At 4/22/18 08:47 AM, MSGhero wrote: Flixel was FINALLY updated to the latest OpenFL version. It's actually been years, and dozens of Flixel's issues on GitHub are being closed for the simple fact that newer OpenFL already fixed them.

Wait, so OpenFL fixed Flixel's issues in OpenFL itself and Flixel just went "oh, it's already fixed, nothing to be done here", or OpenFL fixed Flixel's issues in the Flixel project and it was updated?

At 4/6/18 04:07 PM, Rustygames wrote: Literally didn't even look beyond seeing a 2000 line class. No class should ever be 2000 lines :)

If it's possible, can you go through it again? Removed a helluva lot of redundant code and now the total LOC is around ~1700, or 1500 sloc according to Github (although when I counted it manually it was about ~1100 - maybe it doesn't account for comments?)

Edit: Can't believe it's been more than a month since the last post :\

At 5/29/18 12:20 PM, Gimmick wrote: Wait, so OpenFL fixed Flixel's issues in OpenFL itself and Flixel just went "oh, it's already fixed, nothing to be done here", or OpenFL fixed Flixel's issues in the Flixel project and it was updated?

OpenFL updated to the point where Flixel could easily be compatible with minor changes. I meant that years-old OpenFL bugs present in the old version had already been fixed, but no one using Flixel had access to those fixes. The majority of OpenFL forum posts were people from Flixel complaining about bugs that had been fixed already on the OpenFL side. So after the updates, a solid number of Flixel's remaining GitHub issues were closed since they were really old OpenFL issues.

Edit: Can't believe it's been more than a month since the last post :\

Where is everyone? My commute is so long that I don't have much free time, but I'm moving closer to work in a few months and it's gonna be great.

At 5/29/18 12:20 PM, Gimmick wrote:

At 4/22/18 08:47 AM, MSGhero wrote: Flixel was FINALLY updated to the latest OpenFL version. It's actually been years, and dozens of Flixel's issues on GitHub are being closed for the simple fact that newer OpenFL already fixed them.

Wait, so OpenFL fixed Flixel's issues in OpenFL itself and Flixel just went "oh, it's already fixed, nothing to be done here", or OpenFL fixed Flixel's issues in the Flixel project and it was updated?

At 4/6/18 04:07 PM, Rustygames wrote: Literally didn't even look beyond seeing a 2000 line class. No class should ever be 2000 lines :)

If it's possible, can you go through it again? Removed a helluva lot of redundant code and now the total LOC is around ~1700, or 1500 sloc according to Github (although when I counted it manually it was about ~1100 - maybe it doesn't account for comments?)

Edit: Can't believe it's been more than a month since the last post :\

I don't know if anyone already mentioned this, but Hungarian notation is quite old fashioned and wouldn't get through many code reviews :)

Why does the constructor call init? Couldn't the init code just be in the constructor?

At 6/22/18 04:17 PM, Rustygames wrote: I don't know if anyone already mentioned this, but Hungarian notation is quite old fashioned and wouldn't get through many code reviews :)

Yeah, but that's a stylistic choice that I've come to adopt over the course of this project. Needless to say, it's widely disliked, but there are a few positives about the notation - which, again, are subjective.

At least if you don't like it you can just strip all the text before the underscore and replace with m (eg str_ becomes m_) and it'll become near identical to another widely used style :)

Why does the constructor call init? Couldn't the init code just be in the constructor?

Adobe recommends keeping the body of the constructor as small as possible because constructors are not JITed, and that shifting it into an init() function allows the init function to be JITed. As for whether it's still relevant, I'm not sure - Jackson Dunstan ran some tests, which showed the difference was very slight, but it doesn't do any harm, so I included it.

At 6/23/18 07:15 AM, Gimmick wrote: Adobe recommends keeping the body of the constructor as small as possible because constructors are not JITed

Ah okay legit reason, fair enough :)

At 5/30/18 08:30 PM, MSGhero wrote:

Edit: Can't believe it's been more than a month since the last post :\

Where is everyone? My commute is so long that I don't have much free time, but I'm moving closer to work in a few months and it's gonna be great.

Where is everyone indeed? I've been thinking about this - I guess Twitter is where I see most game developers talking, but it's a very different sort of discourse from the time of forums.

I just found out by accident that my new tutorial got published a few days ago:

https://gamedevelopment.tutsplus.com/tutorials/creating-toon-water-for-the-web-part-1--cms-30447

At 6/23/18 02:30 PM, Rustygames wrote:

At 6/23/18 07:15 AM, Gimmick wrote: Adobe recommends keeping the body of the constructor as small as possible because constructors are not JITed

Ah okay legit reason, fair enough :)

I think you mean leJIT.

At 6/23/18 09:46 PM, OmarShehata wrote: Where is everyone indeed? I've been thinking about this - I guess Twitter

From what I can tell twitter is just full of twits :)

At 6/23/18 02:30 PM, Rustygames wrote:

At 6/23/18 07:15 AM, Gimmick wrote: Adobe recommends keeping the body of the constructor as small as possible because constructors are not JITed

Ah okay legit reason, fair enough :)

I think you mean leJIT.

I am not worthy

...Is everything going to pass the portal nowadays? I know I'm not in the right position to judge submissions, but there are a couple of submission I know for a fact should had been blammed.

This is just my honest opinion, it's like no one votes here anymore.

At 7/2/18 02:43 AM, solitonmedic wrote: ...Is everything going to pass the portal nowadays? I know I'm not in the right position to judge submissions, but there are a couple of submission I know for a fact should had been blammed.

This is just my honest opinion, it's like no one votes here anymore.

i did wonder this also, looking through the obituary logs, seems 10-13 submissions a day get blammed. Which is considerably less than it used to be. I think newer users arent used to or dont understand (or bother to) the blam save system. I'm curious if that will be scraped altogether with the updated review system.

At 7/2/18 09:08 AM, Luis wrote: i did wonder this also, looking through the obituary logs, seems 10-13 submissions a day get blammed. Which is considerably less than it used to be. I think newer users arent used to or dont understand (or bother to) the blam save system. I'm curious if that will be scraped altogether with the updated review system.

NG should have some introductory video for new users so that as soon as they sign up, a vid featuring Tankman or whatever that'll teach newbies some standard when it comes to what should be passed or what shouldn't.

That, or at least increase everyone's voting power so voting 0 feels meaningful again.

At 7/2/18 09:08 AM, Luis wrote: i did wonder this also, looking through the obituary logs, seems 10-13 submissions a day get blammed. Which is considerably less than it used to be. I think newer users arent used to or dont understand (or bother to) the blam save system. I'm curious if that will be scraped altogether with the updated review system.

I haven't touched the portal in a long while... I think I should start again.

At 7/2/18 09:58 PM, MSGhero wrote:

At 7/2/18 09:08 AM, Luis wrote: i did wonder this also, looking through the obituary logs, seems 10-13 submissions a day get blammed. Which is considerably less than it used to be. I think newer users arent used to or dont understand (or bother to) the blam save system. I'm curious if that will be scraped altogether with the updated review system.

I haven't touched the portal in a long while... I think I should start again.

Same. It seems a lot less prominent than it used to. Would be cool if they somehow encouraged users to get back involved in that stuff, it was fun in the old days (there was also a lot more submissions, and less professionals)

At 7/6/18 04:14 PM, Rustygames wrote: Same. It seems a lot less prominent than it used to. Would be cool if they somehow encouraged users to get back involved in that stuff, it was fun in the old days (there was also a lot more submissions, and less professionals)

There was a point a while back where there were hardly any new submissions, and I think that's when I stopped. Seems like there's a fairly good stream of uploads now.

Fuck this shit, use the Discord.

/thread

At 7/19/18 11:24 PM, GeoKureli wrote: Fuck this shit, use the Discord.

/thread

We're trying NOT to make NG dead as it already is. Also...

• Unpublished Submission

Seriously?? This thing got passed?

At 6/24/18 07:09 AM, Rustygames wrote:

At 6/23/18 09:46 PM, OmarShehata wrote: Where is everyone indeed? I've been thinking about this - I guess Twitter

I hate twitter - I think it creates a lot of room for miscommunication and conflict and not a lot gets done.

I'm at StackOverflow: https://stackoverflow.com/users/1348195/benjamin-gruenbaum
And GitHub: https://github.com/benjamingr

I'm also in the JavaScript chat pretty often: https://chat.stackoverflow.com/rooms/17/javascript (along with BoltClock if any of you remember him)

Also holy shit I've been a member of this site for over 15 years and almost half my life :D

Edit: almost

the newgrounds discord got nuked, so that was a bad idea in the end.

At 10/11/18 02:30 PM, Luis wrote: the newgrounds discord got nuked, so that was a bad idea in the end.

Everyone should know by now that newgrounds.com is the only true constant in life.

So anyway, game dev talk

I joined a discord for a kickstarter tactics RPG game that I supported (Arcadian Atlas), and I find it extremely interesting how... "I make an excel spreadsheet when I play Final Fantasy Tactics to optimize my gameplay" a bunch of people in the group are. Not in a bad way, it's just completely opposite how I played and enjoyed the game.

Also interesting is how games in this genre "have" to have certain features. I don't want to pick on anyone, but there was a discussion about if the game was turn-based per character (speed stat matters) or per team. It's per character, and someone commented "good, it makes spells like haste actually do something." In my mind, that makes it seem like tactics games MUST have a haste spell, and it's only useful if turns are per character. As if you can't just not have a haste spell in your game.

I was toying around with a tactics rpg idea where there are no items, no party customization (same couple of characters throughout), hp and mp are unified into one, and all your characters share the same hp-mp bar. I wonder what spreadsheet FFT fans would think about that.

the programming channel in that discord is...…………………….. yeah...…..

At 10/12/18 09:33 PM, MSGhero wrote: I wonder what spreadsheet FFT fans would think about that.

Press X to flee

So apparently the discord I was referring to was an unofficial discord, which is still up,. I didn't know the other 1 existed

At 10/12/18 09:33 PM, MSGhero wrote: So anyway, game dev talk

I was toying around with a tactics rpg idea where there are no items, no party customization (same couple of characters throughout), hp and mp are unified into one, and all your characters share the same hp-mp bar. I wonder what spreadsheet FFT fans would think about that.

Having never play FFT or any of the larger tactics games... I think you should definitely try this, from the sounds of it, I suspect the result would be more casual of a game. And come to think of it Mario + Rabbids did it that way, and I kinda liked it. It was fun to plan the moves of all 3 of my characters at once with no required order, and it was exciting to see what all the enemies would do all at once rather than gradually unfolding as I was going through my strategy. I remember thinking that Mario + Rabbids fixed a whole bunch of stuff I hate about the genre, but in the end I stopped playing because it was just too easy

At 10/15/18 09:41 PM, GeoKureli wrote: Having never play FFT or any of the larger tactics games... I think you should definitely try this, from the sounds of it, I suspect the result would be more casual of a game.

Behemoth's Pit People was nice bc the "attack" commands were removed and was instead determined by your positioning and a bit of chance. There's still strategy, but less boilerplate. I feel that way about leveling, items, and party customization in games (I always stick with the original group unless I cannot).

At 10/11/18 02:30 PM, Luis wrote: the newgrounds discord got nuked, so that was a bad idea in the end.

Set to return XX/XX....

At 10/16/18 03:56 PM, IvanAlmighty wrote:

At 10/11/18 02:30 PM, Luis wrote: the newgrounds discord got nuked

Set to return XX/XX....

20XX

At 10/11/18 02:30 PM, Luis wrote: the newgrounds discord got nuked, so that was a bad idea in the end.

RIP IN PEACE THAT WAS MY HOME AND I MISS IT MORE AND MORE EVERY SINGLE DAY

At 10/17/18 06:29 PM, ninjamuffin99 wrote:

At 10/11/18 02:30 PM, Luis wrote: the newgrounds discord got nuked, so that was a bad idea in the end.

RIP IN PEACE THAT WAS MY HOME AND I MISS IT MORE AND MORE EVERY SINGLE DAY

Hey, Muffin what did I missed? Tell me please what are we lost.

Hello!
I'm here to ask you for advice. Again.

How would you implement the creation of bullet patterns in SHMUPs? I mean, I have a rough idea of how to implement a pattern itself: basically, you have an array of bullet sequences, the as time goes on, you update the weapon and switch from one sequence to the other.
But how do I create the objects for that? Just using constructors is not an option, they tend to have way too many arguments and the code becomes hard to understand and maintain. For now I decided to use the builder pattern, but I don't really like how it looks, either.

So... what would you suggest?

At 12/15/18 08:55 AM, SkyFire2008 wrote: So... what would you suggest?

I've done this kind of thing using text files. My file would look something like

and like the builder in a way, I would have code that parses the text file and fills out the properties of each bullet in a pattern. So the parsing function would handle everything in the way you said the constructor might, but that's typical for parsing functions.

Edit: multiple ways to do it. Sequence works too, but I like having that info outside of the code

At 12/15/18 04:08 PM, MSGhero wrote: I've done this kind of thing using text files. <...>

Yeah, that's a good idea. I'll probably use JSON.