Fr0stbyte's Development Log

Fr0st, for the spherical problem, I know a game, Shores of Hazeron, where the issue is essentially ignored. Textures get squished at the poles, and it still makes it accessable, but you can't do anything on it. IE no building. Nobody complains.

The Doctor wrote:

Fr0stbyte, I need a not-so brief progress report so I can post it on the wiki. Essentially, dazzle with big words and science!

Very well, I can't get into a ton of specifics, but I'm writing a server-based scheduling algorithm which utilizes graph theory to produce a set of legal permutations, dense from the perspective of having one solution per allocated timeslot.

The tricky part comes, as usual, from real-world application. When scheduling a number of time-sensitive resources, I the programmer, can't always know what the logic behind their ordering is. For instance, does it matter if the sequence is going to be A-B-C or A-C-B? Or more complex A must come first, but then B,C,D,E can run in any order, so long as they all run, and the F runs in parallel with the whole thing. The rules can potentially get really specific and really complicated, so I am making the scheduling algorithm read in a sort of text-based grammar similar to regular expressions and use that to build the dense list.

Now the problem becomes how to make this level of control intuitive for the end-user who will need to be the one to set it. I'm guessing non-computer scientists are not going to be comfortable using formal grammars, or even regular expressions. So I'm trying out a visual method based on a tree graph. The branching nodes are rules to follow. Right now there are 3: run in order, run in any order, and run simultaneously. By nesting controls and putting the resources on leaf nodes, you can put together a pretty complicated set of rules really quickly and intuitively. Plus I think it's kinda fun. Once made, the tree is translated into a text string and stored to the database for later retrieval.

That's just one small part of what I've been up to, but the rest is more boring web development stuff. The tasks which require science here are few and far between, so I tend to get fancy when I do them. And if you are wondering what this has to do with Futurecraft, the answer is practically nothing. I said I was going to be really busy for a while and I wasn't making that up.

@Fenway
How would you handle squishing textures at the poles when it's made up of uniform cubes the same size as at the equator?
I suppose it doesn't have to be 1:1, but we still need to address the deformation somewhere in the process.

fr0stbyte124 wrote:

The Doctor wrote:

Fr0stbyte, I need a not-so brief progress report so I can post it on the wiki. Essentially, dazzle with big words and science!

Very well, I can't get into a ton of specifics, but I'm writing a server-based scheduling algorithm which utilizes graph theory to produce a set of legal permutations, dense from the perspective of having one solution per allocated timeslot.

The tricky part comes, as usual, from real-world application. When scheduling a number of time-sensitive resources, I the programmer, can't always know what the logic behind their ordering is. For instance, does it matter if the sequence is going to be A-B-C or A-C-B? Or more complex A must come first, but then B,C,D,E can run in any order, so long as they all run, and the F runs in parallel with the whole thing. The rules can potentially get really specific and really complicated, so I am making the scheduling algorithm read in a sort of text-based grammar similar to regular expressions and use that to build the dense list.

Now the problem becomes how to make this level of control intuitive for the end-user who will need to be the one to set it. I'm guessing non-computer scientists are not going to be comfortable using formal grammars, or even regular expressions. So I'm trying out a visual method based on a tree graph. The branching nodes are rules to follow. Right now there are 3: run in order, run in any order, and run simultaneously. By nesting controls and putting the resources on leaf nodes, you can put together a pretty complicated set of rules really quickly and intuitively. Plus I think it's kinda fun. Once made, the tree is translated into a text string and stored to the database for later retrieval.

That's just one small part of what I've been up to, but the rest is more boring web development stuff. The tasks which require science here are few and far between, so I tend to get fancy when I do them. And if you are wondering what this has to do with Futurecraft, the answer is practically nothing. I said I was going to be really busy for a while and I wasn't making that up.

@Fenway
How would you handle squishing textures at the poles when it's made up of uniform cubes the same size as at the equator?
I suppose it doesn't have to be 1:1, but we still need to address the deformation somewhere in the process.

Have you considered just making a new game? it sounds a lot easier than bending minecraft.

arsln100 wrote:

fr0stbyte124 wrote:

The Doctor wrote:

Fr0stbyte, I need a not-so brief progress report so I can post it on the wiki. Essentially, dazzle with big words and science!

Very well, I can't get into a ton of specifics, but I'm writing a server-based scheduling algorithm which utilizes graph theory to produce a set of legal permutations, dense from the perspective of having one solution per allocated timeslot.

The tricky part comes, as usual, from real-world application. When scheduling a number of time-sensitive resources, I the programmer, can't always know what the logic behind their ordering is. For instance, does it matter if the sequence is going to be A-B-C or A-C-B? Or more complex A must come first, but then B,C,D,E can run in any order, so long as they all run, and the F runs in parallel with the whole thing. The rules can potentially get really specific and really complicated, so I am making the scheduling algorithm read in a sort of text-based grammar similar to regular expressions and use that to build the dense list.

Now the problem becomes how to make this level of control intuitive for the end-user who will need to be the one to set it. I'm guessing non-computer scientists are not going to be comfortable using formal grammars, or even regular expressions. So I'm trying out a visual method based on a tree graph. The branching nodes are rules to follow. Right now there are 3: run in order, run in any order, and run simultaneously. By nesting controls and putting the resources on leaf nodes, you can put together a pretty complicated set of rules really quickly and intuitively. Plus I think it's kinda fun. Once made, the tree is translated into a text string and stored to the database for later retrieval.

That's just one small part of what I've been up to, but the rest is more boring web development stuff. The tasks which require science here are few and far between, so I tend to get fancy when I do them. And if you are wondering what this has to do with Futurecraft, the answer is practically nothing. I said I was going to be really busy for a while and I wasn't making that up.

@Fenway
How would you handle squishing textures at the poles when it's made up of uniform cubes the same size as at the equator?
I suppose it doesn't have to be 1:1, but we still need to address the deformation somewhere in the process.

Have you considered just making a new game? it sounds a lot easier than bending minecraft.

You want him to screw everything he has done so far?
I dare you!

Lol, it's been tempting a few times, but the advantages of using Minecraft outweigh the disadvantages, not the least of which is a preexisting community. When you build an indie game from scratch, you have to build your audience from scratch, too, and that's the trial by fire where most indie developers fail, for no fault of their own. Such is the nature of the beast.
And, since the core of Futurecraft is an MMO, having access to that large community is going to become very important.



Also if it wasn't clear, the scheduling system is something I'm doing for work, not the game. I don't have a progress report for the game because of it. One of these days thing's will settle down though, and I'll be able to get home while the sun is still up and put some real time into the mod.

fr0stbyte124 wrote:

Lol, it's been tempting a few times, but the advantages of using Minecraft outweigh the disadvantages, not the least of which is a preexisting community. When you build an indie game from scratch, you have to build your audience from scratch, too, and that's the trial by fire where most indie developers fail, for no fault of their own. Such is the nature of the beast.
And, since the core of Futurecraft is an MMO, having access to that large community is going to become very important.



Also if it wasn't clear, the scheduling system is something I'm doing for work, not the game. I don't have a progress report for the game because of it. One of these days thing's will settle down though, and I'll be able to get home while the sun is still up and put some real time into the mod.

Since you're sticking to the Minecraft roots... will the planets be cubic?

There are serious, potentially unresolvable issues to every model we've discussed, so there will just have to be a lot of experimentation to see what works best. At least at the start, planets will probably just be flat, 1-sided disks until we can work out the kinks.

Aww no one mentioned my pun :( Well even though I can't read half of the stuff you post here, looking forward to an update! @Ivan Yes that is exactly what I was suggesting.

fr0stbyte124 wrote:

There are serious, potentially unresolvable issues to every model we've discussed, so there will just have to be a lot of experimentation to see what works best. At least at the start, planets will probably just be flat, 1-sided disks until we can work out the kinks.

IF YOU MAKE IT NON-SEAMLESS ALL THOSE PROBLEMS GO AWAY!!!!!


Also when are you gonna release fast culling for 1.3?

hyperlite wrote:

Also when are you gonna release fast culling for 1.3?

Once I get some time to sit down with it. I doubt anyone is more annoyed with the lack of progress than myself.

On the plus-side, it's been mostly planned out now, so implementation shouldn't take more than a day or two. Keep in mind, though, I am awful at time estimation.

fr0stbyte124 wrote:

hyperlite wrote:

Also when are you gonna release fast culling for 1.3?

Once I get some time to sit down with it. I doubt anyone is more annoyed with the lack of progress than myself.

On the plus-side, it's been mostly planned out now, so implementation shouldn't take more than a day or two. Keep in mind, though, I am awful at time estimation.

haha i remember once you gave an honest answer to someones when will it be out statement (i didn't think it was trolling but i could be wrong)... that deadline was around a couple months ago i think.

I am also terrible at making time estimates. For instance, I was working on this one project today for the past 18 or so hours, and probably did 15 the day before. Last week I was positive it would take half a day.

fr0stbyte124 wrote:

I am also terrible at making time estimates. For instance, I was working on this one project today for the past 18 or so hours, and probably did 15 the day before. Last week I was positive it would take half a day.

Well, that's not that bad...

So.... Any progress? Because its been about a month or are you still on your project from work?

If by "progress" you mean going to secret developer parties and getting drunk, yes, we have made TONS of progress.

Professor Fenway wrote:

If by "progress" you mean going to secret developer parties and getting drunk, yes, we have made TONS of progress.

Yo, don't mention those. We agreed.

Tiel wrote:

Aye-Aye memes GO!

I've said it way too many times already. My company is going to our largest trade show of the year in mid-November, and I am the lead developer on two new products. One is already getting half of its features pushed back until December, and the other one is a race against time just to get it presentable. I've been working 12-15 hours a day for the past month, and the next month isn't going to be pretty much the same. I would absolutely love to be working on Futurecraft instead, but the best I can manage right now is to follow the discussion, and answer the occasional easy question.
Eazymc is active on the project right now, but even he has his own situation and prior obligations which come up from time to time.
For the time being, bear with me. It will begin to get better after the show.

It sounds like commercial programming is a very tedious job.

Tiel wrote:

It sounds like commercial programming is a very tedious job.

In my experience, it really depends on where you are within an organization. Frost, being a lead developer for a couple projects, is probably under a lot of pressure by the people signing his paychecks. I at least have the liberty of being sold to other companies for use lol. Basically what I deal with are outside organization who are looking for certain enhancements or products to enhance their buisness. My last 3 day project was just a simple API to allow one company to communicate with another companies server. Took a day to write and 2 days to fully debug, test and get to their QA for implementation. It'll come back 2 or three times with enhancement or bugs and I'll work those out.

Now when I get on what I consider a "real" project (and by real I mean something that pays more than 80 hours, which I'm salary but the company I work for charges our clients by the hour of work) thats when I disappear for a few days and pull a Frost, working long hours per day to complete a project. I don't know what insentives Frost has for pushing out code like some sort of delicious frosting from a tube, but when I push out more, I get monthly insentives for more projects done base on how much money I make the company. Thusly, when it comes time to make decisions, I generally pick code thats going to make money, so unfortunately a game mod comes second.

I know a lot of people want to see this mod come to life, Frost is probably the biggest fan seeing as how he's put a lot of time and effort into the engine, so hang tight people, once the slow programming seasons hit, which mine never actually seems to slow down but I do get days or weeks (rarely more than a week) without a barrage of requests, we'll be able to put more than 5 hours into other projects and this mod will start to shape up.

Very enlightening, eazy.

Thank you! Thank you Mr Obama!

I never thought I'd see the day when the president of these United States would be applauding me. Proudest moment of my life!

So, if fr0st is presenting in a trade show, might it be televised, may we gaze upon the beauty of our glorious leader?

I thought you were an owl ...