EVE General Discussion

This one's pretty much entirely theory crafting related, a couple meeting's - discussion on progress, is there any gross imbalances if we give 5-10-15-25% buff if so bring it up talk it through.

Real work is:
Going through all related values to (ammo type in this case), and making the necessary changes.
Rebuilding/changing values in descriptions.

Could be easily done by two guys in 2-3 weeks. Lots of play testing..




This one's incredibly different and significantly more complex
First you'd have a rhyme/reason to change said battleship, discuss it - pro's/con's and mostly why the change?
If agreed upon of course you would throw it at the art department "we want this" (I don't have a ton of experience with 3d modeling so to do a battleship like in EVE to me seems like a grotesquely massive project that would take me months, I imagine someone miles better then me could hammer it out in 2 weeks.)
Review the model, see what you like / don't like and go back if you want something changed.
Then hitbox's would be done, for clipping and spatial recognition (Ship bumping, pos's, stations, etc)

Continue onto playtesting to make sure the hitbox's fleet right, in a game like eve there would be probably changes to speed/size/signature etc, balancing would come into effect alot here.
adding/updating descriptions/attributes to that item.
Patch team creates patch, deploys it to cdn - players get it and get to whine about it yay~!

Note:: this isn't specifically CCP, but vast majority work along these lines.

//Worked on a smaller MMO about 8 years ago

I would assume the detail would degrade / be out of proportion by doing this and then have to be altered

Shiny!! then as well as my origional question, can we haver bigger battleship models please

hunting through millions of pages of code for specific values and functions and changing them

then going back through the same code to remove the typos and add semi-colons and whatnot so the code will actually run

Lunch with Mittens

a week of testing to determine that the wrong values were changed

repeat steps 1 & 2

Shelf the entire idea because management wants to be able to put on a frilly dress and make dirty emotes at players in stations

drink until the sweet release of death comes


....wait... what was the question again?

Fixed it for you.

Edit

Pity they didn't test Incarna, so we wouldn't have lost all those graphics cards.
/Edit

---

Alterations usually take less time of actual work but can take a lot of time in discussions to find the different impacts. Changing skins is usually a whole new one to have higher resolution.
Many objects such as a ships would have a parent object, so that they only have to specify some values for the statistics of that child object; the new ship. Modules would be much the same.

I think it is new mechanics, such as engine trails and new models with textures that would take the most time. If you want a better idea, go and make some things (sculpted) in Second Life.

Did I volunteer to give my goods somehow from the billions I've taken from dumb pubbies like you? I think your thread demanding an answer into how much time they've spent is a dumb one and their time would be better off invested in making this terrible game a better one instead of trying to getting out an abacus for you. As I said, please quit and unsubscribe forever.

I work in software development, so I may be able to shed a bit of realistic, non-sarcastic light on this.

Changing something like how much damage a weapon does is a trivial change. Check out the file, change the number, build to make sure you didn't accidentally type a o instead of an 0, and check it back in. The bigger part of the change is a little thing called regression.

Once you've changed the number, you have to make sure that any ship that's using that weapon isn't suddenly the only thing that's worth taking into a fight. Yes, something like supercarriers might get through, but it took the players a while to figure out how amazing they suddenly were, build far more than CCP is capable of testing with, and using them to demolish whatever.

Balancing errors are incredibly difficult to catch, though, so you need to do as much testing as possible. However, there are diminishing returns, the more you test the less useful stuff you'll find, especially when you want to get on to building something new, you start to ignore potential errors.

Something bigger than a balance change (like the font) is a fair bit trickier. It's not just a matter of changing some variable from old.ttf to new.ttf. Every single window that has text should be checked to make sure nothing has gone crazy. On top of that, changes to the text wrapping in a window may affect something else in how the window functions, it can be an insane mess that's hard to follow.

Many assume that writing software is like writing a book. Once you've written everything you want, the reader can just figure it out with their imagination. The problem is that a computer has no imagination, so you have to write that too. Also, if you make a single spelling mistake, if it doesn't understand it stop reading until you reprint the book with the typo fixed. If it thinks it understands... who knows what might happen. (*cough* boot.ini *cough*)

Thanks thats pretty enlightening, programing seems like some strange magic (no hat) to me. But you give an idea at how much work, scales up, when you change even trivial things... or at least what i thought was trivial.