|
Author |
Thread Statistics | Show CCP posts - 56 post(s) |
asteroidjas
Rothschild's Sewage and Septic Sucking Services The Possum Lodge
76
|
Posted - 2014.06.12 23:57:00 -
[1] - Quote
EMT Holding wrote:CCP Greyscale wrote:EMT Holding wrote:CCP Greyscale wrote: -- Invention now removes one run from the input copy -- Some special-casing and fuzzing has been done in certain places to make numbers more interesting
For your first point, can I assume that to get a 1 run T2 ship BPC, I now need to use a T1 BPC with 2 runs? Or to get a 9 run T2 BPC, I use a 10 run T1 BPC? If so, what are decryptors for? Is there any point in them? - No, invention literally just removes exactly one run from the copy during the invention job. All invention jobs now need one run. This is happening as we want to normalize copy time at 80% of build time; we've adjusted invention and build times so that end-to-end copy-invent-build times remain roughly constant. 1. If that works how I'm thinking it does, that sounds good. So if I invent off a 10 run T1 BPC, when the invention finishes, I get the T1 BPC back but with 9 runs? That sounds like an excellent midstep towards being able to do batched invention or an invention job for more than 1 run :D 2. Also good. I can understand the need to slightly bend material requirements for smaller/larger things. So, without reading further....does this mean that if i want to continue inventing from the same copy (since you are apparently doing this to keep the overall time constant) that i'll just end up with (on modules) inventing from the 299 run copy which would yield a 9 run T2 bpc? Why would i continue to do this as now that invention has extra costs based on the end product (which is also tacked on again when one builds said end product, per run)...why pay for an inferior invention job?
Am i understanding this correctly? Current TQ method requires a 300 run T1 copy to get a 10 run T2 BPC. (which gets destroyed in the process) How i understand what you are saying is that now we will have the option to continue to invent with diminishing returns on the already used copies? Unless the cost of invention is reduced, if i'm going to invent something, i'll do it with as many runs as i can. |
asteroidjas
Rothschild's Sewage and Septic Sucking Services The Possum Lodge
76
|
Posted - 2014.06.13 00:46:00 -
[2] - Quote
CCP Greyscale wrote:Research time has some confirmed math kinks, looking into them.
Better error messages will ofc be forthcoming!
Quote:We are still looking at this but I can confirm the output runs on invented blueprints is not correct. Will get this fixed next week.
Sooo.....hows that inbeded QA working for you guys? Did nobody even think to attempt a research job before they called this thing ready? How does this stuff get past QA anyways?
|
asteroidjas
Rothschild's Sewage and Septic Sucking Services The Possum Lodge
76
|
Posted - 2014.06.13 02:47:00 -
[3] - Quote
CCP Nullarbor wrote: We are still looking at this but I can confirm the output runs on invented blueprints is not correct. Will get this fixed next week.
Forgive the bluntness...but this (after all the previous "Yeah that is broken, we'll look into it" replies) begs the question...
What of the research portion of this new mess IS correct? What is the point of putting something on the test server that hasn't even been partially tested by the teams involved. Research times are wrong, copy costs are wrong, invention stuffs are wrong..ect...
I understand it is a "test server"....but is that really supposed to mean that nothing has been tested at all prior to it being thrown onto SiSi? Because when this happens...we have to wait yet another week, in the already short time left, to find the actual bugs related to this massive change.
That is my point. In most workplaces these type of deployments would be worthy of disciplinary action. (maybe that is what part of the housecleaning was?)
And no, Beta testing CCP's products is not related to SiSi....TQ gets more than its fair share of fail deployments that are the beta versions of the concepts CCP is trying for. (really amazed and thankful they actually thought about deploying this two weeks ago) |
asteroidjas
Rothschild's Sewage and Septic Sucking Services The Possum Lodge
81
|
Posted - 2014.07.09 03:04:00 -
[4] - Quote
CCP Nullarbor wrote:Something to be aware of, pretty much all of the cost values on Singularity are garbage at this point because they are scaled based on activity, of which there is very little on the test server.
Our own simulations from TQ data put the estimated cost for manufacturing somewhere between 0-15% of the manufactured goods depending on where you build it. Research / Invention costs are a little trickier to calculate, but I have a developer blog coming out next week which will describe the cost formula in detail for those looking to update their spreadsheets. Just for clarification here....
So, this new cost scaling is meant to scale the job cost upwards with increased activity....right?
Yet, you just said the absurdly high costson SiSi are because there is next to no activity on SiSi...right?
Something seems amiss between those two statements. You might want to doublecheck your maths at some point. Maybe. |
|
|
|