On Mon, Oct 20, 2014, Óscar Fuentes wrote: > Adrien Nader <[email protected]> writes: > > > As a project, mingw-w64 wishes to have clang and llvm supported. There > > have been recent reports of troubles with it however but we'd prefer it > > to work. > > Nice. Preferences are a good starting point for achieving things ;-) > > > On the other side, the llvm project needs to do a bit of work too or at > > least to state they want to support mingw-w64 and are ready to at least > > try to not break its support. > > You think about the LLVM project as if it had preferences and stated > goals. Reality is that the community drives what the project does. If > there is no one pushing hard for MinGW-w64 support and contributing hard > work to it, things will stay the same (or change for the worse). > > So far the drive behind MinGW-w64 (or MinGW* at all) is weak. Ruben and > a few others contribute patches, but without a champion within the core > developers community, they are fighting an uphill battle. There is a > group of prominent contributors that would be somewhat pleased if MinGW* > support were officially suppressed. Those developers think that only VS > is relevant on Windows and Clang is not to be taken seriously on that > platform until VS compatibility is achieved.
Definitely. And just for others reading, a few very-publicized links along with quotes from the top of each page. http://blog.llvm.org/2013/09/a-path-forward-for-llvm-toolchain-on.html "Over the past several months, contributors from Google and elsewhere in the community have begun actively working on bringing an LLVM toolchain to Windows in a way that would support and enhance a fully native development experience. This toolchain works with Visual Studio and the Windows development process that Windows developers are already using, making LLVM tools and technologies available to them." http://blog.llvm.org/2014/07/clangllvm-on-windows-update.html "It's time for an update on Clang's support for building native Windows programs, compatible with Visual C++!" http://llvm.org/docs/GettingStartedVS.html "Welcome to LLVM on Windows! This document only covers LLVM on Windows using Visual Studio, not mingw or cygwin." (that one is the top answer for a query for "llvm windows") The current development with the most manpower, including large corporate backings, is definitely for a direct integration with Visual Studio, with executables that replace the ones from VS and mimic them (like a "cl.exe" for compilation). And... reusing proprietary bits from Visual Studio. It's open-source and everything and you can put the effort you wish in it and do a ton of work. But meanwhile there are several people paid to work on it with the sole purpose of integrating with Visual Studio, reusing everything that cannot be replaced directly by something from LLVM, no matter if it's proprietary or not. Good luck is the best thing I can give currently. This is a corporate-driven project and no matter what you do, you risk seeing your work removed because it doesn't please someone from a company. It could happen with any project, corporate-driven or not, but it's more likely when a company is behind it and given the current state, I'm not optimistic. (hello Intel by the way) There could be a fork but no-one would follow, just like only Google has been able to "fork" WebKit. All the other players in the webkit/blink space are completely dwarfed by Google or Apple and it's impossible to do anything not wanted by these. > >> Don't get me wrong, I'm not upset for this question being posted here. > >> It's just that by asking it on clang-users or cmake-users the OP would > >> get more chances of getting help. > > > > I think the discussion is relevant but at least some troubles seem to > > come from clang so the issue is worth raising there too imho. > > Furthermore, it is a good thing to send the message that some people > cares about MinGW* support on LLVM/Clang. Just to be sure: I care about that support but I don't believe upstream wants it. My point is only that if there is no strong technical reason to support it, it should be done. Except that if LLVM upstream actively wants to not be supported, it's going to be difficult to do so. I'm afraid we might well be in this situation. NB: If you are from the LLVM project (I mean this for anyone reading this message, not for someone in particular), please get an _official_ statement that mingw-w64 is something the project wants support for. Only a statement, it's not much work and it'll help going forward a lot. -- Adrien Nader ------------------------------------------------------------------------------ Comprehensive Server Monitoring with Site24x7. Monitor 10 servers for $9/Month. Get alerted through email, SMS, voice calls or mobile push notifications. Take corrective actions from your mobile device. http://p.sf.net/sfu/Zoho _______________________________________________ Mingw-w64-public mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/mingw-w64-public
