ID:2311440
 
Another blip in the schedule hit this week, as further fallout from the events a couple weeks ago. All is well however, and work continues to progress.

I've been working on various bug reports from 512 while also working ahead towards some new feature stuff--some of which might make it into 512. Specifically, I really want outlines and drop shadows to be available for maptext, and this is an achievable goal. I've been modifying the backend HTML/CSS parser to support this, and now that it's in place I have to figure out how to apply the new properties to maptext. That means potentially drawing maptext in stages so filters can be applied to separate pieces, which is the main holdup. I'm sure I'll figure it out.

Meanwhile one of the bug reports from just today appears to be a doozy. The long and short of it is, someone found a situation where the new and improved "chaining" dot operator fails to work correctly. I know exactly why it's happening, but I haven't figured out a perfect solution yet, so that will have to percolate a bit.

The webclient is still getting work done. Visual contents are on the list, as is PIXEL_SCALE, but I also want to get DS's "nice upscale" shader in place for the main map so that when it's zoomed in, it preserves as much pixel detail as possible.

November has come! Today is a special day, but we're also three weeks way from Black Friday. And NaNoWriMo is underway; this year I'm taking the leap again, for the first time since the Great Forum Migration. But if you're not literarily inclined, it's a good time to put 512's features to work on a new game of your own. And while you're working on that game, don't forget about BYOND Membership to support the platform.
I've only briefly tested 513 512 and thoroughly enjoy the filters -- that'll solve some performances issues I posted a while ago having to do with maptext using outlines. Plan on donating in the not-so-distant future.

I do want to share my opinion on something though.

Whenever a new version comes out, I can't help but feel like the time invested in bringing the webclient updated feature support is mostly wasted.

The webclient doesn't seem to be used in any serious development since Severed World's attempt, in which it proved to be insufficient and inferior to DS regardless of many under-the-hood improvements. I have heard of some folks from the SS13 community using it for various things (BYOND key login/validation), but never in the way that it was intended to be: another, more capable method of playing BYOND games.

If support for the webclient was dropped or considerably slowed down, you would be able to put more time into providing features and fixes for the part of the engine the 99% of us actually use.

Do not take this the wrong way though. I know it can be difficult to let go of something one has put so much effort into raising -- much like a parent and child relationship. The webclient seems to be in a limbo and it makes little sense to me for you to spend weeks updating it when it's not used. I would rather you continue doing what you do for DS alone and perhaps begin to conjure up some ideas (could involve the community too) for a new endeavor -- something we'd all put to use.

Would appreciate your and other folks' thoughts on this.
I've only briefly tested 513

In response to Ter13
I'm weak. Corrected.
In response to Ter13
Wow, i just realized how similar Doc and Marty are to Rick and Morty.
Justin Roiland created a short called Doc and Mharti that is basically low budget Rick and Morty. It actually evolved into what we now know is Rick and Morty.

Just... Can't link it here... For three very specific reasons... Pretty sure I'd be breaking a hard rule. Definitely riding the line of good taste. Lummox and Nadrew would probably have a bone to pick with me. And I'm pretty sure that would just absolutely suck.
Welp, can't unsee that. Thanks Ter... ;)
In response to Flick
Flick wrote:
Welp, can't unsee that. Thanks Ter... ;)

You went looking. I warned you.
In response to FKI
The short of it is I don't think the webclient takes time away from development at all. While bugs are shaking out in the main software, it's a useful thing to work on so I'm always keeping on something.
In response to FKI
Actually, the webclient really is the same or better as DS -- it's just that it's more hardware dependent, meaning people with weaker PCs are going to have a harder time with the webclient than they do with DS.

I think the main thing holding back the webclient right now is the shitty pixel resolution that makes everybody's game look like garbage, but that's what he's tackling with it right now. After that's fixed, I fully expect to see Ter go back to tinkering with it (personally, I never stopped -- I just don't work on anything worth mentioning that isn't Feed :p).

ps. This month is also prematurity awareness month.
In response to Lummox JR
Lummox JR wrote:
The short of it is I don't think the webclient takes time away from development at all. While bugs are shaking out in the main software, it's a useful thing to work on so I'm always keeping on something.

A functioning webclient at this point is what I'd like to see over anything else right now. (of course, I want to see bugs squashed just as much. However, being able to have mobile online games using BYOND would be amazing and open up an entire new audience.)
I'm still here to nag about performance... But we need better performance. Splitting map sending and regular game threads gives us 100% (ish) of a core, more than the 70% SS13 is now (after lots of testing, configuration, and controller reworks) capable of using before BYOND steals the rest.

In short, if that made no sense at all, more performance work would be great.
In response to Unwanted4Murder
Actually, the webclient really is the same or better as DS -- it's just that it's more hardware dependent, meaning people with weaker PCs are going to have a harder time with the webclient than they do with DS.

It is not simply based on this fact. It's also been shown (via Severed World testing) even higher-end PCs don't get the greatest performance out of the webclient. I would even contest one of BYOND's strong-suits has been its ability to provide a decent gaming experience for the weaker PC audience, but that's beside my point.




Lummox JR said:
The short of it is I don't think the webclient takes time away from development at all. While bugs are shaking out in the main software, it's a useful thing to work on so I'm always keeping on something.

I get the feeling this is the politically correct answer. Fair enough though; at least you acknowledged.
In response to FKI
Nope, I meant that with complete sincerity. I think a release that adds a bunch of new features needs time for bugs to shake out, and people aren't always quick to report them.
In response to FKI
FKI wrote:
Actually, the webclient really is the same or better as DS -- it's just that it's more hardware dependent, meaning people with weaker PCs are going to have a harder time with the webclient than they do with DS.

It is not simply based on this fact. It's also been shown (via Severed World testing) even higher-end PCs don't get the greatest performance out of the webclient. I would even contest one of BYOND's strong-suits has been its ability to provide a decent gaming experience for the weaker PC audience, but that's beside my point.

It would be great if somebody could pinpoint what is actually causing the slowdown then. My PC is roughly as powerful as an X360 and the webclient runs great for me. Maybe some sort of hard/software conflict?
In response to Lummox JR
Lummox JR wrote:
Nope, I meant that with complete sincerity. I think a release that adds a bunch of new features needs time for bugs to shake out, and people aren't always quick to report them.

Yeah, I cannot argue with that.

I do have one more question though, because I am using Severed World as my main point of reference.

Is it possible that a game of such size is just too much for the webclient to handle properly?

I look back at the webclient work done during SW's testing and it's not like you didn't make a lot of improvements. It makes me wonder whether it's not so much the webclient as it was the game itself. Maybe ideal use of the webclient is in smaller projects?
I strongly suspect smaller projects will see better use of it, sure. Not necessarily size-wise as in maybe the degree of action.