Printable View
What if you could submit your code to a company, and for a small fee (say $20 per app, or some pricing by the amount of code) they would optimize your code for you? Would anyone have a use for that service?
I know that optimization is a real pain in the ass for me, and if I could just write code, get it to function bug free, then hand it off to a pro to double my performance I would easily pay a small amount for that.
Any opinions?
No one competent will even look at your app for $20. Optimizing code takes a long time, you have to analyze the whole structure to make it more efficient.
I really can't see how this would work in any way.
If you're good enough to optimise other peoples code, then I think it's fair to assume that you'd rather eat pins than do it.
Skipping that as a downside, how long would it take you ( As the optimiser ) to find your way around the game ? You're looking at everything from as3 to as1 with onEnterFrames on clips nested 20 deep.
How far do you take it ? Is it using little tricks to squeeze out a ms here and there, or would you be replacing brute force collision routines with smarter ones ?
There could be hours of work just looking at the game and figuring it out how it even works.
Then... the majority of game performance issues are down to the amount of visual data flying around. Say you optimise all the code and send it back to the client, the code may run twice as quick, but due to them using big ass gradient filled vectors it still runs like a turd. Are they going to be pleased with your $20 service, or are they going to think they didn't get what they want ( A quicker game ).
Sorry, but I can't see one good thing about this. I've had to fix / speed up other peoples code ( I even turned down the chance to work on an official Flash Sonic game, which is a dream job, as it would have meant completing someone elses code ) and there is no joy in it at all.
Squize.
I didn't ask you if you wanted to optimize people's code. I asked if you would want someone else to do it.
Please try answering the actual question that was asked. Judging by how much you seem to hate optimizing, it would seem to me that the answer would be yes. Let me worry about the in's and outs, I didn't ask you anything about the technical issues that go into this, just if you would use such a service.
"I asked if you would want someone else to do it... Please try answering the actual question that was asked."
"Any opinions?"
Yeah I posted my opinion, only to get your snotty response back.
In that case, no, as it would be impractical, as can be inferred by my previous post ( Which I didn't realise was so wildy off topic ), and that I wouldn't trust my cats litter tray to someone who charged $20 for a service like that, never mind my / clients code / ip.
Hope that answers your original question.
Squize.
Your opinion is based on an assumption that you know why I am asking this question, not based on the actual question. The practicality of the service is irrelevant, as you are not the one performing the service. If someone offered to wash your entire car, rotate the tires, and check the oil pressure for one dollar, and could be trusted to do this competently, would you refuse because YOU would not do that much work for a dollar? Of course not.
As far as the price point, a larger project would obviously not be $20, and I did say OR price per lines of code. So the impractical (to you but then again every project may not be as complex as what you are doing) price point aside, would you outsource your code to be optimized by someone quickly, professionally, and cheaply, with no worries about piracy or IP theft?
no- because I would have doubt that the optimize guy would have the time and understanding of the code.
Besides that how would you measure your work - just by time would be odd. Or what happens if you actually just cant boost it - because the initial code was already pretty well optimized?
Perhaps if you would compare times (getTimer()) in certain functions and routines you could proove actually improvement.
The thing with other ones code is that that it takes for booth parties just xtra time to understand the code or changes of the other one- I think thats the most skeptical thing. And besides that there will be always some hardcore stuff thats pure garbage in code syntax that it would be a nightmare to fix or would really take weeks to fix it in which case it would be almost better to write it completely from scratch.
What is common however are agencies or people you can hire to assist in your project (I had such a expierence lately) - those people then write frameworks for you or organize the workflow so that many people can work on the code without getting to much into trouble. But that would require a frequent dialogue so that booth parties know the workflow and direction the project is going into.
Hey Squize, remember the days when people could appreciate constructive criticism? It looks like the snotty brats of FK are trying to dethrone you as the flash daddy. :D
By the way, I laughed like hell when I read:
Yeah, most of Squizes games are slow vector filled pieces of junk... Law of the West, LoTw pinball, Phantom Mansion and Prof. Savernoggin, just to name a few. These were all terrible games filled with shotty code, bugs, and it just looks like there was no love involved at all...Quote:
Originally Posted by moony4money
Hey Moony, I wouldn't trust anyone with my code for optimization. $20 would certainly be a wonderful price for such a service, but you get what you pay for. I wouldn't expect more than 1/2 hour of time to be spent on my game (if that) for $20. I'm not certain that anyone could even figure out my game logic in that amount of time.
By the way, Squize and I discuss optimization all the time, so I have to vouch for the fact that his games go through a heavy optimization process already, so that is probably why he would not use a service like this. I'm sure there are some new developers that could use the help though and $20.00 might be a good deal for them.
Quote:
Originally Posted by ImprisonedPride
This is obviously sarcastic, and you obviously have difficulty comprehending the things you read. I never said Squize DIDN'T optimize his code, I said he seems to hate doing it. Which if you read what he wrote, he basically stated that.
And get off the $20. You are missing the forest for the trees. $20 would be some first time flash programmer making a one frame game that has no clue how to optimize. Forget the price point, just focus on the pros and cons of outsourcing optimization. Don't worry about how difficult it is to pick apart someone else code. Don't abscond the virtues of the service with your perceived difficulty of it. I'm not asking you how hard it is to optimize code, or how long it takes, or if YOU would or could do it. I'm asking you if you would want someone else to do it for you.
If the answer is no, then fine. But if your reason is because you think it would be too difficult, then that is a dumb answer. If I asked you if you had a heart problem would you be willing to let someone kill you, cut your heart out of your chest, put someone elses heart in, and bring you back to life, would you say no because that seems SO HARD to do? No, because that's not a reason for refusing anything. And there are people out there that can and do do that, even if YOU can't.
Geeze.
no
"But if your reason is because you think it would be too difficult, then that is a dumb answer"
You're missing the point slightly.
If I said you could be King of the world, don't ask me how, I can arrange it. Now it's going to cost, obviously, but imagine how much time it's going to save you, no having to cut deals with other world leaders, no democracies to overthrow.
K, yes or no ? Simple as that, no questions, do you want to be King of the whole world, yes or no ? Oh, and if you say no, then there's obviously something wrong with you.
I'm being a sarcastic arsehole, I know ( Looks like sarcasm is my response to snotty replies ) but it's a similar theory, just blown up to prove my point.
You can't get defensive when people look at different angles of what you're talking about rather than just all nodding their heads in agreement.
Squize.
No, that's a good point. I am not asking how realistic the scenario is, either, just whether or not it is desired. It is a theoretical question, not meant to be picked apart on a technical level. To answer your question, no I would not like to be king of the world. Bit not because I don't think that you could deliver, that is irrelevant to the question. I would not like to be king of the world because I value my freedom and my free time. I would be giving both of these up in exchange for power, something I have little desire of.Quote:
Originally Posted by Squize
I am not just looking for a yes answer. If someone were to say, for instance, no I would not, because I find optimizing my code a very valuable learning experience that helps me write future code. Then that would be the type of response I am looking for. Yes or no answer that are not focused on whether or not someone would be willing or able to optimize code quickly, cheaply, and efficiently. these types of answers to a theoretical question are irrelevant as they are focusing on the wrong subject. Thank You for your example, though you might have meant it for another purpose, it shows exactly what I AM looking for.
It's a scam to steal source codes of decent games from deserving coders.
Lol, you caught me, I just want your codes! I am also a Nigerian Prince.Quote:
Originally Posted by ImprisonedPride
Most people go through lots of trouble to hide their code, I cant imagine they would be interested to give it out to strangers and even pay for it.
only at places like here,- its ridiculous actually and a reason why this place is drying out.Quote:
Most people go through lots of trouble to hide their code, I cant imagine they would be interested to give it out to strangers and even pay for it.
I assume moony4money was more after code in a general way - not the game -dont steal it type of code.
Lol, I am not after code at all. Between GDC now and E3 this summer, I have a lot of meeting and talking points to go over, and want to get as much feedback on the directions the industry is and will be taking.Quote:
Originally Posted by renderhjs
I'd not be interested in such a service, because I hate it when someone tweaks my code for me.
I like understanding why each line is there, and the simple thought of someone having put it there instead of me makes me totally unsure about its quality. Better performance doesnt mean code I like more ;)
Thanks for the input guys.
lol...wherever do you get your examples from?Quote:
If I asked you if you had a heart problem would you be willing to let someone kill you, cut your heart out of your chest, put someone elses heart in, and bring you back to life, would you say no because that seems SO HARD to do? No,
Let's say it wasn't $20 but more like $200 minimum, and I'm just talking a quote price- the time it would take for an advanced do-it-all coder to look at someones code and make an estimation of how long it would take them to rewrite everything. One blot on the horizon I can see straight away is the spaghetti factor involved when it comes to games, in other words there is much more that goes into games than just code. As Squize has mentioned, a noobie coder might use oversized bitmaps images, vectors, animations that don't fit the framerate- who knows. There's that and a million other tweaks to get the gameplay just right. Maybe he wants you to optimise his code to AS3 when it's been written in AS1 (total overhaul).
Game-coders are the unhailed neurosurgeons of the programming industry, you'll understand that once you investigate more carefully the scope of how much actually goes into a professional game. Most complete game coders who can do anything are guys who were programming when they were six. You would have to find a suitable payscale to match such experience.
It's not such a crazy idea within a framework. I get dozens of mails about tutorials I've written and because- like Render said- the framework was already in place (I designed it) for me knowing what is wrong with their code and how to fix it quickly is relatively simple since I wrote the original code they have manipulated. So say you had a similar person employed as full-time staff and their daily job is fixing games that were built from his own template/s, ie a support for guys in different industries who build a game from your own with maybe a couple of small extras thrown in and unique graphics of their own THEN it might not be such a mad idea to facilitate.
But...a free-for-all service where anyone can just pass along their game and have it completely revamped to a professional level? That's quite unrealistic at a cheap price. In most cases there's no difference between that and someone just coming along and asking you to design a game from scratch.
Honest answer to the founding question:
I would not like this service because...
1. I'm dirt poor, and have to do everything myself, be it math/art/anything
(which as a bonus gives me a little skill in a lot of places :))
2. Yes it would be a valuable learning experience.
Why pay for a service you can perform yourself whilst gaining valuable experience?
It'd be like paying someone to do exercise for you, the only difference being we are exercising our minds here. =\
(and strange examples seem to be popular right now)
3. Most of the time when I code, my brain shuts off and I somehow do it automatically.
I lose this advantage when foreign code is introduced.
Good answer? Yes, no?
By the way, why is people being so grumpy? -.0
New persons only gain respect after they've hung around awhile; they just don't know how friendly us persons can be =)
(yet)
I know not of this "friendly" word you speak of.
I code for the love of coding. If I wanted to pay someone to do the work for me I can't even call the game my own anymore. I am the same as Fil; every single character of code must be my doing.
So I could: be as lazy as can be and create a complete 3D shooter which runs at .0257 frames a second with bloated vectorgraphics, no depthsorting whatsoever and rendering everything including stuff that's offscreen. I could: send this engine over to someone, pay him 20 bucks (or some dirtcheap amount) and then get a Crysis clone running in Flash, which I can sell for a load of sponsorship...
I don't know, I'm kind of leaning towards yes.
They promised to optimize it, they never promised it would run fast =)
Bet if said miracle company could pull off said miracle feat, they'd ask for a cut of the profit though
Oh, and being completely on topic (at least in general about optimization), I did once dream of a program that could optimize for you. The whole works even; in depth.
It's technically possible if you gather enough time or processing power to accomplish the task.
Let's say even with a 100,000 character limit, there's still a finite number of programs that could be written. It's humongous sure, but still finite. This means a computer can theoretically run through every possible scenario, and output the fastest running program that simulates yours within a certain tolerance (sometimes quality is compromised, afterall)
This seems farfetched, unfortunately =\
Unless someone optimized the program's ability to optimize
Eh...
EDIT: I'd say I feel the same way ImprisonedPride, even outside the coding world, I prefer the solution instead of the answer
Those words aren't synonymous right?
Also if you're experienced enough to optimize other peoples code, 90% of the time it's easier/faster/more effective to write the whole thing from scratch than crawling through someones nested movieclip as1 code hell
Quote:
Originally Posted by hatu
NOW you're starting to hit on it. Optimizing code may be the wrong term to use here, more life making the game more efficient. That could be changing all the graphics to vectors and bitmaps as need be, to completely rewriting the code to do the same thing much simpler.
This really is being aimed at beginner and hobbyist coders, not people who do flash games professionally or very technical, robust apps. I guess this is kind of the wrong place to be asking this, as most of the users here are of the more experienced variety, but this is the only flash message board that I know of that responds more than once a month.
Again thanks for all the feedback, your answers really are helpful.
Perhaps you are simply using wrong term. Instead of "optimize" you could say something like "build a game based on my idea and some basic framework for 20$". Its not really optimizing if whole thing needs to be rewritten.
Since it isn't feasable to really optimize someone's engine for him, you can point out what can be optimized and how. More like technical feedback from an expert. I value that.
Then again, there's always Flashkit ;)
A better word might be "enhance."
I think the problem is that you're asking people who's jobs it is to pick apart everything they see, so obviously you're proposal is being picked apart.
On a simple level it would be good to have someone go through a game and improve its performance but when you delve deeper (as Squize has done) it just cant possibly work out. There are just too many issues.
Take a simple example:
Im making a game for BIGGAMESCOMPANY that has to be delivered at gold standard in two months. Now i go ahead and make the game with say two weeks to spare. But i have a problem, it lags like a *****. Now i can either spend that two weeks working out the issues myself and know im going to deliver the game on time because i know my code, or i can give my code to a third party to improve performance for me. The problem is that i and the third party have no idea whatsoever whether they can improve performance and have the game still ready for gold by my deadline. So in other words i would rather labour myself than give my game to someone else without knowing if they can do the same job i can within the timeframe.
You could argue that i could give the game to the third party to look over and give me an estimate but that is going to involve me taking an age to explain potentially thousands of lines of code and how the structure works. Plus if just them looking it over takes a few days then they are going to want to be paid. So i may end up paying for nothing because they decide they cant do the job in the time allowed.
Now you could also argue that if i knew what needed doing then i could give them a brief rundown and then they could decide how long it would take. BUT if i know what needs doing why dont i just do it myself?
The main issue is that NOBODY knows exactly how long any project will take. Even something that seems like a simple change may have a massive knock on effect elsewhere in the engine and everything else needs rearranging.
So in short its just not a possible solution without either or both parties getting burned. Its a massive risk for me because if the third party doesnt get the work done because hes encountered problems then im in huge trouble with the company im developing for. Even if the work does get done, optimizing isnt just a case of tweaking things. It may require an entire overhaul of massive sections of the engine and that often leads to other sections not working. So i may end up with a fully optimized game but upon checking with two days to go to deadline i realize that 50 bugs have popped up because of the changes made.
Its a nice idea in theory but just not feasible in any way.
I wouldn't say "can't". The thing here is that it's just plainly not plausible to expect someone else to be able to understand your way of thinking when you coded line 7317, in the 3rd layer of some fourth level movieclip. Unless you take the tedious task of commenting and explaining theory as to why you code the way you do (but that's just pointless; everyone just accepts the fact that no two people code alike. I take it a step further and never code the same way twice.:))Quote:
Originally Posted by Kakihara
Of course, I'm leading up to a contradiction to that statement from personal experience. I was invited as an add on for a project from Disney down the food chain and was given the grueling task of re-skinning one of the games. The problem was--as other members of that team can confirm--the code I was given was an absolute mess! There was something like 25 layers and each one had a few hundred lines of code on it. There was somewhere in the ballpark of 500 items in the library with no conventional naming system. Despite HOURS of b'tching about it to our team leader, I think we still managed to re-skin it in just over a month, and that was including 15 *extra* revisions they added every time it was submitted "complete".
It's beginning to sound like an echo in this chat, but the idea just isn't solid. The people who would benefit from it most are those people who aren't the best coders but aren't looking to pay money to make money. Those who are willing to pay money are better off optimizing/enhancing/searching their own code.
*cough*Quote:
Originally Posted by ImprisonedPride
Comments?
What are those? :rolleyes:
Some day I'll pay for it, but I almost never comment
And indeed we do need some fresh material on this topic
I think most everyone agrees that "yes" they would like someone to optimize for them, but doubt the feasibility of it
If I were to try and churn a profit, I think I'd devise a program that could streamline the most common of code automatically
Just an example:
Let's say the program goes and finds every line of code that affects output directly (stuff that moves movieclips, drawing api stuff, etc)
It then proceeds to find every variable that affects those lines of code, as they in turn will also affect output.
THEN it finds every variable that affects the variables that affect the lines of code that affect output, etc, etc
If you do that long enough you'll find every line of code that doesn't actually contribute to your program =\
Delete those and your program's guaranteed to run slightly faster
Time = minimal
Effort = zero
Profit = yes
I'm sure there's a million other ways the process of improving your code can be automated, but I'm just not that creative
Point is, if you lower your standards a little, it is possible
Different question:
Would a program, one that makes things a little less monotonous, appeal to you?
I wouldn't mind it =\ about half the time I'm doing optimization of various math related coding
Tell you what might be more feasible is a general support service. You would pay a one-off or monthly fee to have an online mentor;ie an advanced games programmer who will devote an hour a day or so to giving support to your project.
Not doing it for you, but giving appropriate advice for case by case situations and getting you going in the right direction. That would be a godsend for full-time Flash coders with decent programming capabilities that have been given a project and a tight deadline by their company.
Who knows how many times I got stuck and wished I had someone all-knowing like that on msn, there for me to question througout all the AS3 fumbling I had to go through while getting to grips with that language. If I could pay say- $40 and get an hour of solid support from an expert it would be well worth it at times of desperate need. Beats sitting around scratching your head for two days.
Don't say that Kris. If what you say is true, I owe Squize enough to retire on.
Yeh i think a tuition service is a more feasible idea. I get what youre saying IP about people with less experience maybe benefiting from the performance service but for me those people would be doing themselves a massive disservice by having someone else fix their problems. You learn from your mistakes as they say.
I think what the main poster meant was more like "If you are a noob, and can only do simple games.... then WE can optimize it for you"...... otherwise this idea is not practical at all, like many has stated already.... and for just 20$?.. not even 200$ would be enough for the trouble to optimize a code made by most people, as we all usually have our own unique way of coding, and therefore VERY hard for someone else to understand the structure.
Hmmm...yeah ditto for Mr.Malee and Senocular heheeee..Quote:
Don't say that Kris. If what you say is true, I owe Squize enough to retire on.
And imagine how mad rich TonyPa could get off that scheme, he always lends a hand..
I'm sure Tony would be fine without our contributions; the guy pumps out a new mochi-filled game like every other day. (Sadly MA is why I stopped playing his though.)