The same player on official and experimantal server?

[tak]

Hey everybody,
i have gotten a first taste of the experimental server some days ago, when the official server was down for a day (last weekend iirc). I expected both servers to use the same database, however, this was not the case.
I know, working in IT myself, that what i suggest now is a huge change in structure, nevertheless, i think it is the right thing to do.

Both servers need to use the same database.

Every player, and i mean every player i talked to, thinks that having the same character on both the official and experimental server would be an awesome thing to have. In case of downtimes, you could go on playing, train for XP, etc etc. The experimental server would finally see some use and decent testing, the load would be distributed more evenly, maybe even the lag issues would moderate a bit.
I'd love to see soemone from ngd comment on this (of course everyone else as well :beerchug: ).
IMO that would bring the game a step forward. Please share your thoughts on this. Thanks.

[Tyr]

First in the experimental server you get x10 xp gain. Also with some of the changes you have no idea how the structure of the database is modified. The experimental server is the first public test of new features and bug fixes it would be impossible to test if the databases were the same and could cause both servers to crash.

If you follow software devolvement best practices you need a development servers, test servers, and production servers. They are all independent systems and any change starts in the development systems first then is copied to the test systems and finally after QA has approved copied to the production system.

For the lag how are they going to find the root cause if the load is not focused on one server? The only way the lag is going to get fixed if we cause it to happen.

[niclam]

Hello,

I can't possibly explain it better than Tyr (Thank you Tyr, by the way).

Regards

[tak]

First in the experimental server you get x10 xp gain. Also with some of the changes you have no idea how the structure of the database is modified. The experimental server is the first public test of new features and bug fixes it would be impossible to test if the databases were the same and could cause both servers to crash.

If you follow software devolvement best practices you need a development servers, test servers, and production servers. They are all independent systems and any change starts in the development systems first then is copied to the test systems and finally after QA has approved copied to the production system.

For the lag how are they going to find the root cause if the load is not focused on one server? The only way the lag is going to get fixed if we cause it to happen.

Ok, i really didn't make clear what i want. I do not want ngd to use only one database. I want to be able to use my official character on the experimental server. I did not know about the x10 XP gain bonus (why is that implemented anyways?) but that is the only hurdle and that could likely be changed.
Something like a nightly cron job that dumps player data from one database to import in the other could do what i want. Developmental features would get some stress testing that way before making the transition to official. So i still think that having the same playerbase (the same database, yeah it was a crappy explanation of my intention) on both servers
would be beneficial to users and developers alike.

[Tyr]

The x10 xp gain is to help leveling so you can test any class at any level and not take forever leveling.

They do port the official data to the experimental server but how often they do that all depends on what is being worked on. I'm sure the structure of the database is changing all the time and putting in the effort to keep the data mappings is not practical. For the stress testing from time to time there is a server message to go to the experimental server granted it is in Spanish but it is not hard to guess what it means. The hard part is guessing what they need help testing. Then again most of the players speak Spanish so I understand why that is.

[tak]

The x10 XP mod could well be dropped, as you have explained it is a crutchthat would not be needed. There are at best a few ppl on the server, none of whom would take the pain to go through leveling with a character they won't really use, no matter the XP bonus.
I doubt that at this point in development there are any significant structural changes occurring to the database. I am also sure the official database is a subset of the development database, so just importing the values from official to dev should not be that big an issue. There would be a one time effort in writing a script that exports and imports data, afterwards that script would maybe need to be adjusted to include a new value here and there.

But those are implementation details, i do not want to argue about export scripts, only the guys from ngd know the table structure. I would even be willing to write that script if someone asked me to.

I see that at max there are a handful players on the dev server, a few more when the official is down. So the server is atm pretty much useless.
Once again: Having the official character on the dev server as well would be an incentive for ppl to go there from time to time, check out new features and write bug reports. There would be real load on the new features when Ra is down, not just like 60 ppl but 300. All sides win. So please, do not be irritated by my first ass-backwards attempt to explain what i want, i know what i am talking about. Take it at least into serious consideration. Thats all i ask for. Thanks for that and all (further) input.

[niclam]

Hi,

The idea of having the live database copied once in a while to the experimental server wouldn't be a bad one in other circumstances but:
- The experimental server will probably always be just one. The amount of live worlds will increase someday (fingers crossed).
- The experimental server is not just the pre-staging server but our sandbox our well. We usually do a lot of testing there, from quests to new features. While we do this, we tend to free ourselves of having to do support or being careful not to disturb the normal course of the game, this includes recalling all players into an area, spawning dangerous monsters at will, giving items away randomly at users, etc. So, sometimes the less users we have there, the better (don't think the amount of users we have now but in the future, where 30 could as well be 300).
- Our idea is that 'massive' tests on the experimental server are conducted on-demand, this means that we have to be there watching and recollecting statistics.
- Bug reports on the experimental servers are usually taken with a grain of salt because, honestly, there is stuff broken all the time.

That's just about it. And of course, replication is always a big issue when there is a lot of different systems interconnected, so adding one more replication part to worry about it's out of the question with our current budget and personnel (characters could broke if they are copied and we did some structural changes, so it wouldnt be as simple as doing a cron job with database replication).

Regards

[tak]

Hi niclam,
thanks for taking the time and explaining the inner workings a bit :wink: too bad i'll never see my warlock on the server but if its really the developers playground i can understand, i thought you maybe had another box somewhere off limits to the public.
As for having more live servers in the future i doubt that not a moment, i haven't played any game so much in ages, i even bought a new video card :sifflote: i didn't think i'd see another game i'd get so hooked on (the last was Q3, and that was a success wasn't it? :biggrin: )
I'd like to say thanks for making this great game available free of charge for beta testing, and especially say thank you for the linux client.
Keep up the good work, i look forward to seeing this game go gold. Thanks!

[niclam]

Well,

Having another server off limits was on our minds but due to budget constraints and that sometimes we really need the users help in testing new stuff, we decided against it.

Also, please keep in mind that the game will continue to be free even after gold (but will have premium content as well) and the Linux version will continue to exist.

And whenever we are on features at the time of launch (we might be missing a thing or two), it will be the beginning, not the end...

Regards

[Angelwinged_Devil]

Also, please keep in mind that the game will continue to be free even after gold (but will have premium content as well) and the Linux version will continue to exist.

Regards

yay thanks :D, will it then be official ^^? I don't mind helping out a bit if it does, when I learn to code XD. For now, my studies are calling first

[tak]

Well,

Having another server off limits was on our minds but due to budget constraints and that sometimes we really need the users help in testing new stuff, we decided against it.

Also, please keep in mind that the game will continue to be free even after gold (but will have premium content as well) and the Linux version will continue to exist.

And whenever we are on features at the time of launch (we might be missing a thing or two), it will be the beginning, not the end...

Regards

The features you mention lead me to a thing that crossed my mind quite a few times and that could give Regnum another edge over the competition. You have attracted me (and a lot of others, there is a huge thread about it on the forums as well) to Regnum at first because of the native linux client. The linux community is known for the spirit of working together on projects and lending time and energy to become a part of them.
You should IMO use the people you have attracted to help with features of the game, be it with translation (there is a thread about this here, too) or for
example with modeling contests, something where you say: Use these tools, these are the design parameters, create monsters/effects/good looking textures/whatever, there are loads of tasks that could be given to people that would lend the developers time to focus on the basic game engine stuff. I for one would be delighted to have a part in the game,
lend my time and energy for a cause like this.
I do of course not know the legal ramifications, but would not have a problem with restrictive licensing if necessary, or giving up copyright of designed stuff to ngd. I think others would not mind, either. Use the community, Luke :wink:

[niclam]

It's an interesting point what you just said tak.
We have thought a couple of times about having a volunteers programme, that would allow people from outside NGD to participate into the game's development.

The main problems were also what you mentioned: legal issues and copyrights. If we are going to do this, there is no way we can pay the people back, so the licensing would be restrictive as it usually is on this kind of matters.
And last but not least, we need to have somebody on the team as the lead volunteer manager, that can organize and communicate with everyone to check for progress. There are a lot of procedures involved into the content pipeline that makes it difficult to add new stuff without having to worry about things like consistency, testing & QA, balance, etc.

Anyway, Im sure there are plenty of people willing to participate and we will surely have them on the credits so they can show off that on their CV's.
So, maybe after launch where we can hire more personnel, we could approach this.

Regards

[tak]

It's an interesting point what you just said tak.
We have thought a couple of times about having a volunteers programme, that would allow people from outside NGD to participate into the game's development.

The main problems were also what you mentioned: legal issues and copyrights. If we are going to do this, there is no way we can pay the people back, so the licensing would be restrictive as it usually is on this kind of matters.

Yes, it is too bad that legal issues make it hard to implement such exciting possibilities. I hope you will find the money to have a lawyer set up a legal waiver that makes it clear that by participating in development you agree to waive your owner rights, agree to not use content that belongs to other people etc etc. The payment is rather unproblematic IMO, for example for me it would be enough to see "my" monster or to be able to say " i designed this" (maybe a bit of Ximerin would be nice of course :p ) others might naturally agree or disagree.


And last but not least, we need to have somebody on the team as the lead volunteer manager, that can organize and communicate with everyone to check for progress. There are a lot of procedures involved into the content pipeline that makes it difficult to add new stuff without having to worry about things like consistency, testing & QA, balance, etc.

Yes, the additional strain of sifting through (hopefully) lots of user generated content, most of it unusable, and commiting it to the regular process without having someone close by to ask about the details would have an impact on workload. I think it would be worth it overall, though.


Anyway, Im sure there are plenty of people willing to participate and we will surely have them on the credits so they can show off that on their CV's.
So, maybe after launch where we can hire more personnel, we could approach this.

It would really really be great to be able to do that. Having the perspective to be able to do that, at least after launch is something that makes me happy.
I hope you find the man-time, money and consensus of your team to do it, the sooner the better of course. I put my money where my mouth is and tell you that you only need to contact me, i would be excited to have the possibility to participate in your project. I know others who are qualified who feel the same way.