Big Thoughts, Dangerous Thoughts

CuriousNoob
Member
 
Posts: 73
Joined: Wed Jan 27, 2016 09:43

Big Thoughts, Dangerous Thoughts

by CuriousNoob » Thu Feb 02, 2017 21:21

.

Hi All

A huge amount has been achieved, most of which is truly amazing. Past and present contributors deserve massive Gratitude and Respect. The community around these forums is a real asset. There's a lot of bright minds here and a huge amount of eager willingness to expend time to help.

IQ involves seeing the distinctive-essence of things and spotting patterns. I for one have started to notice things around the forums. Unfortunately there are some glaring issues in ''The Minetest Experience 2017.''

TumeniNodes posted interesting and clearly heartfelt comments about MT and the community.

That, on top of a recent post by tinoesroho giving his five-year look-back summary.

I also see quite a few current modding and other threads where the discussion is implicitly questioning the core design and implementation of both the engine and its default game.

Several also have been the cases which essentially said "Moving-On.. so Bye-For-Now" leaving abandoned projects unsupported and incompatible.

Worryingly, even some departures expressing clear negative sentiments in no uncertain terms.

Plus myriad crash-reports which go unsolved.

Or even ignored (cough, cough).

Which tends to skew perceptions and sap goodwill.

I've noticed similar things many times through my years trawling forums for Linux-fix information.

The FOSS (Free Open-Source Software) scene seems particularly prone to certain phenomena.

Honesty can be painful, but it's worth doing.

Maybe it's just New-Year-syndrome.

Or maybe it's genuinely time to ask some Big Questions about the Minetest scene.

Soliciting constructive criticism and informed analysis, considered assessments, pertinent questions.

Thinking Big Thoughts.

Wiser eyes than mine have long-since joined the dots, spotted patterns, and discussed them intelligently.

The 19-point summary of the classic ''The Cathedral And The Bazaar'' is an extremely useful list to consider.

But I would highly recommend reading a critical review of CatB, a useful if depressing reality-check.

Particularly about the politics of mature complex projects, about who knows what and who decides what.

And, crucially, about what is properly publicly documented and comprehensible, such that the Fresh Young Turks can genuinely fork and fix deep design issues.

The centuries when a single polymath could be said to ''Know Everything'' are long gone.

Subject-Expertise and Language are not one and the same phenomenon, but there is clearly a link.

Fluency requires more than mere Proficiency, and penning Poetry still more.

Wittgenstein's assertion that ''the limits of my language mean the limits of my world'' spring to mind when, in computer programming, language is itself the key tool with which a project is conceived and engineered and materialised.

Computer languages continue to proliferate. Maths is a formidable language too.

The raw Capacity of a language to cope with a concept is no guarantee of its Suitability or Desirability.

Real-world considerations demand Efficiency and Elegance.

Sadly none of us is fluent in every language --- neither spoken tongues nor maths nor machine programming.

And few indeed the true poets and original innovators.

The intriguing spherical-view proof-of-concept by Jeija beautifully demonstrates what can be done when you're sufficiently expert in both a subject and its attendant languages.

Here and there I see comments referencing issues in this or that core dependency technology (for example Irrlicht), often accompanied by remarks about ''not knowing enough..'' and mentioning the development languages.

Whilst the metaphor of the ''Power Plant and the Bike Shed'' is both powerful and pertinent here, it's still valid to question everything and to do so regularly.

Is the project being crippled by historical choices of core dependencies and languages used..?

There are plenty of past lessons available to be learned.

Are the devs still asking questions about the fundamental design and architecture or has the current implementation become sacrosanct, immutable and entrenched..?

Infamously there are ''known knowns, known unknowns, and unknown unknowns.''

The smart know the significance of the phrase ''Necessary and Sufficient.''

The wise are honest with themselves about their own limitations.

With the politics of Cathedrals and Bazaars in mind, maybe we should be explicitly sharing notes on ''who actually knows what'' and ''who speaks what'' and be actively looking to fill the gaps.

Might it be sensible to seek external expertise on key elements of the core..?

Who might be needed?

And how might they be co-opted to tackle any deemed-desirable rewrites, splitting and disentangling, rethinking the maths, or swapping to other dependencies?

Every long-standing Linux-user knows about the Marmite figure Mister Torvalds, but is there any such lead-developer for Minetest..?

The Cathedral and its critique recognise the benefit of having the right team with the right Project Leader or Development Coordinator.

Is there even any published ''Road Map'' we can look at, or is this really too much of a hobby-volunteer exercise..?

For diverse reasons I love Minetest and I sooooo want it to be a shining example of what Community can do.

Humans are far from perfect, and FOSS-projects can be fragile flowers.

Can we at Minetest be the exception?

I've said enough for now, and I've plenty stored up, but in the spirit of those recent posts by TumeniNodes and tinoesroho I'm most interested in hearing what others have to say.

In particular it would be nice to read some extensive ''State of Minetest 2017'' posts from the devs themselves.

Comments please.

HTH

.
 

Byakuren
Member
 
Posts: 816
Joined: Tue Apr 14, 2015 01:59
GitHub: raymoo
IRC: Hijiri
In-game: Raymoo + Clownpiece

Re: Big Thoughts, Dangerous Thoughts

by Byakuren » Sun Feb 05, 2017 09:12

Separating each sentence into its own paragraph makes the separation meaningless, and is just as bad as an unbroken wall of text for readability.
Every time a mod API is left undocumented, a koala dies.
 

CuriousNoob
Member
 
Posts: 73
Joined: Wed Jan 27, 2016 09:43

Form or Substance

by CuriousNoob » Sun Feb 05, 2017 11:55

Byakuren wrote:Separating each sentence into its own paragraph makes the separation meaningless, and is just as bad as an unbroken wall of text for readability.

.

Hmm. Okay.

Thank you Byakuren for your feedback, although not the type of ''comments please'' that are most needed.





Ocular/eyesight issues influence my preference for on-screen text.

Whilst certainly reflecting the way I think, the spacing actually helps me navigate as I type longer pieces.





As recommended in exams, and whilst falling short as all humans do, the goal is always compositions which are :

1) Relevant 2) Accurate 3) Complete 4) Clear.

You are asserting that my number 4, Clarity, is irredeemably impaired by its spacing.

Beyond eyesight, I guess our minds are wired quite differently.





Does this extra double-double spacing break the post sufficiently for you to cope and properly parse it..?





What's most saddening is that there are still no comments on its Substance, merely on its Form.

I believe some sports pundits use a censorious phrase ''playing the man not the ball'' and in philosophical circles ''ad hominem attacks'' are generally frowned upon for good reason.

Bezroukov is depressingly correct about the psychology and politics.

FOSS projects are indeed fragile flowers.





Which brings us back to where we were. Does no one have any enlightening comments about the O.P. Content..?

Maybe if the stumbling-block is just its Form, should I re-post below with a few newlines deleted..?

Then will the hoped-for comments start to appear..?




Again though, thank you Byakuren for your input.

And FWIW, I do smile every time I see your post-signature ... it hints at an engineer's perspective set in a world-view tinged by depressing reality.

I suspect that deep-down we share a certain Venn overlap beyond mere enjoyment of Minetest.

Byakuren wrote:Every time a mod API is left undocumented, a koala dies.





HTH

.
 

User avatar
taikedz
Member
 
Posts: 692
Joined: Sun May 15, 2016 11:11
Location: Scotland, UK
GitHub: taikedz
IRC: DuCake
In-game: DuCake

Re: Form or Substance

by taikedz » Sun Feb 05, 2017 18:41

*sigh* well you did put your heart in it, it would be unfortunate not to get any meaningful response.

CuriousNoob wrote:As recommended in exams, and whilst falling short as all humans do, the goal is always compositions which are :

1) Relevant 2) Accurate 3) Complete 4) Clear.


Whilst the metaphor of the ''Power Plant and the Bike Shed'' is both powerful and pertinent here, it's still valid to question everything and to do so regularly.


It also helps to be 1) organized and 2) to the point, lest you even be talking about the refreshments. Your original post would have been given a big red line across every page if my teachers had seen it.

I agree with Byakuren that your post is hard to follow and not really inviting of a response because of 1) bad layout, and 2) lack of any organisation or direction.

So if you want to actually have cogent responses, please do try to lay out your arguments in a strucutred and well-presented manner.

+++

You did however highlight one sentence:

Are the devs still asking questions about the fundamental design and architecture or has the current implementation become sacrosanct, immutable and entrenched..?


Your reference to The Catherdral and the Bazaar would indicate you are aware of software engineering styles and practices. The above question highlighted as it is, thus strikes as odd.

paramat just opened up a question as to whether to change the serialization of mapblocks, and break backwards compatibility. This is a minor change from the individual format's point of view, but has wider repercussions.

If you are asking the devs to call into question the design and architecture, you are essentially saying: are the devs considering yet to throw out this architecture and start anew? An architectural change (even in a subsection) can be immense and costly. We'd probably see the end of minetest as we know it, and have to wait a couple of years before anything playable came from the new design.

So unless the architecture is scrapped, and the whole project re-launched, it is as it is.

They have so far as I can tell done a very decent job of separating game play and mechanics (minetest_game and mod code) from the rendering and data handling engine (minetest C++ binary). Yes there are shortcomings, but we can wonder whether goals can be achieved in the current architecture first before considering massive upheaval.

++++++

TumeniNodes posted interesting and clearly heartfelt comments about MT and the community.


I have just responded to his assertion, since I think this was an attempt to call me on a title he might have judged ill-advised.

But my stance still stands, that when a new player comes in saying things are silly, don't work like they should, and make no sense, to then proceed with ignoring explanations and invitations to read the modding guides and to actually invest a little time into understanding how the state of play is know and why it is, well, they get the responses they do. I think I remained sufficiently to-the-point and non-inflammatory.

I can't really do much when people read things out of context and stop at that.

(EDIT -- the misunderstanding with TumeniNodes is cleared up further down in this very thread)

+++++++

The FOSS (Free Open-Source Software) scene seems particularly prone to certain phenomena. (...) Humans are far from perfect, and FOSS-projects can be fragile flowers.


Every project has this short coming - software projects, hardware projects, government projects, community projects. You just don't see it in the non-open source world or on small ones because they never release before even failing. FOSS projects regularly launch and regularly fail, in the hundreds, every year. But to cite just a few shining open-source projects:

    * Apache
    * nginx -- both web servers going strong
    * NodeJS
    * Firefox
    * Linux
    * BSD
    * Python
    * Java
    * Wordpress
    * MySQL
    * PostgreSQL
    * ...

And those are just a handful of the superstars. Minetest has been going for a few years, and is already a great example of what can be done in FOSS, and hasn't really so far lost speed.

So I see your concern, but I don't think it's yet time to sound the sirens :-)

+++++++++++++++

With the politics of Cathedrals and Bazaars in mind, maybe we should be explicitly sharing notes on ''who actually knows what'' and ''who speaks what'' and be actively looking to fill the gaps.


I understand the sentiment that you are trying to instill here (kudos for that!), but without a concrete example to point at, you're not really helping (alas, despite recognizably good intentions).

At the very least:

    * What specific gaps are you wanting to see filled?
    * Who is becoming sole custodian of what information?

And how might they be co-opted to tackle any deemed-desirable rewrites, splitting and disentangling, rethinking the maths, or swapping to other dependencies?


Again, without pointing to any concrete set of examples and attempting to offer solutions on those fronts, to then open the wider to the commentary, this is all just waving hands in the air (and I would be first to annotate that I have also been known to do this in the heat of an discussion, so it's not something to be sheepish of, just to be aware of..!)

Also, again - swapping dependencies, re-jigging core sections.... yes code refactoring is always important and desirable, and if you would like to contribute to that effort you are more than welcome to ; but outright re-architecturing is too costly right now, especially without listing the explicit concrete gains.

--> What list of gains would you argue to the devs to re-architecture the software?

Suggestions from anybody, of course.

Is there even any published ''Road Map'' we can look at, or is this really too much of a hobby-volunteer exercise..?


It would be nice to have a road map, yes. I would hold back on insinuating that the project were run by a bunch of amateurs... that just doesn't help.
Last edited by taikedz on Thu Feb 09, 2017 11:39, edited 2 times in total.
 

juhdanad
New member
 
Posts: 6
Joined: Tue Jan 03, 2017 21:42
GitHub: juhdanad
IRC: juhdanad
In-game: juhdanad

Re: Big Thoughts, Dangerous Thoughts

by juhdanad » Sun Feb 05, 2017 22:44

I think I am good at using and combining well-known programming algorithms (like Dijkstra's algorithm). But languages without strict types (like Lua) can confuse me.
 

asanetargoss
Member
 
Posts: 42
Joined: Sun Apr 26, 2015 03:10

Re: Big Thoughts, Dangerous Thoughts

by asanetargoss » Mon Feb 06, 2017 04:14

IMO Minetest has really good bones; it just needs refinement. Minetest documentation is good. The engine is well-optimized. It just needs inconsistencies to be corrected, bugs to be fixed, and more fun things to do.

juhdanad wrote:I think I am good at using and combining well-known programming algorithms (like Dijkstra's algorithm). But languages without strict types (like Lua) can confuse me.


Ask questions! We're here to help!
 

User avatar
TumeniNodes
Member
 
Posts: 2835
Joined: Fri Feb 26, 2016 19:49
Location: in the dark recesses of the mind
GitHub: TumeniNodes
IRC: tumeninodes
In-game: TumeniNodes

Re: Form or Substance

by TumeniNodes » Mon Feb 06, 2017 13:00

taikedz wrote:
TumeniNodes posted interesting and clearly heartfelt comments about MT and the community.


I have just responded to his assertion, since I think this was an attempt to call me on a title he might have judged ill-advised.

But my stance still stands, that when a new player comes in saying things are silly, don't work like they should, and make no sense, to then proceed with ignoring explanations and invitations to read the modding guides and to actually invest a little time into understanding how the state of play is know and why it is, well, they get the responses they do. I think I remained sufficiently to-the-point and non-inflammatory.

I can't really do much when people read things out of context and stop at that.



Whoah there taikedz..., nothing of the sort, and no disagreement intended toward your comments regarding the mentioned new-commer.

My words and thoughts in your thread were more directed toward the individual your thread was targeted for (as well as other new-comers), as well as an attempt at adding some deeper insight regarding how MT works, to basically say "If you don't like something, do some work and try to change it to what you do like. And there are plenty of people always willing to help". (meaning, to attempt to add to what you were getting across)

I Apologize if you felt it was aimed at you.
Ich mag keine grünen Eier und Schinken, ich mag sie nicht Sam I Am
 

User avatar
taikedz
Member
 
Posts: 692
Joined: Sun May 15, 2016 11:11
Location: Scotland, UK
GitHub: taikedz
IRC: DuCake
In-game: DuCake

Re: Big Thoughts, Dangerous Thoughts

by taikedz » Mon Feb 06, 2017 13:12

TumeniNodes - ah, then it was my misunderstanding, and I apologise for incorrectly interpreting your intent!
 

User avatar
TumeniNodes
Member
 
Posts: 2835
Joined: Fri Feb 26, 2016 19:49
Location: in the dark recesses of the mind
GitHub: TumeniNodes
IRC: tumeninodes
In-game: TumeniNodes

Re: Big Thoughts, Dangerous Thoughts

by TumeniNodes » Mon Feb 06, 2017 13:59

taikedz wrote:TumeniNodes - ah, then it was my misunderstanding, and I apologise for incorrectly interpreting your intent!

No, no need to apologize, it was my own mistake the way I put my thoughts across. I can clearly see now how it might have been viewed from the other perspective.

I'm trying not to overshadow this thread but, in general... yes, the devs are always thinking about many of the same issues some people post about in these forums... some agree, some don't... this is where it comes into individual taste.

But the devs are obligated to stick to the overall, basic ideology of MT. It is a cubic world... period. While some yearn for more realistic eye candy (related to water and tress), even some of the devs do as well (or maybe one or two).
Some yearn for more realistic functions... (same applies as above)

But at the end of the day... this is a "game". It is fantasy, and it has one basic principle as a basic, design guideline..., whether it be in look, feel, playability, etc.. And that is "cubic".

In addition to this, the original designer/creator... also chose to keep all of it, under a completely "open" license..., which is it's best feature of all..., so, when someone wants something different..., they have full "legal" ability to change what they want, in those areas.
And as an added bonus..., all of this..., is absolutely free of cost. How unbelievably awesome is that? In a world where everyone is out to cash in on anything they do....

To be completely honest, and not just a "theory" in my mind... if Pertuu had really wanted to... he could have probably approached Sony (or other coorps) about potentially purchasing rights to MT.
Is that a far fetched dream? No..., think about the fact that a coorp like Sony, might like the idea of having an answer (counter software) to Minecraft.
Some might view that line and laugh but..., it's really not that unthinkable.

SO..., thank you Pertuu. Thank you to all the devs (official as well as outside), and thank you to all in the community who add their own contributions.

Are there bugs? Yep. Rendering engine limitation? Yep (and so on) Yet there are game out there people are willing to pay even $70+ for which have bugs, limitations, as well as aspects which are not "realistic"... because they are, in the end... not "real".
MT is still, an ongoing development, software. It's not even at 1.0 yet but, is still, or already(in view of how much $ has gone into dev and how much it costs for the user) just as playable and as much fun as MC.
Think about that...

Core (or default) design is discussed regularly... and some even plug away at experimenting with changing it. But this is one area where devs do need to be careful (as far as official releases) due to creating a pandora's box situation.
Anyone is legally able to experiment with the core design, and if some aspects work, set up a PR.
But then one also really needs to take into consideration the very small group of core (official) devs, and remember there is a lot for them to look at on an almost daily basis.. and they all do it free of pay.

Bug fix requests, feature requests, ideas, solid code, not so solid code, complaints, praise, .... do I need to go on?

It's an open source project..., community driven, community maintained, etc.. And it still give MC a run for it's money (pun intended)

SO, when I see posts about realism..., or "it doesn't do this like MC does", etc... I think (as I learned early on)..., then make it what you want. Or at the very least.... try. If all that fails to make you happy... then go throw $29.95 (or is it $39.95 now??) at MC, and complain to MS about what you don't like. (PS you're far more apt to get a satisfactory result, or a more realistic chance of seeing your request happen or to be at least considered with MT)

Which direction is MT going? Same as it has been.... "forward" ; )

I've rambled enough now... filled my quota..., gonna go take my meds now : )
Ich mag keine grünen Eier und Schinken, ich mag sie nicht Sam I Am
 

User avatar
TumeniNodes
Member
 
Posts: 2835
Joined: Fri Feb 26, 2016 19:49
Location: in the dark recesses of the mind
GitHub: TumeniNodes
IRC: tumeninodes
In-game: TumeniNodes

Re: Big Thoughts, Dangerous Thoughts

by TumeniNodes » Mon Feb 06, 2017 14:18

I would have used "edit" but it might throw some who have already read my ramblings off.

I also want to add... that some "core design flaws" stem as far back as the very beginning... (but there is no "blame" to be added anywhere) When this was all created, I do not think it was even intended/ or thought to be something as significant as MT is today.

But, it is.
And at this point, in order to address those unintended/innocent flaws... would require a complete reverse engineering project, and a complete stalemate of thought, and then implementation of a complete rebuild/overhaul from the start, on. And would take a very long time, and an intense amount of work and most importantly... enough interested volunteers to do all of this.

So, with all that in mind... I find the current approach the best solution.... Let's continue with what there is, and strive to improve upon it, with the resources at hand.
And I personally think it has been going fairly well, given all the facts there are to the whole story/adventure. : )
Ich mag keine grünen Eier und Schinken, ich mag sie nicht Sam I Am
 


Return to General Discussion



Who is online

Users browsing this forum: No registered users and 5 guests