ID:2216106
 Feb 21 2017, 3:21 am
|
|||||||||||||
Poll: How often do you use the webclient on a weekly basis?
Login to vote.
Just curious to see how many people use this thing + how much.
| |||||||||||||
0% (0)
0% (0)
3% (1)
3% (1)
24% (8)
69% (23)|
Feb 21 2017, 5:05 am
|
|
Ghost of ET
 |
the what now?
|
I was interested in using it a while back -- and kinda still am -- but for reasons I nor Lummox can explain (in fact, he said he had no issues), it will not load my project properly unless I disable the skin file. I gave up and haven't touched it since.
Just tried it again and same result. The loading bar reaches 100% but the map never updates. I know this much because when I force-close the tab, Dream Daemon outputs that my player has been garbage collected, meaning I connected. | |
If you want to get me a new copy of the source I'd like to take another look and see if I can find anything of note. Also if you have any errors in your JavaScript console when using it, those would be a big help to see.
| |
Sent via pager.
| |
Reasons we don't use the web client in our SS13 fork:
* Godawful performance. * Ban evasion. * Doesn't support all advanced rendering features we need. * It'd be nice if I could even load it without it killing FireFox. | |
PJB3005 wrote:
* Godawful performance. This is the primary reason I hear as to why people won't touch the webclient once they get their project working in it (which can be pretty tricky). The performance is one big sad joke and I really don't understand why it's so dreadfully slow. It's just not worth maintaining if it can't be brought up to speed with DS, which already suffers from its own FPS issues. | |
The performance is only an issue of concern in very large projects and even then its not that bad.
| |
As someone who used and tested the webclient extensively, you're very wrong. It varies from machine to machine, but you're going to see so many FPS complaints if you release your game off-site. It does not perform nowhere near as well as DS--which is also pretty slow.
People off of BYOND expect the game they're playing to be reasonably smooth. Stutter and long pauses that come with BYOND's networking aren't acceptable in 2017. The webclient is a much different beast in that the problems are far worse. | |
I agree, theres some wierd issues with DS aswell where if you lag just a bit or theres a hichup with the network you get all sort of stuttering and long pauses. Its quite depressing.
| |
I honestly think the webclient should have been ditched ages ago. Trying to be flash isn't getting BYOND anywhere.
| |
Well, the idea itself is good and is definitely something worth pursuing if it can be brought up to speed with DS. The communities on websites like Kongregate and Newgrounds are far larger than BYOND's, so being able to publish your game there definitely wouldn't be a bad thing.
To put it in perspective: Eternia gets about 8k plays a month, with a 'play' being defined as a player logging in. We have peaks of about ~100 players. http://www.kongregate.com/top-rated-games?sort=newest Recent games on Kongregate with a fairly decent rating--3.5 stars or higher--get highlighted here. If you look at the ones published within the past month, you're going to find that they have at least 30k plays, triple Eternia's, and the average exceeds that plenty. The majority of these games aren't using Kongregate's kred system either, which guarantees far more exposure (a minimum of 1 million frontpage views). That's just one website of many. Basically, for Severed World in particular, I have a feeling that we'll benefit more from web portals than we will Steam. And regardless, as a developer you want as many places as possible to show off your game: the more options available the better. A lot needs to be done to the webclient before developers should be expected to use things like the JS menus that lock their games into it. Obviously the abysmal performance should be the first priority, and then a way where developers can bypass the login system/ads for a royalty fee (since no web portal will allow you to publish your game with them, making all the above on distribution a mute point as is). I also suggested this to Lummox a few times, but letting trusted developers like Ter13 take a look at the webclient and maybe BYOND in general probably wouldn't be a bad idea. I assume that he's relatively inexperienced in JS since he wasn't sure how to profile in Chrome during the early days of the FPS thread: http://www.byond.com/ forum/?post=1917533&page=2#comment16563682 Not a big deal. Everyone has their strengths and weaknesses, and this project was probably a lot more work than they initially anticipated. But yeah, it's been almost three years, and the webclient still isn't where it should be if it's going to be considered a viable feature of the engine. Naturally I'm a bit bitter since the people working on Severed invested a whole lot of time into it, with the assurance from Lummox & Tom that there's no real reason why the webclient shouldn't perform well with work. That's a lot of time and energy wasted where we went full speed ahead expecting results. It got to the point a few months back where we made the right decision and began porting the game--which was essentially complete at that point--to a different engine. | |
I actually tried to see what would happen if I tried to run my branch of SS13 (FTL13) on the webclient, and it was pretty godawful. Here's the kind of issues I had:
- The parallax was replaced with white space, almost as if it was DS running in software mode. The webclient was running in WebGL mode. - The performance was GODAWFUL - None of the tgUI windows could get past the loading screen - For some reason the thing decided to display the left hand icon way off to the right of the screen - Did I mention the performance was godawful? - Invisibility was broken. | |
The performance being awful in the webclient is both to do with how big of a project SS13 and how awful SS13 is programmed.
TGUI is awful and isn't a big surprise it doesn't work in the webclient. If there are bugs with it, then report them, don't whine about it not working if you're not willing to bring the issues up. | |
It's not. Performance in the webclient is quite literally dependent on the game's FPS and the resolution size. Even at 10FPS and a reasonably small resolution size the performance doesn't cut it; the game's code is mostly irrelevant.
| |
Pixel Realms wrote:
Naturally I'm a bit bitter since the people working on Severed invested a whole lot of time into it, with the assurance from Lummox & Tom that there's no real reason why the webclient shouldn't perform well with work. That's a lot of time and energy wasted where we went full speed ahead expecting results. It got to the point a few months back where we made the right decision and began porting the game--which was essentially complete at that point--to a different engine. It never really made since to me that you started a project expecting performance that literally did not exist at the time. That's one of the weirdest decisions I saw out of Severed. What made it even weirder was that literally every other decision seemed spot on, there was high quality art, programming and gameplay but that original decision plagued the project endlessly. (or at least it seemed to) | |
You know I even think the original idea for the webclient was solid. The potential exposure of having your game on multiple websites is nothing to sneeze at.
But for me personally, The webclient just does not work well. And after taking a look at the poll results, it's apparently not working well for other people, too. | |
Webclient was...I was totally against it.
I vote on taking all webclient related code and erasing it from every system from the universe. Move to SDL, give us real functional portability and finally add the 6+ year old topological sorting to map formats. If the browser is a goal, asm.js has been a thing for what, 5-6 years? WebAssembly soon... Kill that ugly baby! | |