Not even novice programmers 2 cents
Has the funcom team ever thought of changing the game engine to include a prioritised system for loading textures. I think we all agree textures are the problem, and that ping is not. Maybe the information being sent to computers to describe the items and status of a player could be packeted better to decrease actual lag but graphics handling seems to be the problem.
When a player first zones into an area what is wrong with them seeing placeholder textures of a 1x1 white pic. Seriously if I had the option to play like that always I would at times. After this initial zone details of players would stream to the graphics card not get kicked to it all at once... Suppose a data stream cap for loading of these types of asthetic items. And important data such as who is hitting my character and who I am hitting should be prioritised w/o a datastream cap.... if I don't ever get to see pretty textures but live and then after a fight start to get the pretty graphics so be it... I lived, but the other way around is where being annoyed at the game comes from.
As for the idea of capping the number of players in a zone or even fighting area/radius I say a big NO. If I was to be protecting my guild's tower and 100 lvl 200 characters from another guild came storming in I'd think to myself "holy crap".... and then "that was cool as all hell" PLEASE PLEASE PLEASE DON'T LIMIT PLAYERS PER AREA Just use any method to make the players visible and attackable without being graphic intensive till the datastream gives a break for such things to be loaded.
P.S. If anyone knows a method of replacing AO textures all with the same referenced 1x1 texture let me know :)
I'd rather see a white set of polys hittin me in real time than an awesome graphic 1fps and being dead before my keyboard responds to commands.
Feel free to e-mail burns, suggestions and even the before mentioned method to my attention at lung@nyc.rr.com .... I'll be killin that addy in a few weeks anyways. (I won't read replies here)
These times call for caution, and some investment Funcom...
Ok, the theory may be sound, but it's only that, theory.
I definitely did not like the whole "pointing fingers" at low-end computer owners. I have a high-end system. My average ping is about 67. I still experience lag though. Even when my ping is around 80, massive lag occurs occasionaly that can't be associated with my system, since I can do a full blown virus scan of my whole HD and still maintain a 73 ping without lag or other gameplay performance issues.
Funcom, listen up you silly goofballs. Do some research, invest in some better rack systems with hot-swappable components and redundant capabilities, and of course, spend some more time on these patches (aka gameplay sabotage). A good place to start is to make sure ALL of your servers are running on a multiprocessor based system, of course, at this point in the game, if you have a multiprocessor system but the software wasn't originally designed to use it, I guess we're all SOL.
Perhaps instead of implementing a "crowd" control system, how about an alternative. A typical day in Omni-1 Trade (or from AO Chat) consists of people running scripts listing all the stuff they're selling, while they hang out by the grid terminal/bank area (ashtray, keyboard for chat folks). This is what FPS players call camping (amongst other things :) ).
Here's the solution:
Impliment player commerce terminals. The terminal will allow someone, with items to sell, to place them in a terminal and set a price for each item they submit. For every day they have an item to sell in the terminal, the seller pays a small handling fee to Omni-1, the Clans, etc. After that, a potential buyer may access the terminal, look through some different categories so they don't have a couple hundred items to search through, select the category, select a sub-category, then select the item they like and purchase it. The credits are automatically transferred to the seller and that's it. No more campers in Omni-1 Trade or AO Chat running scripts.
It would be a good idea to remove the ability to advertise/run scripts from outside the game through the AO Chat. I mean, is the person really playing if they aren't logged into the world? Is waiting for a response in AO Chat, then logging in for 2 minutes while it is convenient, then logging out and returning to a chat window that's not "in-game" really role-playing? It's like a demented version of pop-up ads for AO. Sure, you can just not "show" those messages, but the servers still have to queue up those scripted sale items and transmit them with all the other packets even if they aren't shown.
My 2 cents, hopefully, they'll collect some interest.
-Crimsus (WHAT'S THAT PRETZEL DOING?!!!)
Are towers meant to be permanent?
If towers are meant to be permanent fixtures requiring 24/7 defending and maintenance, then this will naturally favor the development of massive, international guild structures to support them: ergo greater probability for large wars. If towers are meant to be temporary, operating within the range of a typical login session, then the guild structures dont have to be so massive and things get a bit simpler.
Promote smaller guilds. Make it more expensive to maintain larger guilds. That should help.
Brainstorming - bugs, balance and content issues and possible solutions (Part 1)
Dear Funcom people, dear AO players,
As I want to be constructive as I love this game and want it to succeed, I'll divide this very long article into five subchapters:
I) Why the hell do I write this?
II) Bugs to be fixed
III) Suggestions to improve performance (especially on client)
IV) Balance Issues (continued in part 2)
V) Conceptual Issues (continued in part 2)
I hope the following will be helpful to make AO more fun again.
I) Why the hell do I write this?
I am a GUI and systems programmer, so please excuse if I sometimes speak quite technically, even if I tried to express all technical stuff in a manner that everybody can understand.
I originally didn't intend to comment on the things going on here, I'd prefer only to play for leisure.
But the last days of playing were quite frustrating.
I don't want to whine, but you might be curious what made me stopping playing now to write this, even though I planned to enjoy a long playing evening.
After the last client-side patch the overall program performance has been substantially decreased. I don't know the reason, but it might either be due to debugging or workaround code or due to the modifications of the graphics engine (as mentioned in the patch description). This leads to very bad frame rates when apparently loading new picture elements after zoning, turning, entering new rooms etc.
The introduction of NPC pets appears not to be well balanced yet, this still needs fine-tuning.
Let me explain. The thing that actually made me stop playing because of frustration was that it made it effectively impossible to do solo missions with a sufficient success chance. I died 5 times due to two problems/unbalances due to summoned pets, rendering many hours aof playing effortless by eventually losing any gained experience. After this I was simply fed up. See more below in the suggestions chapter.
And then there was the article by Gaute Godager regarding the upcoming CL system.
Before I discuss it in more detail, let me tell what I saw on RK-3 the day before the usage of beacon warp to bring people to Camelot has been forbidden.
A group of Omnis was going for Tarasque and the pvp zone in the castle.
Some time after they worried how to get their team to Camelot, some had the idea to make a "teleporting team" only for transporting Omnis to the castle. The principle was easy: Engineer in castle, helper in Omni-Ent, the helper invited people, the engineer warped them to camelot, and then they left team for others to be warped after them.
This eventually led to massive crowding, after warping a quite large part of the Omnis being online into the castle and Tara's dungeon. Many people got LD, people entering the zone by feet were unable to do anything useful when the frame rate went far below 1.
I hink this incident was the final trigger for some of the latest developments.
The recent patch mayhem and the CL system discussion together make me really worry where Funcom is going.
It looks to me as if they had no plan what to do, and were in a more or less helpless situation.
This bugged and apparently not balance-tested patch opens up some questions:
Did Funcom not test this sufficiently? I can't believe. It seems more probably to me that something pressed them to make the unfinihed patches, maybe some investor relations? Or was it an internal short-circuit of Gaute? Does Gaute really know the true maturity state of the code? Or is he be lied at by programmers he pressed too much to finish the patch on a fixed day? Questions over questions.
But these are not really questions we customers should have to ask. So let's go on to the next section.
II) Bugs to be fixed
Not all of the listed here are real bugs, in fact many of them can be considered as design flaws.
- "In-Fight-Bug"
Description: One of the more annoying bugs. Almost every AO player already experienced it, it disturbs most after using teleporters.
Solution: Have all monsters clear the aggro flag after using a teleporter. This behaviour would be the most "realistic", as everyone would expect. I have heard that Funcom is of the opinion that the behaviour called by the players "the in-fight-bug" is the intended way things should go. Funcom, rethink please.
-Door Bug
In fact, it's a design flaw. Doors are treated as opaque (not transparent) when calculating what item was being clicked at, even if they are open (graphically transparent, i.e. the area behind is visible).
That is, if you are standing in front of an open door and want to select the mob that is attacking you from the other side of the open door, you (almost) always get the door as target.
This behaviour is unnatural, it could easily be fixed by applying the same transparency calculation when doing the frame display calculations.
However... if it is really true that AO's engine is unable to do the calculations which objects are being obscured by others itself, and leaves these calculations to the video card GPU, as it is hinted by some posters in this forum, then a performance overhaul of the engine is overdue.
This could include adding some cache structures, for example. Such cache memory management would be extremely helpful to avoid excessive disk accesses and to reduce OS swapping. Also the results of the door transparency calculation of the display calculation could be cached for the case that the user clicks onto an object behind the door.
-"Out-of-Range" bug
Very similar to the door bug. Mob with direct line-of-sight to the player is falsely determined as "out of range" because of slight height differences or small things lying between player and mob, resulting player being defenseless against such mobs attacking him until finding a point on same altitude and unobstructed space from where the engine allows him to attack.
It is mostly caused by
a) Mob either standing on slightly higher or lower ground, i.e. different altitude
b) Small things (especially globes, stones and rocks) that lie between player and mob. Not long ago tables in mission rooms caused the same effect. The fact that this got fixed shows that FC is aware of this problem.
The solution of the "out-of-range" bug would be the same with the solution of the door bug above.
-The "Teammate-Is-Invisible"-Bug
This bug is less a bug than a desynch problem. Sometimes teammates cannot be seen even if standing beneath each other. This is a frequent bug in missions, caused by the clients not getting updated correctly about the positions of the team mates.
Even though this bug is easily to work around by walking out of the other's vicinity zone (i.e. until the display of the team mates in the team window disappears) and back in to enforce a resynch, it should be a priority to Funcom.
The solution is extremely simple. Introduce a checksum consisting of the coordinates of the team members in the mission. If a client gets out of synch, then the server can initiate a respective client update.
-The "Didn't get Mission" bug
Sometimes nothing happens after clicking "accept" on the mission terminal, the player does not get the selected mission. This is particularly annoying if somebody was taking the tedious manual approach to find missions with rewards that are needed for upgrading the own character.
Yes, many people do this without clicksaver, many dont even know of clicksaver.
So please don't take this bug easy, FC. Searching for a matching mission is tedious, and losing the time invested due to this stupid bug is annoying.
So, FC, please fix this. It might mean to intoduce some client/server negotiations to ensure that a mission accept request does not get dropped. Shouldn't bee too hard to implement.
O.K. These were the bugs with the biggest annoyance factor.
They should be Funcom's highest priority.
III) Suggestions to improve performance (especially on client)
Now my 2 cents on improving runtime performance.
-First of all, please introduce an client option like "Display all chars in faction/profession uniform", that displays people only in uniform outfits. Maybe blue for Omni, red for clan and grey for Neuts. I'd like very much when the uniform would also indicate the profession, this would make a hinting displaying of weapons unnecessary.
The beneficiary effect of this would be pushing up the frame rates, the more, the lower the client's system is.
For my part, I would prefer an overcrowding system in the form of enforced display of uniform faction/profession when reaching a predetermined threshold of players in an area.
Anyway, I probably would have activated such an option most of the time. I simply dont care much for the clothes and looks of the other players. Instead I am annoyed by low frame rates, so
...
My suggestion also would be a scalable solution, where the user can decide himself how much texture loading lag he is willing to accept. Gaute's approach isn't.
After implementing a such solution, Funcom would gain more time and reduce the pressure to overhaul their graphics engine.
This overhaul of the graphics engine should be Funcom's priority.
I will not go into depth here, this topic has been discussed by several others quite detailed.
This would consist of many steps, the most important of them would be to implement an effective Memory Garbage collection. (i.e. cleaning up unused memory instead of piling up an ever-growing chunk of trashed swap memory). It is completely sick how long the garbage cleanup after a longer AO playing session takes.
Easy solition approach: Implement some caching texture memory lookup management structure with the elements being not used for the longest time being deleted first when physical memory gets low. This would drastically reduce texture load freezes and also help avoid excess swapping.
It is mad that the game loads meshes, textures etc from HD again and again even though they are already in memory. Apparently, currently the Windows HDD cache is the only caching structure in effect here :/
These two things would be quite simple to implement, and would really make a performance boost.
This again could help Funcom to bridge the time until they got their engine adequately optimized, or even make a new one.
(cont'd)
Brainstorming - bugs, balance and content issues and possible solutions (Part 2)
IV) Balance Issues
First of all... AO claims to be a pvp game...
it's quite suicidal to introduce expansions like "Notum Wars" before balancing out the different professions/classes in pvp terms...
This game lives from the multitude of professions. But the professions are far away from having halfway balanced chances in pvp combat.
What might be the consequence? Professions being bad in pvp due to missing inter-class-pvp balancing will be less attractive. Thus the "weaker" professions will become even less attractive. As the time passes, they will be played less and less. But imagine a monoculture of a few professions, while weaker professions, like bureaucrats, for example, might become very rare...
I don't think that is what Funcom wants.
O.K. I think many people will agree with me that this has to be balanced somewhat more than now before the release of "Notum Wars".
It is a must, failure to do so might be catastrophic to the long-term success of AO.
And it should be not too difficult to evaluate and implement appropriate inter-class pvp balance rules/formulae. Really, there is no excuse for not doing so.
Gaute, to solve this, here is my suggestion:
Collect up a team, either or both of experienced pvp beta testers and players. Let them evaluate a fair concept in which way pvp damage should be modified. This could consist in damage or critical chance modifications, or in modified nano effects in pvp combat. The motto should be: Every profession has a chance to survive pvp combat against any other profession in the same level.
This does not mean "equal chances", but as long as you often can tell who will win simply after being told what profession both duelists are, there will be less motivation to play "weak" professions.
Instead it should be the way that wits and skill decides the outcome of a pvp fight more than only the ql and quality of the equipment.
Thinking this way, a viable approach could also consist of non-linear damage recalculations (reductions) in pvp. Then simply grossly uberequipping one's char will no longer guarantee victory, as higher damage output would be reduced more than lower damage output, and this also could be even fine-tuned depending on the class/profession.
And then... with the recent patch a new balance problem...
The new balance issues with the NPC pets are well known. They also pose a problem what could be considered as a bug, but let me explain this later.
To me, it was "enough" after my NT, lvl 82, was killed the 5th time in some hours, again losing all exp, because being confronted by a bunch of big slayerdroids after encountering some pairs of engineer mob in soloing ql 90 missions.
The situations were severed by the "door bug" (described above) making targeting the aggressive mobs/pets difficult.
And then there a phenomenon, which is a design flaw, comes in. As people and NPC can overlap themselves on a coordinate position, you cannot click-select an entity that "engulfs" you this way.
But slayerdroids are very big, and cave mission rooms are small - it is almost impossible to click-select a slayerdroid pet in many cave dungeon rooms. So it can kill you easily before you are able to pull it out to a larger room so you can select it.
Anyway, what really pissed me off and let me stop playing and write this article, was, when the NT opened a door and saw a cross-shaped room with 2 engineer NPC in it. Then the first of them jumped at the NT, while she calmed him. And then two more engineers jumped in from the long cross-shaped room... and 4 big slayerdroids running to the poor little NT in its small room "engulfing" her...
I don't whine, but i find this quite a bit off-balance, at least as compared to the pre-patch state.
It is only a game, i want to to enjoy it and not be frustrated by uncalculable death-and-lose-exp-etc-situations like this.
Suggestions:
-Limit the number of allowable newly spawned NPC pets in a given area, maybe depending also on the team size.
-Reduce the pet's ql generally, but (NPC level x 0.7) might not be the ideal formula. Maybe either this formula or at least the lowest "green" level to the average team player level, whatever is higher. Or such.
It is simply not acceptable to sometimes being confronted by no-survival-possible-situations at random like the described ones, when the NPC pet(s) got randomly a far higher level of the player/the mission.
Again, it is a game only, it should be fun and no tedious challenge for hardcore overtwinkers.
-Reduce the size of the higher level NPC pets so that they become click-targetable in small mission rooms. This should be very easy by changing their display size factor. It would avoid some frustration and also enhance the graphic impression (a slayerdroid from inside in a small mission room does not look impressive at all :P)
I think this would make the NPC pets far more enjoyable.
V) Conceptual Issues
Hmmm... "Content"...
Diagnosis: Desperately missing :/
This is so sad, the game lacks the vividity it's great storyline suggests.
My impression is that Funcom hopes that the expansions will do a miracle and add some "content".
Let's hope, that they make it!
But i fear, that they are acting short-sighted.
This bugged and badly balanced patch should not have been done!
Anyway, as I like the game and would like to help to fill it with "life", with "vividity", so here my suggestions:
Install an event planning team, that should do the following functions:
a) Planning, preparing and supervising minor and major events, these might be
-storyline related events (political conflict)
-event-related (catastrophic event, non-political)
-profession related (profession-specific alternating dungeons) or
-season-related (Halloween etc)
b) Planning and designing event specific reward items (as a materialistic motivation to join the
events)
These should always have some special abilities like the Brock, Nodda and Bobic quest rewards, because these make these quests attractive even for higher level players.
c) Collect ideas for events, and observe the situation on the different servers, to be able to plan and launch events this way, that as many people as possible learn of them and have opportunity to participate
And now to "Notum Wars"...
I know too little about it, but a few thoughts:
Can it be guaranteed that also smaller guilds will still have fun? It sounds like as smaller organizations could be pressed to fusion to survive... maybe this is not what players want to pay for...
Hope this has been thought of and cared for by Funcom in a way that is enjoyable for all sides
I hope Funcom also thinks of all possible abuses that could lead to frustration on the attacked guild's side, by "spamming" attacks etc, as other posters mentioned.
Not an easy job...
A last thing I'd like to ask you:
Please (!) implement a pvp enable/disable flagging like in Neocron. And abolish the stupid "suppression gas" system outside the cities.
When with pvp enable/disable flag (to avoid backstabbing, perhaps only settable in town or with a time delay to warn nearby people), there is no need for suppression gas zones. Anyway, there still could be some mayhem zones left, but no quest area should be mayhem zone.
This in turn would be far more "realistic", by allowing potential duels everywhere where claner and omnis meet in the wilderness, as it should be and also is being suggested by the storyline...
I don't think that anybody would protest against this solution - please implement it ASAP!
Please, Funcom, DO something very soon!
At the moment I must tell, Auno (see www.auno.org) is right... the current state of the game after this patch is so depressing, that I lost the motivation to continue playing...
Really, Funcom, please get your act together, either I might follow him with suspending my subscription :(
O.K. Gaute - I have replied to what you asked and I hope to have helped with my detailed answer, even if you probably expected to read other things than those.
I hope you'll do the right thing(s) now :)
Yours
Katerchen
P.S.: Please also read the articles of Solstyce, Soulmartyr and firedup and all the others I missed, they make good suggestions!