That’s simply, unrealistic. If go route, code will rot, and you will have even more problems long term.
That is perfectly realistic.
Revert to the previous version that works for you, when the bugs are fixed you can update. Simple…
I dont want to disrespect but I dont remember asking to you what I should do.
I just told he has a point.
Don’t belittle me for your reading comprehension problems.
Yeah but his point is totally avoidable as I’ve already explained earlier.
i like how updates work in EPIC.
anyway epic can use even numbers for focus in stability 4.8, 4.10 … and odd numbers (4.7 , 4.9,…). for released the mayority of new features.
Then users who want use a stable release known what versión use, and who want the latests features the same.
Anyway like i say before, i like how versions works now.
I like too the new option for preview release, i not use for my proyect but i can check other things with the release.
Nice epic!
You have got to be kidding. I would much rather have engine releases more often than what you are suggesting. Think of it way. Would you like to do 6 months of work just to find out that it now has to all be tossed out and re-done because there was a fundamental change done to the engine which invalidated your work? Personally, I think you had it right in the old UDK days with a monthly release of the engine.
BTW, may sound strange but I kind of miss UDK. I have lots of good memories of my UDK days…
I dont need to comprehension your advice cuz I didnt ask for your advice.
Just keep your advices yourself until someone asks for it.
The with is if you want to implement anything except a completely on top (like let’s say a new blueprint window or some new nodes).
But as soon as you go for real changes. Like a new approach for UMG to support meshes or a new for the render engine (which will be needed eventually) will absolutely require you touch old code and it’s entirely possible that something old breaks while doing so.
If you only add new functions and never touch old code you will have one monster of horrible software within 3-4 years. Horrible to code for. Horrible to work with. But without new bugs.
Is that really worth it?
I didnt mean that anyway.
Of course you have to touch old code if you have a new on top .
I just told they could make software little bit more low coupling, high cohesion code design pattern that considerably minimizes these issues.
And then a kid comes and gives the greatest advice we never even thought ‘reverting back old version’
Hey everyone, let’s keep the conversation on topic and treat each other with respect. We’re all here for the same reasons 
They release small, and often. Most companies develop way today. It is much easier to catch problems early, and fix them. Rather than releasing a pack of them once a year, and spending months sorting through issues that have compounded, because they were not caught early. Epic is doing the right thing here in my humble opinion.
Hi, Epic!
First of all I want to tell you, that you are doing great job!
But I don’t understand how such subscription-releasing business-model should work fine for Indie-developers. You are releasing new version which coming with new bugs. For small people who spending all time and trying to do something in first project, have no time and strength to deal with bugs and test your engine. So we, people who have stricted funds and time, have no resources to find/report/fix your bugs. Sometimes regular-core features doesn’t work. Like scrollableverticalbox in verticalbox. It just won’t scroll. That’s the basic action for such element and it broken.
So, for us, for small indie-first-project team, we have decided: when prototype will be done, we will take one UE4 version source code, fix all bugs by ourself and will work with that build. No other way, because every release broke something which worked.
And at moment we actually don’t know what should we do with couple of crashes. For example sometimes UE4 crashes on GetActorForwardVector(), or on TArray<>::Add() or crashing on UMG widget compilation from 4.6 crashing and lost dpi curve - Programming & Scripting - Epic Developer Community Forums . I think we have source code and we should dig it by ourself, 'cause your guys telling us “won’t reproduce on our side”. Ok, guys, great. problem haven’t reproducing in your perfect clean project with starter content, but what should other people do, who have a project which have migrated from 4.3 to 4.6.1 and have own structure? How we should work with all problems? Do you creating UE4 to work with clean projects?
Now I undestand why Unity3D won’t reduce their prices. No needs. People will pay for stable release. No one wants to put in buisness plans “Fix/dig engine which we bougth”.
Sorry for my english. Second language.
Last night , I watched Unity3d5 tutorials… nonono…
I’m curious to see Unity3D 5 in 1 year (2016 + 2 update) and Unreal Engine 4 (2016 + 12 updates + tones of new adding & tones great stuff + fix)
Already today, i think , Unreal Engine is much more powerful but need again some great stuff/fix & better documentations/examples/tutorials for the big new users like me .
I bet on UE4
PS : I don’t know if it’s true , but I read in 2015 Unity3D & Modo will be for sale…
Maybe someone will buy Unity3D and will make great subscription offer (Epic you need react? :D)
Ok for modo , we know that Autodesk will buy Modo & add 500 icon in modo :rolleyes:
happy holidays
I think that there are two types of people. Some want to experiment with new features, to study game engine, and so on. But some want to make an actual product using powerful modern tool for professionals. And, for my opinion, last ones have a problem with it. That problem is:
“There is no stable version of engine, and even worse there is no version of engine, which will be stable”. I will try to explain:
For example i have my project, which works almost fine on 4.3. But there is one bug, and i cant finish my project with it. So i wait for new version of engine. And TA-DA! In 4.4 the bug is fixed. But there is new bug with which i cant send my project to production. What should i do with that?
I don`t think that release two times per year will be good solution. The more people will test new features, the more bugs will be fixed. And that is really important.
But i can offer a solution: maybe Epic can assign a stable status for some releases. For example if 4.8 is stable Epic would not stop fixing bugs in it when 4.9 will be released. So in time line we will see something like:
…
4.7 released
4.7.1 hot fix
4.8 released
4.8.1 hot fix - ASSIGN AS STABLE
4.9 released
4.8.2 fix for stable
4.9.1 hot fix
4.10 released
4.11 released
4.8.3 fix for stable
4.11.1 hot fix - ASSIGN AS STABLE
…
And I need a subscription if I want to recive updates for my stable version of engine 
P.S. sorry for my english
If you get constant crashes with such basic things like GetActorForwardVector and T<Array<>::Add() maybe problem is with your code? You have source - you can open utility functions and GetActorForwardVector and look for these crash “callers”. I doubt you will find anything. Try to check existence of variables before passing to function, for example, it may solve a lot of “UE4 is bugged” crashes.
Guys at Epic reacts to any bug, despite it being stupid or something, but if even you can’t help them to reproduce bug - what would you expect? If they can’t reproduce it - they can’t fix it, it’s simple like 2x2.
zeOrb
As I mentioned: I will spend my time and will find a problem by myself. But would other people do it ? No, they just buy Unity or something else which haven’t such bugs. You are talking about reproducing problems by myself. What should I do if is a floating bug? I can’t list clearly steps for reproducing. It just crashes sometimes and sometimes it’s not. I need to build engine and debug it by myself, no problem. But for Indie developer who just starting game development is not an easy task. Epic creates Blueprints which people should use to work without code totaly. So it obvious: they want more people to use that engine. But how people without C++ skills can build/debug engine and find the problem? Just no way. So we have: we want Indie developers, and we can’t give them a stable version of UE4.
I think solution of MadScorp solving all problems. Salute!
Some of you sound like your games are 99% completed but some bug is holding you back…
Seriously, focus on the rest of the game until the bug is fixed, even though most of you speak of bugs that you’ve only read about from others.
The Foundry already bought Modo
Seriosly? https://answers.unrealengine.com/users/8700/alanir.html - yeah almost no bug reports.
I think forum thread is for opinions, not for talking about how we must work.