4 ms·

GCC was originally designed to not be modular so that it wouldn't be possible to integrate it with closed-source code.

by moonchild 6y ago

CyberRabbi 6y ago
Source?
- macintux 6y ago
  Quick searches turned this up: https://lwn.net/Articles/582697/ https://lwn.net/Articles/582697/
  - arp242 6y ago
    Also something like: https://lists.gnu.org/archive/html/emacs-devel/2015-01/msg00015.html https://lists.gnu.org/archive/html/emacs-devel/2015-01/msg00... rms deliberately making GCC (and by extension, Emacs) less useful to "prevent misuse". The entire thread is ... ugh. The funniest part is that it doesn't even matter, because now everyone and their mother has switched to llvm/clang for these kind of things, just as people in that thread predicted.
    - mhh__ 6y ago
      > everyone and their mother has switched to llvm/clang Chris Lattner even offered it to them, but RMS didn't get the email because of the unique way he browses the net
      - spekcular 6y ago
        Do you have a source handy for the missed email thing? I'd love to read a firsthand account.
        mhh__ 6y ago
        I mentioned it once, lost the citation, found it and lost it again. I find I can find I'll link it tomorrow.
        spiralx 6y ago
        https://lists.gnu.org/archive/html/emacs-devel/2015-02/msg00594.html https://lists.gnu.org/archive/html/emacs-devel/2015-02/msg00...
        moonchild 6y ago
        I don't have a citation for the anecdote, but possibly this was the email? https://gcc.gnu.org/legacy-ml/gcc/2005-11/msg00888.html https://gcc.gnu.org/legacy-ml/gcc/2005-11/msg00888.html
        spiralx 6y ago
        https://lists.gnu.org/archive/html/emacs-devel/2015-02/msg00594.html https://lists.gnu.org/archive/html/emacs-devel/2015-02/msg00...
        spiralx 6y ago
        I found it here in an earlier HN discussion: https://lists.gnu.org/archive/html/emacs-devel/2015-02/msg00594.html https://lists.gnu.org/archive/html/emacs-devel/2015-02/msg00...
    - spiralx 6y ago
      I know RMS is often abrasive, but that it is a whole new level of bad faith from him on display there. Plenty of "just asking questions" from one of the other participants as well. Here's another example of RMS on the subject where he told someone he wouldn't accept their code and that they should take it off of the internet and not talk about it: https://gcc.gnu.org/legacy-ml/gcc/2001-02/msg00895.html https://gcc.gnu.org/legacy-ml/gcc/2001-02/msg00895.html
      - CyberRabbi 6y ago
        I don’t see what is so bad faith about his post. His response and his intent are not malicious. On principle he is against proprietary software and he wants to make sure it remains as difficult as possible for proprietary software to exploit GCC for their own means. He doesn’t seem to be rejecting the code or suggests its deletion because he wants to do unnecessary harm to the author here. His intentions seem rational and consistent with his past remarks. As an aside and for comparison, proprietary software companies are just as defensive or more with other companies using their IP without compensation. The compensation RMS wants here is code reciprocity.
        spiralx 6y ago
        His intent might not be malicious, but accusing the other person of doing something to undermine free software that's so bad it should be "cancelled" is not very nice. He certainly isn't advocating for code being free here, it's literally the opposite.
        arp242 6y ago
        rms' entire problem is that he fails to understand there is more than one way to advocate user freedom. My viewpoint, which is also the viewpoint of pretty much everyone in that thread, is that the best way is to write the best possible software possible. Same goals, just different methods. But there isn't even a discussion about that here, it's just a hard shutdown of these type of arguments from rms. Is it any surprise that almost any project rms has any technical influence on has been in decline?
  - CyberRabbi 6y ago
    As quick as the search may be, it isn’t good to search out sources for someone else’s claim because what you find may not be what they are citing and you may assume they are making a point they are not.
not2b 6y ago
Not exactly correct. Rather, RMS resisted attempts to make the compiler more modular based on such arguments, but the initial design wasn't deliberately non-modular. The steering committee had to spend years persuading him not to block improvements that increased modularity, like plugins, and eventually found ways to persuade him.

GCC was originally designed to not be modular so that it wouldn't be possible to integrate it with closed-source code.

CyberRabbi 6y ago

Source?

macintux 6y ago
Quick searches turned this up: https://lwn.net/Articles/582697/ https://lwn.net/Articles/582697/
- arp242 6y ago
  Also something like: https://lists.gnu.org/archive/html/emacs-devel/2015-01/msg00015.html https://lists.gnu.org/archive/html/emacs-devel/2015-01/msg00... rms deliberately making GCC (and by extension, Emacs) less useful to "prevent misuse". The entire thread is ... ugh. The funniest part is that it doesn't even matter, because now everyone and their mother has switched to llvm/clang for these kind of things, just as people in that thread predicted.
  - mhh__ 6y ago
    > everyone and their mother has switched to llvm/clang Chris Lattner even offered it to them, but RMS didn't get the email because of the unique way he browses the net
    - spekcular 6y ago
      Do you have a source handy for the missed email thing? I'd love to read a firsthand account.
      - mhh__ 6y ago
        I mentioned it once, lost the citation, found it and lost it again. I find I can find I'll link it tomorrow.
        spiralx 6y ago
        https://lists.gnu.org/archive/html/emacs-devel/2015-02/msg00594.html https://lists.gnu.org/archive/html/emacs-devel/2015-02/msg00...
      - moonchild 6y ago
        I don't have a citation for the anecdote, but possibly this was the email? https://gcc.gnu.org/legacy-ml/gcc/2005-11/msg00888.html https://gcc.gnu.org/legacy-ml/gcc/2005-11/msg00888.html
        spiralx 6y ago
        https://lists.gnu.org/archive/html/emacs-devel/2015-02/msg00594.html https://lists.gnu.org/archive/html/emacs-devel/2015-02/msg00...
      - spiralx 6y ago
        I found it here in an earlier HN discussion: https://lists.gnu.org/archive/html/emacs-devel/2015-02/msg00594.html https://lists.gnu.org/archive/html/emacs-devel/2015-02/msg00...
  - spiralx 6y ago
    I know RMS is often abrasive, but that it is a whole new level of bad faith from him on display there. Plenty of "just asking questions" from one of the other participants as well. Here's another example of RMS on the subject where he told someone he wouldn't accept their code and that they should take it off of the internet and not talk about it: https://gcc.gnu.org/legacy-ml/gcc/2001-02/msg00895.html https://gcc.gnu.org/legacy-ml/gcc/2001-02/msg00895.html
    - CyberRabbi 6y ago
      I don’t see what is so bad faith about his post. His response and his intent are not malicious. On principle he is against proprietary software and he wants to make sure it remains as difficult as possible for proprietary software to exploit GCC for their own means. He doesn’t seem to be rejecting the code or suggests its deletion because he wants to do unnecessary harm to the author here. His intentions seem rational and consistent with his past remarks. As an aside and for comparison, proprietary software companies are just as defensive or more with other companies using their IP without compensation. The compensation RMS wants here is code reciprocity.
      - spiralx 6y ago
        His intent might not be malicious, but accusing the other person of doing something to undermine free software that's so bad it should be "cancelled" is not very nice. He certainly isn't advocating for code being free here, it's literally the opposite.
        arp242 6y ago
        rms' entire problem is that he fails to understand there is more than one way to advocate user freedom. My viewpoint, which is also the viewpoint of pretty much everyone in that thread, is that the best way is to write the best possible software possible. Same goals, just different methods. But there isn't even a discussion about that here, it's just a hard shutdown of these type of arguments from rms. Is it any surprise that almost any project rms has any technical influence on has been in decline?
- CyberRabbi 6y ago
  As quick as the search may be, it isn’t good to search out sources for someone else’s claim because what you find may not be what they are citing and you may assume they are making a point they are not.

not2b 6y ago

Not exactly correct. Rather, RMS resisted attempts to make the compiler more modular based on such arguments, but the initial design wasn't deliberately non-modular. The steering committee had to spend years persuading him not to block improvements that increased modularity, like plugins, and eventually found ways to persuade him.