Less ambitious, less rubbish 1.12 release

[LunNova]

Tasks:

• [x] Split javassist patching implementation into separate project
  • [x] Split non-minecraft code into JavaPatcher
  • [x] Split minecraft-specific code into ModPatcher
• [x] Mixin implementation in nallar/JavaPatcher. nallar/Javapatcher#2
  • [x] Mixin implementation in Mixin nallar/Mixin#1
  • [x] Mixin transformations in ModPatcherGradle
• [x] Improve performance in ModPatcher by avoiding re-transforming nallar/ModPatcher#10
• [x] Implement old profiling features in separate mod. nallar/TickProfiler#22
• [ ] Start from scratch, bring in some important features. Avoid feature creep
  • [x] Implement variant system (core/threaded for now, should be extensible)
  • [ ] Implement per-world threading
  • [ ] Implement deadlock detector
  • [ ] async/threaded chunkloading improvements
  • [ ] Fix the bloody stupid chunkloading when you world.get with coords an unloaded chunk
  • [ ] Chunkloading cache which actually works
  • [ ] Entity collision performance fixes
  • [x] Fix FMLCommonHandler.getEffectiveSide -> hardcode as always running on server
  • [ ] Avoid slow operations on TE/Entity lists. See TickManager.batchRemove(Tile)Entities
  • [ ] Fix poor World.getPersistentChunks performance

[LunNova]

Quick update:

• Now aiming for 1.12. I'll try to backport a working version to 1.10.2 later if by the time it works 1.10.2 is still popular.
• Working on changing JavaTransformer and Mixin to allow patching small parts of methods instead of just replacing the entire things. Will allow the old XML based JavaPatcher patches to be dropped completely.

[dotModded]

I would wait nallar, 1.13 is a refactor and everything is changing so your mod would be completely broken by then.

[Slind14]

1.7.10 is still more popular than 1.10 :D Gonna be interesting to see if 1.10 is able to establish enough to be around for the long run like 1.7.10

[cpypcy]

Well 1.7.10 is mod Mecca, making TT for any other version is waste of time sort of, since mods don't seem to like 1.10 or 1.12 much. Once we hit final and last java version, i bet there will be mod insanity and every damn mod that is nod dead will move to that version but i don't see which version will be final java one. Still, better to have some TT than no TT.

[magneticflux-]

@cpypcy @Slind14 I'd actually disagree with that assessment at this point in time. I assembled and formatted some mod data from CurseForge to better understand the version situation.

Version	Pages of mods	Page Δ (17d)	Days since last release	Link to search
Total	417	+4	N/A	Link
.
1.7.X	211	+1	1148	Link
1.8.X	104	+0	617	Link
1.9.X	77	+0	464	Link
1.10.X	141	+1	420	Link
1.11.X	103	+2	239	Link
1.12.X	61	+7	14	Link

I believe that there is a large proportion of 1.7.10 mods that have stopped being developed since their release. To back up my statement, I found that out of all mods with 1.7.X versions on CurseForge, only 65 pages of them have been modified at all since the last 1.11.X release in December of 2016 (Link to approximate location in CurseForge list). This tells me that many of those mods are likely to never be updated past 1.7.X and probably have been superseded by a 1.12.X mod or abandoned.

I wrote this mainly because I was curious about the statistics and I'd heard other modders debate it before with only anecdotal information. Obviously, this is only current as of 8/17/2017, but I'd expect the balance to shift even further in 1.12.X's favor in the future with more Forge updates. Hopefully these statistics will prove to be useful to someone in the future.

---

9/3/2017 Edit: I have added deltas to each page count that are current as of 9/3/2017. The deltas do not add up because they are rounded to the nearest page and because mods older than 1.7.X may have changed.

[Shybella]

Do not focus on older versions unless you have some magically way to implement it easier for older versions.

On Thu, Aug 17, 2017 at 2:46 PM, Mitchell Skaggs notifications@github.com wrote:

@cpypcy https://github.com/cpypcy @Slind14 https://github.com/slind14 I'd actually disagree with that assessment at this point in time. I assembled and formatted some mod data from CurseForge to better understand the version situation. Version Pages of mods Days since last release Link to search Total 417 N/A Link https://minecraft.curseforge.com/mc-mods?filter-game-version= 1.7.X 211 1148 Link https://minecraft.curseforge.com/mc-mods?filter-game-version=1738749986%3A5 1.8.X 104 617 Link https://minecraft.curseforge.com/mc-mods?filter-game-version=1738749986%3A4 1.9.X 77 464 Link https://minecraft.curseforge.com/mc-mods?filter-game-version=1738749986%3A552 1.10.X 141 420 Link https://minecraft.curseforge.com/mc-mods?filter-game-version=1738749986%3A572 1.11.X 103 239 Link https://minecraft.curseforge.com/mc-mods?filter-game-version=1738749986%3A599 1.12.X 61 14 Link https://minecraft.curseforge.com/mc-mods?filter-game-version=1738749986%3A628

I believe that there is a large proportion of 1.7.10 mods that have stopped being developed since their release. To back up my statement, I found that out of all mods with 1.7.X versions on CurseForge, only 65 pages of them have been modified at all since the last 1.11.X release in December of 2016 (Link to approximate location in CurseForge list https://minecraft.curseforge.com/mc-mods?filter-game-version=1738749986:5&filter-sort=updated&page=65). This tells me that many of those mods are likely to never be updated past 1.7.X and probably have been superseded by a 1.12.X mod or abandoned.

I wrote this mainly because I was curious about the statistics and I'd heard other modders debate it before with only anecdotal information. Obviously, this is only current as of 8/17/2017, but I'd expect the balance to shift even further in 1.12.X's favor in the future with more Forge updates. Hopefully these statistics will prove to be useful to someone in the future.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/MinimallyCorrect/TickThreading/issues/1248#issuecomment-323137339, or mute the thread https://github.com/notifications/unsubscribe-auth/AEOzKOigW0UWvaesEHike7xuTeHj34-nks5sZHVwgaJpZM4HAeXY .

[Shybella]

I run a 1.5.2 server and as much as I would like to see a TT for 1.5.2 in working order I'd rather see development moving forward with new versions. If TT can recreate some of the magic it did for 1.4.2 and 1.5.2 in 1.10 and newer. That would be heaven.

On Thu, Aug 17, 2017 at 3:10 PM, Nicholas Tow nxtr1114@gmail.com wrote:

Do not focus on older versions unless you have some magically way to implement it easier for older versions.

On Thu, Aug 17, 2017 at 2:46 PM, Mitchell Skaggs <notifications@github.com

wrote:

@cpypcy https://github.com/cpypcy @Slind14 https://github.com/slind14 I'd actually disagree with that assessment at this point in time. I assembled and formatted some mod data from CurseForge to better understand the version situation. Version Pages of mods Days since last release Link to search Total 417 N/A Link https://minecraft.curseforge.com/mc-mods?filter-game-version= 1.7.X 211 1148 Link https://minecraft.curseforge.com/mc-mods?filter-game-version=1738749986%3A5 1.8.X 104 617 Link https://minecraft.curseforge.com/mc-mods?filter-game-version=1738749986%3A4 1.9.X 77 464 Link https://minecraft.curseforge.com/mc-mods?filter-game-version=1738749986%3A552 1.10.X 141 420 Link https://minecraft.curseforge.com/mc-mods?filter-game-version=1738749986%3A572 1.11.X 103 239 Link https://minecraft.curseforge.com/mc-mods?filter-game-version=1738749986%3A599 1.12.X 61 14 Link https://minecraft.curseforge.com/mc-mods?filter-game-version=1738749986%3A628

I believe that there is a large proportion of 1.7.10 mods that have stopped being developed since their release. To back up my statement, I found that out of all mods with 1.7.X versions on CurseForge, only 65 pages of them have been modified at all since the last 1.11.X release in December of 2016 (Link to approximate location in CurseForge list https://minecraft.curseforge.com/mc-mods?filter-game-version=1738749986:5&filter-sort=updated&page=65). This tells me that many of those mods are likely to never be updated past 1.7.X and probably have been superseded by a 1.12.X mod or abandoned.

I wrote this mainly because I was curious about the statistics and I'd heard other modders debate it before with only anecdotal information. Obviously, this is only current as of 8/17/2017, but I'd expect the balance to shift even further in 1.12.X's favor in the future with more Forge updates. Hopefully these statistics will prove to be useful to someone in the future.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/MinimallyCorrect/TickThreading/issues/1248#issuecomment-323137339, or mute the thread https://github.com/notifications/unsubscribe-auth/AEOzKOigW0UWvaesEHike7xuTeHj34-nks5sZHVwgaJpZM4HAeXY .

[Slind14]

Well the latest versions really need it. We just reset an infinity normal server 2 weeks ago. It can still handle +50 players at 40ms, while Age of Engineering/Beyond.. could not even handle 20 players at <80ms during the first weekend. 1.7.10 still performs at least twice as good as 1.10 with less mods.

[LunNova]

I've been working on improving Mixin to support injection of code inside existing methods in a flexible way, including the insertion of return or continue statements, or the modification of local variables.

This support is partially implemented, and the following example works: https://github.com/MinimallyCorrect/Mixin/blob/c55b5e464d0cdf6ffdc25de6f077aae0feccda14/src/test/java/org/minimallycorrect/mixin/internal/mixinsource/MixinSource.java

The aim is to significantly reduce the work required implement new patches and to update old patches.

[Bucky420]

hummm and where are we on this 1.12...?

[LunNova]

We are less ambitious but still rubbish

[Shybella]

Sponge just marked 1.12.2 as a LTS version.