Flower's Dev Diary (Week Of March 18th)

[FlowerChild]

I view your hardcore changes, dealing with Mojang's incompetence, and RTH as "A Savage Journey to the heart of the Minecraft Dream" ...

Hehe...how about:

"Groth of Darkness: The story of Better Than Wolves"

[Six]

Since you're doing the 1.5 port I thought I'd mention one of the lesser known 1.5 changes which will have more relevance with the seriousness of damage in BTW. Flaming zombies attacking other mobs (i.e. Steve or villagers) have a chance to set the mob they are attacking on fire. It makes going to get the last hit on a flaming morning zombie a bit more of a serious affair.

[FlowerChild]

Since you're doing the 1.5 port I thought I'd mention one of the lesser known 1.5 changes which will have more relevance with the seriousness of damage in BTW. Flaming zombies attacking other mobs (i.e. Steve or villagers) have a chance to set the mob they are attacking on fire. It makes going to get the last hit on a flaming morning zombie a bit more of a serious affair.

Yeah, I knew about that, and I've got no issues with that man. Not sure why you'd think I would.

[Xeo]

It's funny all this talk about code optimization takes me back to my discrete mathematics class and Big O. Something that I feel as though Mojang knows nothing about because of the conversation. Oh the memories of thinking through code and finding fun ways to decrease the time it takes to run...

That kind of math is rarely used in programming on this level, or at least I'm not aware of it being used. You care about computational complexity and algorithm design more in lower level programming where you want either something that will run under very tight constraints or are doing something leviathan and don't want it to blow up your computer. People like google or people making a higher level programming language probably spend a lot of time thinking about theoretical running times but in something like a game (that's not trying to push the boundaries of sanity) it's either not worth it or is nightmarishly complex.

However, there are certain things that can be learned from algorithm theory that certainly help in this kind of optimization.

[FlowerChild]

Hehehe...man, I just keep chuckling every time I go into the rendering code. This time, it's all the sanity checks the code is doing to make sure that the textures in questions are really really (really) there each time they're rendered. Nah, we wouldn't want to test that ahead of time or nothing. Why would we when we can just test it every single time the texture is used!?

I can hear the thought process behind this quite clearly: "well, teacher said I should always test for null, so I guess I should do that thousands of time per frame..."

It's kinda like trying to fire full-auto with the safety being engaged with each round :)

Man, I seriously can't wait to go on an optimizing rampage in this code. I think looking forward to that is what's primarily motivating me in getting through this mind-numbing update right now ;)

So far so good though. Got all the mod's items converted over to the new texturing system, and am deep into the blocks now. No nasty surprises along the way, just requires a lot of elbow grease and patience.

My apologies to texture pack artists in advance though, as it looks like there's no reasonable way around breaking up the mod's texture sheets into individual files with the new system. If it's any consolation, it's quite a pain in the ass for me to do so as well.

[savagelung]

My apologies to texture pack artists in advance though, as it looks like there's no reasonable way around breaking up the mod's texture sheets into individual files with the new system. If it's any consolation, it's quite a pain in the ass for me to do so as well.

A month or so ago, I bumped into a Photoshop Action that de-stitched your terrain.png and items.png into 1.5-compatible files. I spent about 15 minutes looking for it, but I couldn't find it :( I'll see if I can assemble one when the BTW texture file names are available.

[FlowerChild]

A month or so ago, I bumped into a Photoshop Action that de-stitched your terrain.png and items.png into 1.5-compatible files. I spent about 15 minutes looking for it, but I couldn't find it :( I'll see if I can assemble one when the BTW texture file names are available.

There's a thread in off-topic where I was asking if anyone knew of such a utility, and an excellent little one that I then used for the mod files was provided.

viewtopic.php?f=8&t=7197

Given that, it's not ripping the texture files that's the major hassle, but identifying and renaming all of them. What I'll probably do is once I'm done setting all the names (I have the extra hassle of associating them in the code as well), I'll post up the default textures in the add-on sections so that texture artists can prepare for the release if they so desire.

On the bright side (there aren't many in this I realize), all mod textures will be identified with meaningful file names after this, so it'll be extremely easy for you guys to tell what does what when creating your textures.

[Graphite]

Wonder how you'll handle the secret features. totallynotsomesecretkindofnewfungus.png? :P

[icynewyear]

Hehe...how about:

"Groth of Darkness: The story of Better Than Wolves"

Now I want to record a documentary about the development and history of BTW...

[FlowerChild]

Wonder how you'll handle the secret features. totallynotsomesecretkindofnewfungus.png? :P

Lol! Good point. I hadn't gotten to that one yet :)

[Xeo]

Wonder how you'll handle the secret features. totallynotsomesecretkindofnewfungus.png? :P

Lol! Good point. I hadn't gotten to that one yet :)

Maybe a hash that's revealed by playing the game and discovering the feature?

[FlowerChild]

Hmmm...so I just noticed that the Forge has announced a series of performance improvements to MC. Coincidence? :)

All right...it's on ;)

[magikeh]

All right...it's on ;)

Yay performance battles!! Now if only we could get Mojang onboard with this...

[ScubaPlays]

Yay performance battles!! Now if only we could get Mojang onboard with this...

Minecraft: Now with chunk loading screens!

[FlowerChild]

Yay performance battles!! Now if only we could get Mojang onboard with this...

Mojang already had their go with the last release which was being promoted as being "better than Optifine".

Now it's our turn ;)

It wouldn't surprise me if the Forge devs had the same "wtf?" response that I did when porting over to 1.5.

[Husbag3]

When playing 1.5 I did experience at least a 10 fps increase when fiddling around in my world. I look forward to any additional changes you make with regards to optimisation.

[xou816]

It's a bit out of the subject, but the FPS in 1.5.X are behaving quite strangely.

Spoiler

Show

In a normal solo, I think I did get around 20 FPS (close to what I had in 1.4.X, if not less), but this was just a world to test out the new blocks, and I've not made any other worlds to test further. However, I played an adventure map with friends and I had around 60 FPS! ...although I hosted the server. What the hell?

Anyway, good luck on your port! :)

[FlowerChild]

Sigh...so I was just considering my upcoming repurposing of nether quartz, and decided to check out the wiki entry to make sure I had all my facts straight when I came across this little tid-bit:

Note that the nether quartz ore block cannot be obtained without the Silk Touch enchantment, and that this smelting is slightly inefficient because with Fortune, you can get more than one quartz per ore. Also, there would be no reason to collect it with silk touch if you were just going to smelt it as it would be more efficient on long excursions to save space and to gather the experience earned from mining the quartz ore for later.

Jesus Christ...what a cluster fuck of muddied design the above represents.

I will of course fix that for BTW, but man...the combination of a mineable resource that basically functions like coal or diamond when you harvest it then also having smelting rules applied to it is just totally frigging clueless when it comes to overall consistency within the game.

No idea what they could have been (or if they were) thinking at the time.

[Solymr]

I will of course fix that for BTW, but man...the combination of a mineable resource that basically functions like coal or diamond when you harvest it then also having smelting rules applied to it is just totally frigging clueless when it comes to overall consistency within the game.

Both coal and diamond ore can be smelted to get the item, so this goes back a long time.

[Azdoine]

I'm pretty sure all the ores have smelting recipies if you can silk touch them.
EDIT: Ninja'd

[FlowerChild]

Both coal and diamond ore can be smelted to get the item, so this goes back a long time.

Arg...wtf. Never realized that before seeing the above.

Nevermind then. It's just symptomatic of ongoing lack of design consistency rather than something new ;)

[savagelung]

I will of course fix that for BTW, but man...the combination of a mineable resource that basically functions like coal or diamond when you harvest it then also having smelting rules applied to it is just totally frigging clueless when it comes to overall consistency within the game.

Not sure if you know this, but whenever they added experience drops to ores and smelting, they also added the ability to smelt ores of all kinds, including diamond, redstone... and for some bizarre reason, coal ore. I actually had forgotten about this until you mentioned it, and I've certainly never wasted durability on a silk touch pickaxe just to smelt the ores.

Edit: Double ninja'd

[CycloneSP]

actually, you were able to smelt all ores the moment the silk touch enchant was put into the game. I remember being able to smelt redstone into redstone dust in the furnace. (I took it out post haste however, due to the fact that you will only get 1 redstone dust per ore) sooooo, yeah, dunno why anyone would want to smelt anything other than gold or iron as with lapis and redstone you are effectively screwing yoursefl over. That, and you would be better off getting the fortune enchant and increase your output even more.

[FlowerChild]

actually, you were able to smelt all ores the moment the silk touch enchant was put into the game. I remember being able to smelt redstone into redstone dust in the furnace. (I took it out post haste however, due to the fact that you will only get 1 redstone dust per ore) sooooo, yeah, dunno why anyone would want to smelt anything other than gold or iron as with lapis and redstone you are effectively screwing yoursefl over. That, and you would be better off getting the fortune enchant and increase your output even more.

You know, I think this might actually just be a bug resulting from all ores being smeltable in the furnace due to the way the code works, rather than an intentional design decision.

I'm thinking it just didn't occur to them when they added silk touch that the furnace would automatically be able to process the previously unavailable ore blocks.

[BlueRavenGT]

You know, I think this might actually just be a bug resulting from all ores being smeltable in the furnace due to the way the code works, rather than an intentional design decision.

I'm thinking it just didn't occur to them when they added silk touch that the furnace would automatically be able to process the previously unavailable ore blocks.

I remember them specifically adding the smelting recipes because of silk touch. All the vanilla recipes are added in FurnaceRecipes.class using block/item IDs, so that explanation wouldn't be very likely.