It's been a long time this issue is troubling me.
There are a lot of dumps that were guessed to be "betas" in the database.
Example 1: When a GBA redump occured, the old dump is sometimes marked as beta on a guess, and recently we discovered that one of those was actually a bad dump, because we know the dumper and he explained it could not be a beta. But what about the other cases where we don't know the dumper? If they're actually bad dumps, are they going to remain forever in the database?
Example 2: When a SNES redump occured, the alternate SNES dump was marked as beta on a guess because it had a good checksum. But I know from my dumping experience it's possible to make a bad dump with good checksum, and we know it's possible to fix a bad checksum with uCON64, a tool frequently used by dumpers. So what if they're just plain bad dumps? Are they going to remain forever in the database?
There are two ways to think about it.
1. We keep all these unverified betas and the bad dumps are never going to be removed because we can't verify something that doesn't exist, can we?
2. We only include certified betas and remove all the unverified ones. If they happen to be redumped in the future we reinclude them.
So far No-Intro used the method 1 and it's just so wrong because what doesn't exist is never going to be verified. I say it's time to mature a bit and use method 2.
Thoughts?
===============================================
Re: Unverified Betas in No-Intro database \ BigFred on 12th May 2008, 07:55 wrote:
Did I miss something? Which one are you talking about?and recently we discovered that one of those was actually a bad dump
As far as I remember all of those were dumped by Eurasia. It is proven they dumped a lot of Betas when doing pre-store releases. Like the Mario Kiosk Demo or the infamous Gremlins dump. Besides - when the data is very different from the retail one and the game is still working fine it can only be a Beta. What I have doubts about are these Power Pro Kun Pocket games. If anyone could provide info about said beta/trial versions I'd be happy to hear about it.
===============================================
Re: Unverified Betas in No-Intro database \ Yakushi~Kabuto on 12th May 2008, 09:19 wrote:
CT Special Forces (Europe) (Beta) was actually a bad dump
I demand to see the documentation, scans, evidences that prove they dumped pre-release versions. Sorry but Eurasia is not a trusted dumper and these proves don't exist, all of their supposedly beta dumps are unconfirmed and will stay unconfirmed forever. So what if they're bad dumps? Are we just going to keep them forever. Is it our goal to preserve unverified dupes forever? I don't think so.It is proven they dumped a lot of Betas when doing pre-store releases.
Face it, we currently use false logic: something exists unless it's proven to not exist.
No! We must use correct logic: something exists because it's proven to exist.
===============================================
Re: Unverified Betas in No-Intro database \ neoforma on 12th May 2008, 10:21 wrote:
Well if there is no prove that a real cart exists, then I believe those unverified betas should be moved to external file. A CSV list with dump info which is supplied along with datafile or a database of unverified dumps just in case those betas will be proven to exist.
But AFAIK MSX dumps are 100% unverified, so they must go. N-Gage datafile is crap (no offence, but who needs verified cracked releases anyway?), GP-32 datafile looks like bad one too.
===============================================
Re: Unverified Betas in No-Intro database \ BigFred on 12th May 2008, 10:34 wrote:
Mario Kiosk Demo is a verified dump by Elfuego. Gremlins shows "s01" in the bootup screen rather than "s02" in the final build. I don't know what it really stands for but many games show this on startup. Looks like some software-revisioning scheme. It is not a verified dump but not a simple bad dump for sure. It is pretty obvious a group releasing games weeks before retail date have access to these versions.I demand to see the documentation, scans, evidences that prove they dumped pre-release versions.
How would they produce so many "bad dumps" and how would they work fine? Did you byte-compare some of their releases? Example given Rayman 3: 13403 differences in these 2.
No it wasn't. The supposed redump was made from a pirate cart which contained Eurasia's dump without the intro and after this discovery it was replaced by a proper redump.CT Special Forces (Europe) (Beta) was actually a bad dump
===============================================
Re: Unverified Betas in No-Intro database \ Yakushi~Kabuto on 12th May 2008, 12:29 wrote:
So it's a hacked intro removed and therefore a bad dump. You're contradicting yourself.No it wasn't. The supposed redump was made from a pirate cart which contained Eurasia's dump without the intro and after this discovery it was replaced by a proper redump.
There is too much uncertainty around Eurasia dumps to keep them in the database, their copier unit could be defectuous, they hacked their releases, etc... their dumps can't be trusted. If one of them is confirmed then fine, we keep it, but we need to get rid of the rest.
Also you still haven't adressed the main issue: how are we supposed to verify a dump that doesn't exist? For how long are we going to keep in our database bad dumps that can not be verified? How do we prove something that doesn't exist doesn't exist? I'd like to hear your answer to that.
===============================================
Re: Unverified Betas in No-Intro database \ BigFred on 12th May 2008, 13:22 wrote:
What does the fact that it was hacked have to do with it being a Beta? You claimed it was no beta but just a bad dump. Many of Eurasia's releases were intro-hacked. Other Beta-dumps in the database as well. Fruit Chase is also hacked like hell but there is no better dump. It's kept because it's the best we have.You're contradicting yourself.
Probably we won't get a redump of most Betas that exist at the current point. So you can go and delete files that obviously ARE Betas (you still didn't tell me how you explain the differences) or just live with it. Imo it's not that important since it's just unfinished software. If you go with cart-only betas you will have to remove them since there is no proof for this being the case with Eurasia's dumps. My suggestion is still listing only officially released Demos. This would make things a lot more transparent since these can be tracked down and documented without too many problems. Everything else is nice to have but could be left to goodxxx.how are we supposed to verify a dump that doesn't exist? For how long are we going to keep in our database bad dumps that can not be verified? How do we prove something that doesn't exist doesn't exist? I'd like to hear your answer to that.
===============================================
Re: Unverified Betas in No-Intro database \ Yakushi~Kabuto on 12th May 2008, 14:19 wrote:
I do not deny there is a high chance Eurasia had access to pre-release versions, but I still think we should drop all unverified betas for the sake of being logical: we know that most of unverified (supposedly) betas are never going to be verified and we should tend toward a 100% verified set. Adding entries which very existence is unproven isn't going to help. Besides most or many Eurasia releases were intro hacked then intro removed, so the rom we now have is likely different than the original beta dump, if it was a dump, because it could be leaked images that were never on a cartridge. Again, many uncertainty when it comes to unknown sources.
I support a cartridge only dat and I believe that we shouldn't include the rom in the dat as long as we don't know if a cartridge exists.
===============================================
Re: Unverified Betas in No-Intro database \ hydr0x on 12th May 2008, 16:11 wrote:
let me propose something here
1) unverified betas that apparently don't differ from the final version: remove
2) unverified betas that have significant differences from the final version: keep, possibly with special tag
3) unverified betas that have no final version at all: keep, possibly with a special tag
4) verified betas: keep
That way you will lower the number of unverified roms but still keep people happy. Of course you could modify 2) to only keep them when we have a picture proof (cart photo) of an existing beta but do not no whether this is the cart that got dumped