Puppy 2.12 Application don't work

[Gn2]

Vector Linux went all through that topic (checking of a burnt CD to original ISO
NO you cannot readily do so = the contents of the burnt ISO are split up into components , some of which are still compressed, others in plain text.
As was stated - the only practical manner is to copy all known good content pf the burn , create new MD5sum hash .
There IS a way of calculating, subtracting some of burnt content so sums match .

[GuestToo]

the iso file is, by definition, the exact byte for byte image of what will be burned on the cd ... if the iso burns without errors, there will be not one single bit, not one single byte that is different between the iso file and the cd

the md5sum of the iso file should be identical to the md5sum of the cd disc, and it will be if the cd burned without errors

if extra padding bytes were added when the cd was burned, the extra padding bytes might be included in the md5sum of the cd disc, which can cause the md5sum calculation to be wrong, even though the cd burned perfectly without errors

another problem that used to occur, was the read-ahead bug ... the cd driver would read ahead on the cd into a buffer, and send the bytes from the buffer to the md5sum program when it asked for them ... but it would read off the end of the cd, and there would be an io error that would stop the program ... i think that has finally been fixed

also, sometimes the size of the cd is wrong, sometimes only by a couple of bytes ... but that is enough for the md5sum program to read a couple of extra bytes from the cd, which will result in an incorrect md5sum ... even a difference of 1 single bit should cause the md5sum number to be different, that's the whole point of the md5sum

so basically, if the md5sum of the iso matches the md5sum of the cd, you are very sure, almost certain that the cd burned perfectly with not even 1 bit wrong

if the md5sums are different, the cd disc probably did not burn properly ... but it might have burned properly, but some other error caused the md5sum to be wrong ... however, it is probable that the cd has errors on it ... even 1 wrong bit will change the md5sum ... however, even if the disc has some errors, the bad bytes may be unimportant bytes ... it would be difficult to know for sure ... generally, if the md5sum of the iso does not match the md5sum of the disc, it's probably a good idea to assume the cd did not burn properly

there is a way to be sure you read exactly the correct number of bytes from the cd disc, by using dd to read the cd and piping it to the md5sum program ... that prevents problems with padding, for example

there are programs that can check the quality of a cd, like cdck and dvdisaster, both available as dotpups ... they read each sector of the cd disk and check whether the sector had errors or not, by using the checksums stored on the cd

Puppy's burniso2cd program has a built in md5sum check to test for errors after a cd disc is burned

bottom line: an iso is an exact image of the bytes that are burned on a cd disc ... if the cd burned properly, the md5sum of the iso file will be identical to the md5sum of the cd disc, always and without exception

[Gn2]

Mirrored > 6de7e4ac9f66ddaaed56924ff2a85ca1 puppy-2.12-seamonkey-zdrv.iso
--------------------------------------------------------------------------
Storage :

-K3b-pup_v2.12final_burn-report.txt
puppy-2.12-seamonkey-zdrv.iso.md5.txt
puppy-2.12-seamonkey-zdrv.iso
---------------------------------------------------------------------------

Code: Select all

 ~/puppy $ md5sum *z*
# 6de7e4ac9f66ddaaed56924ff2a85ca1

#puppy-2.12-seamonkey-zdrv.iso

# puppy-2.12-seamonkey-zdrv.iso.md5.txt
#5929122b2b51a54eb303815b801e7c67

----------------------------------------------------------------------------

Code: Select all

dd if=/dev/hdd | md5sum

104+0 records in
104+0 records out
53248 bytes (53 kB) copied, 83.5665 seconds, 0.6 kB/s
NOTE >> 7b14f5102141fdad4dacffb566930b9d

the cd burned properly, the md5sum of the iso file

= Of the unburnt image file yes
Content of burnt CD, as per Example -(using the suggested method) no

[GuestToo]

this is not safe ... the size of the data on the disc is often reported incorrectly to dd ... for example, if the disc was burned track-at-once, dd often thinks the data is 2 bytes bigger than is actually is, and 2 extra bytes are included in the md5sum, which will not work ... the md5sum will be wrong:

dd if=/dev/hdd | md5sum

something like this should work:

isoinfo -d -i puppy-2.12-seamonkey-zdrv.iso | grep 'Volume size'
Volume size is: 42517
dd if=/dev/hdc bs=2048 count=42517 | md5sum

or this usually works, depending how the disc was burned ... Puppy's burniso2cd seems to work without problems ... k3b may add padding characters that can change the md5sum, if so, you must use the dd method:

md5sum /dev/hdc

again: if you read the same number of bytes from the disc as are in the iso file, the iso file and the disc should have the same md5sum

[Gn2]

K3b output matches mirror but is only the unburnt ISO image check prior to burn .

K3b also has the ability to verify content after burn, generate a full report & offer to save

Yes very good reminder (I have suggested & used isoinfo in some cases to extract more data.) (After some checking - VL - still have old URL link)

As was mentioned B4 in asides .... there is always a solution > just not was the first suggestion was offered.

dd sure is powerful. or powerfully disasterous if any inadvertant syntax error creeps in (all by itself natch)
Nice to see Puppy included what most do not.
-----------------------------------------------------------------------
Off topic- speaking of POWER >toys are getting all too tempting for pocketbook:

http://www.ncix.com/products/index.php? ... cture=TYAN

Best regards