Skip to content
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

extensions rebuild on every build or run #49

Closed
Habbie opened this issue Oct 18, 2014 · 7 comments
Closed

extensions rebuild on every build or run #49

Habbie opened this issue Oct 18, 2014 · 7 comments

Comments

@Habbie
Copy link
Contributor

Habbie commented Oct 18, 2014

which makes cmd-r in xcode take a long time for no reason. Should figure out how to make this better - have the Makefiles actually know about all the invidual targets?

@cmsj
Copy link
Member

cmsj commented Dec 30, 2014

I'm very seriously considering changing entirely how extensions are built, to move them into Xcode. Not because the build speed is slow, but because I've added Crashlytics and all of our crash reports will be useless because they're happening inside an unsymbolicated "internal.so".
I'm not an expert at these things, but it looks like the easiest option would be to have an xcode project per extension, which spits out something like libFOO.dylib and foo.lua, then modify the package paths to look for .dylib instead of .so.

Before I sink a bunch of time into figuring out how to go about producing a working app that requires no changes in init.lua behaviour, what do folk like @asmagill @Habbie @josephholsten and @jhgg think about this?

@josephholsten
Copy link
Contributor

+1 to better metrics

@asmagill
Copy link
Member

While I'm all for better metrics and better debug options, how much does this change the development path of a specific module? When I'm tweaking and refining something and rebuilding 20 times in an hour to get something ready, the overhead of Xcode, even xcodebuild, would be annoying... but if it could be more automated, like a wrapper that I can just drop my files into (maybe a few more steps, but simple) once I'm ready to make it public, then I see no reason not to do so, especially for modules that make it into core.

On Dec 30, 2014, at 7:44 PM, Joseph Anthony Pasquale Holsten notifications@github.com wrote:

+1 to better metrics


Reply to this email directly or view it on GitHub #49 (comment).

@cmsj
Copy link
Member

cmsj commented Dec 31, 2014

@asmagil one option would be to use gyp to generate the Xcode project files. We could also continue to support building via a Makefile I would think.
The answer really is, I don't know because I haven't yet done any work to see how this would have to be done. My hope is that it's just an administrative change and a module wouldn't need any code changes to switch between .so and .dylib.

@cmsj
Copy link
Member

cmsj commented Jan 9, 2015

Managed to figure out a way to get debug symbols into Crashlytics without having to change how extensions are built, significantly. As such, I'm not going to continue trying to get the extensions built as dylib bundles.
I wouldn't object to it if someone else wants to do it, but I no longer have a compelling reason to :)

@mengelbrecht
Copy link
Contributor

Just for reference: I did a prototype transition to gyp as build system here:
https://github.com/mgee/hammerspoon/tree/gyp
In this branch gyp is used to generate Xcode projects for Hammerspoon and all extensions.
Note that this branch is not up-to-date with Hammerspoon/master.

@cmsj
Copy link
Member

cmsj commented Aug 7, 2015

This was closed by #468

@cmsj cmsj closed this as completed Aug 7, 2015
latenitefilms added a commit to latenitefilms/hammerspoon that referenced this issue Jan 26, 2022
…aram-return-notes-newlines

Feature/issue 48 param return notes newlines
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

No branches or pull requests

5 participants