Skip to content

how to make a true binary package?

9 messages · andre zege, Uwe Ligges, Simon Urbanek +4 more

#
I'd like to have a version of a package that doesn't include sources. I
thought that this could be achieved by using binary option in R CMD build,
but in fact it packages source code that could be easily printed once the
library is loaded. Is there an option to avoid visibility of the source? 

Another related question is how to install two versions of the same package,
so that i could load either of them? I am sure this could be done, just
cannot find this in docs.

I am using R 2.12.1 on Red Hat Lunix 

--
View this message in context: http://r.789695.n4.nabble.com/how-to-make-a-true-binary-package-tp3895117p3895117.html
Sent from the R devel mailing list archive at Nabble.com.
#
On 11.10.2011 18:33, A Zege wrote:
R is open source software ...

Anyway, for C sources: these cannot be printed easily in a binary 
package. But then make sure your license is file given you link against 
something under GPL ...
You can install each version of the package into a separate library.

Uwe Ligges
#
On 11/10/2011 12:33 PM, A Zege wrote:
No.  As Uwe said, R is an open source project.  We like to encourage 
others to be open source too.

Duncan Murdoch
#
OK, gentlemen, i agree with you in general. I was not talking about a general
purpose, general use package that one prepares for CRAN. I am sure you are
familiar professionally or can imagine situations where you need to
demonstrate a solution to a specific task without fully disclosing the
details -- sometimes even hard core open source adherents need to sacrifice
desire for openness for some prosaic purposes, like getting paid :).
Compilable languages give an easy solution of a binary code. It sounds like
if one wants true binary, he has to recode in C++. I thought it's possible
in R as well, i thought this was discussed even as a default behavior for
next version of R to make stuff go faster. Maybe not. Thanks, anyway.


--
View this message in context: http://r.789695.n4.nabble.com/how-to-make-a-true-binary-package-tp3895117p3895262.html
Sent from the R devel mailing list archive at Nabble.com.
#
On Oct 11, 2011, at 1:09 PM, A Zege wrote:

            
Those are fairly disjoint concepts - open source doesn't mean you can't get paid. Also the fact that the source is accessible doesn't mean that it is legal for someone else to take it - you define what you allow in the license. Note, for example that Java bytecode can be easily disassembled into readable Java source form and yet there is a lot of commercial software written in Java.
You can compile the functions and strip the source code from the bytecode objects. Such a thing can be serialized without revealing the original source code. However, you can still de-compile it, and R doesn't guarantee that such stripped objects will continue to work, so your mileage will vary and you may want to check if it's what you want.

There are obvious other ways you could use that are much simpler - such as encrypt your sources. Nothing is completely bullet-proof, it's just a matter of how much work it requires to break whichever method you choose.

Cheers,
Simon
#
One of the most secure options, I think, is to provide a 
web-based tool to do the computations.  I know one case where this was 
done.  The clients probably did not even know that R was being used to 
do the computations.  They had an application or a procedure for 
uploading their data to a secure web site, which then returned answers.  
This too can be defeated by determined hackers, but if you are careful 
about your encryption, etc., you can make it quite difficult for them.


       Hope this helps.
       Spencer
On 10/11/2011 10:41 AM, Simon Urbanek wrote:
#
On Tue, Oct 11, 2011 at 6:09 PM, A Zege <andre.zege at gmail.com> wrote:
[deja vu all over again]

 What you seem to want to do is code obfuscation. With C, the
obfuscation is caused by it being turned into machine code
instructions, making exact reconstruction of the C source impossible,
but reconstruction of the code as assembly language and hence reverse
engineering very possible.

 With R, there is no compilation to machine code, so plain R source
code has to be available to the R interpreter [exception: see Simon's
talk of bytecode]. Any encryption you put on has to be decrypted by
the user in order to run it. I've said this a few times on R-help and
maybe R-devel too.

 If someone is offering to pay you, then if you tell them they can see
the source code then they should want to pay you more.

 There's also nothing to stop you putting a restrictive license on
your source code - including clauses like 'if you look at the contents
of any *.R files you are in breach of this license'. These things
might even stand up in court.

 Barry 'Not a lawyer' Rowlingson
#
If you want to hide some code from accidental or casual sight then I have had some success with adding something like this to the .R file for a package (the petals function in TeachingDemos in this case):

.onAttach <- function(...) {
    petals <- petals
    attr(petals,'source') <- "Don't Cheat!"
    assign('petals',petals,'package:TeachingDemos')
}

Though that would only slow an experienced R programmer down by maybe 10 seconds if they really wanted to see the code, it might take an intermediate level R programmer a whole minute to figure out how to work around this.  But it would prevent an honest user that knew of a license or agreement to not look at the code from accidentally seeing what they agreed not to.

You could also include something in the code that checked the attribute and stopped working if it was changed or somehow notified you of the change.  Combining this with the license agreements as others have said might give you additional recourse.

You can see the petals.R file in the TeachingDemos package on R-forge for another example of obfuscating (bleaching) a key piece of code.  Combining the 2 might slow down an expert by a whole 30 seconds, but at least they would not be able to claim that they accidentally saw the code.

Whether this violates the spirit of R licensing or not is another issue.
#
On 12/10/2011 3:57 PM, Greg Snow wrote:
You should be aware that this will no longer work in R 2.14.0, which 
finally drops the "source" attribute.  This would achieve a similar effect:

srcfile <- srcfilecopy("", "Don't cheat")
srcref <- srcref(srcfile, c(1,1,1,12))
attr(petals, "srcref") <- srcref

Duncan Murdoch