My "everything" app

[Mugwump]

@ Baker - please go fuck yourself. Thanks.

Actually, beyond his witty jab, Baker kinda has something of a point. You start lots and lots of very promising projects but we never see you release anything. Worldspawn? I haven't heard of it in quite a while. The map you said you were making for Jam 9? The guys at Func_ asked you where you were at but their question was left unanswered. The simple wad/mipmap tool you said you were making for me? I'm still waiting for it.

The gist of what I'm saying is that you're a bit all over the place. You start something then you start something else and seem to lose interest in the first thing along the way. You have tons of great ideas but your way of doing things is a bit frustrating. In the end, ideas are worthless if they don't come to fruition eventually. I think you may need to be more focused and commit to what you're doing before moving on to the next project. Please don't get mad at me, I'm saying this in all friendliness. You can't blame Baker for losing faith in your promises.

[MadGypsy]

Mugwump

I just want to make sure I understand you. So,what you are saying is, in a world where nobody is paying me to do any of this, it is all done in my free time and it is first and foremost a passion, I can do whatever I want and nobody can tell me jack shit one way or another? Is that about the gist of your message to me?

Of course it is, cause any other message to me regarding my non-profit, no contract, free time, whatever-makes-me-happy work would be silly, presumptuous & privileged.

Regarding the map. I stated on func that I would TRY to enter the jam. There came a point near the end that I knew I would not finish. A week or 2 later I learned they extended the jam but, there wasn't enough time to finish by the time I found out. Then it happened again. If I would have had all that extra time to start with I probably would have finished my map.

It's a shame it turned out like that. I watched all the maps on ArrrCee 's you tube channel and my map was very different from the stuff I saw. I think my map would have been heavily commented on. I went a completely different direction than Quakers are used to.

[Mugwump]

Heh, are you trying to make me believe that whenever you would release that mythical super-duper app of yours, you wouldn't charge anything for it? I think not. I seem to remember you talking about Worldspawn as a future commercial product.

Bottomline is: I think Baker's attitude comes from the fact that you tend to boast quite a lot about your skills but none of us can decide for ourselves whether you really are that super talented programmer you set yourself out to be or just another bragger full of shit. No offense. I personally still believe in you but I can understand why some others don't anymore.

That said, if you can actually finish my wad tool at some point, I'd be more than happy to pay you for your efforts, provided that you don't charge me too much.

[MadGypsy]

"your" wad tool is already in ark and yes ark is free. My engine was NOT a commercial project it was being designed to create commercial games. In other words if you wanted to design a game in my engine you would have to buy a license. From that point it is entirely up to the developer to charge or not or to include ads, dlc, etc.. Of course the engine is free for the people playing the games.

It's very easy to determine what my skills are. Just read the thousands of lines of code I have written spanning numerous languages with technical descriptions in thousands of posts. If that doesn't give you some idea then you have no business judging me in the first place.

I do not care if people do not believe in me. The way someone feels about me or anything about me has zero effect on my reality. It has zero power to make anything true. You could stop having faith that I am male and my junk will still be swinging... so to speak. I share whatever I am doing because I want to not out of any obligation to anyone. At this point everyone's faith should be in the fact that I will always do what makes me happy. If that means you get a wad tool, great. If that means you never get a wad tool, great. I have done much of what I ultimately set out to do... which is ace languages left and right. I have brought to life complex concepts in all of these languages. I have proven to me (the only person that matters) that programming has no boundaries for me. Anything I do with these skills from this point is little more than showing off.

I'll give you some insight. Go all the way to the beginning of me here and read everything. You will notice that my projects get substantially more complicated as I go. This is because I kept getting better. I have spent 6 years here studying and practicing within numerous environments and numerous languages. I have arguably forgotten more than most of you will ever know regarding it. You say "you don't finish projects" I say "that project is too simple to me now and I need something harder."

[Mugwump]

if you wanted to design a game in my engine you would have to buy a license

How is that not commercial?

You say "you don't finish projects" I say "that project is too simple to me now and I need something harder."

I hear you but that's precisely why I was talking about frustration earlier. You keep on teasing us with stuff that would be properly amazing if it was released, but then you move on to bigger things and we never get to play with all that cool stuff. Would it really be such a chore to finish a project, release it, THEN move on to a bigger project based on the previous one? If anything, with your current method you never get the satisfaction of finishing something, and we never get the satisfaction of appreciating your work. You're like a musician who keeps refining one song over and over again while the others are making albums. It may very well be the best song ever made but no-one ever gets to listen to it.

"your" wad tool is already in ark

Maybe, but I don't wanna wait one more year (or ten, or a lifetime) for a mega-app doing everything and then some that may or may not get to be finally released depending on your continued interest. I needed it last year for my ZinCity map project, you told me you would strip that wad tool from Worldspawn for me but then I never got my hands on it. This ZinCity project is on hold until you release that simple wad tool because the existing wad editors just don't do what I need.

[MadGypsy]

@ your mod on hold.

That's no excuse. You do not need wad tools to edit images. Black and White are palette colors. MSPaint could do the job. Treating your textures as external could also do the trick. Basing your progress on someone like me is a cop out.

@ wad tool again

You sort of fucked yourself on this. I gave you something like 2 days to respond to me. I was on fire and ready to do your thing but you left me in limbo. 2 days may not seem like a lot to you, to me it's 48 hours of "I need to do something. I need to do something. I need to do something.....". You ironically waited the almost exact amount of time for me to replace my fire to help you with some other fire. I don't regret it. I wound up doing really deep studies of HaXe and officially adding that language to my skill set. ::Makes one-hand-other-hand scales:: specific customized wad tool for one person ~ deep knowledge of yet another language...there's no contest that my time was better spent.

Let me ask you...let's assume you were a calligrapher. Let's assume you came here and said "I am going to draw the best letter A you have ever seen." Now let's assume while you were drawing the A you realized something that applied to all letters. Let's also say that you started testing that idea on other letters. Let's finally say that you realized spending all your time on an A was counter productive and even wasteful as a CALLIGRAPHER (ie not an A drawer)...

Would you still waste your time on an amazing A as opposed to...I don't know...a sentence? After that sentence or maybe even during it would it never occurred to you that a font would be far more productive?

Your complaint is that I don't finish sentences and letters because I'm more concerned with a font. And you go on to explain how other people can not benefit from me trying to complete a font because, for whatever reason, you would rather have a letter or sentence...both of which you can have when I complete the font.

You'll have to excuse me that I am fully prepared to say "fuck your A" at any time I decide that it is too boring or wasteful for me to complete. I move forward. Forward is where even higher learning resides. Staying stuck on an A just to say I drew a fucking A is retarded.

If people want to act like some form of faith they possess has any bearing what-so-ever on what will or will not happen that's really only their problem. It isn't going to stop me from doing what I do in a way that I feel is the best for me. All the complaints and opinions in the world are a complete waste of energy. I am going to do what I do regardless of anyone or anything else. You know I'm dead serious and telling you facts.

I can only imagine how far behind I would be from where I am regarding programming if I had actually sat down and finished many of the lackluster, niche ideas I had. I don't need to finish a wad tool to know that I can. The same can be said for many things. And while I'm not finishing some small idea I'm learning things that give me better ideas. I know that none of you know this but the last rewrite of my engine was entirely a practice run for applying true design patterns from top to bottom. The engine in my view was entirely an afterthought. My goal was to take something huge and figure out where to put my state machines, factories, etc.

Baker style programmers don't know shit about this stuff and will procedurally program themself into an infinite loop of spaghetti dependencies and claim it's good code simply because it doesn't seem to break.

"Hey look I finished my garbage that works but will break if you change even one dot."

[Mugwump]

I fully understand your stance on perspective and all that, but IIRC at one point you said Worldspawn was mostly finished. I reiterate: would it really have been such a chore to commit to finishing and releasing it before moving on to Ark? Hell, you even took the extra time to create a badass logo! How was that in any way relevant to your programming endeavours?

And if your only goal is to exclusively "do what you do in a way that you feel is the best for you", why do you even bother giving the world a glimpse into what you do in the first place? Especially if you slam the window shut when what you do becomes something else before completion. Meaner people than me could easily scream "attention whore", which contradicts your claim of not giving a shit about others' opinions.

You sort of fucked yourself on this. I gave you something like 2 days to respond to me.

I don't remember the specifics but if I took 2 days to reply it's probably because I was busy with RL stuff and lacked the time to check QuakeOne. I do have a life outside my computer, you know.

You do not need wad tools to edit images.

I know that, thank you. You don't seem to remember it but editing images was not my problem. It was the fact that the current wad editors available blend the colors when generating the mipmaps from the main textures. I ended up with a bunch of greyscaled pixels in my mipmaps while the full-sized textures were strictly black and white, and that didn't look good ingame. I asked the guys @ Func_ if a wad editor or any other image manipulation program could do what I wanted (either by not blending colors or by allowing me to manually edit the mipmaps) but got no positive answer, so basically you're my only hope, my Obi-Wan...
Let me be the Luke who will rise and become a great Jedi mapper thanks to your involvement. Pretty please with sugar on top? I promise you won't die at Vader's hands.

I sincerely hope that Ark gets released before you realize you'd be better off taking over the world or something, but if inbetween your important things you could sometime in the not-too-distant future take a day or two to whip up that wad tool and ship it in my general direction, I'd be really grateful.

Oh BTW, what has become of that map you started for Jam 9?

[MadGypsy]

I finally externalized the data system for Ark's object driven elements. I have written plenty of string to object parsers before but this one definitely takes the cake. It fully supports ALL AMF3 spec features. This means you can write objects in text format and they successfully get compiled to a byte object. That byte object can then be imported into ark and become the properties for just about anything. I also tweaked my graphics API so that it can compile interface elements with nothing more than the byte object. The old system had me inventing each element manually and then the proper object gets dumped in which styles everything. The new system simply invents the element based on the object. This was made possible when I realized that there are only a handful of possible things you can do in an element. Mostly, click or click and drag. Instead of hard coding things like scroll bars any element can simply be assigned the necessary generic behaviors.

In flash, Vectors are used for stuff like UVs, trisoup, etc. In the past I had no good way to denote that something was a Vector within one of my parsed objects. Pretty much everything can be handled with type inference due to it's delimiters or lack of them but, Vector and Array have the same syntax. My solution was simple.

ex Array
myArray:[1,2,3]
ex Vector
myVector<int>:[1,2,3]

By simply adding a type tag after the name a vector is easily denoted BUT nesting Vectors becomes a problem because just doing this [ [...], [...] ] doesn't leave any way for you to tag the nested Vectors as a Vector type. My solution to this was also simple.

ex Nested Vectors
myVector<<int>>:[ [1,2,3],[4,5,6] ]

By simply wrapping the tag in another set of delimiters (as many as you need) my system will maintain proper typing to the nth nested vector

AMF3 supports the Dictionary class. The only reason I could see ever needing this is if you just have to have an Object that can have an Object as member names. In the past I had no good way to denote that something was a dictionary cause for the most part a Dictionary is just an Object with an extra feature. I solved this by stealing a delimiter from PHP/HaXe. Dictionaries are written just like objects but name:value becomes name=>value...and more specifically it actually becomes {}=>value (you can use dictionary like an object with string names but the main point of Dictionary is Object names)

ex
myDic:{ {name:"john", age:28} => 762354623574265 }

My system is so determined and universally accepting that you can actually do this

myDic:{ {name:"john", age:28}<int> => [762354, 62357, 4265] }

and it works perfect. That would give you a Dictionary with an Object name and a Vector.<int> value

---

This was a lot of work but, I am glad that I did it. I took years of parsing techniques (most of which I invented or discovered on my own) and polished the shit out of them. The adobe version of SQLite was modified to accept AMF3 format as a type. Ark (actually renamed PaperWeight) relies heavily on SQL databases. My next step is to form the connections from my object parser to the database. I have a lot of ideas and ground to cover but, the idea that appeals to me most right now is the ability to write objects that completely replace SQL for table creation. In a nutshell, the unparsed (or parsed) Object will represent the structure for a table/database and creation of these things will be derived from the members of the object. The SQL api will essentially be completely black boxed for creation purposes. Hopefully it will be black boxed for query purposes, as well but, I need to do a lot of consideration on how an Object can be used to retrieve data. Sticking in an object with a bunch of null fields waiting for population seems like a terrible way to go. I want a more solidly typed system than that.

[MadGypsy]

Mugwump

@jam9 map
It's just sitting on my hard drive. I opened it up and looked at it last week or so but, I didn't feel compelled to do anything with it.

@why do you even bother giving the world a glimpse into what you do in the first place?
cause I like to share my ideas and various progresses

@Meaner people than me could easily scream "attention whore"
I guess if that's all they have going on. There was an IRC channel devoted to hating me. I felt special and enjoyed the fact that there was a group of people focussed on a guy that paid them absolutely no attention at all. In short, Meaner People can say whatever they want it will have zero effect on everything and anything.

@would it really have been such a chore to commit to finishing and releasing it before moving on to Ark?
My engine is IN Ark. It is still only 60% done or whatever but, it is a major part of the possibilities of my project. I didn't ditch my engine in any conceivable way. I studied harder, reviewed my engine, realized where some of my implementations were stupid or nooby and began redesigning it with some seriousness. When I started worldspawn I knew next to nothing about that end of 3D. Now I know enough that I have rewritten much of Away3D. Away3D is cool, it is also bloated and nasty. Some of the Away3D code is beneath my skill level. I don't like junky junk

@but if inbetween your important things you could sometime in the not-too-distant future take a day or two to whip up that wad tool and ship it in my general direction
This is doubtful. My inbetween times are spent knitting (lol, dead serious). I'm getting pretty good at it. I saw a bunch of knitted stuff where the stitches created ellaborate designs like vine-work and what not. I had an idea that, if I understood enough stitches to be able to say "this will make a left curve" or "this will make a 45 degree angle" (etc) then I could essentially draw with stitches. Each stitch is a pixel so-to-speak. I'm hardcore with my knitting, son. I can crochet too but, it's boring to me. My girl and I were talking about learning weaving next.

[MadGypsy]

Ive been amplifying the possibilities of my Object parser.

my example object script:


for one I created a reader. It is the only way that I can be sure that the objects I am describing are what is being created. The below image is a print out of my parser reading the created object back to a script.



As you can see, my print out is fucking identical to my script, with one awesome exception. In my script I purposely made the mistake of adding trailing commas to the last members of my objects in player_lib (the commas after the avatar members). In the printout those commas are gone. I purposely put them in my script because it is a common mistake to add those and it would generally mean you just broke your script. My parser does not crash on trailing commas. As a matter of fact you could put 1000 trailing commas and it will ignore them all. My reader however does not make "human" mistakes and will never accidentally add a trailing comma.

All this is well and good but there is more happening here. You may notice that the names in the player_positions dictionary look odd. This is because they are being printed as "pre referenced". In order for dictionary to work to it's full potential you need external references of the objects that you intend to use as names. Dictionaries use direct equality so just making an object that is identical in values to one used as a dictionary name will not be accepted. This is because one Object is [email protected] and the new Object you created is like [email protected] Although the same in value they have unique identities and that makes them not equal. In lieu of this a system needed to be created where dictionary names could be made as a reference in a hand written script.

If we take my same script and reparse it with REFERENCES on and then print, we get this


As you can see all reference markers have been replaced with their actual object. Whereas this may appear to be a duplication it is not. It is essentially a reprint of the original reference object.

In the print the members of some Objects are in a different order than the original script. This does not matter at all. Objects are never sorted (one reason they are very fast). With Objects the only thing that matters is the existence of vars. You will notice my iterable types maintain their order as they should.

You may also notice that my hex values never get converted back to hex. This is because after parse there is no way to know if a number was a hex. It also doesn't matter as the number is the exact same value as the hex.

What is not illustrated is that my read cannot add comments back in and every string gets double quoted cause there is no way to determine if it was single quoted in the original.
However, quotes do not matter because ALL forms of quoted string are treated as literal. My method for retrieving strings is very complex and favors ACTUAL first and last quote as opposed to "first quote ~> next unescaped quote". This means you only have to escape quotes in a string if you want to type what my system considers a close quote delimiter (('|")[\]\},]). Opened doesn't matter cause my parser will find actual open delimiter before it finds the one in the string and won't be looking for open delimiters after that til it finds it's close delimiter.

The next and final step will be to add my math expression parser in. For my graphics api it is easier in some cases to write a math formula instead of the product of the formula. Like certain divisions of PI for rotation. By creating an delimited expression type I can easily allow and pinpoint expressions. Technically ~ references could even be used in the expressions, creating more dynamic possibilities.

[nahuel]

whoah, nice stuff MG, i think you are become a better coder everyday

[MadGypsy]

Thanks, nahuel! This is something like my 5th Object parser but it is the first one where I decided to basically lock it to AMF3 specs. There are some possibilities that don't make sense for a hand typed version ... Like Dictionary and ByteArray. My very example is not likely to be hand typed data but, it's in the specs and my parser follows the specs.

My parser script with all comments and unnecessary whitespace (other than new line) removed comes in at just over 300 lines. It's pretty aggressive for being so tiny. However add back in the comments and whitespace and it is over 600 lines . I document everything.

[Dutch]

@why do you even bother giving the world a glimpse into what you do in the first place?
cause it's a hobby. Why post pics of my classic truck on the chevy forum if no one else gets to drive it? Cause it's a hobby. Other people enjoy checking it out. I enjoy sharing it. I enjoy checking out their rigs. We bullshit about our rigs. The only difference here is gypsy is on a high programming level in a language no one else here understands, so you can't expect a very stong two-way conversation. Just check out his progress and realize how cool it is. Appreciate the hobby.

@gypsy

I like the name PaperWeight. Don't know if you're going in this direction, but the combination of your interface and the engine could lead up to a Unity/Unreal-esque platform. If you were in the same weight class as those, the monetary gain from licensing could be substantial.

[MadGypsy]

Dutch

I changed the name cause there is ironically a movie that just came out named Ark. I know the 2 are completely unrelated but, when I saw the commercial I immediately thought "nope." and changed the name.

----

I did something last night that I should have done long, long ago. I broke my parser script into 2 scripts. It became real clear to me (this time) that my parser is a set of generic parse functions and a set of mutator functions that utilize the generics to a very specific degree. I split these things up. Next step is to rewrite my expression parser with the generic class as it's base. It should go really fast cause my old expression parser is very similar to my generic parser. Actually, my expression parser methods are like the 0.9 version to my generic parsers 1.0. It almost seems that without even trying or being aware of it my parse system has been upgrading itself on every rewrite towards a noticeable direction. When I rewrite stuff like this I just start over. I don't worry about the last way I did it or attempt to reuse anything. I like to see what system I will create by the seat of my pants so-to-speak. Well, by observation, I'm just creating tighter versions of the last system.

That's good IMO. In the end, instead of having an object and expression parser that are independent of one another I will end up with those things in a way that utilizes some base parsing methods. What I'm trying to say is instead of 2 parsers I end up with 2 implementations of 1 parser. Upgrades to the base system will upgrade all extensions together.

@monetary gain
Ya know, it would be nice to create something that generated some income and I certainly have ideas regarding how income could be generated from my project but, I honestly don't really care if I make dime. For me, the reward is more in line with successfully implementing my ideas. I could make 1000s of dollars and it would pale as a reward in comparison to simply thinking something and then bringing that idea to life in a completely working manner. Then there are the other bonuses that are automatic due to the action. An example would be back when I was working nightly on worldspawn and I have 2 guys pointing me to C code to help me write my AS3 code while explaining a system I do not understand to somehow use to work in an API I am unfamiliar with. Yet we all saw worldspawn move forward. I'm sure I'm not the only person that can dump himself in a completely unknown world and manage to make something out of information that is just as alien as the unknown world but, it's rewarding to be one of those people.

[MadGypsy]

I mentioned that quotes were a complex system and I wanted to illustrate why. The below image pretty much explains it all in comments. To give you a bit more guidance than that ~ lines 29 through 32 are all that is necessary to parse FOUR delimiter groups. Lines 48 through 58 are what is necessary to parse quotes, which are considered ONE delimiter group.