LGPL VP6 decoder released - Page 8

Sirber · 22nd April 2006, 18:43

Isn't java kind of easy to decompile?

Liisachan · 22nd April 2006, 18:52

Sirber: yes. but if you do that, you can't get the original variable names. So, that's not the case if we believe that lawyer.

soresu · 22nd April 2006, 18:54

So, if true this would mean that any reverse engineering of it is not actually bound to silence by the DMCA?

CEC · 22nd April 2006, 20:10

Can you imagine what might happen if someone put the code of their encoder on the net???????

They are nearly crying with the fact that their decoder code is out!!!

siddharthagandhi · 22nd April 2006, 20:56

What rjamin is doing is probably not legal, and he shouldn't do it, but striking him isn't going to stop it. Everyone already knows the link and as long as this thread is viewable anyone can still acess the code, so that won't stop anything.

MfA · 22nd April 2006, 22:30

Googling a bit variable names seem to be actually included in unobfusciated byte code (or at least they were at one point in time).

dimzon · 22nd April 2006, 23:11

Quote:

Originally Posted by Liisachan

Sirber: yes. but if you do that, you can't get the original variable names.

Debug information...

shark37 · 22nd April 2006, 23:15

Quote:

Originally Posted by BoNz1

You might remember a little while ago that On2 had some Java applets on their website to demonstrate VP5/6.

TrueMotion Streaming Java Applet 1.1

Unfortunately webarchive page refers to .jar and .avi files located on www.on2.com -- see <applet...> fragment of page source below

[HTML]<applet archive="truemotion.jar,on2_speex.jar" code="On2.VPXDecoderApplet.class" width="240" height="144">
<param name="InputFileName" value="http://www.on2.com/applet/potter240-300.avi">
<param name="Seconds" value="50,60">
<param name="soundSystem" value="JSND">
<param name="JumpToURL" value="http://www.on2.com/sales.php3?qs1=Java%20Licensing">
<param name="StartImage" value="http://www.on2.com/applet/intro.jpg">
<param name="EndImage" value="http://www.on2.com/applet/end.jpg">
<param name="PostProcessLevel" value="0">
<param name="timeout" value="10000">
<param name="credits" value="http://www.on2.com/applet/credits.txt">
</applet>[/HTML]

But...
with a little help of Google anyone can still take a look at "truemotion.jar"

http://www.google.com/search?sourcei...uemotion%2ejar

From
"G o o g l e's cache of http://www.martyrtv.co.uk/index_files/ as retrieved on 5 Jul 2005 23:55:54 GMT."
one still can download http://www.martyrtv.co.uk/index_files/truemotion.jar

Quote:

Originally Posted by Liisachan

Sirber: yes. but if you do that, you can't get the original variable names.

Interesting that there was debug version of the applet -- "truemotion_debug.jar" unavailable now

"G o o g l e's cache of http://manzanaroja.com/on2/test/ as retrieved on 19 Aug 2005 11:09:02 GMT"

bratao · 22nd April 2006, 23:28

In java decompile you get the original variable names !
This not happen, only if the developer use some type of obfuscator..
But the normal compile , in a decompile you get the original names !

bratao · 22nd April 2006, 23:31

Update>
I download truemotion.jar and the file VP6Decoder.class in essence is the source found in cvs..
But this code is obfuscate , you cant get the original variables names..
"Obfuscation by RetroGuard - www.retrologic.com (author: Mark Welsh, markw@retrologic.com)"

but exist a debug version, its possible that this version is not obfuscated !

rjamorim · 23rd April 2006, 01:09

Quote:

Originally Posted by siddharthagandhi

What rjamin is doing is probably not legal, and he shouldn't do it, but striking him isn't going to stop it.

Why would I be stricken? Did I break forum rules?

Quote:

Originally Posted by BoNz1

Ok, so I have a little theory as to where the code came from, since nobody seems to know and On2 seems far to embarrassed to say anything, you might guess that it's pretty funny. And I think it is. And I have a good hunch it isn't just a theory...Since a lot of people who are more in the know than I am think this is where it came from.

You might remember a little while ago that On2 had some Java applets on their website to demonstrate VP5/6. Unfortunately, for On2 they never required anyone to sign any license of any sort. The code did not come from someone hacking the company network or a disgruntled employee. It came from those applets according to what I have heard.

Right. I actually just got this e-mail:

Quote:

Hi, I've been following the discussion about libvp62. I just want to
give you the story how I think this all happend. Short version: this
code is most likely a legit reverse engineering job. Long version: on2
has distributed some java applets with vp5 and vp6 decoding
capabilities. These applets were availible from their homepage without
any shrinkwrap license. One version can be found here:
http://www.martyrtv.co.uk/index_files/truemotion.jar

The nice thing with java is that it is easy to RE, just try it on a java
class with jad. The vp6 applets were obfuscated with Retroguard, the vp5
applets wheren't that obfuscated. On
http://multimedia.cx/eggs/category/r...n2duck/page/1/
can you read abit more regarding the vp5 applets.

I have several versions of the diffrent applets and also some source. I
compared the source of libvp62 and one of the vp5 sources. Although not
a perfect match some functions looked really really similar.

Giving someone fairly experienced with videocodecs and with all this
information, the probability of libvp62 being a valid RE work is quite high.

So we have on2 claiming that the source is from them, I would say it's
from RE work on their binary java applets.
They then claim license breach, well they distributed them by themselfs
without any need to agree on some license. This can
be veryfied by looking at their site from archive.org.

I think this info invalidate their claims of the source beeing stolen.

And then the question about complying with their request. I would like
to see their response to this information, and the best way to get a
proper response is by keeping distributing the source. But I wouldn't
blame you for not wanting the hassle and I can tell you I would stop
distributing the source.

I can back all my claims with source or other information if you doubt
some of my claims.

Sounds very well explained and thought out. I guess that's "reasonable doubt" for you...

Edit: I omitted the e-mail author's name because I don't know if he wants to be identified. He is probably reading this thread anyway, so he can identify himself if he so wishes.

Who knows, maybe he is "ztoon padli" :P

rjamorim · 23rd April 2006, 01:17

Quote:

Originally Posted by Doom9

Reminds me of SCO. Let's recall Sigma and XviD.. the XviD team properly backed up their claims, and in no way would publishing the matching original source code constitute open sourcing proprietary code.. the code would still be protected under copyright law unless the code comes with an open source or other license that allows free use.

That actually made me wonder about the lawyer firm's understanding of open source, source code and software development as a whole. I'm afraid I know more on this subject than mister Berger (or, at least, they expect me to know very little)

Quote:

So, if I were in this situation, I'd honor the request, but insist that irrefutable proof be provided within a certain period of time. When being accused of aiding in copyright infringement on mere heresay, if you comply (under threat of a lawsuit nontheless) courtesy also goes the other way.

More on that as I reply to Nic's post...

Quote:

Originally Posted by dragongodz

i had to laugh at this example because its so far different than this case it beggars belief. hows it different ? people can read and compare the words. now if someone wrote a story and Hemingway came out later and said it contains large parts of 1 of his works which people have not read and are not allowed to what then ? would you still believe Hemingway or would you(or a court) ask that someone be allowed to read those parts to compare ?
sorry but that statement is either (1) a joke,(2) coming from someone who doesnt have a clue about the comparison or (3) someone purposfully trying to make a smokescreen so they dont have to prove anything. since its from a lawyer i will leave that up to yourselves to decide.

Pathetic, really. I wouldn't hire a lawyer firm like that to defend me from accusations of jaywalking.

rjamorim · 23rd April 2006, 01:22

Quote:

Originally Posted by Nic

@rjamorim: I would comply. They have been respectful enough now and to be honest a VP62 decoder source code isn't going to be much use to anyone if On2 will try and shutdown/takedown anything made from it. I have a feeling they're not going to let the issue drop, IMHO it isn't worth your efforts rjamorim.

I agree. It's no use keeping distributing this code (which is pretty useless to non-programmers) if they plan to FUD any potential user - FFMPEG, ffdshow, MPC, GStreamer or the like. I guess anyone able to take this code and create a clean room reverse engineering out of it already got it. So, I'll probably just remove it from RareWares tomorrow, as a major overhaul of the site is planned (a whole section will be moved to ReallyRareWares) and leave some notice, maybe instructions to download it from The Pirate Bay

More than 200 people downloaded it. That's a good enough amount of people to make sure this code won't simply vanish.

Liisachan · 23rd April 2006, 01:56

@shark37
Nice job

I'm not sure, but I'm feeling this way:

If truemotion_debug.jar is (or was in the past) in public, and freely accessible via HTTP (meaning, without even EULA)--then, technically, what was apparently done is a reverse engineering in a broader sense, and is generally legal.

dragongodz · 23rd April 2006, 02:07

rjamorim - when you email them that you are removing the download it would be interesting if you pointed out how silly the Hemigway comparison was and see what(if?) response to that aswell.

actually looking around the places where On2 devs have been posting about this i see they seem to have stopped. i am guessing all the calls for proof from lots of places is something they are just not willing to do and can not really argue against. as its been said before, to ask for some kind of proof, even to just select individuals, is not unreasonable. unfortunatly they seem to have decided to now leave it to the lawyers.

vlada · 24th April 2006, 02:25

Hi,
I think that I just let you know this: In my country is reverse engineering explicitly legal. Any license agreement not written in Czech language is obsolete. SW patents do not exist here. Your own code is copyrighted and that's enough.

Sirber · 24th April 2006, 02:30

Quote:

Originally Posted by vlada

Any license agreement not written in Czech language is obsolete.

Isn't it kinda racist?

[edit]

Also, if my software is GPLed, and the license is in english, it's not valid in your country?

[edit 2]

Here you go: http://www.gnu.org/licenses/gpl.cs.html

rjamorim · 24th April 2006, 02:59

Quote:

Originally Posted by Sirber

Here you go: http://www.gnu.org/licenses/gpl.cs.html

Well, that GPL has no value. The FSF themselves state that only english language versions have value

In vlada's case, in theory, since he can't understand the license (or, even if he understands, it would have no legal value as legal documents must be written in the county's current language), he must refuse to accept it (and therefore not use the software/source code/whatever). YMMV, according to your country's laws, of course.

Actually, that's why software such as Photoshop comes with EULAs in several languages

Aaaaaanyway...

To comply with Herrrrr Berger's polite but pitiful requests, I removed VP6's source code from RareWares.

But wait! You can still obtain VP6 source code from RareWares!
http://www.rarewares.org/files/others/VP6_src_legal.zip

Ohhh, poor pitiful me...

Sirber · 24th April 2006, 03:08

Hum.... that seems to be legal now

MfA · 24th April 2006, 03:12

The license being invalid doesn't really matter much to the GPL, without the GPL you can't redistribute the source code period.

22nd April 2006, 18:43	#141 \| Link
Sirber retired developer Join Date: Oct 2002 Location: Canada Posts: 8,978	Isn't java kind of easy to decompile? __________________ Detritus Software

22nd April 2006, 20:10	#144 \| Link
CEC Chaos Energy Creator User Join Date: Sep 2005 Posts: 49	Can you imagine what might happen if someone put the code of their encoder on the net??????? They are nearly crying with the fact that their decoder code is out!!! __________________ A FIREFOX SEARCH ENGINE FOR DOOM9'S FORUM

23rd April 2006, 02:07	#155 \| Link
dragongodz .... Join Date: May 2002 Location: Australia Posts: 2,797	rjamorim - when you email them that you are removing the download it would be interesting if you pointed out how silly the Hemigway comparison was and see what(if?) response to that aswell. actually looking around the places where On2 devs have been posting about this i see they seem to have stopped. i am guessing all the calls for proof from lots of places is something they are just not willing to do and can not really argue against. as its been said before, to ask for some kind of proof, even to just select individuals, is not unreasonable. unfortunatly they seem to have decided to now leave it to the lawyers. __________________ Narrator: And of course, with the birth of the artist came the inevitable afterbirth - the critic. (History of the World part 1)

24th April 2006, 03:08	#159 \| Link
Sirber retired developer Join Date: Oct 2002 Location: Canada Posts: 8,978	Hum.... that seems to be legal now __________________ Detritus Software

22nd April 2006, 18:52	#142 \| Link
Liisachan 李姗倩 Lǐ Shān Qiàn Join Date: Nov 2002 Posts: 1,340	Sirber: yes. but if you do that, you can't get the original variable names. So, that's not the case if we believe that lawyer.

22nd April 2006, 18:54	#143 \| Link
soresu Registered User Join Date: May 2005 Location: Swansea, Wales, UK Posts: 196	So, if true this would mean that any reverse engineering of it is not actually bound to silence by the DMCA?

22nd April 2006, 20:56	#145 \| Link
siddharthagandhi Go Nero Digital Join Date: Jan 2006 Location: Edison, NJ Posts: 466	What rjamin is doing is probably not legal, and he shouldn't do it, but striking him isn't going to stop it. Everyone already knows the link and as long as this thread is viewable anyone can still acess the code, so that won't stop anything.

22nd April 2006, 22:30	#146 \| Link
MfA Registered User Join Date: Mar 2002 Posts: 1,075	Googling a bit variable names seem to be actually included in unobfusciated byte code (or at least they were at one point in time).

22nd April 2006, 23:28	#149 \| Link
bratao Registered User Join Date: May 2005 Posts: 146	In java decompile you get the original variable names ! This not happen, only if the developer use some type of obfuscator.. But the normal compile , in a decompile you get the original names !

22nd April 2006, 23:31	#150 \| Link
bratao Registered User Join Date: May 2005 Posts: 146	Update> I download truemotion.jar and the file VP6Decoder.class in essence is the source found in cvs.. But this code is obfuscate , you cant get the original variables names.. "Obfuscation by RetroGuard - www.retrologic.com (author: Mark Welsh, markw@retrologic.com)" but exist a debug version, its possible that this version is not obfuscated !

23rd April 2006, 01:56	#154 \| Link
Liisachan 李姗倩 Lǐ Shān Qiàn Join Date: Nov 2002 Posts: 1,340	@shark37 Nice job I'm not sure, but I'm feeling this way: If truemotion_debug.jar is (or was in the past) in public, and freely accessible via HTTP (meaning, without even EULA)--then, technically, what was apparently done is a reverse engineering in a broader sense, and is generally legal.

24th April 2006, 02:25	#156 \| Link
vlada Registered User Join Date: Apr 2003 Location: Czech Republic Posts: 752	Hi, I think that I just let you know this: In my country is reverse engineering explicitly legal. Any license agreement not written in Czech language is obsolete. SW patents do not exist here. Your own code is copyrighted and that's enough.

24th April 2006, 03:12	#160 \| Link
MfA Registered User Join Date: Mar 2002 Posts: 1,075	The license being invalid doesn't really matter much to the GPL, without the GPL you can't redistribute the source code period.