A plea for direction.

BobbyBonsaimind
Member
 
Posts: 97
Joined: Tue Apr 14, 2015 19:32
GitHub: RobertZenz
IRC: Robert_Zenz
In-game: Bobby

A plea for direction.

by BobbyBonsaimind » Wed Feb 10, 2016 22:52

I've started working with Minetest over a year and a half ago. It took me quite some time to get up to speed with Lua and Minetest, but roughly a year ago I released the first code into the wild and have been constantly improving it. I've also made a few pull requests and tickets, but have been reading through most of them. I am a gamer, but have not really played in Minetest, I'm mostly using the Lua API to spring my ideas to life. So that is where I'm sitting and from where I see the project, from somewhere behind the Lua API.

What do I see when I look at it from the Lua API? An awesome project which could itself position with little effort to be the FLOSS voxelgame engine. It already is, basically, because there is no other project like Minetest, there is no other engine that allows you to get so easily up to speed and start extending and building voxelgames. I say that again, I looked hard, there is no project like this! Minetest sits alone in this category and no other project seems to even try to provide something similar.

What do I see when I look at it from the community? A lot of people that create, this is awesome. You can never have enough people churning out content, because the more people are involved, the more ideas are brought up, the more innovation happens, the more awesomeness happens.

What do I see when I look at it from the GitHub page, the IRC channel? Chaos. Just chaos. Pull requests are sitting there for no apparent reason, arguments lead to no end and forks are springing up.

There are many pull requests which just sit there, for no reason, without response or action. Why is this bad? Because these were made by someone, sometimes from outside the project, someone who took the time to read through the code and improve it. Ignoring such work is worse than simply throwing it away. Denying and closing a pull request leads to anger in the worst case, ignoring a pull request creates frustration. The contributor does not know what is going on, if there is something wrong, and sooner or later simply gives up because nobody seems to be interested in the contribution.

Now on to forks, is a fork bad? No, not at all, not if there is a good reason to fork. But in the last time quite some forks have sprung up that are either hostile or completely unnecessary because they incorporate only small changes that sit unanswered in the pull requests section. Such forks are bad, not in mindset, but in terms of the community. They create fragmentation which is not necessary, avoidable and ultimately bad for the community. Down the road you'll end up wanting two features which reside in two different forks, then you'll have to choose, or worse, create your own fork incorporating both changes. All this leads to a lot of work for the fork maintainer, work that could be avoided or at least would be better applied to Minetest itself.

Now on to the tickets and features. From reading many tickets and pull requests I realized one thing: Different developers of Minetest have different goals for the project, sometimes even contradicting each other. So what gets added to Minetest and what does not is a game of who has the longer breath and can rally more support of other core developers behind them. Some areas seem to be a constant uphill battle for the developers, that also leads to frustration sooner or later.

Now, why am I so eager to have Minetest a generic voxelgame engine instead of a modable game? Because there is need for it. Everytime I take a look at the Minecraft community I see need for a FLOSS alternative that is easier to mod (Minecraft might have a lot of mods, but it is also absolute shit when it comes to modding) and there is no other contender in this area, I looked, there is none. The next best things are a few extensions to proprietary game engines which are not even close to the easiness and feature richness that Minetest provides. Minetest can easily and with little effort fill this niche, all it needs is a little push in the right direction.

I'm not saying that we should target the Minecraft community, fuck no, let us stay away from that toxic stew as far as we can, but I bet there is a "small" group within the Minecraft modding community which would jump ship within the blink of an eye if given the chance.

Of course there is an egoistic angle to it, I've spend over a year working on something that I'd hate to flush down the drain. And the only way to make sure that Minetest is a good, solid base is to make sure that it embraces its strengths as an engine.

What am I proposing and requesting?

Establishment of a clear mission statement: A clear mission statement will help reduce friction between core developers and also between core developers and contributors. It will also give a clear direction to Minetest, a better public representation and it serves as a goal which the whole team can work towards, together.

Establishment of a better organization: There is currently no way to solve disputes or someone who makes sure that certain things receives attention. There does not seem to be anyone who has the whole project in mind.

Improvement of the current handling of pull requests: Currently a pull request needs to have the approval of two core developers, but there is no list showing who is a core developer and who is not. There is no indication without checking the comments if it has one or two votes for it. To improve this situation I suggest an extension to the way how pull requests are handled:

* There should be a publicly visible team of the core developers.
* The core developers which feel responsible for the pull request (meaning, it falls in their area of expertise) should assign it to themselves. This allows the contributor to see who is the contact for any further questions and also allows the core developer to easily see what needs attention, as they can query their assigned pull requests.
* A pull request which gets the first approval is tagged with the "reviewed" tag, to indicate that the request has been reviewed by one core developer.
* A pull request which gets the second approval is tagged with the "approved" tag, to indicate that it can be merged at any time.
* Optionally, if a developer wants to merge an "approved" pull request, it should be tagged "merging", to make clear that somebody is now merging it, as that removes the need to inform other developers about the merge in IRC.

This would dramatically improve the feedback a contributor gets to their pull request. Additionally it allows the core developers to easily query pull requests by these tags and see which need a second review, which just need to be merged or what needs attention.

Working through the backlog is equally important, because there are currently so many pull requests that it is easy that some might get lost in between. Such a big backlog should be avoided in the future, even though that means quite tedious work to keep the list clean.

I am aware that all of this is a lot of work and we're all here on our spare time, but I believe that this is a good start and a good way forward.
 

User avatar
benrob0329
Member
 
Posts: 1192
Joined: Thu Aug 06, 2015 22:39
GitHub: Benrob0329
In-game: benrob03

Re: A plea for direction.

by benrob0329 » Thu Feb 11, 2016 00:47

Agreed.
 

User avatar
DatJohnDoe
Member
 
Posts: 22
Joined: Mon Feb 01, 2016 16:46
IRC: DatJohnDoe
In-game: DatJohnDoe

Re: A plea for direction.

by DatJohnDoe » Thu Feb 11, 2016 03:16

0.0 My Gawd Much Awesomeness, Very True
Do or do not, Therw is no try - Yoda
Luke... I am your Father - Darth Vader
 

User avatar
Evergreen
Member
 
Posts: 2131
Joined: Sun Jan 06, 2013 01:22
GitHub: 4Evergreen4
IRC: EvergreenTree
In-game: Evergreen

Re: A plea for direction.

by Evergreen » Thu Feb 11, 2016 03:22

I couldn't agree more.

I don't see any reason to make minetest only playable as a minecraft-like game since that already exists. Of course you can make that type of game for minetest if it were to become such an engine, since there is clearly a demand for it (including me), it is just that I don't want to be restricted to making a game with minecraft-like mechanics. For example, see what happened when Kenney tried to make such a game for minetest.
"Help! I searched for a mod but I couldn't find it!"
http://krock-works.16mb.com/MTstuff/modSearch.php
 

paramat
Member
 
Posts: 2662
Joined: Sun Oct 28, 2012 00:05
GitHub: paramat

Re: A plea for direction.

by paramat » Fri Feb 12, 2016 02:07

> There are many pull requests which just sit there, for no reason, without response or action.
> Ignoring such work is worse than simply throwing it away.

These are common complaints but there are reasons. We are very understaffed but there are lots of contributors. PRs are always looked at, never ignored, if there is no comment added it may be a dev is neutral or has nothing to add. Often a PR does not have enough approval to be merged but also not enough disapproval to be closed, so it is kept open to continue to be seen and considered, this is kinder than rushing to close it.

> quite some forks have sprung up that are either hostile or completely unnecessary because they incorporate only small changes that sit unanswered in the pull requests section.

I'm only aware of MinetestHD, which was forked after much discussion and argument on certain PRs.

> Different developers of Minetest have different goals for the project, sometimes even contradicting each other.

That's unavoidable with more than 1 dev.

We already have Celeron55's 'roadmap' that started as a post in his blog and was continued as a forum thread:
http://c55.me/blog/?p=1491
https://forum.minetest.net/viewtopic.php?f=7&t=9177
There's some discussion with Celeron55 here https://github.com/minetest/minetest/issues/3476

The lack of a more specific direction is not necessarily a bad thing. sometimes a system that evolves organically in tiny steps with no purpose can become beautiful, such as nature. Minetest is by Celeron55's own definition 'chaos development' by 'a random bunch of lunatics' but seems to be very popular, it was recently condsidered (perhaps wrongly) the best open source game by a major open source website.

> There is currently no way to solve disputes or someone who makes sure that certain things receives attention. There does not seem to be anyone who has the whole project in mind.
>Working through the backlog is equally important.

Until recently Celeron55 was defined as the person who solved major disputes, he is still considered the person to do this. Occasionally he steps in and stops certain PRs. When active hmmmm is very much an organiser. Sometimes we have a PR reviewing session to go through them all, although not so often recently due to being very understaffed.

> there is no list showing who is a core developer and who is not.

https://github.com/orgs/minetest/people Except Calinou and Rubenwardy who are 'trusted non core devs'.

The rest of your suggestions for handling PRs are already done in a laid-back informal way. We had a more structured system a while back but we relaxed it.

Overall you're being rather negative and have the wrong impression of a few things.
 

User avatar
jp
Member
 
Posts: 705
Joined: Wed Dec 18, 2013 09:03
GitHub: kilbith

Re: A plea for direction.

by jp » Fri Feb 12, 2016 09:42

Not to mention we have to subject our PRs to an abstract divinity called "The Spirit" who determines the right way forward.
 

User avatar
prof-turbo
Member
 
Posts: 518
Joined: Mon May 07, 2012 17:02

Re: A plea for direction.

by prof-turbo » Sun Feb 14, 2016 00:15

PRs are always looked at, never ignored

That's right, they're simply forgotten.

I totally agree with you BobbyBonsaimind. And that's one of the reasons that made Freeminer appear.
If this system was functionnal at this moment it would be much better IMO.
 

BobbyBonsaimind
Member
 
Posts: 97
Joined: Tue Apr 14, 2015 19:32
GitHub: RobertZenz
IRC: Robert_Zenz
In-game: Bobby

Re: A plea for direction.

by BobbyBonsaimind » Tue Feb 16, 2016 18:03

I'd also ask everyone who is interested in this, to state their opinions, ideas and suggestions in #3476.
 

User avatar
xeranas
Member
 
Posts: 99
Joined: Fri Feb 05, 2016 11:06
GitHub: xeranas

Re: A plea for direction.

by xeranas » Tue Feb 16, 2016 20:10

paramat wrote:Until recently Celeron55 was defined as the person who solved major disputes, he is still considered the person to do this. Occasionally he steps in and stops certain PRs. When active hmmmm is very much an organiser.


So who currently responsible for development guidence?

paramat wrote:We are very understaffed

Then focus and clear direction is needed more than ever.

BobbyBonsaimind wrote:I'd also ask everyone who is interested in this, to state their opinions, ideas and suggestions in #3476.

Who will be responsible for reviewing these suggestions and validating/invalidating them? According to Minetest Game development direction one of core developers says that thread is a mess and nobody reads them anyway [1]. Then why bother?
 

BobbyBonsaimind
Member
 
Posts: 97
Joined: Tue Apr 14, 2015 19:32
GitHub: RobertZenz
IRC: Robert_Zenz
In-game: Bobby

Re: A plea for direction.

by BobbyBonsaimind » Tue Feb 16, 2016 22:04

> Then why bother?

I'll be damned if I didn't at least try. I have very little hope, though.
 

BobbyBonsaimind
Member
 
Posts: 97
Joined: Tue Apr 14, 2015 19:32
GitHub: RobertZenz
IRC: Robert_Zenz
In-game: Bobby

Re: A plea for direction.

by BobbyBonsaimind » Tue Feb 16, 2016 22:17

> We are very understaffed...

In my opinion, that is a symptom of the problems I've tried to outline.

> PRs are always looked at, never ignored, if there is no comment added it may be a dev is neutral or has nothing to add.

#1438 #1439 #1489 #1562 #1645 #1699 That's just from page 6, I don't see a reason why any of these need to linger around in the state they are.

> Often a PR does not have enough approval to be merged but also not enough disapproval to be closed, so it is kept open to continue to be seen and considered, this is kinder than rushing to close it.

I never said that anything that does not accumulate enough approval within a certain time should be closed. What I did try to say is that these need to be looked at on a regular basis, and not just having them lay around until they are dead. On the other side, if there would be a clear goal, actually closing pull requests which do not fit would be an option.

> I'm only aware of MinetestHD, which was forked after much discussion and argument on certain PRs.

There is also BlockPlanet and FreeMiner (not so recent, though).

> That's unavoidable with more than 1 dev.

Damn no!

> We already have Celeron55's 'roadmap' ...

Yeah, that's a TODO list.

> The lack of a more specific direction is not necessarily a bad thing.

I thought I made clear that that is a problem.

> sometimes a system that evolves organically in tiny steps with no purpose can become beautiful, such as nature.

Evolution requires by design that something better comes forward, continually. In nature, everything that is not better than previous iterations will eventually wither and die. Also, it is less chaos than one wants to believe (there are quite a few underlying rules, like that the best systems for distribution of water, warmth or energy wins).

> The rest of your suggestions for handling PRs are already done in a laid-back informal way.

So take it as proposal to do them not in a "laid back" but explicit way.

> Overall you're being rather negative and have the wrong impression of a few things.

I'm frustrated, that's a difference.
 

celeron55
Member
 
Posts: 430
Joined: Tue Apr 19, 2011 10:10

Re: A plea for direction.

by celeron55 » Thu Feb 18, 2016 12:40

I now took a new look at the current arguments about this overarching issue, which I agree is an issue.

I don't feel like there is anything in particular that I could do outside of personally starting to handle all PRs. There is no mission statement in this world that would help those PRs RobertZenz just linked.

The issue seems to be that people who can do stuff just aren't motivated to spend the time to do stuff. That certainly is the case for me, and I have no reason to assume it's any different for others.

However, I certainly would be interested to see RobertZenz propose a mission statement, along with some kind of a hypothesis on how the particular mission statement would change the status quo. RobertZenz: You have more power than you think; you just need to use it.
 

User avatar
Casimir
Member
 
Posts: 1101
Joined: Fri Aug 03, 2012 16:59

Re: A plea for direction.

by Casimir » Thu Feb 18, 2016 17:07

There are actually a lot of things that come together and give the impression of "chaos". But those are several small problems, development is not as broken as it seems.

I don't have that much insight in the development, but those are some things I noticed. For each point I also propose a solution, but that's just the first thought that came to my mind.

Reported issues are piling up and there it does not look like it will stop. This has several reasons.

1. What actually counts as issue? Minetest is a game where everything is possible, logically people have an overabundance of ideas of what to do with the game. Most of those ideas are great, but it also takes someone to implement them. And as soon as something is added to the game a bunch of followup ideas is growing on that base.
If you look at the feature requests on github, there are many things that are just ideas on what to add. And several people agree that it is a good idea. Then people forget about it and it stays an issue on the list. As I said, implementing every each and on of them would only create more possibilities for more ideas. At the same time we can not close them because they are relevant suggestion and everyone wants them.
The proposed solution: Rethinking what the issue tracker is for. With so many issues we don't have a clear view of what problems to address. For my projects (not only coding but other things as well) I usually order the issues in three categories: Bugs, Things to Improve and New Features. You can further differ those. New Features can be divided into things that are essential for the roadmap and things that are nice to have. If you are a creative person, the last one has the most entries, but those distract from the roadmap, and from the bug to be fixed and improvements to be made. So I propose to make a list, e.g. in the wiki, to collect those good ideas, but to have the things we would like to have separated from the things that have to be addressed (e.g. bugs). The issues could then be closed and reopened when there is someone who likes to implement this.

2. Of course someone would have to go trough those issues and do the work of writing it into the wiki and closing those. This leads to another thing I noticed: Very few developers feel like closing issues. Often there is a vague agreement that this will not be added or that it is not a bug. But then nobody wants to do the decision of closing it up. I guess this is to enable other to continue discussion. Or of not doing a one sided decision.
A proposed solution would be to give out "jobs". E.g. One dev is selected (maybe randomly) to care about some kind of issues. A janitor so to say. So there could be one job to care about the feature idea wiki mentioned above. That person then could close those issues without having to ask for full consensus before. You can still complain afterwards and reopen things. In the same way there could be one person assigned to bike shed discussion, and one to confirm bugs, and one to close dublicates, and so on. And it would rotate every month or so. The idea is that those people then feel responsible for those things and have it easier to step up and make a decision.

3. Another thing I noticed is that many issues are discussed on irc. But the result of the discussion never get out onto the github. It would help a lot if in irc when one feels like a decision is reached someone just asks: "So we agree that [something]?" and then goes to the issue and writes just a few sentences. "We discussed this on irc and the consens is that [something]".

There is also one general problem. When you have some part of code where there are some bugs hiding and you add a feature on top of the existing code. Then it can happen that those bugs actually get fossilized, or become part of the new code. Fixing those bugs then gets much harder, because 1. you have to look through more code and 2. the new code might depend on those bugs. So one thing I do in my project is, to make sure that the ground on which I build is solid. If I have bugs in some area where I want to add a new feature, I always try to fix those first. And in extreme cases I won't add that feature then. It could be made a requirement for big pull requests, that first one has to look at the existing bugs and check if any of those have to do with the new code. But I also understand that I am a very rigorous person in that regard and that not everyone likes to work this way.
Last edited by Casimir on Thu Feb 18, 2016 17:20, edited 1 time in total.
 

User avatar
Gael de Sailly
Member
 
Posts: 475
Joined: Sun Jan 26, 2014 17:01
GitHub: Gael-de-Sailly
IRC: Gael-de-Sailly
In-game: Gael-de-Sailly

Re: A plea for direction.

by Gael de Sailly » Thu Feb 18, 2016 17:13

I completely agree with your first post, Bobby. Many people think like you, but in fact no one has really said it before you.

BobbyBonsaimind wrote:What do I see when I look at it from the GitHub page, the IRC channel? Chaos. Just chaos. Pull requests are sitting there for no apparent reason, arguments lead to no end and forks are springing up.

There are many pull requests which just sit there, for no reason, without response or action. Why is this bad? Because these were made by someone, sometimes from outside the project, someone who took the time to read through the code and improve it. Ignoring such work is worse than simply throwing it away. Denying and closing a pull request leads to anger in the worst case, ignoring a pull request creates frustration. The contributor does not know what is going on, if there is something wrong, and sooner or later simply gives up because nobody seems to be interested in the contribution.


The devs can't be everywhere. There is over 2 PRs per day in average. But, with a little organization, they could be more efficient.

BobbyBonsaimind wrote: * The core developers which feel responsible for the pull request (meaning, it falls in their area of expertise) should assign it to themselves. This allows the contributor to see who is the contact for any further questions and also allows the core developer to easily see what needs attention, as they can query their assigned pull requests.
* A pull request which gets the first approval is tagged with the "reviewed" tag, to indicate that the request has been reviewed by one core developer.
* A pull request which gets the second approval is tagged with the "approved" tag, to indicate that it can be merged at any time.
* Optionally, if a developer wants to merge an "approved" pull request, it should be tagged "merging", to make clear that somebody is now merging it, as that removes the need to inform other developers about the merge in IRC.

On MinetestForFun's server source code, issues and pull requests are assigned to someone that will fix them. Look at how we use it. We can also filter per assigned people. After discovering this feature, I've been surprised to see that minetest doesn't use it. It should. That's really not difficult to use: 5 seconds and 2 clicks to be assigned to a PR. And it could help a lot.
Using tags to indicate whether a PR has been reviewed is a good idea too.

celeron55 wrote:The issue seems to be that people who can do stuff just aren't motivated to spend the time to do stuff. That certainly is the case for me, and I have no reason to assume it's any different for others.

Yes, but reviewing the PR does not take as much time as coding it, in many cases. People that code and submit pull request are motivated. But I can understand that the developers have also their life and can't/do not want to do stuff. The only solution I see is to invite more people to the dev team. But that may be more complicated to manage, and I guess that's also a matter of trust. If we let too many people modify whatever they want in Minetest, that will really be the chaos.
Very busy this year too, so do not expect me to be very active on the forum or in game. But I'm not about to drop Minetest forever :)
 

BobbyBonsaimind
Member
 
Posts: 97
Joined: Tue Apr 14, 2015 19:32
GitHub: RobertZenz
IRC: Robert_Zenz
In-game: Bobby

Re: A plea for direction.

by BobbyBonsaimind » Fri Feb 19, 2016 19:59

> However, I certainly would be interested to see RobertZenz propose a mission statement, along with some kind of a hypothesis on how the particular mission statement would change the status quo.

When I wrote #3476 I thought that would be enough, yes. In the mean time I've come to the conclusion that it is not, hence the addition of how to change the handling of pull requests. If you'd be really interested, I believe that I come up with a definitive plan for this all, yes. Will take some time, though, as I'm a little bit short on time over the next week and a half, and such a thing will need time and input from others.

> RobertZenz: You have more power than you think; you just need to use it.

I don't know what you mean.
 

User avatar
Gael de Sailly
Member
 
Posts: 475
Joined: Sun Jan 26, 2014 17:01
GitHub: Gael-de-Sailly
IRC: Gael-de-Sailly
In-game: Gael-de-Sailly

Re: A plea for direction.

by Gael de Sailly » Fri Feb 19, 2016 21:10

BobbyBonsaimind wrote:
celeron55 wrote:RobertZenz: You have more power than you think; you just need to use it.

I don't know what you mean.

I don't know either. For me you're one of the few Minetest players that really use their power.
Very busy this year too, so do not expect me to be very active on the forum or in game. But I'm not about to drop Minetest forever :)
 

celeron55
Member
 
Posts: 430
Joined: Tue Apr 19, 2011 10:10

Re: A plea for direction.

by celeron55 » Sun Feb 21, 2016 16:54

Gael de Sailly wrote:
BobbyBonsaimind wrote:
celeron55 wrote:RobertZenz: You have more power than you think; you just need to use it.

I don't know what you mean.

I don't know either. For me you're one of the few Minetest players that really use their power.


Well, you are using it. I said that more from a standpoint of being concerned that you would stop.
 

cheapie
Member
 
Posts: 304
Joined: Mon May 14, 2012 00:59
GitHub: cheapie
IRC: cheapie
In-game: cheapie

Re: A plea for direction.

by cheapie » Sun Feb 21, 2016 20:37

The problems I have with the project and related things from my perspective as a more-or-less end-user are slightly different.

The biggest one is currently the whole "let's add this, this, and this, and who cares about performance" thing. The experience of playing on VanessaE's creative server (my favorite one) is a great place to see this. My computer is not at all slow (FX-9590 and Radeon R9 270X), but the incredibly vague "Initializing nodes" step takes on the order of 30 seconds. Enabling the minimap makes it take over twice as long. Once that finally finishes, I'll often head over to one of the more urban areas to be greeted with something on the order of 20 FPS at a view range of only 100. For a game with graphics as simple as Minetest, that's something that should be getting 75-100 FPS on this computer. This is part of why I barely play MT any more - it runs so badly on any computer I try it on.

Unfortunately, this brings me to the second one. Most attempts to contact anybody, submit a bug, etc. feel about as futile as communicating with a large company. There might be a reply at first if I'm lucky, but it quickly turns to nothing as none of the devs usually seem interested in communicating at all.

...and in regards to that performance aspect earlier, completely nonsensical decisions like this one (I'd call them much worse things if the devs couldn't hear me, by the way) almost seem to point towards MT running fine on the devs' computers but nobody else's, and of course, nobody cares.
 

User avatar
rubenwardy
Member
 
Posts: 4500
Joined: Tue Jun 12, 2012 18:11
GitHub: rubenwardy
IRC: rubenwardy
In-game: rubenwardy

Re: A plea for direction.

by rubenwardy » Sun Feb 21, 2016 20:58

cheapie wrote:The biggest one is currently the whole "let's add this, this, and this, and who cares about performance" thing. The experience of playing on VanessaE's creative server (my favorite one) is a great place to see this. My computer is not at all slow (FX-9590 and Radeon R9 270X), but the incredibly vague "Initializing nodes" step takes on the order of 30 seconds. Enabling the minimap makes it take over twice as long.


Dreambuilder? Homedecor is massive, let alone dreambuilder. That's not the Minetest project as such, that's a third party subgame / mod set.

cheapie wrote:Once that finally finishes, I'll often head over to one of the more urban areas to be greeted with something on the order of 20 FPS at a view range of only 100.


Sounds like the meta data program, where meta data is repeatedly resent causing massive FPS drops. There's a PR in progress.

cheapie wrote:Unfortunately, this brings me to the second one. Most attempts to contact anybody, submit a bug, etc. feel about as futile as communicating with a large company. There might be a reply at first if I'm lucky, but it quickly turns to nothing as none of the devs usually seem interested in communicating at all.


I disagree, people usually respond to me. How have you tried contacting people? A combination of IRC and Github, the problem is that people may not know about what you're asking, so don't respond.

cheapie wrote:...and in regards to that performance aspect earlier, completely nonsensical decisions like this one (I'd call them much worse things if the devs couldn't hear me, by the way) almost seem to point towards MT running fine on the devs' computers but nobody else's, and of course, nobody cares.


A fixed view distance causes quite a big performance increase, apparently.
 

User avatar
AnxiousInfusion
Member
 
Posts: 146
Joined: Sun Aug 02, 2015 05:43
GitHub: AnxiousInfusion[GitLab]
IRC: AnxiousInfusion
In-game: AnxiousInfusion

Re: A plea for direction.

by AnxiousInfusion » Tue Feb 23, 2016 03:01

Mostly related, I find it extremely frustrating that contributions by RealBadAngel are shot down so often. He's one of the few still creating new features that users will actually notice and so what if FPS is reduced by some insignificant percentage? It's not like we're all runing 386 with 3DFX as evidenced by my hardware polling thread.
 

cheapie
Member
 
Posts: 304
Joined: Mon May 14, 2012 00:59
GitHub: cheapie
IRC: cheapie
In-game: cheapie

Re: A plea for direction.

by cheapie » Tue Feb 23, 2016 03:22

AnxiousInfusion wrote:so what if FPS is reduced by some insignificant percentage?

It's already too low to begin with. Any drop from that is too much.

AnxiousInfusion wrote:It's not like we're all runing 386 with 3DFX

Sure as h*** runs like it most of the time.
 

User avatar
benrob0329
Member
 
Posts: 1192
Joined: Thu Aug 06, 2015 22:39
GitHub: Benrob0329
In-game: benrob03

Re: A plea for direction.

by benrob0329 » Tue Feb 23, 2016 14:03

But now that we have a split renderer, more shaders can get added without effecting older/slower hardware.

So I'm hoping *cough*expecting*cough* that at least some of RBAs pull requests will start to be commited.
 

BobbyBonsaimind
Member
 
Posts: 97
Joined: Tue Apr 14, 2015 19:32
GitHub: RobertZenz
IRC: Robert_Zenz
In-game: Bobby

Re: A plea for direction.

by BobbyBonsaimind » Tue Feb 23, 2016 21:07

I've started working on the requested documents, however they are in a very early stage. I've just committed the first draft of the mission statement, I'm not quite happy with the others right now, so I will keep thinking them over.

https://github.com/RobertZenz/minetest- ... -statement

Feedback is obviously welcome.
 

User avatar
rubenwardy
Member
 
Posts: 4500
Joined: Tue Jun 12, 2012 18:11
GitHub: rubenwardy
IRC: rubenwardy
In-game: rubenwardy

Re: A plea for direction.

by rubenwardy » Tue Feb 23, 2016 21:14

 

BobbyBonsaimind
Member
 
Posts: 97
Joined: Tue Apr 14, 2015 19:32
GitHub: RobertZenz
IRC: Robert_Zenz
In-game: Bobby

Re: A plea for direction.

by BobbyBonsaimind » Thu Feb 25, 2016 23:17

I've just committed all drafts with my ideas. That means, theoretically the repository is now holding all the ideas I have on how to improve the current situation. But I'm afraid that it actually boils down to "people need to get stuff done and they need to do this in a more organized way".

Also, this whole matter has started to occupy my thoughts lately, up to the point that it becomes distracting and annoying. So I will say this only once: I've told you my opinion twice, I've told you what I'd like to see changed and what I believe is a good change and now I also got that all into a (hopefully) nice format. I've started to stop caring if you act on it or not, I will not argue about any of this any longer. You have here all I can provide you with on my own, I'm open for discussing improvements and even providing a helping hand.

I guess there is no nice way to put this nor a way which makes me not look like a damn self-absorbed martyr, but to get this straight, I will not chase after you for any of this.
 

PoignardAzur
Member
 
Posts: 30
Joined: Fri Jun 19, 2015 17:23

Re: A plea for direction.

by PoignardAzur » Fri Feb 26, 2016 11:20

I agree with everything you said. As far as I am aware, Minetest is an horrible failure at any decent goal I could imagine for it. If it really is the best open-source game there is to find, then the open-source game making community sucks (as far as I'm aware, this isn't true).

I think the best thing that could happen to this game is serious competition. Right now people who want a good voxel game play minecraft. The people who play minetest want special things only it can offer (free access, easy modding, distance from Microsoft, high performance for slow PCs, etc), and as long at it will be the only game providing those, it will keep stagnating.

Again, if a better game is to be found, then minetests needs to have competitors. Not some minetest clone with two (or even two thousands) lines of code changed and a few dozens of items added for "realism". No, minetest needs serious competitors made by competent people who are ready to start from scratch, provide alternative experiences (because you can only do so many variations on "dwarf-fortress-style survival" before people stop caring) and create their own, finished product.

Needless to say, I intend to do exactly that as soon as I have the time and the resources for it. I expect to start working on it in the following months, and to produce my first results before the end of summer. I expect I will fail. But I'm still going to fucking try.
 

User avatar
jp
Member
 
Posts: 705
Joined: Wed Dec 18, 2013 09:03
GitHub: kilbith

Re: A plea for direction.

by jp » Fri Feb 26, 2016 12:41

This is a pattern in any FOSS game. Yes, the organization sucks.

There's no cohesive team aiming at the same goals and communicating full-time at the same place like at Mojang. Minetest core-team is (mostly was) like a can of crabs thinking to pinch at each other due to conflictual visions. We also suffer alot from a poor communication due to different timezones, lack of availabilities, absence of real meetings to get better consolidated, and the sole chatty place full of interferences that is IRC for conducting the devel.

The team itself is not as good as professionals from gaming industry as well. We're managed by intermittent hobbyists with very personal tasting, post-teenagers looking for learn or hardcore programmers who aren't even gamers and hate playing. Those people aren't paid and naturally they doesn't feel very responsible, get bored or afflicted by a burn-out sooner or later.

Minetest missed and keep missing opportunities to be funded by public or private entities to ensure a better care on the long run. Nonetheless things keep progressing, and are not likely to definitely stop in the near future. But you shouldn't ever theorize too much or thinking to "professionalize" our organization and leave it be a chaos momentarily ordered by a random bunch of lunatics ;)
 

User avatar
prof-turbo
Member
 
Posts: 518
Joined: Mon May 07, 2012 17:02

Re: A plea for direction.

by prof-turbo » Fri Feb 26, 2016 23:29

No, minetest needs serious competitors made by competent people who are ready to start from scratch, provide alternative experiences (because you can only do so many variations on "dwarf-fortress-style survival" before people stop caring) and create their own, finished product.


This is why this was created. If you want to help, don't hesitate :) https://github.com/ElementW/Diggler
 

celeron55
Member
 
Posts: 430
Joined: Tue Apr 19, 2011 10:10

Re: A plea for direction.

by celeron55 » Mon Feb 29, 2016 21:49

That's a whole lot of stuff; I wasn't really expecting such a bunch of documents - I guess time will tell if this is useful or not.
 

User avatar
TumeniNodes
Member
 
Posts: 1335
Joined: Fri Feb 26, 2016 19:49
GitHub: TumeniNodes

Re: A plea for direction.

by TumeniNodes » Mon Feb 29, 2016 22:54

I realize I am very new here, so hope my input is not over-stepping. One option to try (if others agree) may be to contact Canonical (the structural and financial backbone of Ubuntu). Maybe they would be interested in some small involvement. I say this because Minetest is a popular Linux game and also because it involves having good software / hardware support. Some aspects of Minetest put graphics cards etc to the test.
There are also a few gaming groups / companies on the Linux platform which are climbing the ranks in the industry which has AMD and others looking at it more seriously than ever.
There are numerous options and possibilities to address these issues and concerns. One of the biggest means of support in the linux world is donations. They really add up, especially when the product is quality. And there is always an endless source of people willing to help, etc..
I think this (Minetest) is a quality product. One can speak for Mojang but, they also took a product someone created and made available to the public for free, made some minor tweaks then slapped a copyright on it (the microsoft / gates business plan) and are making a killing. And I have to be honest, Minetest smokes Minecraft in quality, especially considering the wide difference in financial support.
I was part of a dev team for a popular Linux OS and I have seen fighting first hand. In the end, we all left (most of us), and that project still lived on and the fork failed because, well..., it failed. regardless of all the talent involved. Just very little organization, we had one dev who was business savy who did his best to keep it all cohesive, but one against many proved fatal to our project. I left because it turned into a strictly core product with very little need for graphics.
I got off point here but as I say, this is definitely not a hopeless project, so don't fret. Bickering happens, it is unavoidable. Sometimes some stiff rules need to be applied to development procedures and on devs (even contributors who are not supporting nor updating their work which has been committed) and then the consequences need to be applied if rules, requirements are not met after some soft warnings
Flick?... Flick who?
 

Next

Return to Minetest General

Who is online

Users browsing this forum: Bing [Bot] and 35 guests

cron