Garrison I & II working version(s),
 

Since there is some interest in fixing this game, and assuming I am not mistaken in thinking there is simply no way I can play either game from my Gotek as if it was the retail version, here is the opportunity to discuss the potential issues with the IPF's or / and with the available cracked versions.

As I said earlier, none of the cracked versions would allow me to pick up different characters, all characters look the same when entering the game area. They shoot their respective weapons, but each of them look like the mage. Hit F1, mage appears, hit F2, mage appears, F3, F4, etc. mage, mage, mage :lol .

Additionally, the screen flashes (once) in red when I attempt to play from original, be it Garrison I or Garrison II, HFE v1 or HFE v3. I suppose that amounts to "uh oh, not good!" in Gotek language :) . And before you ask, yup, I am running the latest version of FlashFloppy (v3.29 stable).

Having looked into this, you need 1 meg of memory for the character graphics to change.

Because of the age of the game, i'm guessing it assumes and only detects for slow ram.

Have you tried this configuration?

OK, can confirm that the original works fine, but the cracked version by HQC doesn't change character graphics at all, so this is a problem that requires fixing.

Depends on how you define "the original". The original image (the IPF) works fine in WinUAE, the actual floppies from the boxed version would probably work fine on any bog standard A500, but IPF compliancy of these games is lacking to be used as such in Gotek (else, it would work from my Gotek after conversion to HFE v1 and / or HFE v3, which is the normal course of action).

That's the issue affecting the cracked version of Garrison I. But there's more, Blizzard version of Garrison II is affected by the exact same issue. Also, there is some slight confusion with the numbering of the ADF files of that version of Garrison II. On the file server, the only second disk (2 out of 3) image available is flagged as hit by a virus, so you're supposed to use the third disk as the character disk. You'll figure it out ;) .

My miggy doesn't bear any slow ram. By flipping the metal switch located at the back of my machine, I can only choose between 0.5 MB chip and 1 MB chip. By default, I leave the switch on the 1 MB position. If the game needed slow ram vs chip ram type of fixing, the ADF's would fail to load completely upon 1 MB chip ram detection presumably. That's the kind of games ross has been eager to fix these last weeks, Street Rod I, Wayne Gretzky Hockey I etc.... Garrison doesn't belong there IMHO.

A couple more remarks :

• The boot disk is interchangeable e.q. I was able to play Garrison II using the (1 of 2) ADF from Garrison I.
  
• It is true that one would be able to play the original image properly assuming his machine (emulated or actual) would bear 0.5 MB slow ("additional") ram
• The fact that the original image won't work from a Gotek is an issue for those who play exclusively from converted IPF's (a minority of users, but still). Having said that, even if the floppies were reworked from scratch by the SPS team, it wouldn't help in solving the look alike characters for those who lack slow ram in their machine.
• My guess is that the game must have been developed on a A2000. When the main dev. handed the final version over to Rainbow Arts, they realized in shock that the memory requirements would ward off the unexpanded A1000 & A500 crowd, so they asked Andreas to do something about it. And the result is the rather grating issue described, no slow ram detected means no distinctive looks for the character you choose, which is an even more distracting limitation when 2 peeps enter the dungeons simultaneously :guru .

To sum it up, if any of the usual suspects wanted to face the challenge, he (or she :) ?) would have to obtain a raw image of the three floppies (boot disk, scenario disk I, scenario disk II), and then he would have to code a new crack, and then he would have to fix the game to enable to distinctive looks of each character upon detection of any type of ram > 0.5 MB chip, be it 0.5 MB chip + 0.5 MB slow, 1 MB chip, or otherwise...

Maybe it could be the right time that I fix this game for a one-disk version (it is another one that I started and never finished...).

The file system is really wasteful, with its 8192 bytes per cylinder (yes, this is also the granularity :rolleyes).
I already have somewhere the main program decrypted and with its relocation table (so making a version that works in any memory configuration is trivial).
And I also have the tracks extracted from the original IPFs (if I remember correctly the Blizzard crack has bad sectors, at least regarding Garrison I)

It could also be an opportunity to try a virtual filesystem in ram with compressed cylinders with a new packer.
Among other things, Garrison II could benefit from sector deduplication since many parts are shared.

Well, I'll take a look, at least to understand where I was with the code :)

By all means carry on, I cant muster much enthusiasm for it :great

To tell you the truth it's just to try the "cylinder compressor".
Let's see if anything decent comes up ;)

:great

the strange aspect of Garrison is that the Disk 1 is standard AmigaDOS, while disk 2 use 2 custom formats : The Rainbow Arts format type 3 and the Rainbow arts 4489 protection track located on track 2. The HFE conversion seems to fails on it.

Am I right in thinking the US version(s) are not protected? Seem to recall I have a US copy of either Garrison or Garrison II that is just ADOS.

Have checked and my copy is Garrison (US) that's plain ADOS, uploaded ADF's of both disks to The Zone!

This is a tampered version.
It contains the Blizzard crack (with intro removed) and the HQC's ADOSed 2nd disk.

What? Some git has copied over my originals :D The case has a sticker saying "$42.95 from The Computer Cellar" so it is a US release, unless the sticker was copied too :)

Well that explains that, so I guess ALL copies were protected then so SPS #0800 and SPS #0610 are the only known releases?

No idea why your 'original' is in this condition, but I assure you that is a cracked version.

Also, the second disk isn't exactly like the HQC one (very similar but wrong audio data in a sector?), but it's not even the 'corrupted' Blizzard one in TOSEC either (as I wrote in a previous post there are bad sectors).
I didn't check if they fixed the sectors or added another bad ones (taking here and there..).

Anyway there will soon be a proper version with data from original game :)

Hmm... I wonder if it was an officially *fixed* cracked version for the US Market? There seems to have been 'cracked' games released commercially before so maybe this is one of them?

Well, unless someone also has a US version then we will know, but best just wait for the ross SuperMegaDeluxe 1 Disk release of all 4 disks :lol

Could be..
But they could have done it better.. I checked, in disk 2 those "new" ADOS sectors with different data are in fact corrupt.

Well if it's a copy that's probably why it's corrupt, probably the original owner copied a cracked version over his original?

Doesn't matter now, you'll release a 100% working version :great

I was browsing eBay the other day, and I noticed there was a rare boxed release of both games comprised of 3 disks, sold there by some French man, check it here. It is comprised of a boot disk and two scenarii disks (and a leaflet). Garrison I & Garrison II used to be sold separately, but also as a pack apparently, in which case the included boot disk could be used to load either scenarii disk in the same way. My point is SPS is missing this three disk release (that's the one Blizzard attempted to crack incidentally). Don't miss your chance this time, guys :D (just joking, I am not affiliated to this seller).

I just bought this compilation. The seller lives 5 minutes from my flat and i met him already for other games.

Maybe it's the right time to sort out the various versions (which are probably a single one, the usual v1.02).
If the disks are in good condition, a comparison can be made with the available IPFs for the raw data.