "confirmation bias," but these people failed to do any actual testing. I did a little bit and came to the opposite conclusion.
It's sometimes easier to play it safe than to test, which is the case here. I mean, who has decades of controlled EPROM testing with significant sample sizes to refer to? What sounds safer: taking a risk based on possible but untested assurances or not taking the risk based on possible but untested warnings? If being overly cautious is still just a matter of covering a window with a sticker or tape, I'm a lot more confident advising that for obvious reasons.
An expert discussing the functionality of the EPROM is not speaking from the perspective of preserving data that only exists on that EPROM. Without that perspective then the EPROM will work fine and if it fails the data can almost always be written back onto it.
No one's life/health is on the line here, but "I did it and it was fine and it is usually fine therefore it is always fine" is about like assuming that smoking doesn't cause lung cancer (I have an anecdote about that too). Just because you smoked several times and may know people who got lung cancer without smoking ignores the mountains of testing that proves it. Who is doing that testing here? No one. That is why we should exercise caution. If we're wrong, it's no big deal.
If curious about that lung cancer anecdote: Not even a year ago I witnessed a completely ignorant person respond to his daughter's pleas to stop smoking after his mother was diagnosed with lung cancer. His response was "that don't got nothin' to do with it. My great grandma smoked every day and lived to [90-something] without ever getting lung cancer." Guess who was diagnosed with lung cancer by the end of the year? That same yokel. His mother died earlier this year. His daughter is my niece. Thankfully, I'm related through her mother.