So, I stole the concept of Teleporters from Microsoft's Terrarium, which was primarily a way to get people excited about learning C# and .NET. People ran the Terrarium app on their machine and authored creatures in C# that competed against others. A creature was essentially a C# program that could be safely executed on other machines with restrictions imposed by a .NET sandbox. The sim program displayed the environment and creatures in a 2D world similar to DB. When they ran internet mode, creatures teleported in and out to other machines peer-to-peer using a server based location service. All the organisms were hand coded. There was no mutation or evolution unless you built it into your organism yourself.
The Terrarium guys hid the teleporter very delibertly. They wanted to make the bar for storming other sims very high. Now, the senses a Terrarium creature had available to them were a little more high level than DB - instead of physical eyes, creatures were given an automatically updated array of all the other creatures around them with refvar equivalent info for each plus distances and such every cycle. This approach focused the authored code more on creature behaviour than on interpreting senses. One has to write a lot of bot DNA in DB to look around and keep a high resolution list of all the bots around them. Never been done to my knowledge. Takes a bunch of cycles, narrowing eye widths, spinning, etc.. Not so in Terrarium. So, it was easier in Terrarium to write creature code that tracked everything it could see over time, cycle to cycle and did stuff based on that as creature authors had so much more info provided to them magically each cycle. Even though it was not evolved, you saw some pretty cool anticiaptory behaviour w.r.t. anticipation of other's movements, fleeing, hearding, attacking, coordinating, hiding behind plants and so on. The bar for writing a creature that could survive even a few thousand cycles got pretty high over time since all the succesful creatures had super sophisticated logic for threat and prey determination, how much nrg to use to fight or flee, whether it was worth it or not based on the other's nrg levels and so on. You didn't fight if you were unlikly to win. You fled to fight another day and much code was devoted to makign this determination. Would you win? Should you fight or flee? What does it cost to flee and how far and will you be chased? That kind of thing.
But that said, a Terrarium creature's code was not that much more higher level than DB bot DNA. And in Terrarium, people were able and did indeed build hand-authored organisms that could locate and storm the invisible teleporter. They did this by keeping track of the creatures around them and watching when organisms disappearred that should not have I.e. their nrg level was too high for them to have starved that cycle, they weren't being attacked and so on.
Now, I should point out here that in Terrarium and DB, there is no gaurentee that two bots going through a telepoter right after each other end up in the same destination sim. As the number of sims in Internet mode increases, this probability declines. But bots can still play the probablities to their advantage.
As far as the meat of your question, I have no philosophical problem making teleporters visible. DB is not Terrarium. It would not even be that much work to do so. I will say though that as it is, invisible randomly moving telporters favor multibots or at least single bots that form clusters for periods of time even if they can't see the teleporter. Probability will favor them since they will occasionally get teleported as a cluster and a cluster (up to 50) is guarrenteed to arrive all together in the same sim. They can then split apart and fight as individuals or not. Thus bots can do things today to increase their chances of a successful en-mass attack on another sim even without teleporters being visible. The .totalmyspecies and .totalbots sysvars can easily be used to know when you have conqerred a sim or when you've just been teleported somewhere where you are now in the minority. Changing behaviour accordingly - e.g. to form clusters or not or some other behaviour - based on this should be easy.
I do think that while I'm okay with making the outbound teleporter visible or easier to find, I don't think I like inbound teleporters that are visible, fixed or don't randomly distribute incomers as this would make it too easy to defend a sim from incursions. Terrarium worked pretty much the same way I've coded it for DB as far as random dispersal of incomers, though the radius of random materialization around the telepoter was perhaps smaller than the entire field as it is now in DB. I could narrow this some so that incomers had a higher priority of finding others that teleported in at about the same time, but I would not want to make this too small lest it be too easily defended.
I could make private teleporters work over the internet. That way, if two or N people wanted to connect their internet sims with fixed telporters that did not randomly distribute, they certainly could, instead of or in addition to the internet mode teleporter.
As far as finding their "home" sim goes, we need to discuss what "home" means. Physics I was built for? Environmental elements I recognize and was built to leverage? I.e. shapes with defensable cracks to hide in? It would not be difficult to allow a bot to know when it was in the sim it's ancestors were initially released in. But again, things like .totalmyspecies can serve today as a poor man's way to change behaviour based upon relative population in a sim. Perhaps a better strategy is to provide a little more global info to bots so they can perhaps guage such things as veggy density or morphological costs in the sim. If forming a tie in one sim costs nothing and in another costs 100, well, perhaps bots should have a way to know that so they can change their behaivour accordingly.
My plans for internet mode are the following:
In the near term, say the next month or so, I want to finish 2.44 and release it. This means getting internet mode to a stable point with a solid, consistant, core set of features that work well. Mostly, this is plumbing stuff. We need to move to a different FTP server or perhaps have the ability for users to override the hardcoded one so when we do move, peopel don't have to rev the exe (though I don't plan to go anywhere). Not sure I will test full on private internet modes for 2.44, but I want to do something here. Similarly, there are some cheats that need cleaning up. For example, virus immune bots should not be teleported, internet telporter properties should not be changable, etc. I think I'll leave the Disable Mutation option enabled. Mutating can be seen as an advantage or not and bot authors can choose to release their bots with or without the ability to mutate. There are some more internet mode graphs I want to add, including population of each sim but I don't want to get too crazy here. It won't scale for one thing and part of me thinks that other sims should be a mystery at a certain level. Part of the fun. Then there are some other little tweaks and bugs of course, but nothing too huge. I do like Shvarz's idea of making Intenet mode the default for first time newbie start ups, perhaps with nothing but some null plants. But most of what 2.44 will be is complete today.
Longer term, I do want to provide for different internet modes. People should be able to play in a multi-bot only mode or a no-mutation mode or a zerobot-descended only mode or a mode with enforced standardized sim settings and costs. That kind of thing. But I agree with Nums that a single mode everybody plays in should be the focus in the near tem. I also want to tie things into the web site better so its easy to see who is on top in what mode for how long at what population and so on. We could run weekend contests, etc. Then I think I'll go back to adding generalized environmental richness for awhile e.g. shapes that bots can break apart, physics and costs that can vary by sim region, egrid stuff, lakes and rivers, rocks and sand. Anything we do that enriches a stand alone sim will also enrich internet mode.
And of course, suggestions and feature requests as always where reasonable and feasable...