This is the mail archive of the mailing list for the GCC project.

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: Exhaustive Instructions for Toolchain Generation

On 05/10/17 22:16, R0b0t1 wrote:
On Wed, Oct 4, 2017 at 3:33 AM, David Brown <> wrote:

R0b0t1, you might not realise this but CodeSoucery is a major
contributor to gcc and other gnu tools.  Individuals and companies pay
them for their services - to put together tested, qualified and
documented bundles of development tools, for support, for porting gnu
tools to new architectures, for supporting the latest chip families, for
fixing problems, etc.  The work they do runs right back to gcc -
patches, fixes, improvements, mailing list support, release management,
etc.  Take a look at the gcc patches mailing list, or release lists, or
lists of contributors and maintainers.  You will see CodeSourcery email
addresses all over the place.

You make your own choices about where you spend your money or not.  But
if you use gcc, remember that you are not just using the work of RMS,
and a bunch of generous idealistic volunteers - you are also using the
work of companies such as CodeSourcery, Redhat, Intel, Google, and many,
many others.  You might choose not to pay CodeSourcery for their work
here, but if you really are going to be "respectful", then you should be
thanking them rather than condemning them as "the misery of closed
source software".  They are a fine example of how cooperation between
commercial entities and free software is supposed to work.

(This is not a business solicitation, I have no affiliation with
CodeSourcery.  It just bugs me when people are disrespectful or
insulting to individuals or companies that are trying to do the right


I find it hard to care about someone's position or affiliation but
instead choose to care about what they do and how they act. If it was
Sandra's intent to ask me for free work, then I am not sure how that
qualifies as "the right thing." Per my latest response to her, even if
that is what she meant to do, my actions were not meant to be an

I apologize if they were interpreted as such, but I would caution
against associating one's work with oneself too closely.

That companies contribute to open source software and that those
contributions are useful is not something I ever wanted to dispute,
but at the end of the day those companies make money by restricting
access to information. There was no indication the information I was
asked to provide would ever benefit the GCC or open source community
at large.


I don't know what information was being asked for, nor how it would be used, so I cannot comment on the particular case. I can merely say that as a general rule, a company like CodeSourcery lives and dies by its cooperation with the free and open source communities. Whether this particular piece of information or work would have directly benefited gcc, I have no idea - but if it could be of use to free and open source development tool projects, it would have ended up there. Yes, CodeSourcery aims to make money. That money, in part, then pays for more development and free support of the GNU tools. No one claims it /all/ goes there, but certainly a solid portion does. And yes, to make that money, CodeSourcery may restrict access to some information. That mean that paying customers get early access to code, or non-FOSS libraries, or additional tools, or detailed certification and testing reports. I am at a loss to see how any of that could be seen as "wrong" - though I can see why you might not want to share information or contribute to their work without direct assurances that it will be published freely. (Noting, however, that /you/ are benefiting directly from code and information that CodeSourcery has given freely, without asking you for anything in return.)



Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]