This actually makes sense... kind of. It's mostly some regexp magic.
This actually makes sense... kind of. It's mostly some regexp magic.
Perl is like brainfuck, but taken seriously.
I sense a lot of Lua fanboyism in this thread. The man likes Python, who's to stop him from using it? I personally don't like Lua, but I wouldn't waste posts criticizing it just because someone decided to implement it in their program.
Show me some code to expose a C++ class to Lua, then tell me it's not complex or tedious.
ramble ramble ramble ramble ramble ramble ramble ramble ramble ramble ramble ramble ramble whitespace ramble ramble ramble ramble ramble ramble ramble ramble ramble ramble ramble ramble ramble ramble ramble versioning hell ramble ramble ramble ramble ramble ramble ramble ramble ramble ramble ramble ramble ramble ramble ramble ramble ramble ramble ramble ramble ramble ramble ramble weird syntax ramble ramble ramble ramble ramble ramble ramble ramble ramble ramble ramble ramble ramble ramble ramble ramble ramble ramble ramble ramble ramble ramble ramble ramble ramble ramble ramble ramble ramble ramble cpython is slow ramble
This is true for all of Facepunch.
it seems like his first preference was Ruby but he couldn't get it to link.
Edited:
Not that I blame him - MRI's API might be better than most other languages' APIs, but the whole thing is still a pain in the ass to embed
=~ wat
Still not as bad as Malbolge.
My experience with python has been a bunch of syntax "gotcha"s whenever I pick it back up for a script of some sort (I don't use it a ton), but other than that, I've found it a fine, and even at times, a fairly nice language for doing huge variety of tasks without the hassle most languages have.
I mostly use it for automation. My favorite script I wrote was this script that would pull all the images from a an fp thread, while simultaneously playing the set that had been downloaded in a slideshow. I also wrote one that mirrored all the assets in my release and debug folders. Oh, and I wrote a script that made my laptop beep at me when it became unplugged (the battery only had like a 10 second life after unplugged)
It makes about as much sense as embedding perl, only slightly less because I can't see how you could have an inline console for a language which relies on whitespace to function.
I've got nothing against Python, it's a great language. But as an in-game scripting language?
I use Python from the command line (as in using its the interactive python) all the freaking time, even to write complex code. Never ran into a problem with whitespace.
Seriously, unless you're mixing tabs and spaces (WHY WOULD YOU DO THIS?), you'd have to go out of your way to mess that up.
:'(
Edited:
inline regex
There's quite a big difference between the python interactive interpreter and a typical game console.
"Randomly" generated (from a seed)
Edited:
Looks a bit nicer with 8x8 tiles
This is true. I would have much rather have used ruby but embedding ruby was impossible due to lack of documentation.
Did you look at V8? It's API is nice.
Regarding Udacity, I haven't taken Python before (I've learned the basics of C++ and I've gotten pretty far into C#). Do you think I can skip the 101 course and take CS212 and CS262?
The Google V8 API is horrendous. I've looked at a few scripting languages and it's by far the biggest and messy to implement.
I don't like javascript any more than python, unless I can get ruby working I'm not going to switch
Oh? What others have you used?
Hello, moving blocks.
What's better than a Raspberry Pi? Two Raspberry Pis, of course!
Hoping to have Chromium running within a few days, getting screwed by toolchain issues at the moment.
Oh you bastard
I read your last post a bit too late to reply - but isn't 200MB of RAM waaay too low for chrome? My chrome is currently using 320MB for 1 Facepunch tab and 1 Stackoverflow tab (-8MB for one extension, and -10MB for the about:memory tab).
It'll eat memory if you have it to speed things up, but it can do surprisingly well on less memory. It'll just keep less tabs in memory, so when you switch to them, it has to reload them.
I seem to remember that you can change the memory model chrome uses, which can make it use substantially less memory (at the expense of speed and reliability)
>:(
i used to run Chromium(the browser) on a 128mb tablet(SmartQ 5) (kinda from the before the tablet hype) it ran a ripped down version of Ubuntu and ran stable up to 5/6 tabs performing better then the browsers specifically made for low ram usage.
I Also ran the Chrome beta under windows XP at 512 MB for a few months being the most efficient browser scaling perfectly with the amount of ram you have available.
Procedural terrain generation is more difficult than I thought so I'll just learn the basics of normal terrain generation for my Ludum Dare entry.
At least you've gotten a delivery date estimate :|
Turns out, my graphics card was running in 2d mode for the entire duration that i've been running the particle simulator
It really runs at 30 ms/frame for me with 2000 particles
Just made a huge leap in the development of our voxel engine by adding octree optimalization to the ray tracer. Still kind of unoptimized code, though, so I believe we can still gain some performance.
Worldsize 13x13x13 chunks (each 32^3 blocks). Don't mind the landscape artifacts. Still got some data-fetching bugs to iron out in the ray tracer.
Worldsize 5x5x5 chunks.
And don't mind the huge grass walls either. They currently just signify the end of the known universe.
Where did you get that from anyways?
I'm guessing you got the second batch.
wake up, see scripting language api discusssion
yes
Edited:
if you're having problems, toss me a message on steam and we'll work out your issues
Edited:
any language can look confusing if you intentionally make your code unreadable
Edited:
from what i've seen, api complexity amongst a few choice scripting languages is some what like this:
in order of high-level to low-level oriented complexity:
Code:Lua C API /- (Fast production, high turn out rate level) Ruby C API \- (Fast production, high turn out rate level) Python C API V8 API (Why-would-you-ever-consider-using-it level)
Turns out with 3000 particles, 10 ms is spent processing particles on the graphics card and 40ms is SFML playing with itself compiling circles. I'm going to have to swap to opengl at some point
Edit:
Fixed it, now 6000 particles runs at 40 ms/frame. A large time of that though is still spent with sfml not doing anything
i never said he shouldn't use python i was just saying i don't like the way it looks