3 ms·

Proguard can also apply optimizations while it obfuscates. I think a good JVM will eventually do most of them itself, but it can help code size and warm-up. I'm

Proguard can also apply optimizations while it obfuscates. I think a good JVM will eventually do most of them itself, but it can help code size and warm-up. I'm guessing as JVMs get better and everyone is less sensitive to file sizes, this matters less and less.

mort96 1y ago
And there's no way to do only the optimisation part? Surely you could optimise without messing up class and method names..?
- the_hoser 1y ago
  One of the biggest optimizations it offers is shrinking the size of the classes by obfuscating the names. If you're obfuscating the names anyway, there's no reason that the names have to be the same length. "hn$z" is a heck of a lot smaller than "tld.organization.product.domain.concern.ClassName"
  - voxic11 1y ago
    Yeah in some ways the obfuscation and mappings are similar to minification and sourcemaps in javascript.
    - mort96 1y ago
      And minification in JavaScript only reduces the amount of bytes that has to be sent over the wire, it doesn't improve runtime performance.
      - voxic11 1y ago
        According to the v8 devs it also can increase parsing performance > Our scanner can only do so much however. As a developer you can further improve parsing performance by increasing the information density of your programs. The easiest way to do so is by minifying your source code, stripping out unnecessary whitespace, and to avoid non-ASCII identifiers where possible. https://v8.dev/blog/scanner https://v8.dev/blog/scanner
        mort96 1y ago
        Sure, but that's also just in the category improving loading a bit. It doesn't have anything to do with runtime performance.
  - mort96 1y ago
    So we're not talking about runtime performance, but some minor improvement in loading times? I assume that once the JVM has read the bytecode, it has its own efficient in-memory structures to track references to classes rather than using a hash map with fully qualified names as keys
    - saagarjha 1y ago
      The names need to be stored somewhere because they are exposed to the program that way
      - mort96 1y ago
        They have to be stored somewhere, but they don't have to be what the JVM uses when it e.g performs a function call at runtime. Just having the names in memory doesn't slow down program execution.
        saagarjha 1y ago
        At runtime this is going to be a branch instruction yes
    - swiftcoder 1y ago
      Proguard was heavily influenced by the needs of early Android devices, where memory was at a real premium. Reducing the size of static tables of strings is a worthwhile optimisation in that environment
      - mort96 1y ago
        Okay but we're talking about Minecraft on desktops and laptops, where the relevant optimizations would be runtime performance optimizations, no?
        swiftcoder 1y ago
        Probably, but proguard tends to bundle the whole lot together

mort96 1y ago

And there's no way to do only the optimisation part? Surely you could optimise without messing up class and method names..?

the_hoser 1y ago
One of the biggest optimizations it offers is shrinking the size of the classes by obfuscating the names. If you're obfuscating the names anyway, there's no reason that the names have to be the same length. "hn$z" is a heck of a lot smaller than "tld.organization.product.domain.concern.ClassName"
- voxic11 1y ago
  Yeah in some ways the obfuscation and mappings are similar to minification and sourcemaps in javascript.
  - mort96 1y ago
    And minification in JavaScript only reduces the amount of bytes that has to be sent over the wire, it doesn't improve runtime performance.
    - voxic11 1y ago
      According to the v8 devs it also can increase parsing performance > Our scanner can only do so much however. As a developer you can further improve parsing performance by increasing the information density of your programs. The easiest way to do so is by minifying your source code, stripping out unnecessary whitespace, and to avoid non-ASCII identifiers where possible. https://v8.dev/blog/scanner https://v8.dev/blog/scanner
      - mort96 1y ago
        Sure, but that's also just in the category improving loading a bit. It doesn't have anything to do with runtime performance.
- mort96 1y ago
  So we're not talking about runtime performance, but some minor improvement in loading times? I assume that once the JVM has read the bytecode, it has its own efficient in-memory structures to track references to classes rather than using a hash map with fully qualified names as keys
  - saagarjha 1y ago
    The names need to be stored somewhere because they are exposed to the program that way
    - mort96 1y ago
      They have to be stored somewhere, but they don't have to be what the JVM uses when it e.g performs a function call at runtime. Just having the names in memory doesn't slow down program execution.
      - saagarjha 1y ago
        At runtime this is going to be a branch instruction yes
  - swiftcoder 1y ago
    Proguard was heavily influenced by the needs of early Android devices, where memory was at a real premium. Reducing the size of static tables of strings is a worthwhile optimisation in that environment
    - mort96 1y ago
      Okay but we're talking about Minecraft on desktops and laptops, where the relevant optimizations would be runtime performance optimizations, no?
      - swiftcoder 1y ago
        Probably, but proguard tends to bundle the whole lot together