GCC association with the FSF

[Gabriel Ravier]

On 4/9/21 1:48 PM, Pankaj Jangid wrote:
> Gabriel Ravier via Gcc <gcc@gcc.gnu.org> writes:
>
>> RMS is not indispensible because he does not contribute to GCC and
>> doesn't bring much to it, and otherwise takes more away from it. If
>> you were to remove all of Ian, Jonathan, Joseph and Nathan you would
>> be removing ~13% of active contribution to GCC (counting in
>> commits). If you also remove all the major contributors that are from
>> corporations (counting a major contributor as someone with 10 or more
>> commits), you're removing ~63% of active contribution. If you also
>> remove the major organizations contributing to GCC, like Adacore and
>> the GDC project, you're removing ~18% more of active contribution,
>> meaning you're left with 19% of active contribution. While I do not
>> doubt that all of the contributors that would remain are talented
>> individuals, GCC would undoubtedly, in the best case, heavily suffer
>> from the loss of 3 to 4 fifths of active contribution and become much
>> less appealing as a compiler, and in the worst case simply die
>> out. While each of the individuals forming any of those groups aren't
>> indispensable, as a group, they certainly are indispensible to GCC
>> unless you think GCC can really survive with 3/5 times less
>> contributions to it.
> What is this man? Are you trying to compute the probability of survival
> a project? You forgot to count me. I am one of the users of GCC. If
> there are no users then the project is dead; however heavyweight the
> maintainers are.
>
> And let me also tell you the truth. I have looked at the list of
> maintainers and the steering committee for the first time, when this
> thread was started. My reason for sticking to GCC is FSF and associated
> cause. Not the above list of people. Those who are not connected with
> the cause have already started migrating to the competing tools.
>

If you have an enormous exodus of maintainers that takes away 4 fifths 
of maintainers, then there is a very high probability that the project 
will simply die by essentially all measures except for such asinine ones 
as "there is still at least 1 user of it" (under which, say, Version 6 
UNIX is not dead as there are still computers running it and people 
using it), as a GCC with much less people maintaining it would, over 
time, become very unattractive as a tool for actually making programs, 
as progress on developing it would become very slow. It would, for 
example, lose any C++ users that want to use anything beyond C++17 or 
the partial support for C++20 that GCC has right now.

While I am not saying that the amount of maintainers is directly tied to 
the survival of a project, I would certainly say that a project with 
near to no maintainers without which it cannot compete with competing 
projects (for example, Clang) /will/ die off.

The only ones that would remain would be those that would use GCC 
despite its enormous shortcomings for the single and only reason that it 
is licensed under the GPL, and those would be rather rare compared to 
the amount of people that use GCC right now. I am not saying that they 
are just a few dozen people or something like that, but GCC would become 
a shadow of its former self without any other support.

I would say that under those circumstances GCC would become about as 
popular as Turbo C or other antiquated tools like it, and I would 
certainly hope one would consider Turbo C to be a dead compiler, despite 
the fact that it still has at least 1 active user. While I don't think 
this outcome is likely, it would become likely if every single 
corporation and organization involved in the development of GCC suddenly 
retracted support for it. Do you really think GCC could remain 
competitive compared to compilers like Clang or MSVC if development on 
it was 5 times as slow, and if distributions like Fedora and Ubuntu 
started to migrate to LLVM, or even maybe straight up removed GCC from 
their repositories ?

PS: Of course, this is completely implausible, and it is almost certain 
that this will never happen, but you're implying that GCC can perfectly 
survive without any support from corporations: I am simply telling you 
what would happen if all those corporations actually stopped to support it