Proof of concept mod compiler (Now with GUI!)

[tuxie]

https://github.com/magnusjjj/mcbuildatron

 

Update: Made a GUI

 

 

I made a tool that downloads the source code for open source mods (there is a list of recipies for a bunch of them included) from github, adds gradlew to the directory if required, and builds it.

Its.. not that polished right now. Built it mostly to see how much of the modpack that I am assembling could actually be built automatically .

[tuxie]

Tiny bumb for huge change. Hope you will forgive me.

[loordgek]

why?

[tuxie]

Why what?

[loordgek]

why doing this

[tuxie]

To make it simpler to set up a dev environment for open source mods, so you can help out when you notice something wrong without spending half of your energy setting up a dev environment and all your settings just right. You download this, run the run_me_first.bat, then the gui.bat, and bam, a list of mods that are actually open source. You press the mod you want, you hit the red button, it clones the repo, sets up gradle, and compiles it.

 

 

<my lifes story>

Oh! Well, I was making a modpack, and after going through every single mod from another major modpack, finding the license requirements for each and every single one and going ever so slightly insane...

 

I wanted to see how many mods I could actually compile. I run my own game servers, and every once in a while a mod absolutely shits itself, and throws an exception when the server starts. I wanted to make a tool that would let me, since I am a person with.. not a lot of energy.. with the click of a button set up a dev environment for that mod so I can help out fixing the bug instead of just setting those 'delete all entities that cause exceptions' settings.

 

I then figured it might be usefull for someone else. So I tried to make it into a simple command line dealio. Type in 'builder name_of_mod', and it downloads the sources, sets up gradle, and compiles it for you.

 

Then I figured that I wanted to fuck around with making a GUI, and, *patting own shoulder*, it turned out decent-ish.

</my life story>

[MoeBoy76]

This seems like a lot more work than simply forking a project and setting up a dev environment yourself, while you may have a list of mods that can be seen on one screen without the use of a quick browser/google/github search, it will never be complete, to plus you just download the master, what if i wanted to say make a PR fix for BoP 1.8.9? Your tool is going to get me the default branch which is currently sitting on 1.11.2

[loordgek]

Oh! Well, I was making a modpack, and after going through every single mod from another major modpack, finding the license requirements for each and every single one and going ever so slightly insane...

for that there is curseforge

[tuxie]

Yup. Hence the 'proof of concept' in the title . Its incredibly, incredibly, incredibly easy to add new recipes, and to share said recipes.

 

For the rest? Yeah, totally with you on that. It lacks a toooooooon of features. Just a ton of them. I'll probably add them as I need them, but right now all it is is a request for comments and for others to play around with. Its not all that usefull right now. I plan to add a bunch of shortcuts, and try to be somewhat newbiefriendly. The goal, right now, is just to be able to build as many mods as possible as simply as possible, and to fuck around with it a bit .

 

Oh! Well, I was making a modpack, and after going through every single mod from another major modpack, finding the license requirements for each and every single one and going ever so slightly insane...

for that there is curseforge

 

There is not . Curseforge has about a 50% hit rate when it comes to mods and their licenses. Source: I spent hours and hours compiling a list of 261 mods, their licenses, where to find the source... :'). The google spreadsheet of doom and despair.

This is not a solution to that, its just the backstory behind my bout of insanity~

[kreezxil]

I really like this idea and sincerely hope you keep working on it.

 

For all the naysayers, disbelievers, and what have yous, sheesh, even the car you take for granted started with a wheel and an idea. Don't shit on his idea just because all you see is a door handle.

 

Most people who would even bother to post a proof a concept would simply post a bunch of words describing a grandiose idea, this guy went and produced a working program.

 

Good job!

[tuxie]

On 2017-02-01 at 0:10 PM, kreezxil said:

I really like this idea and sincerely hope you keep working on it.

 

For all the naysayers, disbelievers, and what have yous, sheesh, even the car you take for granted started with a wheel and an idea. Don't shit on his idea just because all you see is a door handle.

 

Most people who would even bother to post a proof a concept would simply post a bunch of words describing a grandiose idea, this guy went and produced a working program.

 

Good job!

Thanks <3. Your comment made me have a bit more energy.

 

Just wanted to post some progress for the latest version due in a couple of days :).

132 recipes at the moment, 21 failing. Better list, more things in the edit window...

Setting a common gradle cache directory, so if you launch multiple builds it doesn't hammer the forge website so heavily.

 

Most recipes that won't compile is due to missing libraries. Some mods want you to drop a jar or two into a specific directory. Not sure how I want to implement that, given that most of those are manual (say, thaumcraft's API and the crazy licensing there). Might just be a popup explaining what to do :).

 

That, and most other errors are web servers that are down. Seems like some of the forge api servers are having a hard time lately? Timeouts are crazy often, but if you try and try again everything works just fine.