MYDUL avaialbe for recovery, another choise of DUL.
AnySQL (d.c.b.a)
I have sucessfully recover some data for two customers with MYDUL,
also I don't know whether they have another recover method, it now my
pleasure to announce that MYDUL is ready for recovery just as DUL,
except for LOB column and compress table/IOT.
AnySQL from China
http://www.anysql.net/en/
http://www.anysql.net/en/mydul.html
fitzj...@cox.net
We've discussed your 'tools' before, when you were altering datafile
headers to attempt making datafiles usable for clone databases by
bypassing the provided mechanism from Oracle for such tasks. And we've
discussed other 'tools' you've created which won't work until you
supply much needed information outside of the installation
instructions. I would be a complete fool to accept, on YOUR word
alone, the quality of this 'hack' you've managed to produce. So would
anyone else with but a scrap of sense. Using this 'tool' of yours will
violate any support agreements between the end-user and Oracle
Corporation, a situation I would NEVER endorse.
Play all you want with YOUR OWN databases. Put THEM in jeopardy by
violating YOUR OWN licensing and support agreements. Do NOT expect
anyone with a shred of responsibility to follow in your misguided
footsteps.
David Fitzjarrell
AnySQL (d.c.b.a)
rows out from datafile.
fitzj...@cox.net
AnySQL (d.c.b.a) wrote:
> I have discard that tool, this tool is just as DUL, to directly read
> rows out from datafile.
And DUL is proprietary Oracle software you've attempted to duplicate
which violates any licencing agreement you may have had with Oracle
Corporation. I would not suggest to anyone that your 'tool' is worthy
of use. You have a habit, apparently, of knowingly and willingly
infringing on Oracle's copyrights on intellectual property and then
asking others to share in your crime by requesting they put to use
these so-called products you cobble together. Far be it from me to
even CONSIDER such an offer, as I'm not in the habit of writing huge
settlement checks to large corporations whose software has been
illegally and incompletely duplicated. I also find it disheartening
that you have found willing accomplices in your 'product' testing, most
likely due to a severe unwillingness to pay Oracle to do the job that
you convinced them your product could complete. Reverse-engineering
ANY part of ANY Oracle software is ILLEGAL. You, obviously, couldn't
care less, else why would you continue writing such programs in clear
violation of all Oracle licencing agreements, agreements which you
willlingly accepted only to disregard them in your quest to rob Oracle
of revenue?
It matters not that you fancy yourself a software guru; the truth in
this matter is that you have attempted to reverse-engineer software
unique to Oracle Corporation, and have done so illegally. That you
have provided a similar functionality in your 'product' is grounds for
a lawsuit of which I would not care to be a recipient, so I shall pass
on your generous offer to join you in prison.
David Fitzjarrell
AnySQL (d.c.b.a)
knowledge of oracle block format. Spent a lot of time on it, illegal?
Jim Kennedy
"AnySQL (d.c.b.a)" <any...@gmail.com> wrote in message
news:1127790265.0...@g49g2000cwa.googlegroups.com...
> Why I want to reverse DUL, I just wrote one of mine according to my
> knowledge of oracle block format. Spent a lot of time on it, illegal?
>
knowledge doesn't mean it is legal.
Jim
fitzj...@cox.net
AnySQL (d.c.b.a) wrote:
> Why I want to reverse DUL, I just wrote one of mine according to my
> knowledge of oracle block format. Spent a lot of time on it, illegal?
Most certainly it is, as stated in the Oracle licence agreement:
"(d) prohibits causing or permitting the reverse
engineering, disassembly or decompilation of the Program(s); ..."
You have, in your infinite wisdom, reverse-engineered DUL. You have,
therefore, violated the terms of the Oracle licencing agreement and, as
such, are subject to prosecution.
Have a nice day.
David Fitzjarrell
Mladen Gogala
> And DUL is proprietary Oracle software you've attempted to duplicate
> which violates any licencing agreement you may have had with Oracle
> Corporation. I would not suggest to anyone that your 'tool' is worthy
> of use.
Even if the tool was seaworthy, DUL is not a simple tool. With DUL, one
usually gets a senior specialist from Oracle's tech support unit, which
helps not only if one doesn't know what he or she is doing, but it also
helps in case of an experienced DBA with years and years of experience. I
can personally vouch for that. I wouldn't dare to use DUL without a
support specialist. Using a public domain, "DUL-like" tool is a
professional suicide.
Jim Kennedy
"Mladen Gogala" <gog...@sbcglobal.net> wrote in message
news:pan.2005.09.27....@sbcglobal.net...
they are doing and are invaluable in those situations where things went
horribly awry. In our case we were unable to recover the data. This was
not a limitation of the tool. The data was erased and over written.
This tool you have written might be a nice academic exercise, but I wouldn't
make it privately or publicly available. Oracle has much deeper pockets
than most people. I wouldn't want to be on the other end of a cease and
desist order.
Jim
AnySQL (d.c.b.a)
Actually, I don't know whether I can open this utility. I wrote the
complete code by myself, no one line copied from another one's, how I
do research of oracle block?
I wrote a tool to do os dump of binary file, just like od, and the I
issue the "alter system dump datafile ..." command, then compair and
found the location, after multiple times, I can locate the most of key
bytes in block, Oracle should use a relative simple rule to format
block to keep high effectivity, so after tried many times I found out
some rules.
And until now, I did not how to use oracle's dul, just viewed some
dul document.
And this is complete a personal staff, how about the jDUL & DUDE?
AnySQL (d.c.b.a)
Thanks for you notification!
I did not use any reverse tool and try to find the logic in oracle
dul binary. And I have interesting in the illegal issue, can you give
me more information about what's is illegal and what is legal?
Thank you!
Jim Kennedy
company did) Oracle, installed it and electronically poked it to see what
the program (Oracle) produced. Just because others do it does not make it
legal. Consult a lawyer.
Jim
Jonathan Lewis
news:1127793538.8...@g43g2000cwa.googlegroups.com...
That introduces an interesting point for many
of the Oracle 'gurus'. When Steve Adams
writes something about how the redo writer
works and how to set the best size for the
log buffer and files, has he "caused or permitted
the reverse engineering" of part of the code.
When I describe how the hash join mechanism
works, or how the cost based optimiser does
its arithmetic have I done the same.
When Cary Millsap explains what's in a trace
file, and where tkprof gets it wrong, has he done
the same.
Where can you draw the line ?
If someone gives you a raw datablock and says:
The information on this block is from a
table with N columns of types x,y, and z
you might be able to write a program to decode
the dump without knowing the block came from
an Oracle database.
If you then claimed you could dump the block with
a self-consistent image of all committed changes as
at a particular SCN, then I think you would be on
very shaky ground, because you would have had
to apply very detailed knowledge of the operation
of the software that produced, and reads, the block.
This is all just conjecture, by the way, but I can't
help thinking that the people who explain Oracle
internals (correctly) might technically be in breach of
their licence agreements.
--
Regards
Jonathan Lewis
http://www.jlcomp.demon.co.uk/cbo_book/ind_book.html
Cost Based Oracle - Volume 1: Fundamentals
On-shelf date: Nov 2005
http://www.jlcomp.demon.co.uk/faq/ind_faq.html
The Co-operative Oracle Users' FAQ
http://www.jlcomp.demon.co.uk/appearances.html
Public Appearances - schedule updated 4th Sept 2005
Mladen Gogala
> This tool you have written might be a nice academic exercise, but I wouldn't
> make it privately or publicly available.
I wrote a simple DBA tool in PHP5, not a public domain version of DUL.
You can see for yourself on my web page.
fitzj...@cox.net
You undestand how DUL operates, you set about to write a tool to
perform the same function, you have DONE so (such as it is), and you've
named your tool in a manner consistent with how Oracle has named
theirs. You have ALSO advdertised your product for public use as an
alternative to Oracle's proprietary software. To me this is a clear
breach of the licence agreement. Your advertisement in this newsgroup
("MYDUL avaialbe for recovery, another choise of DUL."), although
misspelled, clearly indicates you intend this 'product' be used as a
replaceemnt for Oracle's licensed software. Your act differs from
those of Jonathan Lewis, Cary Millsap, Thomas Kyte and others as they
do not compete with Oracle in the marketplace with 'clone' products,
they disseminate information; you obviously ARE attempting to compete
with Oracle and have decided to do so by poorly duplicating a
proprietary utility and offering it for use without qualified support.
I would SERIOUSLY consider talking with an attorney to discover the
extent of your violations. And until this matter is resolved I'd stop
advertising this 'tool' of yours to prevent any further legal action
which may arise.
David Fitzjarrell
Joel Garry
Jonathan Lewis wrote:
> <fitzj...@cox.net> wrote in message
> news:1127793538.8...@g43g2000cwa.googlegroups.com...
> >
> > AnySQL (d.c.b.a) wrote:
> >> Why I want to reverse DUL, I just wrote one of mine according to my
> >> knowledge of oracle block format. Spent a lot of time on it, illegal?
> >
> > Most certainly it is, as stated in the Oracle licence agreement:
> >
> > "(d) prohibits causing or permitting the reverse
> > engineering, disassembly or decompilation of the Program(s); ..."
Violation of a contract term isn't necessarily illegal, it just exposes
you to lawsuits for a remedy. I wouldn't want to be in China when
Oracle convinces them to crack down...
http://www.signonsandiego.com/uniontrib/20050926/news_1n26china.html
> >
> > You have, in your infinite wisdom, reverse-engineered DUL. You have,
> > therefore, violated the terms of the Oracle licencing agreement and, as
> > such, are subject to prosecution.
Putting it that way makes it sound criminal, which it usually isn't,
although you can technically say a civil case is being prosecuted.
> >
> > Have a nice day.
> >
> >
> > David Fitzjarrell
> >
>
> That introduces an interesting point for many
> of the Oracle 'gurus'. When Steve Adams
> writes something about how the redo writer
> works and how to set the best size for the
> log buffer and files, has he "caused or permitted
> the reverse engineering" of part of the code.
Yes.
>
> When I describe how the hash join mechanism
> works, or how the cost based optimiser does
> its arithmetic have I done the same.
Yes.
>
> When Cary Millsap explains what's in a trace
> file, and where tkprof gets it wrong, has he done
> the same.
Yes.
>
> Where can you draw the line ?
Very simply, the line has been drawn by various governments to make it
technically illegal to do common software engineering tasks. In
addition, lawyers for all large software vendors have done the same
thing contractually for jurisdictions where they can.
>
> If someone gives you a raw datablock and says:
> The information on this block is from a
> table with N columns of types x,y, and z
> you might be able to write a program to decode
> the dump without knowing the block came from
> an Oracle database.
>
> If you then claimed you could dump the block with
> a self-consistent image of all committed changes as
> at a particular SCN, then I think you would be on
> very shaky ground, because you would have had
> to apply very detailed knowledge of the operation
> of the software that produced, and reads, the block.
>
>
> This is all just conjecture, by the way, but I can't
> help thinking that the people who explain Oracle
> internals (correctly) might technically be in breach of
> their licence agreements.
There is simply no legal way to do it, certainly under US law. I am
not a lawyer, so take any legal opinion I may give with much seasoning.
http://www.kaner.com/pdfs/reveng.pdf - even the "strings" command is
reverse engineering. http://www.eff.org/IP/FTAA/tpm_issues_eng.pdf
DMCA criminalizes the copyright issues. It would be ironic if people
who explain things incorrectly would be on more solid legal ground, but
reverse engineering doesn't have to be correct, just attempted.
Should we worry about getting slammed by Oracle? Most of us, not
generally. Guru's noted above might want to check with actual lawyers
(and let us know what they say). I don't think Oracle would actually
go after anyone like them (except Oracle employees in some situations,
maybe), but lawyers are paid to anticipate and defend against the worst
case.
I believe Oracle _should_ give access to the details at all levels
through documentation (and included instrumentation), but I don't think
that is going to happen, since the withholding of information is
valuable to the corporation. I think the current situation in general
sucks, and we haven't seen yet how badly it can suck.
I am salivating in anticipation of your book, Jonathan! Please don't
let any of this legal mumbo-jumbo get in the way.
jg
--
@home.com is bogus.
RIP Donald James Yarmy 1923-2005 AKA Don Adams AKA Maxwell Smart.
"Sorry about that Chief..."
jks...@gmail.com
Thank you Jonathan for your concise thoughts on this.
I was thinking along the same lines, that the activity
that many of us engage in while trying to find what makes
Oracle tick could be construed as reverse engineering.
Creating 10046 and 10053 trace files could be considered
reverse engineering, but it is apparently not viewed that
way by Oracle.
eg. the symbols in 10046 traces are explained in a document
on MetaLink, which is available to anyone with a support
contract.
http://metalink.oracle.com/metalink/plsql/showdoc?db=NOT&id=39817.1
I don't know if reverse engineering a data format (such as in
myDUL) is equivalent to reverse engineering software.
Many utilities exist that can read MS Office formats, some of
them are open source - this seems analagous to me.
Personally I would not care to have Oracle's attorneys emailing
me, and would try not to push the envelope in such endeavors.
Jared
fitzj...@cox.net
Such traces are provided by Oracle without the use or construction of
third-party tools. As you've stated the explanation of the trace
symbols is provided by Oracle Corporation; Oracle provides these traces
to enable a DBA to properly tune queries and the database. This, in my
mind, is a far different situation than knowingly writing a
replacement, albeit incomplete, of DUL.
> I don't know if reverse engineering a data format (such as in
> myDUL) is equivalent to reverse engineering software.
>
And you may be correct, however I wouldn't want to be on the receiving
end of any communication from the legal department of Oracle
Corporation. DUL is a proprietary software tool used, in the most
extreme cases, to attempt data recovery from corrupted datafiles. This
tool is also accompanied by a trained technician who can properly
operate it. If such a tool requires specially educated personnel I
certainly wouldn't trust any DBA to use it, much less trust a
second-rate copy in anyones hands.
> Many utilities exist that can read MS Office formats, some of
> them are open source - this seems analagous to me.
Reading the contents of a Word document is far different, to me, than
reading the contents of a corrupted, proprietary-format datafile,
especially with a tool that is incomplete, as admitted by the
programmer. Lacking the ability to recover LOB data, comporessed
indexes and IOTs certainly leads me to wonder why I should trust such a
tool to correctly read any other data in a corrupted file. The OP has
claimed his tool is "ready for recovery", yet misses some data
structures, meaning tables with LOB columns will be incomplete upon
'recovery', IOTs will be missing as well as compressed indexes. To me
this is a failed attempt at reverse-engineering DUL; it is also an
accident waiting to happen. Using this 'tool' to 'recover' corrupted
datafiles would violate any service agreements between Oracle and the
party using said 'tool' as it is NOT Oracle's DUL utility. There is
nothing good about this work as I see the situation. Possibly I am
wrong about the legality of this software and the process used to
create it; from what I have read thus far I do not believe I am. That
notwithstanding, this 'tool' is a hazard for anyone to use, whether it
be legal or not. I certainly wouldn't trust it to recover my data.
>
> Personally I would not care to have Oracle's attorneys emailing
> me, and would try not to push the envelope in such endeavors.
>
I wholeheartedly agree.
> Jared
David Fitzjarrell
yon...@yahoo.com
>
> I have sucessfully recover some data for two customers with MYDUL,
> also I don't know whether they have another recover method, it now my
> pleasure to announce that MYDUL is ready for recovery just as DUL,
> except for LOB column and compress table/IOT.
Hi, AnySQL,
I'd like to make some suggestions to you, if I may. As other nice folks
point out, there may be a legal issue with MYDUL. I suggest you rename
the tool to anything not having the word "DUL". Secondly, clearly state
wherever appropriate that the tool is built by studying block dumps and
trial-and-error experiments, not by disassembling either Oracle's DUL
tool or Oracle server software. Lastly, clearly state that this tool
should be used when no documented recovery method is applicable and the
cost of using Oracle's DUL is prohibitive for business, and that in
order to keep Oracle's support on the recovered database, this database
must be exported and imported into another database (remind people of
the same trick that a database with a dropped column can be recovered
by altering data dictionary, as long as it's subsequently exported and
imported; credit to an Oracle expert well known in this newsgroup)
If you email me, I have a few more suggestions. In the meantime, keep
up your good work and continue to read messages in this newsgroup with
an open mind, giving more weight to the authors with better reputation.
Yong Huang
yong321ATyahoo.com
Sybrand Bakker
>I'd like to make some suggestions to you, if I may.
Don't forget, AnySQL, your tool is crap and using it will make your
'customers' loose support.
If you don't want to end up bankrupt and/or in jail, do not put any
weight on 'yong321' silly opinions.
--
Sybrand Bakker, Senior Oracle DBA
AnySQL (d.c.b.a)
my tools, anyway this tool is just a personal software.
Thank you.
Jonathan Lewis
news:1127882329....@z14g2000cwz.googlegroups.com...
>
>
> (remind people of
> the same trick that a database with a dropped column can be recovered
> by altering data dictionary, as long as it's subsequently exported and
> imported; credit to an Oracle expert well known in this newsgroup)
>
A (not very minor) correction to the 'trick'.
If you want to do this, then you export the entire
database BEFORE you hack the dictionary.
THEN dump to file the table data that reappears.
(Note - an export at this point may not be safe, as
it may carry some definition error if your hacking of
the data dictionary missed something). Whether you
do an export or another form of extract, you need to
check that the data extracted is actually correct, and
that (for example) any triggers are associated with
the correct columns.
Then you create an entire new database and import the
entire export, drop the incorrect version of the table,
and reload the table.
Alternatively:
You create a backup of the database (or a self-consistent
set of tablespaces). Hack the backup, dump the table
from the backup. Drop the table from the original and
reload the copy. This is probably going to be quicker in
most cases - but some people may have a problem finding
somewhere to put the copy of the database.
Both strategies, of course, requires you to stop people
using the database for the duration (to avoid changes
to the critical table), and both require you to have a
method for checking that the data from the dumped table
is correct, and that all the triggers, indexes, RI etc. are correct
after you've rebuilt the table.
Any other strategies leaves you with a data dictionary that
may not be self-consistent, and therefore may crash,
corrupt, or breach the security of your system at some
future date. Either strategy is messy and long-winded - but
you don't want a production database which has had the
sys tables updated by hand.
jks...@gmail.com
>
> > Many utilities exist that can read MS Office formats, some of
> > them are open source - this seems analagous to me.
>
> Reading the contents of a Word document is far different, to me, than
> reading the contents of a corrupted, proprietary-format datafile,
> especially with a tool that is incomplete, as admitted by the
> programmer. Lacking the ability to recover LOB data, comporessed
> indexes and IOTs certainly leads me to wonder why I should trust such a
> tool to correctly read any other data in a corrupted file. The OP has
> claimed his tool is "ready for recovery", yet misses some data
> structures, meaning tables with LOB columns will be incomplete upon
> 'recovery', IOTs will be missing as well as compressed indexes. To me
> this is a failed attempt at reverse-engineering DUL; it is also an
> accident waiting to happen. Using this 'tool' to 'recover' corrupted
> datafiles would violate any service agreements between Oracle and the
> party using said 'tool' as it is NOT Oracle's DUL utility. There is
> nothing good about this work as I see the situation. Possibly I am
> wrong about the legality of this software and the process used to
> create it; from what I have read thus far I do not believe I am. That
> notwithstanding, this 'tool' is a hazard for anyone to use, whether it
> be legal or not. I certainly wouldn't trust it to recover my data.
>
Something else that comes to mind involves the PC BIOS.
Years ago some companies decided to create BIOS emulators.
Phoenix for instance.
When hiring engineers to create the BIOS code, the candidate engineers
had to be interviewed by attorneys to ensure that the engineers were
'virgin', that is, they had never had any involvement with IBM BIOS,
and had in fact never even seen the code.
Producing a BIOS that emulated IBM's was legal under these conditions,
as they were not reverse engineering the code, but producing new code
that emulated the behavior of the IBM BIOS.
Jared
Frank van Bortel
I thought originally it was called Duinen's UnLoader, after the
writer, Benard van Duinen. He used to work for Oracle, don't
know if he still does. Why not contact him?
--
Regards,
Frank van Bortel
Top-posting is one way to shut me up...
Sybrand Bakker
wrote:
> anyway this tool is just a personal softwar
So why are you advertising it? Evidently to gain profit sooner or
later.
Joel Garry
Just because metalink or the docs say how to do something, doesn't
cancel licensing restrictions - nothing stopping you from installing
partioning with no license, for example. That's fair enough, but the
licensing and laws are abusive on their face.
The practical effect boils down to Larry's mood at the moment. If he
can maintain a 40% margin over the next few years, his mood might
remain pretty good and nobody has anything to worry about short of
temerarious theft of intellectual property.
jg
--
@home.com is bogus.
jks...@gmail.com
That is some spurious reasoning.
Partitioning is a feature that is licensed separately from the basid
database software, and that is made clear.
There is no licensing pertaining to the use of 10046 traces.
There are a number of things available on MetaLink that are not
documented anywhere else by Oracle, but you are free to use them.
Scripts, papers, howto papers, etc.
They may not be supported (use them at your own risk, don't open a TAR
asking how to use it, etc), but they are placed there to use.
Jared
Anurag Varma
So according to Mr. David Fitzjarrell, the tool "Shareplex" by Quest is
Illegal?
It does do similar reverse engineering.
Anurag
fitzj...@cox.net
Apparently you missed this statement of mine in your 'skip-reading' of
the thread:
"Possibly I am wrong about the legality of this software and the
process used to create it"
Thus, I have admitted I don't know everything. Were you to read the
entire thread you'd have realised this and possibly not posted such a
remark.
David Fitzjarrell
Joel Garry
??? How do run a 10046 trace without running Oracle? How do you run
Oracle legally without accepting a license?
>
> There are a number of things available on MetaLink that are not
> documented anywhere else by Oracle, but you are free to use them.
> Scripts, papers, howto papers, etc.
>
> They may not be supported (use them at your own risk, don't open a TAR
> asking how to use it, etc), but they are placed there to use.
I've driven cars at, um, over 75 mph on the highway. Does that mean
I've violated my driving license privilege? The government placed the
road there for me to use!
The point wrt Oracle is, the licensing and laws involved would prevent
people from doing normal things if followed precisely. That is because
they are written by lawyers to have a strong position in any possible
case. What is lacking is law to protect the other side (that would be
us programmers and users). Beyond that, I think there should be
compulsion for software vendors to guarantee software do what it
claims. But that's just me.
I just used OEM to see:
SELECT property, intcol
FROM sys.exu8ccl
WHERE cno = :1
ORDER BY colno
Guess what! I'm seeing something from an Oracle binary! Because I'm
far too lazy to bother with a 10046 trace for something so simple. I'm
explicitly violating my license! neener, neener Oracle!
Well, it's not really my license, it belongs to somebody else... and
Oracle says they don't have a contract because the license is really
with another vendor. Oh wait, they don't exist anymore. And somebody
else _is_ giving money to Oracle for the license. But Oracle can keep
the money and say there is no support. Right?
jg
--
@home.com is bogus.
Freke Dorothy Fluck Lane
3000...@gmail.com
> Hi all:
> I have sucessfully recover some data for two customers with MYDUL,
> also I don't know whether they have another recover method, it now my
> pleasure to announce that MYDUL is ready for recovery just as DUL,
> except for LOB column and compress table/IOT.
>
> http://www.anysql.net/en/
> http://www.anysql.net/en/mydul.html
bucuo bucuo