Ada source code obfuscator
in [ADA]
|
Prev: operations on Directory
Next: Access idiom
From: okellogg on 19 Jan 2008 18:28 Does anyone know of a free obfuscator for Ada source code? (I hit a memory thrashing bug in FSF gcc-4.[1-3] that requires a large amount of proprietary code to reproduce. As soon as I remove stuff from the test case the memory hog goes away. I'd like to post a bug report to GCC bugzilla but would need to obfuscate the source code.) Thanks, Oliver
From: Simon Wright on 20 Jan 2008 06:26 okellogg <okellogg(a)freenet.de> writes: > Does anyone know of a free obfuscator for Ada source code? No, but J-P Rosen's AdaSubst see http://www.adalog.fr/adalog2.htm) might form a basis?
From: anon on 20 Jan 2008 10:23 Ada was designed not to allow "obfuscation" of its code. The only way to protect your proprietary code is to use a close source version of Ada like Green-Hill, IBM Ada compiler which is based on Ada 95 spec. Compile the program and only distribute the binary version. For Adacores GNAT version compilers, unless you are using GNAT PRO, you must supply the source code with the binary. See the FAQ page at Adacore. It states that all code create by the GNU/GPL version of GNAT must be license under the GPL. Which means that the source must be made available to the users of the code. In <bce765d6-c3be-4849-9911-e89608a11880(a)v46g2000hsv.googlegroups.com>, okellogg <okellogg(a)freenet.de> writes: >Does anyone know of a free obfuscator for Ada source code? > >(I hit a memory thrashing bug in FSF gcc-4.[1-3] that requires a large >amount of proprietary code to reproduce. As soon as I remove stuff >from the test case the memory hog goes away. I'd like to post a bug >report to GCC bugzilla but would need to obfuscate the source code.) > >Thanks, > >Oliver
From: Gautier on 20 Jan 2008 12:09 I'm afraid you may have misunderstood Oliver's need. His Ada compiler seems to produce memory-leaking code on a very specific situation, which involves proprietary code, and he wants to report this bug with source code so that the compiler maintainers can reproduce this problem. It is not about distributing source code to customers. And if his compiler was Green-Hills or IBM, the situation would be exactly the same: Green-Hills or IBM would say: "Are you sure it is a bug of ours ? Could you please send an example of sources which shows this problem ?". The obfuscation needed here consists in replacing each identifier like "Send_money", "Launch_missile" or "The_solution_to_all_problems",... by I29684, I29685, I29686,... G. anon wrote: > Ada was designed not to allow "obfuscation" of its code. > > The only way to protect your proprietary code is to use a close source > version of Ada like Green-Hill, IBM Ada compiler which is based on > Ada 95 spec. Compile the program and only distribute the binary version. > > For Adacores GNAT version compilers, unless you are using GNAT PRO, > you must supply the source code with the binary. See the FAQ page at > Adacore. It states that all code create by the GNU/GPL version of GNAT > must be license under the GPL. Which means that the source must be > made available to the users of the code. > > In <bce765d6-c3be-4849-9911-e89608a11880(a)v46g2000hsv.googlegroups.com>, okellogg <okellogg(a)freenet.de> writes: >> Does anyone know of a free obfuscator for Ada source code? >> >> (I hit a memory thrashing bug in FSF gcc-4.[1-3] that requires a large >> amount of proprietary code to reproduce. As soon as I remove stuff >>from the test case the memory hog goes away. I'd like to post a bug >> report to GCC bugzilla but would need to obfuscate the source code.) >> >> Thanks, >> >> Oliver ______________________________________________________________ Gautier -- http://www.mysunrise.ch/users/gdm/index.htm Ada programming -- http://www.mysunrise.ch/users/gdm/gsoft.htm
From: Larry Kilgallen on 20 Jan 2008 14:21
In article <47938035$1_5(a)news.bluewin.ch>, Gautier <gautier(a)fakeaddress.nil> writes: > I'm afraid you may have misunderstood Oliver's need. His Ada compiler seems to > produce memory-leaking code on a very specific situation, which involves > proprietary code, and he wants to report this bug with source code so that the > compiler maintainers can reproduce this problem. It is not about distributing > source code to customers. And if his compiler was Green-Hills or IBM, the > situation would be exactly the same: Green-Hills or IBM would say: "Are you sure > it is a bug of ours ? Could you please send an example of sources which shows > this problem ?". The obfuscation needed here consists in replacing each > identifier like "Send_money", "Launch_missile" or > "The_solution_to_all_problems",... by I29684, I29685, I29686,... But to get a rapid solution from a vendor often requires providing a _simple_ reproducer. There is often no easy way to do that. |