19 Mar, 2009, David Haley wrote in the 141st comment:
Votes: 0
I think that if we start doing something like this, we'd need to rethink a lot. Also, I'm not sure why you'd want p2p MUD listing; the MUDs themselves don't really care about other MUDs, right? The idea is for them to provide information to a source that players go to for finding MUDs.
19 Mar, 2009, elanthis wrote in the 142nd comment:
Votes: 0
But… but… but it's FUN!!!

I once saw a graph showing the productivity of programmers ranging from the "I don't care about software it's just a job" type of people to the "omg software is so cool" type of people. The graph was a big curve, being near zero at the left, because the people who don't care about software don't put effort into design or quality, and usually never even bothered to learn how. It was also near zero at the far right, because those people like to spend a great deal of time creating a complex and interesting solutions for problems that nobody is having, just because designing and implementing those solutions is entertaining.
19 Mar, 2009, Cratylus wrote in the 143rd comment:
Votes: 0
Starting to sound like a personal feud.
21 Mar, 2009, Cratylus wrote in the 144th comment:
Votes: 0
Where can I score me a telopt crawler to play with, if I do not want to
download and install Scandum's client?

-Crat
http://lpmuds.net

PS Also, even if we ditch the idea of it being ok to accept data for
multiple muds from one node, it's worth discussing that the protocol
doesn't seem to take it into account at all, allowed or disallowed.
I'd just like to know what the preference is, and if it's to be disallowed,
then it should be recommended how crawlers ought to handle that.
21 Mar, 2009, Kline wrote in the 145th comment:
Votes: 0
Crawler I don't know, but if you're willing to look at TinTin++ as more than "Scandum's client" it supports showing debug. z/cMUD also will show telopt debugging, and additionally let you create triggers based on telopt codes.
21 Mar, 2009, Lobotomy wrote in the 146th comment:
Votes: 0
Cratylus said:
By allowing queried
sites to provide information not only about
themselves but also other muds, what this sets
up is the capability for a peer to peer listing network.

What happens when Mud A starts disseminating false or misleading information about Mud B?
22 Mar, 2009, Zeno wrote in the 147th comment:
Votes: 0
Is LEVELS the max levels a player can obtain or the real max levels (including Imms)?
22 Mar, 2009, Kline wrote in the 148th comment:
Votes: 0
I reported levels as "highest effective player level", in terms with gear, etc. My game has a mortal tier (5 classes x 80 levels each), remortal (2 classes x 80 levels each), and a final "Adept" tier of 20 levels. So do I report 580 levels? Even though most of those are lateral? Or 180, counting only one class per tier? I didn't either, I reported 120, because levels in the first tier are 1:1 in your primary class, and second tier are 4:1 as "psuedo levels". So you end up with NPCs, gear, and players all reporting level 120 end-game.
22 Mar, 2009, KaVir wrote in the 149th comment:
Votes: 0
Perhaps there should be MULTICLASSING and REMORT options…

What happens if certain mud listing sites decide they want to display additional information? Will muds simply be able to add new fields, with other crawlers ignoring data they don't care about?
22 Mar, 2009, Scandum wrote in the 150th comment:
Votes: 0
Updated the spec to avoid any confusion:

"CLASSES"            Number of player classes, use "0" if classless.
"LEVELS" Number of player levels, use "0" if level-less.
"RACES" Number of player races, use "0" if raceless.
"SKILLS" Number of player skills, use "0" if skill-less.


KaVir said:
Perhaps there should be MULTICLASSING and REMORT options…

REMORT is a Rom term I think, MrMud has what is called 'Reincarnation' which is a much more sensible term if you ask me. Not sure what other muds call it.

KaVir said:
What happens if certain mud listing sites decide they want to display additional information? Will muds simply be able to add new fields, with other crawlers ignoring data they don't care about?

That's the general idea, I'm trying to keep the variable list as brief as possible and to avoid making it diku centric so as not to piss off the tiny branch.

Kline said:
So do I report 580 levels? Even though most of those are lateral? Or 180, counting only one class per tier?

Your call, I think the main value of this variable is to separate the level based muds from the level-less muds.
22 Mar, 2009, David Haley wrote in the 151st comment:
Votes: 0
Scandum said:
avoid making it diku centric so as not to piss off the tiny branch.

If your world view is that Diku is the vast majority and everything else is just a tiny branch, it's no surprise that what you produce is so Diku-centric! :stare:

Scandum said:
Your call, I think the main value of this variable is to separate the level based muds from the level-less muds.

If that's your goal, you should have chosen a boolean.
22 Mar, 2009, KaVir wrote in the 152nd comment:
Votes: 0
Scandum said:
REMORT is a Rom term I think, MrMud has what is called 'Reincarnation' which is a much more sensible term if you ask me. Not sure what other muds call it.

http://en.wikipedia.org/wiki/Remort#Remo...

"Remorting (also known as rebirth, ascending or reincarnating)".

It's been a common term among Diku derivatives for the last 15 years or so, and I'm fairly sure the earliest muds with remort weren't ROM derivatives.

Also worth a read: http://www.mud.co.uk/dvw/whatisremort.ht...

David Haley said:
Scandum said:
avoid making it diku centric so as not to piss off the tiny branch.
If your world view is that Diku is the vast majority and everything else is just a tiny branch, it's no surprise that what you produce is so Diku-centric! :stare:

I believe he's referring to TinyMUD and its derivatives, possibly including LPmud (which drew inspiration from TinyMUD).
23 Mar, 2009, Cratylus wrote in the 153rd comment:
Votes: 0
Lobotomy said:
What happens when Mud A starts disseminating false or misleading information about Mud B?


I'm still concerned about this and would like to hear folks' opinion
on dealing with muds poisoning crawler tables. Y'all know we have mud
owners immature enough to try to knock rival muds off listings
with dirty tricks. How will they be prevented?

-Crat
http://lpmuds.net
23 Mar, 2009, David Haley wrote in the 154th comment:
Votes: 0
The problem is only relevant if we allow MUDs to give out information about other MUDs in the first place… I thought we weren't doing that? What's the argument in favor of that?
23 Mar, 2009, Cratylus wrote in the 155th comment:
Votes: 0
David Haley said:
The problem is only relevant if we allow MUDs to give out information about other MUDs in the first place… I thought we weren't doing that? What's the argument in favor of that?


I disagree. It's trivial to set up a dummy mud with poison info.

-Crat
23 Mar, 2009, KaVir wrote in the 156th comment:
Votes: 0
The mud listing sites could use name and IP address as a form of authentication. Unless these both match the listing you've added to their site, the data would be ignored.
23 Mar, 2009, Cratylus wrote in the 157th comment:
Votes: 0
KaVir said:
The mud listing sites could use name and IP address as a form of authentication. Unless these both match the listing you've added to their site, the data would be ignored.


That's a natural solution, and since listings sites have contactable admins, it's easy enough
to petition a change in that "auth pairing" when your mud moves.

My concerns in general here are three:

1) If this protocol is just for current and established listing sites, that is sufficient. However,
it's my opinion that what MSSP does is make it possible for anyone to be a listing
site, such that folks are no longer chained to listing sites in bed with competitors. I think
this decentralization is implicit in the protocol. If I'm crazy to think that I hope someone
will set me straight, otherwise I think it's worth considering that "how do I notify the
MSSP crawlers that I've moved but I'm really me" causes problems.
It's not a huge deal, but since we're at the stage where we can fix things before
they're broken, maybe it can be dealt with best now.

2) I forgot what 2 was for.

3) If we're only allowing a mud to provide info about itself for security reasons, then
I think the ip and port values ought be dropped, since the only thing we know for
sure about the mud is its ip and port (that we connected to) and believing other
data for that is a vector for poison.

-Crat
http://lpmuds.net
23 Mar, 2009, KaVir wrote in the 158th comment:
Votes: 0
Cratylus said:
1) If this protocol is just for current and established listing sites, that is sufficient. However, it's my opinion that what MSSP does is make it possible for anyone to be a listing site, such that folks are no longer chained to listing sites in bed with competitors.


It does, but the method of validation is really up to each individual listing site. If I register an account with TMC then I should be able to add my listing (under the control of that account), enter a few essential details, click the MSSP box, and let it update on its own.

On the other hand, another listing site might not have any concept of accounts. Instead it may just grab addresses from anywhere/everywhere and attempt to connect, generating its own list. In this case it's certainly possible for someone to poison the listings, but that's the problem of that particular site.

Cratylus said:
3) If we're only allowing a mud to provide info about itself for security reasons, then I think the ip and port values ought be dropped, since the only thing we know for sure about the mud is its ip and port (that we connected to) and believing other data for that is a vector for poison.


I don't think it hurts to send that data, even if it's redundant. It's also worth remembering that some muds support multiple ports, and I could even see muds sending multiple IP addresses (perhaps as a means of providing automatic forwarding for listing sites which don't use accounts).
23 Mar, 2009, David Haley wrote in the 159th comment:
Votes: 0
To be sure I understand, you're talking about somebody going to some site, putting in the name "Crat's MUD", and pointing it to address "CratHater.com" – and filling the list with bad information. Is that the problem we're trying to solve? If so, how is that any different from somebody doing the same thing by hand anyhow, without MSSP?

I don't understand where this "poison" is supposed to come from if a MUD reports a "poisoned" IP/port. That would amount to the MUD poisoning itself, since it's the one sending the response. What very specifically are we trying to guard against?
23 Mar, 2009, KaVir wrote in the 160th comment:
Votes: 0
I believe he's talking about people creating additional listings with the same name but different data, to confuse people who are looking for a specific mud. So I might open nine new muds all called "Legends of the Darkstone", with all traffic being redirected to my mud. People looking for your mud then have a 90% chance of reaching mine instead.

People could do this by hand as well, but mud sites would simply reject them. A purely automated system wouldn't be able to make such decisions though.
140.0/292