From: Joerg Schilling on
In article <p4_1j.27930$Zn.13509(a)edtnps90>,
Unruh <unruh-spam(a)physics.ubc.ca> wrote:

>cdrecord has nothing to do with paths. cdrecord and wodim simply take bytes
>off the disk and burn them onto the dvd. The path names are a concern of
>mkisofs. That is what assembles the stuff you tell it to burn and makes a
>single iso file. That is what contains path names, etc. Now it sounds like
>you did not tell it to use the appropriate rock-ridge/joliet extention to
>make the iso.

In summer 2006, I fixed more than a dozen serious bugs in mkiosfs. These
bugs have been in mkisofs since the time before I became the maintainer.
For this reason, these bugs are all still in genisoimage.

Many of the bugs fixed in mkisofs are related to Rock Ridge path handling
with deep directories (> nesting level 7). Some of these bugs caused the
well known "cannot sort directory" messages and graft points did not work at all
if the target directory had a nesting level > 7. Important other bugs caused
wrong link count for directories and hard linked files as well as incorrect
permissions for directories.

These bugs have all been detected when I added support for correct hard linked
files to mkisofs. I did run many tests and after I was ready with fixing bugs,
I was finally able to create a 100% correct copy of a UNIX root filesystem.
This is something you definitely cannot do with genisoimage because genisoimage
is based on a 2 year old mkisofs.

My tests show that mkisofs now finally works as documented and as I did get
some minor bug reports (that have been fxied a year ago), I am sure that
other people did test mkisofs too. If there was a problem left over, I did
definitely get a bug report long ago. I know that k3b prefers the original
cdrtools because the original software does not have known problems.

It is extremely improbable that the claims from "Ian Rawlings" are correct.

In contrary to the "wodim maintainers" I listen to my users and if there are
real problems, I fix them as soon as possible. The claim from "Ian Rawlings"
do not even prove a problem.

--
EMail:joerg(a)schily.isdn.cs.tu-berlin.de (home) J�rg Schilling D-13353 Berlin
js(a)cs.tu-berlin.de (uni)
schilling(a)fokus.fraunhofer.de (work) Blog: http://schily.blogspot.com/
URL: http://cdrecord.berlios.de/old/private/ ftp://ftp.berlios.de/pub/schily
From: Ian Rawlings on
On 2007-11-24, Joerg Schilling <js(a)cs.tu-berlin.de> wrote:

> In contrary to the "wodim maintainers" I listen to my users and if there are
> real problems, I fix them as soon as possible. The claim from "Ian Rawlings"
> do not even prove a problem.

Jorg, despite you being a bit illogical in this thread, I would prefer
to use mkisofs as I do think you've done a lot of good work over the
years and it's a shame that the fork and resulting disagreement has
happened. From your attitude however I strongly suspect that you
have played a large part in the disgreements.

Sadly mkisofs did not work with brasero on gentoo while genisofs
worked fine. I did not have these problems with gnome's built-in
burner or with gnomebaker while I was using mkisofs, they worked fine
with mkisofs, it was only brasero that has a problem with mkisofs.
Using genisofs works fine with brasero, despite what you say. I have
two DVDs that burned fine with it.

Maybe genisofs doesn't work with lots of other things but it worked
when building dvds with brasero while mkisofs didn't so something odd
happened there. You can claim I'm wrong until you're blue in the face
but it's a fact. It may well be that brasero is working around
problems with genisofs and this may be breaking interaction with
mkisofs. I suspect that the problem was with the interpretation of
filenames and graft points fed to mkisofs by brasero, I can only
assume that it's doing something that works with genisofs but not with
mkisofs. Perhaps brasero is tested with genisofs but not with
mkisofs.

I do not intend to investigate as I have no desire or need to. If you
feel strongly that mkisofs should conquer all then I'd suggest that
instead of ranting on usenet, you look into it yourself. I in the
meantime will continue to use brasero and genisofs as *it works*, no
matter what you claim.

--
Blast off and strike the evil Bydo empire!
From: Unruh on
Ian Rawlings <news06(a)tarcus.org.uk> writes:

>On 2007-11-24, Joerg Schilling <js(a)cs.tu-berlin.de> wrote:

>> In contrary to the "wodim maintainers" I listen to my users and if there are
>> real problems, I fix them as soon as possible. The claim from "Ian Rawlings"
>> do not even prove a problem.

>Jorg, despite you being a bit illogical in this thread, I would prefer
>to use mkisofs as I do think you've done a lot of good work over the
>years and it's a shame that the fork and resulting disagreement has
>happened. From your attitude however I strongly suspect that you
>have played a large part in the disgreements.

>Sadly mkisofs did not work with brasero on gentoo while genisofs
>worked fine. I did not have these problems with gnome's built-in
>burner or with gnomebaker while I was using mkisofs, they worked fine
>with mkisofs, it was only brasero that has a problem with mkisofs.
>Using genisofs works fine with brasero, despite what you say. I have
>two DVDs that burned fine with it.

You have been asked a number of times to provide details. Exactly what did
not work. You seem to claim that when you told Brasero to place a bunch of
files in a directory under that same directory on the cd, some of the files
ended up under that directory, and some ended up a level higher.
This sounds like a bug with brasero as I certainly have never seen anything
like this happen. It may also have been a bug with you, where you somehow
when you transfered the files into brasero, you told it to put half of the
files into the subdirectory and the other half not. But since you provide
zero details, it is impossible for anyone to decide between various
possibilities, including that you are lying.

>Maybe genisofs doesn't work with lots of other things but it worked
>when building dvds with brasero while mkisofs didn't so something odd
>happened there. You can claim I'm wrong until you're blue in the face
>but it's a fact. It may well be that brasero is working around

No, no facts have been displayed. Your interpretation of unknown events has
been on display, and we have no idea how reliable that interpretation is.
If what you say is true, that is either a serious bug in mkisofs, in
brasero, or in genisofs. Eg, if you told brasero to put the files half into
the subdirectory and half into the main one, and brasero /genisofs put all
intot he subdirectory, then that is a bug in brasero/genisofs.


>problems with genisofs and this may be breaking interaction with
>mkisofs. I suspect that the problem was with the interpretation of
>filenames and graft points fed to mkisofs by brasero, I can only
>assume that it's doing something that works with genisofs but not with
>mkisofs. Perhaps brasero is tested with genisofs but not with
>mkisofs.

Maybe all kinds of things. Give us more details.



>I do not intend to investigate as I have no desire or need to. If you
>feel strongly that mkisofs should conquer all then I'd suggest that
>instead of ranting on usenet, you look into it yourself. I in the
>meantime will continue to use brasero and genisofs as *it works*, no
>matter what you claim.

What is he supposed to look into? You have provided no details. Noone
knows how to duplicate your problem. Provide those details or have us
believe that you are lying or have gone off your rocker for some weird
reason of your own. With the amount you have written here you could have
provided tons of details. But you have supplied none.



>--
>Blast off and strike the evil Bydo empire!
From: Ian Rawlings on
On 2007-11-25, Unruh <unruh-spam(a)physics.ubc.ca> wrote:

> You have been asked a number of times to provide details. Exactly what did
> not work.

I've said repeatedly that I'm not trying to get someone to fix it or
provide a bug report. If you feel so strongly, experiment with it
yourself.

--
Blast off and strike the evil Bydo empire!
From: Joerg Schilling on
In article <gK92j.9554$Ji6.3355(a)edtnps89>,
Unruh <unruh-spam(a)physics.ubc.ca> wrote:

>You have been asked a number of times to provide details. Exactly what did
>not work. You seem to claim that when you told Brasero to place a bunch of
>files in a directory under that same directory on the cd, some of the files
>ended up under that directory, and some ended up a level higher.

This something I did not read before, but it is still not helpful unless
the mkisofs command line and the source directory structure is known.

>This sounds like a bug with brasero as I certainly have never seen anything
>like this happen. It may also have been a bug with you, where you somehow
>when you transfered the files into brasero, you told it to put half of the
>files into the subdirectory and the other half not. But since you provide
>zero details, it is impossible for anyone to decide between various
>possibilities, including that you are lying.

Graft points have not been written by me but are the last enhancement
added by the original mkisofs Author Eric Youngdale. If I find a problem
with graft points, I check the documented behavior and try to let mkisofs
match the documented behavior in order keep users happy. This may of course
let people fail who follow the observed behavior of old mkisofs versions
instead of the documented behavior.

>>Maybe genisofs doesn't work with lots of other things but it worked
>>when building dvds with brasero while mkisofs didn't so something odd
>>happened there. You can claim I'm wrong until you're blue in the face
>>but it's a fact. It may well be that brasero is working around


>No, no facts have been displayed. Your interpretation of unknown events has
>been on display, and we have no idea how reliable that interpretation is.
>If what you say is true, that is either a serious bug in mkisofs, in
>brasero, or in genisofs. Eg, if you told brasero to put the files half into
>the subdirectory and half into the main one, and brasero /genisofs put all
>intot he subdirectory, then that is a bug in brasero/genisofs.

Without facts, it is impossible to decide whether there is a bug or whether
thr "problem" is just a result of incorrect usage.


>What is he supposed to look into? You have provided no details. Noone
>knows how to duplicate your problem. Provide those details or have us
>believe that you are lying or have gone off your rocker for some weird
> reason of your own. With the amount you have written here you could have
>provided tons of details. But you have supplied none.

Claiming not to have time but at the same time writing long text looks somehow
contradictory, except if there is no real problem to describe.

--
EMail:joerg(a)schily.isdn.cs.tu-berlin.de (home) J�rg Schilling D-13353 Berlin
js(a)cs.tu-berlin.de (uni)
schilling(a)fokus.fraunhofer.de (work) Blog: http://schily.blogspot.com/
URL: http://cdrecord.berlios.de/old/private/ ftp://ftp.berlios.de/pub/schily