|
CHDMAN
#290895 - 07/04/12 09:50 AM
|
|
|
http://rbelmont.mameworld.info/?p=739
Quote:
Regarding CHDMAN in 0.146u2
Due to a series of bugs, CHDs created with CHDMAN versions prior to the one in 0.146u2 could differ between different OSs (XP vs Vista vs Win7 vs Linux vs Mac OS X), different service pack patchlevels, and even different PCs with the same software load. In almost all cases the SHA1s would match and the data and metadata should be safe, but if you attempted to join a torrent your files would not match. Additionally, there was a possibility of corrupt CD and GD-ROM CHDs being created.
For these reasons, you should re-convert all of your CHDs from the “old” v4 versions using the newest CHDMAN. Note that ClrMAMEPro and chdman -verify are unable to detect if your CHDs are problematic in this way, do not use them to determine if you may have the problems.
Thanks to MAMEWorld user jmak for debugging these problems.
So is it safe to convert everything to V5, or should we keep some things at V4?
If so, is there an automated tool/script/gui for Windows to do the conversion?
|
GroovyMAME support forum on BYOAC
|
|
|
Re: CHDMAN
[Re: krick]
#290899 - 07/04/12 12:36 PM
|
|
|
You will end up with wrong sha1 of course for the known bad dumps...
|
|
|
redk9258 |
Regular
|
|
|
Reged: 09/21/03
|
Posts: 3968
|
Loc: Troy, Illinois USA
|
|
Send PM
|
|
|
Re: CHDMAN
[Re: Roman]
#290915 - 07/04/12 05:47 PM
|
|
|
So if I understand correctly, once converted correctly everyone's files should have the same CRC32/ MD5 / SHA1 values. I am talking about the chd file and not the contents of the chd file.
Does this match anyone else's?
Path: D:\MAME\chds\area51\ Name: area51.chd Size: 498022497 CRC32: b353301d MD5: ef3b1b835540f1fa00e37b98b0623a17 SHA1: f568066ec4039a615757281e590629f84bf98f7e
Can someone post a list with the correct values?
|
|
|
Tafoid |
I keep on testing.. testing.. testing... into the future!
|
|
|
Reged: 04/19/06
|
Posts: 3137
|
Loc: USA
|
|
Send PM
|
|
|
Re: CHDMAN
[Re: redk9258]
#290917 - 07/04/12 06:03 PM
|
|
|
> So if I understand correctly, once converted correctly everyone's files should have > the same CRC32/ MD5 / SHA1 values. I am talking about the chd file and not the > contents of the chd file. > > Does this match anyone else's? > > Path: D:\MAME\chds\area51\ > Name: area51.chd > Size: 498022497 > CRC32: b353301d > MD5: ef3b1b835540f1fa00e37b98b0623a17 > SHA1: f568066ec4039a615757281e590629f84bf98f7e > > Can someone post a list with the correct values?
v4 and v5 will always be different externally because they use different compression by default and they will be different sizes. Even v5's can be different between other externally if different compression methods are used. So, it stands to reason that whomever chooses to make v5 should make them default compression schemes as to match up to other peoples' v5 CHDS. Internally, they should be the same because the hashes are based on the actual data contained inside.
|
|
|
|
Re: CHDMAN
[Re: krick]
#290922 - 07/04/12 06:58 PM
|
|
|
> So is it safe to convert everything to V5, or should we keep some things at V4?
Try it, but nobody on the mame team has admitted to doing so.
|
|
|
B2K24 |
MAME @ 15 kHz Sony Trinitron CRT user
|
|
|
Reged: 10/25/10
|
Posts: 2663
|
|
|
Send PM
|
|
|
Re: CHDMAN
[Re: Roman]
#290923 - 07/04/12 07:03 PM
|
|
|
> You will end up with wrong sha1 of course for the known bad dumps...
Easiest way I found was to batch convert all to V5 without deleting all the V4s, then audit with clrmamepro, which will remove the ones that changed SHA1 on conversion, then simply manually move back a V4 in it's place to be 100%
|
|
|
|
Re: CHDMAN
[Re: ]
#290929 - 07/04/12 08:04 PM
|
|
|
actually rb did on his blog
|
|
|
Tafoid |
I keep on testing.. testing.. testing... into the future!
|
|
|
Reged: 04/19/06
|
Posts: 3137
|
Loc: USA
|
|
Send PM
|
|
|
Re: CHDMAN
[Re: Roman]
#290931 - 07/04/12 08:37 PM
|
|
|
> actually rb did on his blog
http://rbelmont.mameworld.info/?p=739 ??
I don't see him mentioning to convert all your CHD's to v5 as a mandatory act - just that if you have converted already, make sure to do it again with the latest (0.146u2) CHDMAN, using the v4 sources if you want v5 the entire collection. Just because he said you CAN does not mean it is required (or mandatory).
The stance is as before - v4 is fine, v5 can be used but is not a requirement for all CHDS currently. Consider v5 as a "bonus round".. you get more score, but you generally don't advance in the game overall :P
|
|
|
|
Re: CHDMAN
[Re: Roman]
#290941 - 07/04/12 09:57 PM
|
|
|
> actually rb did on his blog
I don't believe he has actually done it himself.
|
|
|
|
Re: CHDMAN
[Re: Tafoid]
#290943 - 07/04/12 09:59 PM
|
|
|
yeah yeah....forget about the word "everything"
|
|
|
|
Re: CHDMAN
[Re: Tafoid]
#290986 - 07/05/12 10:54 AM
|
|
|
> v4 and v5 will always be different externally because they use different compression > by default and they will be different sizes. Even v5's can be different between other > externally if different compression methods are used. So, it stands to reason that > whomever chooses to make v5 should make them default compression schemes as to match > up to other peoples' v5 CHDS. Internally, they should be the same because the hashes > are based on the actual data contained inside.
and Arbee's post on his blog is about the fact that (due to some uninitialized variables in the source) with 0.146 chdman slightly different compression could have been achieved even with the same settings on different OSes or in different runs on Windows. Now this is fixed, and everyone should get the same result if they use the same compression
|
|
|
redk9258 |
Regular
|
|
|
Reged: 09/21/03
|
Posts: 3968
|
Loc: Troy, Illinois USA
|
|
Send PM
|
|
|
Re: CHDMAN
[Re: etabeta]
#290987 - 07/05/12 12:00 PM
|
|
|
> and Arbee's post on his blog is about the fact that (due to some uninitialized > variables in the source) with 0.146 chdman slightly different compression could have > been achieved even with the same settings on different OSes or in different runs on > Windows. > Now this is fixed, and everyone should get the same result if they use the same > compression
That's why I posted the results of the area51.chd that I converted. I used 'chdman copy -i area51.old -o area51.chd'. I did not specify compression and let chdman take care of that. I would expect that everyone using chdman 0.146u2 would get the same checksum for area51.chd.
|
|
|
R. Belmont |
Cuckoo for IGAvania
|
|
|
Reged: 09/21/03
|
Posts: 9716
|
Loc: ECV-197 The Orville
|
|
Send PM
|
|
|
Re: CHDMAN
[Re: Roman]
#290994 - 07/05/12 04:47 PM
|
|
|
> actually rb did on his blog
No, I'm just saying that it's probably actually safe to do so now (excluding the optical discs marked BAD_DUMP, of which a great number will hopefully go away in u3).
|
|
|
|
Re: CHDMAN
[Re: R. Belmont]
#291005 - 07/05/12 08:24 PM
|
|
|
yeah yeah...as I already posted, I've added a "all" by myself...and yes...I know that most get replaced anyway
|
|
|