Closed curiousdannii closed 7 years ago
Thanks, I will try to look at this later in the week.
I have started looking at this. It's good, but I want to adjust the way the formats module is laid out. So it's not merged yet. Assume main.js, apphooks.js, and formats.js are in flux until tomorrow.
Of course. I never expected it to be more than a starting point :)
I realised after I committed this that Git also needs blorb_gamechunk_type
to be set. In my WIP I stopped using a common emglken_options
function and just made specific ones for Git and Hugo.
And thinking for the future, would you be willing to give me commit access for me to update the VMs? For anything other than a routine update of VM-internal code I would still make pull requests.
I've merged a cleaned-up version of this.
It all works from "npm start", but when I build an app bundle, the Hugo interpreter stops working.
Uncaught TypeError: Cannot read property 'buffer' of undefined at ReadFileContext.fs.readFile [as callback] (/Users/zarf/src/lectrote/dist/Lectrote-darwin-x64/Lectrote.app/Contents/Resources/app/emglken/hugo.js:53:41) at FSReqWrap.readFileAfterOpen [as oncomplete] (fs.js:336:13)
Can you take a look at this?
It's possible that I'm not copying the correct files into the distributed emglken directory. Currently:
emglken_dispload.min.js
git-core.js.mem
git.js
hugo-core.js.mem
hugo.js
versions.json
There's also the line
game_options.memdir = 'hugoem';
...which disappeared somewhere; I'm not sure if you dropped it on purpose.
Your cleaned up version looks good!
The .bin files are needed as well.
The memdir
option has changed into the dirname
option, to match the __dirname
Node variable, and because it's used for more than just memory now.
And when you're satisfied it's stable, I'd appreciate if you could do a release soon so that I could get more feedback on whether the Hugo undo buffer is large enough.
Thanks. I needed one change in the emglken code -- see https://github.com/curiousdannii/emglken/pull/2 . With that, Hugo works.
Release is up.
I've centralised all the format and engine details in formats.js. This reduces some redundancies, and also makes functions like
game_file_discriminate
andload_named_game
much more systematic.I've also added Git for testing purposes. It chooses the first listed engine, so Quixe unless you manually switch the order. It would be very easy to make
launch_game
choose another engine though.I don't think it will take much more to move to a single play.html (for all except inkjs maybe). I'd like to do that for the VMs I'm providing at least.