Top Down Learning

[USER=“4894”]Tim Hobson[/USER]
[USER=“2247”]Richard Hinckley[/USER]
[USER=“7”]Jeff Wilson[/USER]
[USER=“197”]Tom Shannon[/USER]

Epic is busy reworking the docs and adding resources such as Unreal Academy at the moment. So maybe its a good time to offer a couple of suggestions about the Docs in general and about adding to those that still feel very incomplete, such as Intermediate / Advanced Multiplayer & AI etc…

**#1. **The millionth question about BP Comms & Casting was asked recently. Its always a popular subject that never seems to run out of road. I wonder why that is. :stuck_out_tongue: I passionately believe topics like this are better learned through practical projects. Something which Epic seems to shy away from. For example, what happened to offering an equivalent to the UDK-Necropolis / UDK-Jazz demos in UE4? A substantial game project that’s BP based etc… In general, why aren’t there more practical projects where beginners can learn lots of abstract concepts in one place and then watch them all play out. A form of Top-Down learning… This bottom up approach of Typed-Docs / Wikis and Video-Tutorials is fine, but it often leads to fragmented learning. So a shout out here to devs like Jacky and others for their practical contributions. It feels like Epic sees Community-Tools as a bit of an inferior learning space… But you know what, its actually a sleeping giant… :wink:


#2. When it comes to more advanced topics such as Multiplayer and AI, many devs quickly find they’re sort-of on their own, and that’s a missed opportunity. Why not invert that… Why not start out by creating the most ‘Advanced-Tutorials’ possible then work backwards… If Epic runs out of time / resources then there’s always leftover tutorials on YouTube / Vimeo anyway to fall back on… What I’m saying is… Why reinvent the wheel? Focus first on ‘Advanced Tutorials’ for AI & Multiplayer that include ‘best practice’… After those are nailed, work on simpler things… Because to run out of resources leaving advanced topics unfinished is unforgivable… :eek:

You are right, it is pretty hard to get started on the UE4 (or at least it was for me) because the info is scattered all over the internet and lets admit it, it is 2018, nobody wants to read a wall of technical text with a bunch of lifeless pictures (yeah I’m talking about the doc). I think EPIC has done great with the live-streams, but also making video tutorials is excessively time consuming. I know that from experience.


^^ This **^^ **

The Docs rarely cover exceptional cases either. A few ‘screaming warnings’ or ‘quick list of assumptions’ would go a long way. One case that appears on AH a lot is the Widget-Interaction-Component (WIC). The Docs fail to cover best practice when using the WIC for in-world UMG interaction. So in effect, the ‘Interaction’ settings act as a kind of TRAP instead. For every 10 usage cases there’s 9 where the WIC doesn’t work as expected or acts unreliably. But this is a tool that could use a few SCREAMING WARNINGS not to use standard Trace-Channels, as nearby moving actors with attached Collision-Box-Volumes can silently ‘eat’ all the WIC traces!

This is where the Docs / Wikis / Video-tutorials are often a FAIL as there’s just not enough correct usage or caveats included. Overall that’s where Livestreams are better, as they often contain more caveats and insights that are missing from the Docs. But of course, Livestreams are infrequent and long, so details get missed. Plus Livestreams can’t include every use case either. The recent stream from ‘Zak and his Slider’ was interesting for what it didn’t cover, as much as what it did.

Hints come in a later Livestream with some blunt honest feedback from the SixFootGames guys… Eh, where are the Docs for UMG-Sliders and Gamepads??? :stuck_out_tongue: <Tumbleweed> Its just left hanging. Personally, I couldn’t wait any longer for Zak’s cancelled Livestream. So I just created a ‘trick-slider’ by scaling an in-world widget and filtering input from a Gamepad button to size the slider and ‘fake the effect’. It doesn’t look as neat as Zak’s slider… But you can’t argue with the simplicity, versus 2 Livestreams (one official and one unofficial) to try and nail this for any UE4 users watching.

This is where Epic making their own complete sample projects would solve heaps of problems. Because there will never be enough Livestreams or Wikis or Docs or Unreal-Academy videos to cover every use case. But game demo projects can get users closer than any of the other options combined! However, I’ve always had the impression that Epic engineers get by – by shouting over to their C++ colleagues in another cube and asking:

That sums up the Docs… All the caveats on usage of BP nodes are just left out. Its up to users to figure out the magical gotchas behind all the BP nodes. By trial and error or by delving into C++ source. But if everyone has to do that, why even bother with Wikis and Docs and Video-Tutorials at all… They’re not enough of a parachute for UE4, that’s for sure. So if ultimately all roads lead to C++ to really learn how things work, why invest the time to make what can only be flawed and incomplete Docs / Wikis / Videos… Is it just a PR exercise???.. :stuck_out_tongue: :D… And what happens when the C++ road still isn’t enough and you still need Docs…:eek:

In terms of sample projects… there are quite a bit… what am I missing. The recent ActionRPG is really cool…shooter game, is pretty decent. They even have some mobile stuff. What specific project ideas does this thread have?



Blueprints only! Please re-read the original post Teak!
Your examples are not BP only… They’re part / full C++.

The mini-game projects from Devs mentioned above:

  1. 3D space shooter working game demo from Jacky…
  2. RPG game demos from GameDev005 & AlFlakky…
  3. Radar + Procedural templates from CoquiGames.
  4. Stormrage256’s many and varied game mechanics …
  5. 100’s of other Community game demos / templates…

Another point in favor of offering Blueprints-only based Game-Demo / Projects as opposed to focusing on Docs / Wikis / Videos-Tutorials or Mixed C++/BP-Projects etc… **I****s that **BP game demos are more likely to survive over time… Why? - They can be upgraded between engine versions easier… Sometimes even automatically… Whereas some of the other options above go stale much quicker. This applies to C++ game demos especially (example thread here)


Using Animated-GIFS alongside Docs / Wikis offers huge advantages…
Having more like this for Object-Collision-Traces etc would be so helpful.

What about the example project called BLUEPRINTS. What about CONTENT EXAMPLES. Have you reviewed them? Great stuff.

Most folks tend to use C++ and BP… so, wouldn’t it make sense to highlight that? Besides, there is a ton of BP in those projects to get an advanced how-to look.

When I started I had no issues learning BP. I learned from YOUTUBE tutorials… their current example projects…friends… reading articles…even purchases on the marketplace.

As for your other point…agreed, makes sense.


PS: BP only example (a bit older): Multiplayer Shootout | Unreal Engine Documentation


Agree, those are useful resources. But they’re not complete game demos from beginning-to-end like UDK-Necropolis or Jacky’s ‘Viper’ game demo etc… I’m talking about complete game demo projects for those who like dissecting things, or working top down and taking stuff apart etc. Would it hurt to have a few official Blueprints-only projects that are FPS / TPS-RPG / 3D-Space-shooters etc etc… Just sayin… :slight_smile:

That’s debatable. It’d be interesting to see some stats on this. Out of the hundred or so Indies I know, most don’t use C++, unless you count using ‘other people’s C++’ i.e. Plugins etc. But hey if you’re part of a larger Indie team or the dedicated coder or use C++ daily at work etc, your reality is going to be different, I can respect that. So nothing against C++… I’m just pro-simplicity or anti-complexity which is what Blueprints offer. Why is that important? Well, shipping an Indie game is like completing a feature film. There are so many layers to get right yet so many pitfalls waiting to trip you up. So the simpler you can keep the tools the better. Plus, when it suits Epic they like to sell the idea that Blueprints* are all an Indie needs to make games*. :stuck_out_tongue:

And all of those sound great… I’m just asking for a few more complete sample projects like we had in UDK etc…

Lost me there… Which makes sense… More Docs with Anim-Gifs or BP projects tend to have greater longevity?

I wonder how much of that still works anymore though, for example after the recent online subsystem changes etc…
But its a good example of a demo that could be a flagship-project if updated often or every single engine version.
But what happens is, there’s no f/t staff assigned to this. So things rot and multiplayer is just too important for that.
For example the promises made in this thread, whatever happened there? Just left to die and be forgotten about…
If there was even just one flaghip Blueprint-based complete multiplayer game demo, it might get regular updates. :wink:

Sorry, should have been clear… your #2.