The case of unreliable read of fake CRC err. sector

All general stuff about SPS, like philosophy and mission - discuss it here!
Petari
Posts: 32
Joined: Fri Aug 05, 2011 9:56 am

The case of unreliable read of fake CRC err. sector

Post by Petari » Fri Sep 09, 2011 12:47 pm

This may be interesting issue and proof about Kryoflux capabilities:
I imaged recently about 50 floppies. In case of game Voyager (age of disk is now 22 years) - it has Copylock on tr 0, sector 6 (slower load and fake CRC error) : I needed to repeat imaging of it with different floppy drives, as first images worked not. Only fourth drive did read it correct, and game started. I examined images and found that exactly mentioned slow sector was not readen correct by first 3 imaging. But imaging SW reported not problem as got CRC error. All it would be OK, as it was obvious at first and short testings that image is not OK. However, what with titles which have copy protection with delayed effect, or just checking for protection at some later game stages ? We can not be sure that everything is OK without playing game completely, what may be very time and nerve consuming.
I did imaging not with Kryoflux, what I have not yet. It was with Pasti, but I think that it is not so relevant here. The question is: how Kryoflux stream image can give reliable info about such errors ? Is it possible to say with confidence that such sectors are readen correct, or to say that there is error for sure. Of course that real cause of bad reading is usually condition of floppy, or not so good floppy drive (bad head align maybe) .

User avatar
IFW
Posts: 2634
Joined: Mon Nov 08, 2010 2:42 pm

Re: The case of unreliable read of fake CRC err. sector

Post by IFW » Sat Sep 10, 2011 2:37 pm

Yes, you can tell for sure for most things whether something is a good read or not, but it requires complex analysation usually not available and not even possible during dumping a disk due to he amount of computing required.

Petari
Posts: 32
Joined: Fri Aug 05, 2011 9:56 am

Re: The case of unreliable read of fake CRC err. sector

Post by Petari » Mon Sep 12, 2011 9:53 am

Forgive me, but what I see from your reply is that you are not sure, and/or not much experienced with this.
I expected that it requires complex analysis. And meant that you can determine it from image, and not some advanced user.
According to this, it is still something what requires playing game from copy until end. Of course, if it is the real goal of SP :D .

User avatar
mr.vince
Posts: 1871
Joined: Tue Oct 05, 2010 5:48 pm

The case of unreliable read of fake CRC err. sector

Post by mr.vince » Tue Sep 13, 2011 11:07 am

Forgive me, but why would we support other tools here? We have tech and tools to verify dumps made with KF.

I don't see why we would need to discuss credibility of particular SPS members or the project in general.

User avatar
IFW
Posts: 2634
Joined: Mon Nov 08, 2010 2:42 pm

Re: The case of unreliable read of fake CRC err. sector

Post by IFW » Tue Sep 13, 2011 12:04 pm

No, it does not require playing.
Please see the email I sent to you.

Petari
Posts: 32
Joined: Fri Aug 05, 2011 9:56 am

Re: The case of unreliable read of fake CRC err. sector

Post by Petari » Tue Sep 13, 2011 4:19 pm

I just try to point on potential problem. Hope that you will sort out all them by time. Not talking about them would be big mistake.

Petari
Posts: 32
Joined: Fri Aug 05, 2011 9:56 am

Re: The case of unreliable read of fake CRC err. sector

Post by Petari » Tue Sep 13, 2011 4:23 pm

mr.vince wrote:Forgive me, but why would we support other tools here? We have tech and tools to verify dumps made with KF.
I don't see why we would need to discuss credibility of particular SPS members or the project in general.
I did not ask to support other tools. I told what I experienced. I asked clear and concrete reply. I did not talk about member credibility, but about potential problem. Please, don't try to say that you have solution for everything. You should reread my previous posts here :)

User avatar
IFW
Posts: 2634
Joined: Mon Nov 08, 2010 2:42 pm

Re: The case of unreliable read of fake CRC err. sector

Post by IFW » Tue Sep 13, 2011 4:29 pm

Actually, we do have a solution for that :)
As I said in my email, you need both the clock and the data bits read and check for encoding errors - a peak shift, ie read error would normally cause an encoding violation.

User avatar
mr.vince
Posts: 1871
Joined: Tue Oct 05, 2010 5:48 pm

The case of unreliable read of fake CRC err. sector

Post by mr.vince » Sun Apr 15, 2012 1:55 pm

Guess Rosalyn is a SPAM bot.

Petari
Posts: 32
Joined: Fri Aug 05, 2011 9:56 am

Re: The case of unreliable read of fake CRC err. sector

Post by Petari » Fri Jul 27, 2012 1:05 pm

IFW wrote:I am not sure what you are referring to...
This discussion was taken offline as it was based on misunderstandings - so no need to be amazed by anything :lol:
After some time passed, I can respond. I don't think that there was misunderstanding, at least not from my side. Maybe you understood that I don't respect your work and knowledge. I'm actually like what you do, and your knowledge about floppy technic is far above mine.

What I try to do is to make some kind of different approach, view of all this. I'm sure that accessing only from technical side is not good way. Simple, whatever good is your knowledge and SW, only prectice and thorough tests can prove that everything is OK. All troubles are because copy protections. Without them, we could image things easy, fast, with less space usage etc. Not exactly true - there are troubles because publisher's shallow work too.

First question here is: what form of preservation is the best ? I'm absolutely not sure that is is exact copy of original. It is best of course in some aspects. But even 25 years ago I preferred 'crack'- here I mean my crack. I always 'cracked' originals I bought. Not to enjoy breaking law, to spread SW, but to use copies instead originals, or in case of manual protections to avoid bothering. Additionally, I improved usage my adding RAMdisk and simlar, sometimes cheats etc.
I neer went deeper in floppy protections, simply because all it interests me not much.
Now I don't use floppies at all. And my 6 floppy drives all are in not too good shape, Yes, you can buy new for some 10 Euros, but it is not really good. Or some older, DD drive - little better, I guess. But whatfor ? I have everything imaged for long time, verified, stored on multiple medias.
10 years from now, my old Sinclairs, Commodores, Ataris will barely work, or so unreliable that it will be annoying.

All in all, we need good images, and not option to write originals onto floppies. I talk it from aspect of home user. Of course, on some higher level, exact copy is necessary - who can assure working machines and floppies in good shape. All it costs money and time etc.

Ideal would be to have source codes, unprotected originals, which may still exist in many cases, just almost nobody seems to be interested to spread or arvhive them . Does SPS has some base of SW in such form ?

Another: forcing legal , fair usage today seems pretty outdated. Nobody cares about sites with tens of thousands pirate copies (planetemu, game atchives). Law makers are far behind what happens. Why prohibit something what is not possible to control ? Why not making all cheaper SW older than 20 years to freeware ? It is anyway not in sales, so nobody will loose a penny. There are more arguments, but try to keep this post shorter than Apathosaurus neck :D

So, I think that forcing legal way here makes more damage than benefit. It just results in less people interested to submit images. So, at end community is one who will be harmed, by loosing more titles.

Then, I think that despite all believes about error detections in images, you should go on gathering multiple copies of titles. Just for case of missed errors. And errors may happen not only during imagings, as we can see. Additional argument is possibility to build 100% good image from multiple sources with errors - what I made several times.
Additionally, more people involved would mean more testings.

Post Reply

Who is online

Users browsing this forum: No registered users and 1 guest