From what I've seen and experienced, many of the D&D players also share quite some Computer science backgrounds. Thus, I propose a thread to help the devs over there, and share ideas on how to:
a) prevent any exploitation to the CB-System by scripting it
and
b) how to store characters efficiently so they don't bring up a space issue at all.
maybe they could use some shoring up on that. :-)
ps: please don't post if you are not an IT Professional or do have some expertise on the matter.
Assuming 100k sized characters, it would take 11 days of it constantly creating a new character every 10 seconds to hit 1 TB. No human can make a full character in 10 seconds let alone have all the pages load up giving the options it would be painfully obvious that its a bot.
That sort of activity will already be monitored by the network admins so there really is no issue to be fixed.
I hope their new project fails utterly.
The true issue isnt how much storage it takes its the fact that wotc doesnt want to have people share accounts and came up with that. I can understand them wanting to protect against malicious attacks and abuses but after I thought about it more it just seemed silly.
If they want to take it fully online then they need to be prepared to handle with everything that comes with it.
Someone already did the math on this.
Assuming 100k sized characters, it would take 11 days of it constantly creating a new character every 10 seconds to hit 1 TB. No human can make a full character in 10 seconds let alone have all the pages load up giving the options it would be painfully obvious that its a bot.
That sort of activity will already be monitored by the network admins so there really is no issue to be fixed.
And that was with 100 computers working on it.
Why on earth would I give ideas to support this heinous business decision? They won't get any of my money, much less my ideas.
I hope their new project fails utterly.
Why on Earth would you post in this thread then?
The character storage will be much less of an issue once the export feature appears. Then you can just export, save locally, and delete from the cloud.
First, setting a lofty character limit of something like 1000 characters per account:
1) Prevents the DoS scenario
2) Provides ample storage for any reasonable user
3) Still results in a trivial amount of data to store on WotC end
Also, the DoS scenario can be effectively mitigated without even needing to limit the number of characters, by limiting the number of characters created within a set interval. There are many ways this could be done, either at the database or web server. Here at work we're partial to using Fail2Ban for this - it monitors web access logs, and if certain thresholds are exceeded (i.e., more than X account creations by IP address Y within Z minutes, transaction X occurring too many times from Account Y within Z minutes, etc), then the IP or account is blocked temporarily.
The main problem is that non of us has any insight into the actual design of the system and therefore we can only work with probably flawed or incorrect assumptions. Some things may be purely a business decision imposed by the devs' superiors at WotC or even Hasbro, so there's little they may be able to do about that.
Anyway, regarding your points.
a) Scripting/DoS attacks
My guess is that they simply use some webservices to access the rules engine and query/store data from/to the backend. In that case we can do absolutely nothing to help them because they need to fortify their services against attacks. Also, there are tools that easily allow you to disassemble Silverlight sites and applications, so unless the CB application is signed with a strong certificate and that certificate is checked at each and every log in, the CB itself could be reworked into an attack tool.
b) Storage
This is probably one of the business decisions imposed on the dev team. Some users think they are smarter than WotC/Hasbro and try to teach them how cheap storage in Amazon's cloud service or something else was. Do they really think the guys at WotC are too stupid to read numbers? There was very likely a specific reason why WotC decided not to use such a means of storage.
We also don't know the assumed load the service will generate on the backend. While zipping the characters was a good way to save storage space, it will create additional load on the servers with each load and save. Because we don't know the capacities of the backend we can't really say if that would make or break the service regarding response times and usability.
a) and b) combined are likely the reason why there will never be unlimited storage. A handful of cracked accounts could not only make the service unusable for everyone else (response times) but also wreck havor with the storage. Therefore - at least at launch and for some time after that - a hard limit is logical.
Google didn't start with 7.5GB free storage in their email service either (hint: they started with 1GB) so I really can't blame WotC for using a conservative number. Also, many people seem to be using the CB for things it wasn't meant to do. The CB was meant for players to build the characters they play. It was never meant as an NPC management tool. It wasn't meant as a graveyard for all possible character builds. For a single player 20 active characters would be really extraordinary. Anything else is likely using the tool for things it wasn't meant to do.
Why on earth would I give ideas to support this heinous business decision? They won't get any of my money, much less my ideas..
Seconded
"The CB was meant for players to build the characters they play. It was never meant as an NPC management tool. It wasn't meant as a graveyard for all possible character builds. For a single player 20 active characters would be really extraordinary. Anything else is likely using the tool for things it wasn't meant to do."
That's quite an assumption. I don't think I've ever seen anyone at WotC or elsewhere say that the CB wasn't to be used to create NPCs. I imagine doing that very thing was used as a selling point when it first came out, and I'm sure there are a lot of DMs who believe that was one of the things it was 'meant for'.
No it's not. It's simple reasoning:
- Fact #1: In 4e NPCs follow different rules than PCs. This is true for character creation and for game play.
- Fact #2: The CB doesn't allow you to model a character as defined in the NPC rules in the DMG.
Ergo: The CB wasn't meant to create and manage NPCs.
Explain why there will be any space or bandwidth issues? Half the subs will be gone anyway.
Regardless of how many subs are gone, both storage space and bandwidth are finite resources.
Explain why there will be any space or bandwidth issues? Half the subs will be gone anyway.
Regardless of how many subs are gone, both storage space and bandwidth are finite resources.
I doubt they are in the realm of being exceeded.
Explain why there will be any space or bandwidth issues? Half the subs will be gone anyway.
Regardless of how many subs are gone, both storage space and bandwidth are finite resources.
I posted an analysis of this over in the "20 character limit?" thread. The total cost would be pennies/subscriber assuming each subscriber used the builder a crazy amount.
- Fact #1: In 4e NPCs follow different rules than PCs. This is true for character creation and for game play.
- Fact #2: The CB doesn't allow you to model a character as defined in the NPC rules in the DMG.
Ergo: The CB wasn't meant to create and manage NPCs.
Wrong. NPCs are defined by what the DM wants, and nothing else. A lot of DMs like to use things like the CB to add different NPCs since it had basically an NPC generator (like the free ones for 3.5) built in. In ewvery version there have been different rules for PCs and NPCs, but a few slight alterations to what the CB generated made quick NPC stats. I've seen quite a few posts from other DMs that also use this method. It's not the 'Player Character Generator', it's the 'Character Generator.'
hey people, can't you read?
this thread is not for your nerdrage. discuss stuff where it ought to be discussed - in the proper threads.
Give the tools an "upload your own character" entry or "load character from url" entry, and we get to upload our chars and make changes, and print them in their new form. The character wouldn't be saved in the webtool, but we could export it again and redo the process again in case.
Not user friendly but the best I can think of in terms of having more options without asking WotC to store the characters for us.