-
Notifications
You must be signed in to change notification settings - Fork 141
rmlint: fix the build on older Darwin systems #573
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Conversation
instead of MAC_OS_X_VERSION_10_7 use 1070 |
e9133b8
to
7ee224e
Compare
@sahib Could you please take a look at this PR? |
@vassilit I can't check as I have only macOS |
@RayOei There is a lot of older hardware around, including powerpc, which is not supported in 10.7+. Even 10.4 has some active users. MacPorts supported it until recently and still supports 10.5+, NetBSD’s |
|
@barracuda156 привет/你好。︁we have in
Do you know if old versions of MacOS X do define |
The Apple doc is a bit vague.
But it also explains why you should use a different approach. |
@RayOei do the actual state of this branch (with compat in .h) compiles normally on your mac ? |
Currently compiling fails but not limited to this branch only. Not sure what has changed but |
@vassilit Looks like OS X does, while 68k systems do not, so combo is supposed to filter out macOS < 10.0. Apple doc reference: https://developer.apple.com/library/archive/documentation/Porting/Conceptual/PortingUnix/compiling/compiling.html |
Which ones exactly ? Because some of them are normal. |
@barracuda156 ok thank you. Does my modification to put those headers in .h works for you ? I have a iBook G4 (ppc 32) but it's running Linux, not macOS X anymore. I also have Intels macbook pros but they are running Linux as well. So I cannot test any of those branches. |
Agree there. Been trying to get my old mac 2015 Monterey (macOS 12) to a state it will install and compile and a lot has been deprecated. Including |
All of them 😁 As I said: at this point I suspect my mac itself because of the update to It did show me this particular comment in
May need to have a look there, and mark it properly with a |
@RayOei Systems from 10.5 up can be in a very decent shape if used correctly (obviously, a single-core G4 will be slow, nothing to do about it). I have a lot of stuff building and working on 10.6 on powerpc. It is decently fast, aside of some ridiculously huge test-suites. |
The «Not available before 2.64» are warnings that could be ignored. They are still used properly. See below.
It's addressed in #707. Basically, 64-bit compilation already requires Glib 2.64. The 32-bit code does not, so we wait a bit before imposing this requirement to 32-bit users (as PPC are, actually). |
I don't even know where to get macOS from, if I where to install it again on my iBook G4. The laptop is running well on Linux however, but it is seldom used. |
Ok, so I am able to compile and run on macOS 12 (Monterey) with current master.
I am aware of that but I didn't have those warnings before and scons didn't complain and bail before (that is: with
👍 I now see: cool 😁 |
I won’t spam here with detailed info LOL, but my GH page has everything needed, more or less.
I can try tomorrow and let you know. |
Have you been able to test it? |
@vassilit Could you please specify which commit I should test? e5e0396 this one? |
HEAD of barracuda156:rmlint-darwin, i.e. e5e0396 |
Unfortunately, that fails:
Here I just used e5e0396 without adding patches from my side. |
I've rebased to master (I see you have GLib warnings). |
Could you try again ? After we merge that one, I release right away 2.10.3.
Thank you for testing ! |
Almost, but linking fails due to a duplicate symbol (I think Linux is lax on that, but macOS linker is stricter):
|
Linux also have Could you try again ? |
From 4a6b81c everything works now, thank you!
|
This PR fixes
rmlint
build for 10.5.8 and 10.6.It contains two part, which come from:
P. S. There is also another problem related to how SCons build system works – trying to build a FAT binary on Leopard fails, for example, because arch flags get ignored. I solve it like this (for Macports): https://github.com/macports/macports-ports/pull/15003/files#diff-d43b1faa4b64984e308f2d41fddfd09fb559868da27208ff41417d636a4c7d20
I don’t know how to make a similar fix in the source code that gonna work without Macports assistance.