Splitting the [a] from the Commodore Amiga - Games - [ADF] dat

Aral · 11 February 2011, 08:55

This has been a hot topic for many years and it has reared its head up at TOSEC again. Should the Commodore Amiga - Games - [ADF] be split into 3 dats or remain as one giant dat?

The thinking is that there could be the following:

Commodore Amiga - Games - [ADF]
Commodore Amiga - Games - Alternate Dumps - [ADF]
Commodore Amiga - Games - Bad Dumps - [ADF]

What do you all think? Tell us your thoughts...

TCD · 11 February 2011, 09:13

If the remaining non-[a] versions are checked as working that would be a nice improvement

Certainly worth to make a seperate bad dumps DAT

Retroplay · 11 February 2011, 10:47

Sounds like a pretty good idea to me.

mai · 11 February 2011, 20:48

Quote:

Originally Posted by Aral

This has been a hot topic for many years and it has reared its head up at TOSEC again. Should the Commodore Amiga - Games - [ADF] be split into 3 dats or remain as one giant dat?

The thinking is that there could be the following:

Commodore Amiga - Games - [ADF]
Commodore Amiga - Games - Alternate Dumps - [ADF]
Commodore Amiga - Games - Bad Dumps - [ADF]

What do you all think? Tell us your thoughts...

Can you explain, why splitting the dat, is it too big?
Why creating more and more seperate Amiga dats?
First of all, i dont like the idea, but maybe you have good arguments.
My reasons against seperate dats.
-Often it is unsure, which set is the original and which ones are alternates.
-There are still sets, which are named as [b] but in true the set is not bad at all.

Ian · 11 February 2011, 20:55

Who decides which one is the alt version?

Big-Byte wrote something years ago to find out which one contained least changes but no one was really interested.

Crashdisk · 11 February 2011, 22:39

Quote:

Originally Posted by Ian

Who decides which one is the alt version?

me ^^

+1 for me :
Commodore Amiga - Games - Bad Dumps - [ADF]

-1 for:
Commodore Amiga - Games - Alternate Dumps - [ADF]
this is not an absolute science

BippyM · 12 February 2011, 12:29

I am on board with a [b] dat for each group section (apps, PD apps, Games, PD Games, Demos etc)

I don't agree with the [a] dat

I think there should be a knackered adf dat as I have loads of ADF's that cannot be loaded or looked at at all.. This would remove these

PandMonium · 12 February 2011, 13:55

I'm probably the person with less knowledge about Amiga around here but i've discussed this a bit and still maintain my point. Agree with bad dumps (or even over and under) for some dats like Amiga and Nintendo.
As for alternates, i don't find it too useful, the problem is that the Alternate flag is used wrongly most of the times.

TCD · 12 February 2011, 14:51

Quote:

Originally Posted by PandMonium

As for alternates, i don't find it too useful, the problem is that the Alternate flag is used wrongly most of the times.

Can you explain a bit more how it should be used?

Crashdisk · 13 February 2011, 18:13

Example :

Apocalypse (1994)(Virgin)(Disk 1 of 3)[cr HLM][a]
=> Apocalypse (1994)(Virgin)(Disk 1 of 3)[cr HLM][f memory GOD]

Apocalypse (1994)(Virgin)(Disk 1 of 3)[cr HLM][a2]
=> Apocalypse (1994)(Virgin)(Disk 1 of 3)[cr HLM][f memory Sphinx]

Aral · 14 February 2011, 13:18

What should happen to the [a] files if they are not split out of the dat.. This is close to the most [a] of one single version
BoBo (1988)(Infogrames)(FR)[a14].adf

And what a fine game it is to have 15 versions of it :P

Lucky Yoshizuki's Unrenamed - Commodore files havent worked there way into the dat yet or there may be 400 versions of BoBo

Another way to look at this dat is should there be a stop point of [a] versions like [a2] for instance or maybe just have one [a]. As someone said earlier how do we work out which is the good copy and which is the [a]. This dat has now grown so big that it just may not be possible to split much out of it in case mislabeled broken files are kept in the dat as good ones.

mai · 14 February 2011, 18:18

At least there should be a limit of [a] sets.
Personally i dont add any [a] set, where we already have much of these.

PandMonium · 16 February 2011, 16:12

Quote:

Originally Posted by TheCyberDruid

Can you explain a bit more how it should be used?

No problem, will try to make it brief. Also, sorry for taking so long but i've been a bit busy / tired with other stuff.

The [a] flag - alternate, is used to distinguish different dumps from similar 'original' software. For example, imagine i developed a checkers game for some old, obscure system, it would be renamed as: "Checkers (1974)(Rambo, John)". At the same time i decided to release the exact same game but the AI / difficulty was easier or you played with blacks instead of the white pieces. These small changes would result in different dumps, even if the game was the same, so this would end up named as "Checkers (1974)(Rambo, John)[a easier]".

The point above describes the normal, expected use of the alternate flag as i see it. To separate different, valid dumps of the same software. A set with [a] is not worst than the other, usually the alt is just the one discovered / renamed after, given that, there is no valid reason to separate or remove them.

The Problems with alt flag

The flag was highly abused / used in the past (and currently), because it was not well understood and was easier to solve things with it.

1) Alt flag is (AFAIK) for original dumps, when the sets are different due to user action there is a flag named [m] - modification to be used. For example, playing a game in unprotected media may lead to highscore updates, save games and other changes in it. After that, when dumped the set is different than original and should have [m highscore] or [m savegame] instead of [a]. Other modifications also apply (from fixes, cracks, translations...).

2) Poorly renamed sets, sometimes the software is indeed diferent, different versions, publishers, release years and so on but the renamer didn't noticed it. This can make the difference between 2 sets:
"Checkers (1974)(Rambo, John)" and "Checkers (1974)(Rambo, John)[a]"
or
"Checkers v1.0 (1974)(Rambo, John)" and "Checkers v1.1 (1974)(Rambo, John)"

3) Problems when dumping some media (like magnetic disks, analog(?)), in these cases it seems that each dump will probably result in slightly different set, unfortunately. I don't have a lot of knowledge about this but if it happens the number of sets must be indeed limited.

Concluding, the alt flag is indeed useful and should be used with caution. I wouldn't expect a lot of them per dat. The reality is a lot different, this flag is present in 15% of our sets and i suspect most of them are caused by what i described earlier. Crashdisk also gave an excellent example of it. Many times it is harder (or impossible) to figure out what makes a set different, it could be a bad dump, problems with dumping process or some real difference. When renaming tons of images you get tired and when seeing the same game for the tenth time, it is way easier to just put it there with a [a6] than to discover with it is an alt (eg.: [a final boss]) or that it is indeed a different version, or modified by some user (eg [m savegame]).

Hope i made it clear, bad dumps and alternates are REALLY different cases and shouldn't be treated equally.

TCD · 16 February 2011, 18:01

Thanks for the explanation PandMonium

I see the point now and will have a closer look at the differences before I use [a].

PandMonium · 16 February 2011, 18:03

No problem, is just my view about it. Feel free to help out with constructive critics and by explaining me a bit more about point 3 if you know :P

Crashdisk · 16 February 2011, 18:59

really good explanation!
For my part, some [o] or [u] are true useless junk

mai · 26 February 2011, 20:35

An example to decrease number of [a] sets:

Current:

Quote:

BuilderLand (1991)(Loriciel)[cr QTX]
BuilderLand (1991)(Loriciel)[cr QTX][a]
BuilderLand (1991)(Loriciel)[cr QTX][a2]

thereafter checking and renaming:

Quote:

BuilderLand (1991)(Loriciel)[cr QTX]
BuilderLand (1991)(Loriciel)[cr QTX][h VCS]
BuilderLand (1991)(Loriciel)[cr QTX][b checksum error]

...no [a] anymore.

PandMonium · 26 February 2011, 20:53

Awesome example mai!
That was exactly what i was talking about, there are a ton of alts but are just caused by lazy renaming :P

Crashdisk · 27 February 2011, 11:07

this is unfortunately not always practicable, the differences are sometimes very minor ex: a icon placement or worst, the disk structure

absence · 29 April 2011, 20:40

In some cases, e.g. [a5] can be more "clean" than the non-[a] version, so uncritically splitting [a] dumps from the rest (causing them to be harder to find as people focus on the "good" set) is probably not a good idea.

There are however plenty of useless alternative versions, e.g. one version has unused sectors filled with zero, another has them filled with incremented longwords. Also quite a few bad non-dos dumps are flagged as [a] instead of [b] because crackers often didn't bother with checksums, so it's hard to detect the errors (in retrospect, doscopy+ wasn't such a great idea

).

Cleaning up the mess and simply deleting the useless alternative versions would be ideal, but probably not practical as the amount of work is huge and requires in-depth technical knowledge (i.e. extract the loader code and use it to inspect differing areas of the disks to determine which one has corrupt data).

11 February 2011, 08:55	#1
Aral TOSEC-PIX Curator Join Date: Apr 2006 Location: Australia Posts: 104	Splitting the [a] from the Commodore Amiga - Games - [ADF] dat This has been a hot topic for many years and it has reared its head up at TOSEC again. Should the Commodore Amiga - Games - [ADF] be split into 3 dats or remain as one giant dat? The thinking is that there could be the following: Commodore Amiga - Games - [ADF] Commodore Amiga - Games - Alternate Dumps - [ADF] Commodore Amiga - Games - Bad Dumps - [ADF] What do you all think? Tell us your thoughts...

11 February 2011, 09:13	#2
TCD HOL/FTP busy bee Join Date: Sep 2006 Location: Germany Age: 46 Posts: 31,525	If the remaining non-[a] versions are checked as working that would be a nice improvement Certainly worth to make a seperate bad dumps DAT Last edited by TCD; 11 February 2011 at 11:28. Reason: added 'a' ;)

16 February 2011, 18:59	#16
Crashdisk Moderator Join Date: Jun 2009 Location: France Age: 46 Posts: 1,982	really good explanation! For my part, some [o] or [u] are true useless junk Last edited by Crashdisk; 16 February 2011 at 19:04.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Dynamic screen splitting/joining in a game: Was it possible on Amiga? Did it exist?	lilalurl	Retrogaming General Discussion	17	17 December 2009 20:04
Commodore Amiga - TOSEC DAT Collection (20070314_CM)	DamienD	Amiga scene	13	31 January 2008 18:16
need a .txt file of the latest official games .adf TOSEC .dat	Overdoc	project.TOSEC (amiga only)	19	01 November 2006 15:37
ADF's n' DAT's n' stuff	[idoru]	Amiga scene	8	20 December 2002 18:37

11 February 2011, 10:47	#3
Retroplay Lemon Curry ? Join Date: Sep 2004 Location: Denmark Age: 49 Posts: 4,079	Sounds like a pretty good idea to me.

11 February 2011, 20:55	#5
Ian Global Moderator Join Date: May 2001 Location: Derby, UK Age: 46 Posts: 2,287	Who decides which one is the alt version? Big-Byte wrote something years ago to find out which one contained least changes but no one was really interested.

12 February 2011, 12:29	#7
BippyM Global Moderator Join Date: Nov 2001 Location: Derby, UK Age: 48 Posts: 9,355	I am on board with a [b] dat for each group section (apps, PD apps, Games, PD Games, Demos etc) I don't agree with the [a] dat I think there should be a knackered adf dat as I have loads of ADF's that cannot be loaded or looked at at all.. This would remove these

12 February 2011, 13:55	#8
PandMonium Registered User Join Date: Sep 2008 Location: Portugal Posts: 408	I'm probably the person with less knowledge about Amiga around here but i've discussed this a bit and still maintain my point. Agree with bad dumps (or even over and under) for some dats like Amiga and Nintendo. As for alternates, i don't find it too useful, the problem is that the Alternate flag is used wrongly most of the times.

13 February 2011, 18:13	#10
Crashdisk Moderator Join Date: Jun 2009 Location: France Age: 46 Posts: 1,982	Example : Apocalypse (1994)(Virgin)(Disk 1 of 3)[cr HLM][a] => Apocalypse (1994)(Virgin)(Disk 1 of 3)[cr HLM][f memory GOD] Apocalypse (1994)(Virgin)(Disk 1 of 3)[cr HLM][a2] => Apocalypse (1994)(Virgin)(Disk 1 of 3)[cr HLM][f memory Sphinx]

14 February 2011, 13:18	#11
Aral TOSEC-PIX Curator Join Date: Apr 2006 Location: Australia Posts: 104	What should happen to the [a] files if they are not split out of the dat.. This is close to the most [a] of one single version BoBo (1988)(Infogrames)(FR)[a14].adf And what a fine game it is to have 15 versions of it :P Lucky Yoshizuki's Unrenamed - Commodore files havent worked there way into the dat yet or there may be 400 versions of BoBo Another way to look at this dat is should there be a stop point of [a] versions like [a2] for instance or maybe just have one [a]. As someone said earlier how do we work out which is the good copy and which is the [a]. This dat has now grown so big that it just may not be possible to split much out of it in case mislabeled broken files are kept in the dat as good ones.

14 February 2011, 18:18	#12
mai Registered User Join Date: Feb 2008 Location: Federativnaya Respublika Germaniya Posts: 4,994	At least there should be a limit of [a] sets. Personally i dont add any [a] set, where we already have much of these.

16 February 2011, 18:01	#14
TCD HOL/FTP busy bee Join Date: Sep 2006 Location: Germany Age: 46 Posts: 31,525	Thanks for the explanation PandMonium I see the point now and will have a closer look at the differences before I use [a].

16 February 2011, 18:03	#15
PandMonium Registered User Join Date: Sep 2008 Location: Portugal Posts: 408	No problem, is just my view about it. Feel free to help out with constructive critics and by explaining me a bit more about point 3 if you know :P

26 February 2011, 20:53	#18
PandMonium Registered User Join Date: Sep 2008 Location: Portugal Posts: 408	Awesome example mai! That was exactly what i was talking about, there are a ton of alts but are just caused by lazy renaming :P

27 February 2011, 11:07	#19
Crashdisk Moderator Join Date: Jun 2009 Location: France Age: 46 Posts: 1,982	this is unfortunately not always practicable, the differences are sometimes very minor ex: a icon placement or worst, the disk structure

29 April 2011, 20:40	#20
absence Registered User Join Date: Mar 2009 Location: moon Posts: 373	In some cases, e.g. [a5] can be more "clean" than the non-[a] version, so uncritically splitting [a] dumps from the rest (causing them to be harder to find as people focus on the "good" set) is probably not a good idea. There are however plenty of useless alternative versions, e.g. one version has unused sectors filled with zero, another has them filled with incremented longwords. Also quite a few bad non-dos dumps are flagged as [a] instead of [b] because crackers often didn't bother with checksums, so it's hard to detect the errors (in retrospect, doscopy+ wasn't such a great idea ). Cleaning up the mess and simply deleting the useless alternative versions would be ideal, but probably not practical as the amount of work is huge and requires in-depth technical knowledge (i.e. extract the loader code and use it to inspect differing areas of the disks to determine which one has corrupt data).

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)