It is currently Sun Dec 15, 2019 8:37 am

All times are UTC




Forum locked This topic is locked, you cannot edit posts or make further replies.  [ 38 posts ]  Go to page 1, 2, 3  Next
Author Message
PostPosted: Fri Oct 22, 2010 2:10 pm 
Offline
Obduction Backer

Joined: Tue May 09, 2006 8:06 pm
Posts: 703
Location: Virginia, US
To recap: the MOULa servers up and running, we have a working client, and we have the means to produce our own MOULa ages. However, there's currently no way to test these ages in a multiplayer MOULa environment.*

That's why Rich/RAWA and Mark/Chogon suggested that the next step is to develop test server software: to complete the pipeline for fan-created ages on the MOULa servers. Fans would download, install and run the software, enabling age developers a save haven to test their ages without interfering with the main server.

The problem is that there's currently not enough time, staff nor money to cover the development.

The time and staff issue may temporarily disappear. Stoneship and its first revision have shipped, and testing for iRiven is wrapping up. And since there have not been any calls for beta tester for the next chapter of MagiQuest, that is probably on hold for the time being.**

So it comes down to finding the funding source. Here are some suggestions.

Option 1: Withdraw money from the MOULa server funds to cover development. This would dip the CAVCON level down to 2 - or even 1. If this happens, I would pledge to make a donation every week until it goes back up to at least CAVCON 3.

Option 2: Ask for separate pledges to cover development. Each week, RAWA's CAVCON updates could include an additional metric: an indicator of the total funds received to cover test server development. They don't need to disclose an amount - just a percentage. When people make a PayPal donation, they can include the phrase 'test server' in the comments. That will signal Cyan to put those funds into the server development account.

Option 3: A combination of the top two options. Half the money comes from the MOULa server funds, and the other half will come from separate pledges.

What are your opinions on these options: which ones would you support? If none of them are appealing, what other way could funds be obtained to cover development of test server binaries?

---

* The PlasmaClient might be able to be extended so that it could test a MOULa age, but it would be offline/standalone only. And while Deep Island and other shards do host fan ages in a multiplayer environment, it is using the URU:TPOTS code base, not MOULa.

** This needs confirmation from Cyan/Creative Kingdoms staff.

_________________
KI #: 1299


Top
 Profile  
Reply with quote  
 Post subject:
PostPosted: Fri Oct 22, 2010 3:29 pm 
Offline
Obduction Backer

Joined: Tue May 09, 2006 7:39 pm
Posts: 726
Location: Cullman, Alabama
I think separate donations would be the way to go. To touch the CAVCON donations for anything else than the MOULa server, although it is a related need, might alienate some donors. I would gladly split my monthly donation between the two, increasing the total a bit, of course.

_________________
MOULagain KI: 00042417 00032066


Top
 Profile  
Reply with quote  
 Post subject:
PostPosted: Fri Oct 22, 2010 5:39 pm 
Offline
Former MystOnline Moderator

Joined: Fri Nov 10, 2006 3:05 pm
Posts: 4203
Location: 56°2'26", -3°20'28"
Personally, I'm inclined to think that Cyan will go down the route of Option 1. CAVCON 4 suggests there isn't enough going into the pot to fund any worthwhile content development, even if resources do become a bit more available, so what do you do with the balance left over after keeping the current servers paid up? Just keep it sitting doing nothing particularly useful for a while longer in anticipation of something dramatic happening, or put it to some use that will actually provide a "benefit", like setting up a test server?

I'm guessing that within Cyan's Amazon account, pulling in another couple of hosts to create a test (or rehearsal) server isn't going to be a big deal either in terms of cost or setup effort. But there will be some cost to running the additional server, and that may mean that the sustained CAVCON 4 we've had will drop to CAVCON 3 and I take Wafna's point that some people might not agree that a test server is a good use of their donation :?

I'm also a bit sceptical over how easy it will be to arrange for distinct donations to fund a test server: Again I'm guessing but maybe the effort of setting up the necessary accounts, modifying webpages, etc., isn't going to be too appealing when they could just use the pot that's already there.

_________________
Image Mac - MOULagain KI#00004826 00004289
In the interests of the environment, this post has been constructed entirely from recycled electrons.


Top
 Profile  
Reply with quote  
 Post subject:
PostPosted: Fri Oct 22, 2010 5:51 pm 
Offline
Obduction Backer

Joined: Tue May 09, 2006 8:06 pm
Posts: 703
Location: Virginia, US
Mac_Fife: in my OP, I was thinking that Cyan should not run a test server, due to the continued funds required to support it. Rather, this post was about funding the development of test server binaries that would be passed on to the fan base - who would run their own servers.

But that assumed that the funding for developing such binaries would be much less than what will be required to sustain a high-capacity test server.

_________________
KI #: 1299


Top
 Profile  
Reply with quote  
 Post subject:
PostPosted: Fri Oct 22, 2010 6:01 pm 
Offline

Joined: Thu Aug 03, 2006 10:55 pm
Posts: 625
I don't think a test server will be high-capacity until Cyan has a policy in place for submitting ages - there aren't many of us working on server-connected projects. That being the case, it really doesn't make sense for Cyan to run a server yet. Not until they have a submission process.

I would like to see a binary released so I don't accidentally take down the auth server someday working on PC.

_________________
MOULagain KI #: 66990

When I was your age, we rocket-jumped up hill both ways in boiling lava.


Top
 Profile  
Reply with quote  
 Post subject:
PostPosted: Fri Oct 22, 2010 6:20 pm 
Offline
Former MystOnline Moderator

Joined: Fri Nov 10, 2006 3:05 pm
Posts: 4203
Location: 56°2'26", -3°20'28"
Yeah, OK Robert, I didn't pick up what you had in mind from the OP.

And I suspect your assumption on relative cost is probably a reasonable one. I'd guess the binaries are pretty much there already, I'm just not sure on how they'd go on licences. Until such time as they get round to Open Sourcing the server code, I imagine Cyan would still want to have some EULA restricting how the binaries are used and what is served, i.e. ensuring that they're only used to test Fan Ages.

As PaladinOfKaos mentions, irrespective of "third party" test shards, I would think that Cyan may want their own rehearsal server for anything they're thinking of adding to MOULa. On the other hand, they might decide that MOULa stays as it is and all fan stuff remains on fan run servers. Who knows at his point?

_________________
Image Mac - MOULagain KI#00004826 00004289
In the interests of the environment, this post has been constructed entirely from recycled electrons.


Top
 Profile  
Reply with quote  
PostPosted: Fri Oct 22, 2010 6:35 pm 
Offline
Obduction Backer

Joined: Mon May 15, 2006 2:02 pm
Posts: 818
Location: Switzerland
Robert The Rebuilder wrote:
That's why Rich/RAWA and Mark/Chogon suggested that the next step is to develop test server software: to complete the pipeline for fan-created ages on the MOULa servers. Fans would download, install and run the software, enabling age developers a save haven to test their ages without interfering with the main server.

Can you point to where they suggested that? The one mention of a test server I remember is here, and as far as I understand, that server was to be run by Cyan, not by fans, and used for testing clients, not fan ages.

Robert The Rebuilder wrote:
testing for iRiven is wrapping up

Where did you get that from?


As to your question, I would very much like to have the option to donate specifically to the open-source effort (I’ll include a test server in that) and would gladly give more than I currently do for keeping MOULa up. (Of course, in that case, I’d also appreciate some reporting of what is being done using my money.)


Top
 Profile  
Reply with quote  
 Post subject:
PostPosted: Fri Oct 22, 2010 6:57 pm 
Offline

Joined: Thu Mar 11, 2010 5:25 pm
Posts: 158
Location: Aberystwyth, Wales
Personally, I don't want the binaries, I want the source code. If we had the source code, then fan teams could build better software, for example with better lag compensation so people don't warp all over the screen!

I would like to see a little more transparency on the Cavcon meter. Numbers of registered is fine and all, but a tally of how much was donated would be nice. The developer of Dwarf Fortress does this, for example. If people were willing to come forward publicly with how much they donated, you could do this without any Cyan time whatsoever. :P

I would donate if I knew how much was in the pot.

Again though, Cyan have said that they're very reluctant to go to Cavcon 5. That's understandable, as 5 means that Cyan at that point is almost obliged to work on MOUL which draws staff off their money-making projects. Which is why we need to see how much money is actually going into the donation pot. Sitting at Cavcon 4 does not mean that eventually we will hit 5. To hit 5 needs a BIG amount of money to justify the cost of a developer's time and then some. Add into this that Cyan don't have a great deal of people left who are experienced with Plasma and you have the current situation, I feel.

In case anyone's wondering, the average game developer makes about 4-6k a month. Add in the fact that that developer is working on something which isn't going to make a profit, maybe ever, and the number goes up.

_________________
I am a member of Team OSCAR. I make content for MOULa. Help OSCAR here: http://forums.openuru.org/viewforum.php?f=103


Top
 Profile  
Reply with quote  
 Post subject:
PostPosted: Fri Oct 22, 2010 7:46 pm 
Offline

Joined: Thu May 11, 2006 5:22 pm
Posts: 1814
Location: California
Robert's idea that fans run the server is probably the best 'operational' solution. A test server tends to crash and need restarts as well as posting data about the crash. There is going to be a higher level of effort required to run the test server.

Unless Cyan can say they have the time and staff to devote to it, I think we would be adding another road block by asking them to run the test server. Plus how they accept ages for testing is going to be a whole big thing, which I think is less of an issue for a community/fan run test server. But, I suspect this is something that Cyan wants to do and we may not be given the option.

Getting binaries might be the only solution available to move things forward now. I suspect we will find bugs that are suspected to be in the server code. Not having the source is going to hamstring efforts to deal with those. Plus binaries while they will help the reverse engineering efforts, they require those efforts.

I think Mac_Fife is right that with binaries Cyan will want an EULA. If they are going to take the time to write a EULA, why not devote that time and effort to providing us the open source and content licenses and turn the community loose? I can't image how most of the open source issues would not have to be addressed for a server EULA, so...

I suspect they can't license the server binaries, EULA or other, because the working code will have Oracle and other proprietary code that can't be licensed even in binary form...

/me waves 'Licenses Now' sign

_________________
Nalates - GoC - 418 - MOULa I: Nal KI#00 083 543, MOULa II: KI#00 583 875Nalates 111451 - Second Life: Nalates Urriah
Guild of Cartographers Image


Top
 Profile  
Reply with quote  
 Post subject:
PostPosted: Fri Oct 22, 2010 8:28 pm 
Offline
Former MystOnline Moderator

Joined: Fri Nov 10, 2006 3:05 pm
Posts: 4203
Location: 56°2'26", -3°20'28"
Nalates wrote:
I suspect they can't license the server binaries, EULA or other, because the working code will have Oracle and other proprietary code that can't be licensed even in binary form...

I don't know about that. Partly based on a comment about a password expiry a while back and my own expectations, I suspect that the server code only runs SQL queries against the Oracle RDBMS: It'd be up to the host to provide the Oracle installation. Sources might only be needed if you want to adapt the configuration (like changing installation path names or retargetting to a different RDBMS). However it did cross my mind that within the server code there might well be other third party products included that make licensing difficult, for example there may be a "soft" load balancer that's bought in from someone else (I know MOULa has something like that because the Amazon load balancers aren't responsive enough, but I don't know if it's "home-brew" or proprietary).

Source is obviously the preferred option. But binaries would be a good start, if that can be made to work.

I don't think it's particularly clear whether RAWA's post was meant to imply that the test server was only for client application testing: The context was certainly in response to a client issue, but I took a wider interpretation from it as testing client side stuff is probably lower down the order than testing fan ages in terms of the number of people expected to participate.

_________________
Image Mac - MOULagain KI#00004826 00004289
In the interests of the environment, this post has been constructed entirely from recycled electrons.


Top
 Profile  
Reply with quote  
 Post subject:
PostPosted: Fri Oct 22, 2010 8:32 pm 
Offline

Joined: Mon Apr 26, 2010 5:37 pm
Posts: 58
Once again, lots of chatter but no reliable information on which to base it. We have no idea how much in donations has been collected to date, how much it currently "costs" Cyan to run the servers, how much of that has been "spent" and how much is in "reserve".

Without this information available, the balance of this conversation is meaningless speculation.

_________________
...I seem to recall that It was paraphrased by a fan somewhere in these forums that RAWA once might have said that they didn't have a firm release date -- shortest would be 1 year, longest would be 3,4,8 or 10 years.


Top
 Profile  
Reply with quote  
 Post subject:
PostPosted: Fri Oct 22, 2010 8:49 pm 
Offline
Obduction Backer

Joined: Wed Oct 14, 2009 4:14 pm
Posts: 1765
Someone opened that proverbial 'Can 'O' Worms' again..didn't they...I'll wait this out a few screens B-4 I speak (yell) too loudly..

Believe me...I watch this stuff..It does affect a bunch of peeps..

_________________
Moul(A)II Charura KI#296707 Char Gearz KI#600002 Teri Dactyl KI#600568 Chickopee KI#601018 Cannon Belle KI#601422

_____________________________
How Many Times Does A Myst Player Play Myst Before A Myst Player Decides To Play Myst Again


Top
 Profile  
Reply with quote  
 Post subject:
PostPosted: Fri Oct 22, 2010 11:18 pm 
Offline

Joined: Tue May 09, 2006 12:33 am
Posts: 1182
Location: British Columbia, Canada
There is work being done on 3rd party servers, so that we won't be dependent on Cyan to release theirs.

Dustin has already released some server code, which should be enough to run a local (single-player) server with the MOUL client.


Top
 Profile  
Reply with quote  
 Post subject:
PostPosted: Mon Oct 25, 2010 1:51 am 
Offline
Obduction Backer

Joined: Tue May 09, 2006 8:06 pm
Posts: 703
Location: Virginia, US
The most recent post by Cyan regarding a test server is here from RAWA - but in regards to setting up a server. Chogon had not called out specifically working on a test server; the closest is his acknowledgement at the end of this thread that a test server and other suggestions were "good ideas."

Regarding iRiven: release date is suggested here. Call for testers ended over a month ago (see this announcement).

A release of fully functional (i.e. multiplayer) fan-run MOULa server binaries would be fantastic, be it from a 3rd party or from Cyan. I'll take whatever comes first. [Side note - it would be great if the Talcum server could be released separately without Drizzle, so that it can be discussed openly on these forums - as PlasmaClient is.]

If Cyan manages to have some time available for doing something MOULa related, I'd vote for releasing the server binaries. [Open source server software would be ideal, but unlikely to come any time soon due to reasons discussed above and elsewhere.]

_________________
KI #: 1299


Top
 Profile  
Reply with quote  
 Post subject:
PostPosted: Mon Oct 25, 2010 11:55 pm 
Offline

Joined: Thu Mar 11, 2010 5:25 pm
Posts: 158
Location: Aberystwyth, Wales
Binaries are really frustrating, though. :( I'll take binaries over a longer period of waiting, but source code would be really nice. It's not as if Cyan can't give out the sources - we do have the Ki and an emote source-wise.

In other news, Talcum (fan-made MOULa server software) is actually pretty easy to set up now, although the only thing you can do at the moment is wander around Relto. Here's hoping that more development is seen there.

_________________
I am a member of Team OSCAR. I make content for MOULa. Help OSCAR here: http://forums.openuru.org/viewforum.php?f=103


Top
 Profile  
Reply with quote  
Display posts from previous:  Sort by  
Forum locked This topic is locked, you cannot edit posts or make further replies.  [ 38 posts ]  Go to page 1, 2, 3  Next

All times are UTC


Who is online

Users browsing this forum: No registered users and 1 guest


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum

Search for:
Jump to: