I have three iOS projects:
- The first is 35k Swift LOC, Swift 2.1 (or 2.0?) Compiled using Xcode 7.2.3
- The second is 15k Swift LOC, Swift 2.3, compiled using Xcode 8.2.1
- Third - 15k Swift LOC, Swift 3.0, compiled using Xcode 8.2.1
Compilation time for all of them is terrible (10 + minutes for a clean build), and development on each of them slows down the scan.
I already tried:
- Using
Whole Module Optimization
- Change
C Dialect
for default compiler - Analysis of the build time for the project and optimization of bottlenecks (several methods that required from 6 to 10 compilation with array concatenation, nil-coalescing operators, etc.) that shaved like 15 seconds are an absolutely imperceptible difference
HEADERMAP_USES_VFS = YES
- A few others that I don’t remember now
I am working on a Mac mini Late 2014 with an SSD, 8 GB of RAM and a 2.6ghz i5 if that matters.
Does anyone know what can cause this, what to do with this or even workarounds? From what I read in SO and elsewhere, this seems like an unresolved compiler issue.
EDIT: Yes, I use Cocoapods for dependencies. I will try to move to Carthage and see what happens.
source
share