Miscellaneous defects in the engine

  • #1, by gregor-schweitzerTuesday, 09. June 2020, 18:47 4 years ago
    High guys,

    this is gonna be a long shot, so sry in advance for your time but here I go.

    I'm working with a very small team on a point-and-click adventure situated in lovecratian town Innsmouth. It's supposed to be a commercial game with tons of backgrounds, charming characters, video Sequences, and a thrilling story. 
    (pics attached to give you some impression) 

    As you can probably see we are not ungifted but when it comes to working with the engine, we are always stumbling over the most grotesque errors.

    I will start with two... well... questions, from which one might be easier to solve the other one is more of the philosophical kind. 

    1.) Recently I was experimenting with the sound system... Well, it looks a bit tricky and seems a lot inconvenient than something rudimentary as sound must be, but since I'm quite familiar with that sector I was able to handle it. Still don't know whats the sense or difference in sound containers and sound busses, but maybe one of you knows!?
    NOW comes the actual question. When I was working on somethin totally different in the engine and tested several times in the editor everything worked normal and as it was supposed to. Half an hour later when I show my progress to my partner the speech sound of several characters is overdriving and way too loud. Best thing for the sound of one character for one line normal and the next line overshoot in volume. Just for the protocol: I didn't changed in between that NOTHING that has remotely to do with the sound system.

    2.) Now comes the tricky philosophical question:
    There are so many  weak points and and instabilities in the engine and of course we now that the originators of the engine are sure putting a lot of efford to make it more stable. Anyways, after every update a lot of differnt things DON'T seem to work anymore and the way before the update was done. We experienced that now several times. 
    Now here it comes. How do the creators of the engine expect that a project of 200 BGs can be finished, if after EVERY update I have to check EVERY single background through test simulation and than have to check back with pro's here in the panel or consult an external professional. That is humanly impossible. Does anyone have similar struggles throughout the game developing process or knows what the creators of the engine are thinking when releasing bad updates? I mean so many of the failures after an update are on so rudementary levels... Does anybody has an idea? We would really care to know!

    With kind regards 
    The GM Factory


    7 Posts

  • #2, by MateuszWednesday, 10. June 2020, 16:54 4 years ago
    A for point two - it's true, every new incarnation introduces it's own set of new bugs and sometimes breaks things that were working earlier. Based on our experiences, if you find version of engine that works for you, STICK TO IT and forget about updates if you care about your own sanity.  Yes, you may miss some new, useful features, but you won't miss long hours of guesswork "whats broken now".


    59 Posts

  • #3, by MachtnixWednesday, 10. June 2020, 18:17 4 years ago
    But you can't go back - usually. You have to load your game into the new version, test it and hope it works. But you will recognize new bugs often after two days using it without problems - and your progress is gone... 
    That's why I use still 4.2.5 - I can jump to a newer version but never back... I'm not sure how the 5th versions work together - I did't try it.

    Btw: the game is already bought. smile It's great.

    Thread Captain

    1097 Posts

  • #4, by smakovecWednesday, 10. June 2020, 19:18 4 years ago
    First of all, your game looks great. Remind me of the first two Broken Sword games and I like that. Also, I am a big fan of Lovecraft so this looks like a excelent game to me.

    I have tried many different engines and I think Visionaire studio is easies and best adventure game engine on the net.
    But as for the updates. I was updating regularly. I know my game was so far bug-free as I have put a demo online and didn't have any complaints. So I continue to work on the game. After a  few updates and a month of development, I started to play a game from start and I have found so many bugs: conditions and variables missing, animations not working properly, actions not working as before.... I took my two months to check and fix everything (and at that time I would get so frustrated and almost decided not to continue with the further development.)
    Then I looked back at updates and saw that I don't have any benefits from them, but only problems with bugs. The new audio system is implemented for some time now and there is still no easy to understand tutorial. There is one example and some explanation on wiki page but that didn't help me at all.
    I was very excited about new visual scripting language expecting something easy to use as the one in the Gamemaker engine. I hope some people will benefit from Ilios because there was a lot of time and effort put into it, but I don't even know to start using it. Again there is no easy to understand tutorials.

    So my decision is: I am satisfied with the engine so far as it is and I will not update further until my game is finished.


    8 Posts

  • #5, by LebosteinWednesday, 10. June 2020, 22:24 4 years ago
    In my eyes Visionaire is dead. There was no further development during the last two years, not even bugs are fixed. Please make it Open Source! This would be a chance for it to be further developed and improved...

    Key Killer

    621 Posts

  • #6, by SimonSWednesday, 10. June 2020, 23:42 4 years ago
    These calls for Open Source annoy me. As if Open Source magicly transforms software to the best thing that exists. That is not the case. There need to be  people to keep the code quality, manage coders, make plans, acquire help, these are normally paid. But these are details, who cares for details. Second, there are no halfway capable programmers that would work on it if it was Open Source. Where would they come from ? People, who know how to do it, work on their own engines or go for the big engines. And they are normally paid. 

    Also I wrote around 100000 lines of code in around 7 years for this engine. I would need to document that and such, won't happen.

    As I'm currently the only active developer, things like "there was no development during the last two years" really get to me. Seems I wasted all my weekends for nothing and I could have just done nothing.

    I am not making money from this. I do this because I like doing it.

    So now for bugs. It's easy if I don't change anything I don't break anything. But then there is also no progress. 5.1 had a lot of changes to the core and I'm now always fixing changed behaviour when I'm alerted to it. Problem is, nobody tests the beta versions, noone gives me feedback and tells me here are bugs. All expect that to magically happen on its own. And then I get some sweeping things like "it's full of bugs", it's really frustrating for me. Often small changes have big impacts, that I can never foresee. 

    Of course I can understand that is frustrating when things don't work as they did before, but often the behaviour before was incorrect and I corrected it, then some things that built on the incorrect behaviour don't work anymore.

    This is currently a phase, that the engine goes through to modernize the code and increase flexibility. After the phase the updates will not change so many things anymore. For this time I need the cooperation of y'all to tell me of bugs and give me feedback.

    Thread Captain

    1578 Posts

  • #7, by MachtnixThursday, 11. June 2020, 03:55 4 years ago

    I am not making money from this. I do this because I like doing it.
    The problem is: Visionaire costs 500 Euro as a professional licence. Okay, the single licence is cheaper and it's worth. But it must be stable! My opinion is: there are too much big changes (3D characters, shaders, ...) at the same time. It's a good idea to make the core better and faster. The release 5 should start with a new engine, but with working features. If there are bugs in simple things like sound or position of animations or wrong layouts it's not easy to explain why. Visionaire is great to make 2D adventures. Not more not less. It's a tool. If I want to make an open world with fightimng, shadows, collisions, levels and lights I will use Unity or s.th. else. It's okay to sell a working simple engine to make simple adventures as simple as before. And now Visionaire is switching to an usually English thing with features I don't need like all the others and the unique advantage of a German program is gone. With only one (?!) developer Vis have to (in German: den Ball flach halten...). That's only a statement.

    Thread Captain

    1097 Posts

  • #8, by SimonSThursday, 11. June 2020, 09:07 4 years ago
    We sold 0 pro licenses last year. That's a license for people who really need support and got some money to get it and/or want console releases. You're paying for my support time with that license.

    It's stable. We are talking about small bugs, that nevertheless, I understand, can be frustrating, but then tell me and I fix it. After a release I'm fixing bugs for a week. Alert me to it and I fix it. But I need your cooperation. I can't magically make bugs disappear before they even appear. Unless someone has an example for more bugs ? And I often get bugs, that stem from things that the user themselves made wrong. "Bug: My sound vanish after 30s" - There's an action in the ved stopping them after 30s.

    3D characters got here in version 4.2. Shaders got here in version 4.2.

    It's not that I want this to be the next Unity, it's about increasing flexibility. Let's add an example, interfaces were not built hierarchical in the engine, now some want to apply rotation and scaling to it, if I now try to build that I create a big pile of bad code. So I changed the structure so that transform is part of every entitity. That's not about collision, fighting, shadows, levels. That's the main source of all problems currently.

    But I can't now choose to not do these features, because I can't take responsible for the huge pile of code, that will in fact introduce even more bugs and glitches. And you're probably not suggesting to not develop anymore ?

    If the suggestion was the rework everything for a 5 version and reset to 0, noone would have used that and it would not have been out yet.

    All parts except the visual scripting part are translated to German. And visual scripting is for programmers, consistent naming is key and that's English. Every time I'm in Excel and adding formulas it annoys me because it's not the normal programming style, but localized. So they change , to ; to allow numbers with , and also all functions are named differently... Makes a mess.

    Anyone have bugs here, before I start building the next update ?

    Thread Captain

    1578 Posts

  • #9, by gregor-schweitzerThursday, 11. June 2020, 11:58 4 years ago
    I think we get the difficulty of the picture and I'm rather the type of person that says, let's work with what we have and not get caught to long into things that should have been be better considered in the past. You say I pay for your support with the full commercial licence? I take you at your word, because that's the license I bought. And I'm willing to report from now on every failure that comes in my way. But my point stands ... You know better than anyone else that game development is a process that goes over years. So when you reach a certain point in progress you can't just say "let's do the update and dump all the progress we made" and then proof and test over hundreds of rooms over again. And from the actual point of view I have no chice bu to go with the better-dont't-update-strategy (using version 5.0.9 at the moment) unless I say screw hundreds of hours and weekends I put into that.


    7 Posts

  • #10, by afrlmeThursday, 11. June 2020, 12:32 4 years ago
    I think one of the main points here is that you really shouldn't immediately update as soon as a new version comes out, & if you do decide to update, then at least make sure you make a backup of your current ved file & test the new version on that thouroughly before continuing to work on it, because as I'm sure most of you that have any programming experience will know, if you change/fix something then you will often find that 2 new bugs appear in its place - it's kind of like a multiple headed hydra where you chop one of its heads off & 2 new ones grow in its place.

    Like I said, ALWAYS backup your ved before upgrading. You should also be able to install the upgraded version of VS into a new folder instead of overwriting the current build seeing as it downloads the installation filed, which means you can have the last version & current version installed, which means if you backed up your ved that you can carry on working from where you left off.


    7278 Posts

  • #11, by SimonSThursday, 11. June 2020, 12:57 4 years ago
    Write to me at discord, I help lots of people there and that takes me lots of time. I'm here to help as far as I can. And we had some conversations via mail before as far as I remember.
    If you're happy at 5.0.9, no reason to update. But I can't do fixes for that unfortunately. There are many people keeping it at 5.0.9.
    Are the bugs really that gamebreaking, that it means you dump your whole game by updating ? Could you give me some examples ?
    Oak Island is in the works for over 7 years, I know how it is and it's part of my testing, I also test Daedalic games (got around 7 with source code). But there are always cases I miss.

    The files can be downgraded and if you would loose much work in it, I'll be happy to do it for you.

    Thread Captain

    1578 Posts