Programming in general is not like the workshop. First, you need the knowledge, which is obvious. Second, you need to come up with an idea of what you want to do if you’re programming a game/app. Third, you need to split your time correctly and pay close attention to what you’re typing in, because 1 line can make a critical issue in the code/specific algorithm you’re trying to make. If you don’t split your time correctly, you’ll get exhausted at some point so you’re gonna make more mistakes and the code will be poorly written and won’t work properly.
Although I dislike the rate in which we’re getting new patches, it’s still better to wait a month or 2 for the next patch. It gives you more time to adapt to the balance and feeling of the game post patch, and it gives more room for the next patch to be a high quality patch (although it doesn’t usually happen and we don’t really get that many great patches, but still).
Absolutely trusting a company i know has experience over people that have no experience is not the same thing as trusting everything they say or do, but good job on you for completely missing the point and just seeing an opportunity to crap on someone you don’t agree with.
watch out buddy, this is the forums. If we raise our expectations to unreasonable levels its the devs faults for not following throw.
I actually 100% agree with you most of what you’ve said so far in this post. Sadly the forums will never except that anything is good and shall always complain about the smallest details
OMG THE NEW HERO DOESN’T REPRESENT MY MINORITY, FIX THIS IMMEDIATELY OMG THE GAMES BALANCE ISN’T GOING IN THE WAY I WANT IT TO, FIX THIS IMMEDIATELY OMG I DIED TO A STRAY HANZO ARROW, NERF IMMEDIATELY
as a few examples
Have a good day! You don’t see many people actually put forth good arguments these days
They did not simply splittet SR in 3 seperate categories, the rewrote the whole system.
This means, what data get tracked, how to derive meaningful information out of it, it can even mean rewriting parts of the engine if they use something like routes, positioning etc as data.
The amount of contextual Data per match is insane.
You could say at least 1/3 of the games code is Matchmaking
It takes long, very long to develop such a complex system, test it, fine tune it.
If it goes live and it’s not working as intended it will cause serious problems in the long term
Pretty much. I understand that the forums are generally a place where people go to vent their frustrations, but there is a difference between that and the vitriol that gets tossed around here. It is mind-numbingly annoying sometimes.
and 4 posts up I was crapped on for having a good argument too lol
First they have to test to see what kind of effects strict 2-2-2 role lock will have. Is it too limiting? How does it affect the existing SR system? Will we need a different SR system for each class? How will matchmaking be determined under this new system? How can we optimize that? Does the existing code support the necessary changes, or will we need to modify/rewrite it altogether? And different tests/analyses/calculations would need to be conducted to understand how it affects end users, particularly to understand matchmaking times and if they’re significantly adverse to the end user experience, if certain features implemented would help mitigate those difficulties, etc. This isn’t something that simply happens overnight, but over multiple rapid prototype/testing build iterations, design deliberations, etc. Working out all these nitty-gritty details is essential to build a good foundation for the implementation, otherwise you can encounter major issues along the way that can frustrate the development process.
Then comes the actual implementation, which depending on what they’re changing can involve many months of work. In this case major rewrites to their complex international matchmaking system in order to fit into the existing code infrastructure, or an entire ground-up rewriting if the existing framework proves inadequate to support these features (this might even involve the creation of entirely new dev tools for monitoring/interfacing with these systems). You’ll also need to build the GUI that players interface with and how it connects to back-end information on different parts of the system (for example, not just how players get into matches under the new system, but also career profile information for their new separate SR rankings in different submenus and even online on the playoverwatch website).
Then there’s code reviews, build testing, quality assurance, security clearance, multi-platform certification, public messaging coordination, etc.
If this were a tiny dev team of a handful of people working on a smaller project then yeah, something like this could all happen on the fly really quickly, but we’re talking about massive, highly sensitive, sprawling inter-related programmatic chunks of an international multiplatform game. If you’re a kid then maybe a year feels like forever, but for an adult getting such a task done and done right a year is actually pretty good in programming terms.
Shifting from discussing why it took so long, what was the point of saying that? To guilt those resistant to into liking it? To make it seem like they were listening this whole time and just couldn’t announce it yet? As an excuse for them to hide behind from accusations of lacking content? To make it seem higher quality?
It just seems like a weird thing to say in a video where you’re just supposed to be describing how the system will work.
I’m sure Jeff doesn’t understand how the system works. He is not senior developer,
If you worked in the industry or studied computer science you would know how complex such systems are and it’s impossible to describe it in words
I do not think anyone can really appreciate or get a good sense for is this an accurate statement of a year without having done the work before. Particularly in regards to programming or working for a larger corporation.
My experience have never really involved anything with say high level applications running on an OS such as what this game must do. However, having worked on very low level interfaces from SystemVerilog in FPGAs, to writing test code for firmware applications on an ARM processor or DSP, and some projects into C# or Python in a windows environment I have to say the effort involved shouldn’t be taken lightly.
The order of magnitude from start to finish seems about right for a serial process within a company. The actual amount of man hours and labor put into the process might not be like 8760 hours in a year, but I can certainly see this taking a number of iterations and testing that spawned a few months while also working on possibly other things.
Absolutely trusting a company i know has experience over people that have no experience is not the same thing as trusting everything they say or do, but good job on you for completely missing the point and just seeing an opportunity to crap on someone you don’t agree with. 
