Limelight Forums

The amount of backlog in the suggestions area isnt the issue, the issue at hand for LL isnt one of content, but player choice and freedom.

No amount of development will solve that issue.

It has to be solved by the team as a whole, and it involves shrinking the rules list, and reducing the heavy handed nature of enforcement, and the microscope level analysis of PRs that result in punishments for a .5s infraction.

(Apr 30, 2020, 12:53 AM)BasicallyMental Wrote: [ -> ]

(Apr 30, 2020, 12:03 AM)Ben Wrote: [ -> ]

(Apr 29, 2020, 11:37 PM)BasicallyMental Wrote: [ -> ]Woah hold the front door there bud. You're assuming MASSIVELY here that I all of a sudden want MESKA size updates to be pushed, that that is what the community wants. Half of this thread is asking for rule changes, which has a small percentage to do with the dev team.

This is not me yelling "mush mush", it is me stating what contributes to the fact people find the server boring. I know it takes time, and funnily enough, I know they are also normal people with lives. However. Constantly accepting suggestions and leaving them for years is not helping the sitaution.

So what's the suggestion to improve on this? Sure, we could remove a ton of the backlog of suggestions and only keep the most wanted / essential items, however, we've still got to tend with a small dev and contributor team to be able to actually complete these. The only solution I'm seeing so far is taking on more of them, but they're not exactly coming out of the woodwork at the moment.

A solution to a backlog, is to add more to it? Simply close suggestions and state "We're going to start working through the current approved suggestions. Once we have reduced the amount to a sustainable level, suggestions will be reopened".
It's like me saying me engine is broken, so instead of stopping putting fuel in it, find the issue and drain the fuel out, i'm just gonna keep adding more until it works

Devs like the easy suggestions :) that’s why most don’t get done. Or better fixing the bugs that are pushed without being checked before going live.

(Apr 29, 2020, 08:14 PM)BlackDog Wrote: [ -> ]I'm fairly sure many will agree with this statement:

Limelight was at its best, when its city was like Baghdad at its worst.

The chaos bred friendships, RP and fun.

Back when LL first came out was some of the best time's I've had on the server. We had our run-ins, especially when Nickname was on, but damn it was an overall fun experience. I can't remember the specifics, but some of the rules on LL were looser than on FL at the time and it was a splendid time.

(Apr 29, 2020, 11:22 PM)BasicallyMental Wrote: [ -> ]If you want to post a suggestion, go to the suggestions sub-forum, and post it. Allow it to join the current 26 pages of approved suggestions. Through the 4 years this community has been a thing, we have made suggestions on how we would enjoy playing the server(s) more, but sadly these suggestions have not been implemented yet, some dating to be nearly 4 years old.

I've been (very slowly) working through the oldest suggestions. Some are a little too big for me to tackle atm. It doesn't help that it's easier to make suggestions than to implement them. Also, it's 26 pages thank you very much >:(


Suggestions are not what keep the server ticking. QoL ones for sure, but many of them would not really affect player count. I'd argue hodge podging more suggestions in without regard to performance is more of an issue than letting them sit and gather dust.

(Apr 29, 2020, 11:31 PM)Montyfatcat Wrote: [ -> ]I'm not sure you understand quite how difficult it is to push out a workable update on the scale of MESKA. It's not that the dev team doesn't care, they wouldn't be doing it if they didn't care, it's that they are a small group. They've finally finished the backend, the most difficult bit of something on this scale, and it now means that we will be able to see MESKA updates within this year, assuming that there's no stumps with the modules.

Even small updates can be huge pains. One thing many people miss is that all of the content on the server has probably been the cumulative work of over 50 developers. I'm not talking exclusively about the gamemode because the server is not exclusively the gamemode. The vehicles have been made by what, 4-7 people? The gamemode itself over time has probably had at least 10-20 people working on it over its history if you include contributors (whose code does make into the gamemode). This has lead to many different styles of code and assets which can lead to undefined behavior and inconsistent experiences.

GMod does not have a debugger or a good toolset for debugging. This can make finding simple bugs mind numbing. When I was making the injury info plugin, which is rather simple (when planned correctly), I spent probably 3-4hrs debugging some stupid issues. What was the issue? The creator of the citizen models didn't make the models correctly. A plethora of other suggestions are about vehicle handling. Vehicle handling is somewhat consistent for an author's vehicles, but 2 vehicle creator's idea of good handling will be completely different. That's why the dodge caravan goes 100+MPH and the Tahoe goes 40-60 MPH. It's unfortunately not that easy to edit the handling of vehicles either.

Hell, I encourage anyone who wants to take a crack at it to download the VCMod Vehicle Handling editor. No coding knowledge needed. If you fix a vehicle with atrocious handling or one that has had an approved suggestion about it, send it my way. I'll test it and get it added.

(Apr 30, 2020, 01:01 AM)BlackDog Wrote: [ -> ]The amount of backlog in the suggestions area isnt the issue, the issue at hand for LL isnt one of content, but player choice and freedom.

No amount of development will solve that issue.

It has to be solved by the team as a whole, and it involves shrinking the rules list, and reducing the heavy handed nature of enforcement, and the microscope level analysis of PRs that result in punishments for a .5s infraction.

I agree completely. There's some development work that can help, but think about it this way: the server had a higher pop before all those suggestions were conceived. CityRP has had a stage before LL even existed where content was stagnant and pop counts were still rather high.

(Apr 30, 2020, 03:19 AM)Brin. Wrote: [ -> ]Devs like the easy suggestions :) that’s why most don’t get done. Or better fixing the bugs that are pushed without being checked before going live.

You're speaking from a position of ignorance. Do we like easy suggestions, or do we want to just get them done so they don't pile up? If I can do a suggestion in 1hr instead of a different one that would take 2 days, should I just not do the 1hr one because it's too easy? What does that even mean? There's straight up no point in not doing an easy suggestion and just getting it out of the way.

Also, as of March 2019 all updates have been required to be tested on beta. We don't get paid if it doesn't get tested. We don't have a full Q/A team so sometimes things get missed, especially when players use things in a way we never thought of. I've personally added code to the beta server to make it easier for devs to test updates. I made a command to add bots to the server which is probably mainly used by me since I'm in a diff time zone

(Apr 30, 2020, 03:19 AM)Brin. Wrote: [ -> ]

(Apr 30, 2020, 12:53 AM)BasicallyMental Wrote: [ -> ]

(Apr 30, 2020, 12:03 AM)Ben Wrote: [ -> ]So what's the suggestion to improve on this? Sure, we could remove a ton of the backlog of suggestions and only keep the most wanted / essential items, however, we've still got to tend with a small dev and contributor team to be able to actually complete these. The only solution I'm seeing so far is taking on more of them, but they're not exactly coming out of the woodwork at the moment.

A solution to a backlog, is to add more to it? Simply close suggestions and state "We're going to start working through the current approved suggestions. Once we have reduced the amount to a sustainable level, suggestions will be reopened".
It's like me saying me engine is broken, so instead of stopping putting fuel in it, find the issue and drain the fuel out, i'm just gonna keep adding more until it works

Devs like the easy suggestions :) that’s why most don’t get done. Or better fixing the bugs that are pushed without being checked before going live.

That is quite a selfish statement to make.

I think one of the biggest issues that people don't understand is about the development team. They don't work 9-5 and are independently contracted. I know one is currently working a job and is working at university full-time. I know others have full jobs that they are still required to work even with country-wide lockdown around the world.

The "easy suggestions" you talk about are done because they are able to be done like that so we are getting updated content. Those that require additional or further development need time. Everyone on the development team is here by their own choice and we have had developers and contributors come and go because of peoples ungratefulness or lack of understanding on how to give constructive criticism which is one of the reasons the development team is at the current numbers it is now. If I remember correctly, there are only three full-time Lua developers (excludes any contributors). One if busy with a full-time university studying and the other is with a job.

With that said, the content is not an issue. It is the limitations from the rules which about 70% of people here have already stated. There are too many restrictions on what is what - Especially again with the unwritten and what defines what in terms of the rules. What I'd suggest personally from the 10 rule suggestion that was written a long time ago, is do it for a weekend. Hold all current PRs and try a weekend with the 10 only rules. If it is successful, convert into that mindset. If not, then there should be a discussion on it.

I honestly think the problem lies outside of development. The development team is fantastic, and really - always has been. We've been very lucky on this community to have great content, great developers and really smart people coding in new stuff.


Is there processes that can be improved in backlog management, system architecture, prioritization, etc. I'm sure there is - but that's not the problem LL has. The lack of fun definitely isn't due to lack of content. As Blackdog pointed out:
 

Quote:It has to be solved by the team as a whole, and it involves shrinking the rules list, and reducing the heavy handed nature of enforcement, and the microscope level analysis of PRs that result in punishments for a .5s infraction.

He's right. The issue stems from staff and higher ups who determine what is going on in LimeLight. As I stated, until the staff figure out what type of community they're trying to make here it's pointless to argue antics. They're labeled as semi-serious RP but the rules, regulations, HR guidelines, etc. lean far towards Serious RP. 

It's not just on the server, it's also the staff. For whatever reason there's this weird formal style of management on the staff team, like I remember being an admin wasn't fun... If I did anything goofy I'd get an uber-corporate style slap on the wrist. It's not like "bro stop doing that, you're triggering this one dude" and then a conversation about it, it was always like "You are aware of the employee handbook, rule 143.2, you are not allowed to spawn any explosives on the server past 8pm GMT. This behavior is unacceptable and is not reflective of our extremely high standards, this is your first strike". (obviously a made up example, but that was the style). And then there was a secret HR section where all your violations were recorded. Then they started implementing staff minimum requirements, and I'm like bro wtf I'm volunteering my time. 

So I assume there's still this weird way-too-formal staff/HR culture going on which probably trickles down into the community and rules, etc. I remember thinking it would be great if LL could think of themselves more as a business and have direction and long term purpose, but they said "No we're just a gmod community", which eh, it's a matter of preference where you want things to go really. But I do kind of find it funny that they were so anti thinking of LL as a business opportunity, yet they had the most extreme business-like HR regulations. 

But think of it, how are you supposed to have light-hearted fun as an admin when you're going to get called out by HR and have extreme scrutiny whenever you do anything considered fun or edgy *Gasp*.

This tread has simply turned into the biggest argument I've ever seen.

*I am not saying this to insult or be rude.

The FPS is hugely problematic.

The FPS is pretty decent compared to other server. Nonetheless it’ll get better :p

The FPS used to be much worse when we had M.A.D weapons at FL back in the day. Who gives a shit, it was still more fun than this.

(Apr 30, 2020, 10:45 AM)Luvbunny Wrote: [ -> ]The FPS is pretty decent compared to other server. Nonetheless it’ll get better :p

The community I use to play on, who recently started back up which I checked out enables me to maintain 100 FPS when the server is at a low player count (Around 20 players) And it drops to around 60 when they have their peak.

I also hopped onto Fearless (Who I see have no intention to ever fix my corrupted account, whereby my hours and inventory fluxuate in and out of existence and has been reported by me since just after the split) I was getting around 40 FPS in the city when they had 50 players on. And let it be known, there was more props, signs and entities spawned in on main street, than LimeLight typically has. And VCMod was running with a day and night system. 

I was hitting 12 fps on LimeLight when there was 20 people on .-.

I understand though, how difficult it is to fix. The gamemode is some janky arse gamemode that was ported from Gmod 12 to Gmod 13, made well over 10 years ago. Its no easy endeavour to optimise, and the dev team has made great strides to improve the FPS. I distinctly remember the doors being so poorly optimised, and if I'm not mistaken that has been fixed. 

But I do also remember it running better on my R9 280x back in 2014, than it does now on my new hardware. Then again, back then the F1 menu was that really sexy generic Gmod Menu, with the default derma colors. And the HUD was, basic as all hell but hey. The new hud costs me upwards of 30 FPS. To drive, I have to turn it off. 

I like how it looks, but it just has to be turned off sometimes. That and, for some reason there are points around the map where lag is almost guaranteed. And this is true for any map it seems. Rockford ran as poorly as EvoCity, and I can tell you EvoCity is not a poorly optimised map by any means. I'm almost curious to see how well, or badly the New York map we are switching to runs. 

3D Skyboxes can tank performance, and that map has a 3D Skybox constantly rendering to compensate for a reduced view distance. Also its broken up in ways similar to that of Rockford. I could be wrong, but I dont think a map change is going to really help. the community has constantly been going between maps with no success. 

V2D>V33X>V5P>V4B1>Rockford>True North>V4B1 Its never ending. And now we are changing again, and it does in my opinion take time out from actual development. The gamemode doesn't just work with any map, you have to painstakingly set up the points of spawns, petrol stations, NPCs, road names etc. Sure, it is easier than making a new update that adds content, but bygod is it more repetitive. The thought of doing it triggers ADD that I do not have.

Edit: I was just browsing, we should also not forget that Devs have to focus on things like bugs.

(Apr 30, 2020, 11:14 AM)Andromeda Wrote: [ -> ]I understand though, how difficult it is to fix. The gamemode is some janky arse gamemode that was ported from Gmod 12 to Gmod 13, made well over 10 years ago. Its no easy endeavour to optimise, and the dev team has made great strides to improve the FPS. I distinctly remember the doors being so poorly optimised, and if I'm not mistaken that has been fixed. 

But I do also remember it running better on my R9 280x back in 2014, than it does now on my new hardware.

I like how it looks, but it just has to be turned off sometimes. That and, for some reason there are points around the map where lag is almost guaranteed. And this is true for any map it seems. Rockford ran as poorly as EvoCity, and I can tell you EvoCity is not a poorly optimised map by any means. I'm almost curious to see how well, or badly the New York map we are switching to runs. 

3D Skyboxes can tank performance, and that map has a 3D Skybox constantly rendering to compensate for a reduced view distance. Also its broken up in ways similar to that of Rockford. I could be wrong, but I dont think a map change is going to really help. the community has constantly been going between maps with no success.

Door text straight up got removed. It did help FPS a little bit.
Your GPU is almost completely irrelevant to how GMod runs. GMod will crash before most modern GPUs are the performance limiter. It's more nuanced than that ofc. GMod is a CPU bound game. Processors with faster single thread performance will be superior, even if they're 'slower'. Most modern CPUs take advantage of threading to gain performance increases. GMod, being from 2004, uses 1 thread. That's why gmod_multicore_test 1 increases FPS by quite a bit.

HUD is being looked at. In terms of map lag spots, it's because Source is not designed for open worlds. Many modern games use something called frustrum culling where only what the camera can see is what is rendered. Valve's approach, which worked very well for the games they made, is using something called vis leaves which are baked into the map. If you can see a visleaf, everything in that visleaf might get rendered (even if it is behind you!). This causes a bit of a lag when switching from one open area (Nightingale road) to another (Lake).
This is why maps like Southside and Union City can have much higher FPS: by sequestering the city into blocks, you limit what you can see.

3D skyboxes aren't a huge issue. If they affect your FPS then just run 'r_3dsky 0' in console and it's gone

(Apr 30, 2020, 04:49 PM)goigle Wrote: [ -> ]

(Apr 30, 2020, 11:14 AM)Andromeda Wrote: [ -> ]I understand though, how difficult it is to fix. The gamemode is some janky arse gamemode that was ported from Gmod 12 to Gmod 13, made well over 10 years ago. Its no easy endeavour to optimise, and the dev team has made great strides to improve the FPS. I distinctly remember the doors being so poorly optimised, and if I'm not mistaken that has been fixed. 

But I do also remember it running better on my R9 280x back in 2014, than it does now on my new hardware.

I like how it looks, but it just has to be turned off sometimes. That and, for some reason there are points around the map where lag is almost guaranteed. And this is true for any map it seems. Rockford ran as poorly as EvoCity, and I can tell you EvoCity is not a poorly optimised map by any means. I'm almost curious to see how well, or badly the New York map we are switching to runs. 

3D Skyboxes can tank performance, and that map has a 3D Skybox constantly rendering to compensate for a reduced view distance. Also its broken up in ways similar to that of Rockford. I could be wrong, but I dont think a map change is going to really help. the community has constantly been going between maps with no success.

Door text straight up got removed. It did help FPS a little bit.
Your GPU is almost completely irrelevant to how GMod runs. GMod will crash before most modern GPUs are the performance limiter. It's more nuanced than that ofc. GMod is a CPU bound game. Processors with faster single thread performance will be superior, even if they're 'slower'. Most modern CPUs take advantage of threading to gain performance increases. GMod, being from 2004, uses 1 thread. That's why gmod_multicore_test 1 increases FPS by quite a bit.

HUD is being looked at. In terms of map lag spots, it's because Source is not designed for open worlds. Many modern games use something called frustrum culling where only what the camera can see is what is rendered. Valve's approach, which worked very well for the games they made, is using something called vis leaves which are baked into the map. If you can see a visleaf, everything in that visleaf might get rendered (even if it is behind you!). This causes a bit of a lag when switching from one open area (Nightingale road) to another (Lake).
This is why maps like Southside and Union City can have much higher FPS: by sequestering the city into blocks, you limit what you can see.

3D skyboxes aren't a huge issue. If they affect your FPS then just run 'r_3dsky 0' in console and it's gone

Yeah. My CPU used to be an FX-8320, I upgraded the a Ryzen 1700x. I have the Multicore command always running but its not substantially different. Or it hasn't been lately. As for skyboxes and map optimisation, I was aware of the methods used, as a mapper myself I was really impressed with Southside with the amount of detail, and I'm curious to see how well it runs, or doesn't run. As I can't actually tell if it will be good. I really want to take a look at it in Hammer, but I can't be arsed. Its got so much going on, I really hope its as optimised as I think it is. As for 3D Skyboxes, when I played SantosRP? Or whatever that gamemode was called that had a short life span, (or at least I've not seen it for a while now) They had a option to turn off the skybox on their map. And it substantially improved FPS, I never understood why. The skybox in my maps never affected FPS, but then I had never tested them out with lots of people. Could get 300 fps in my map on sandbox mode. Twice what I could get in EvoCity. 

Largely the choice of map has always been a focal point of any 'server-wide reinventions' and virtually every time the wrong elements are focused upon.
Identifying the clear limitations the server has, i.e a literal lack of players, it makes sense to have a small map which is large enough to house variety and offer space, but small enough to bring players together.

If you look at the original maps, V2D and V33X - the common theme (despite being similar maps) is that the city was condensed, but was where most of the action was. Police etc did patrol in the outer areas, and there were more secluded areas - however it wouldn't take too long to see anyone.

All of the past maps since the change from those two have neglected this aspect - if anything the idea has been focused on a map with as much variety for RP as possible, which tends to mean very large/spread out.

Step one, in my opinion, is to revert to a map such as V2D or 3x, which doesn't hurt optimisation. 


---

Step 2 is to bring back agressiveRP as mentioned previously - Regardless of your want for roleplay, and how annoying it was when you were roleplaying for a super agressive rper to just run in and 'ruin' you RP - it forced you to take actions and often forced action.

AgressiveRP also facilitated clanRP, which was a huge factor of why the server was so popular. (i.e big clan events happen weekly - each time would bring 5-15 clan members on the server, thus boosting population). Alongside this, the events usually brought interesting and dynamic changes to the server layout, or often brought new RPs. (I.e country RP, a clan having a base and protecting government etc. This alone would bring hours of content - both for those defending government (or their own base) and those trying to raid. The rule of justifying raiding forced agressiveRPers to make a storyline that puts them against government, and then allowed the government/clan to find ways to get rid of them.

Honestly i've had hundreds of hours of excitement on this server, which through my memory seems to completely stop around the time of Rockford.


---


AgrRP was always the backbone of cityRP for the majority of players- Who honestly wants to roleplay a shopowner for hours on end aside from getting roleplaypoints - And honestly that's fine. It gave interesting builds, and changes to people agressiveRPing and those passiveRPing were often trying to gain cool skins/items for their usual agressiveRPs (i.e gasmask of early days). 


The clan system, which was talked about in depth and planned over two years ago now, would likely have solved much of this issue.


---

And finally, the economy is honestly screwed up. Making money is far too easy now and is worthless - so why would anyone buy moneypacks if they have more than 100 hours on the server? (which is the majority of sustained playerbase.) - if you're looking to increase income, one of the best ways would be to reinvent the whole economy of the server - which most players would certainly not want (considering time invested, let alone real money invested.)

So the chance of this has kind of gone - However having the system of seperate clan bank, and clan inventory would have gave a new life to the economy as players would essentially be starting from scratch for their clan roles ...

Anyway, lots of valid points being made- but as usual they are the same points being made over they years which haven't truly been corrected.

I actually really agree with the suggestions on the first page. There would however need to be trade offs so that you can just go around being a cunt because there is no problem with killing lots of people randomly.

Staff could be moved over to some sort of role where they instead monitor the roleplay and advise on what would be the most suitable thing going forward with the situation for example.

(Apr 30, 2020, 08:49 PM)SpaghettiFTW Wrote: [ -> ]I actually really agree with the suggestions on the first page. There would however need to be trade offs so that you can just go around being a cunt because there is no problem with killing lots of people randomly.

Staff could be moved over to some sort of role where they instead monitor the roleplay and advise on what would be the most suitable thing going forward with the situation for example.

The ideal method would be: Reduce the rules, reduce scrutiny, let players have their freedom and choice for an extended period of time to monitor how the community reacts to it and what they do.

The same was done with the racism rule changes, and all the doomsayers were eventually proven wrong as the server did not descend into memey chaos.

There does not need to be a "trade off". If some guy is being a cunt to you, RP it and get back at him, the way things used to be done.