[knewsom]

Assuming just an AR hack, what is stop a decker from repeatedly trying to put a mark on a host (Hack on the Fly) and if it fails, then performing a Hide action and then Jack Out, then immediately trying again? I am thinking of a decker in a vehicle, driving around the location of the host, trying to put a mark on the host until successful.

There is the repeat action penalty on page 49.  Is there anything else? If not, then I am uncertain why there is even a roll for Hack on the Fly other than getting information from hits, as you can always Jack Out and try again.

[brasso]

Lots, actually:

#/ Repeat action penalty (as you observed)

#/ Overwatch score, then convergence

#/ The host gets a mark on the hacker, on a failed HotF

They'd probably be on full alert by the time the poor hacker actually gained access! It really doesn't matter where the hacker is in meatspace.

They don't really need to make a hide roll either, because so long as they haven't actually entered the host, the host can't do anything to the hacker until they actually go in. I'd probably say that there was a pre-launched IC for each time the hacker had gotten a mark on them by the host, when they finally went in. So if they get in on their fourth go, have 3 lots of IC waiting when they enter

If this isn't enough, the host might even call up a hacker of their own to see off the PC, but this would need to be reasonably high rating to warrant this. Or maybe a GOD decker might be called out?

YMMV

[Jack_Spade]

Repeat action doesn't apply to attack actions of any kind.
Restarting your deck will remove any mark your failed attempt will have created - no chance of tracing
But you'll probably still be presented with a host on full alert and Patrol IC on the lookout. You'd better have high sleaze to stay hidden.
If its a specialized host with very sensitive data they might put the thing you are after into the archive and so make it impossible for you to get to it without a deep run.

[忍]

It's late at night. You're resting from a long day at work. You hear your door's deadbolt jiggle as if someone is trying to unlock it. Naturally you go to the door to see what's going on, however when you check no one is there. Confused, you go back to relaxing. Seconds after sitting back down the deadbolt jiggles again. Something is off and whoever is out there wants in. You keep checking only to find no one present.

When that person finally manages to pick that deadbolt, if they don't expect a sizable greeting they aren't lasting long.

[Bushw4cker]

It's late at night. You're resting from a long day at work. You hear your door's deadbolt jiggle as if someone is trying to unlock it. Naturally you go to the door to see what's going on, however when you check no one is there. Confused, you go back to relaxing. Seconds after sitting back down the deadbolt jiggles again. Something is off and whoever is out there wants in. You keep checking only to find no one present.

When that person finally manages to pick that deadbolt, if they don't expect a sizable greeting they aren't lasting long.

That was a Great Analogy btw!

[Hobbes]

With a failed AR HotF attempt, its entirely possible for the security spider or whoever to infer that "The call is coming from Inside the House!". 

Your turn, fail HotF, target gets a mark and before the hacker gets an action the target could make a Matrix Perception check.  From that check it's pretty easy to infer that the hacker is really close from the lack of a noise penalty if nothing else.

The correct response to a failed AR Hacking check is to start running most of the time   

[Gingivitis]

On a failed HotF, the host gets a mark on you.  You cannot perform a Hide action on against a target with a mark on you (pg 240).  Without a Reboot, you will always be on their radar.  You can be sure that the Host Spider or Trace IC will be looking for you eagerly.  They might even place a second mark on you to Trace Icon to your physical location.

Remember that Jack Out does not clear marks.  You could Reboot instead but you loose all your other marks that you have acquired thus far...which maybe you needed.

Also, while IC must stay inside the host, nothing keeps a spider there.

[firebug]

Constantly rebooting can relieve most of the problems if you don't mind it taking a while...  Until you enter, as most people have said, the Spider will probably have manually launched a few IC, and be on the lookout for you personally.

Since this is presumably a host that does not need to be accessed constantly (AKA, it's not public), the Spider would probably pull up the login history of the host.  If you enter, he'll see "Oh, someone entered right now?  After a bunch of failed hacking attempts?  Idiot."  You will come up as a legit user, but that doesn't negate your incredibly suspicious timing.  You could even still be running silent, but the Spider will order the Patrol IC to look specifically for you, so expect to have only a handful of actions before you're found out.  So, no time to do a Matrix Search.

If you're curious as to my logic about the login history...

First we have to assume it's not a public host.  If it was, you wouldn't need to hack your way in, you can just ask for a mark and get one.  Only one though.  So since it's not public, it's a restricted one.  Like a host at a corporate facility.  Employees are potentially logged in to this host, but the thing is, they probably all connect to the host in order to "clock in" and don't log out until that shift is over.  Since this isn't the host for some shitty retail store if it's actually difficult to hack (unless you're just trash at hacking) the shifts aren't scattered out through the day.  It's likely there's two or three eight-hour shifts; "day shift" "night shift", "overnight shift" etc.  Logging in outside of the times when all the employees for that shift are rotating in/out is extremely obvious as someone who doesn't belong there, even if you hadn't alerted all the matrix security to a potential attack.

So, as 忍 said, expect to just be destroyed_banana'd once you're inside by the twitchy, alerted security and IC.

[Hobbes]

So, as 忍 said, expect to just be destroyed_banana'd once you're inside by the twitchy, alerted security and IC.

And the HTR team that is headed your way.  AR means you're close, and mechanically, after each failed attempt the target gets at least one action before you can re-boot.  They will do a Matrix Perception check and it will show you're within 100 yards.  That will go badly for you.

Honestly if there is a better than average chance of failure a player may be better off with a Brute Force Action rather than HotF since failed Attack actions don't alert the target.  Just sayin.

[DigitalZombie]

You could use the "cry wolf!" tactic though.
Step 1: Load up on programs that interferes with real world tracking of your physical location.
Step 2: hack the host and fail miserably
Step 3: repeat. While having as much surveillance on the host and its users as possible
Step 4: the host might now change its IC setup to something than BLACK IC! All the way- to minimize dmg on corp drones.
Step 5: after x weeks. Get your street sam to make a failed HotF as normal. Edge HotF yourself into the host.
Step 6: ?
Step 7: profit

[firebug]

So, as 忍 said, expect to just be destroyed_banana'd once you're inside by the twitchy, alerted security and IC.

And the HTR team that is headed your way.  AR means you're close, and mechanically, after each failed attempt the target gets at least one action before you can re-boot.  They will do a Matrix Perception check and it will show you're within 100 yards.  That will go badly for you.

Honestly if there is a better than average chance of failure a player may be better off with a Brute Force Action rather than HotF since failed Attack actions don't alert the target.  Just sayin.

AR doesn't mean you're close.  There's no difference in range between AR and VR when it comes to hosts, as you can explicitly enter them from anywhere.  Noise from distance could theoretically be an issue...  But NR is so cheap, it won't be.

Brute Force wouldn't be any better.  If you're likely to fail, you'll start taking ping after ping of matrix damage until your deck is fried, then have to repair it, taking even longer.  If it goes bad enough, it could take days of trying before you succeed.

You could use the "cry wolf!" tactic though.
Step 1: Load up on programs that interferes with real world tracking of your physical location.
Step 2: hack the host and fail miserably
Step 3: repeat. While having as much surveillance on the host and its users as possible
Step 4: the host might now change its IC setup to something than BLACK IC! All the way- to minimize dmg on corp drones.
Step 5: after x weeks. Get your street sam to make a failed HotF as normal. Edge HotF yourself into the host.
Step 6: ?
Step 7: profit

Unlikely the host would change its IC.  If they were too dangerous to the employees in the first place, they wouldn't use them.  Plus, Black IC is usually the last to launch.  No need to launch all of them until you see the attacker.  See what I said above about login history.  Your Street Sam won't stop that from revealing that you've entered.  Finally, if spending Edge was all you needed to get in, why not just do that in the first place?

[Hibiki54]

I'm making the assumption that the Decker repeating the hack is also resetting their deck after every attempt.

When the Host gets a mark on you, it will launch IC and attempt Host Convergence. This will also alert any Spider on the Host. By the time they have a chance you get to the hacker, they will already be gone and their deck reset. The hacker CAN try HotF as many times as they want and repeat the process. However, once they are successful they will be inside a Host that is on high alert from multiple hacking attempts and a Spider actively searching through the strands of matrix data for anomalies.

[Tym Jalynsfein]

Hosts probably suffer thousands of Hack attempts on a daily basis. That is the cost of doing business in the 2070's... Every Corp/Megacorp is engaged in industrial espionage on a grand scale, so Hacking will be almost as commonplace as ordering a mochasoy.

I do not see excessive alerting of the system in that case, because they will constantly be on a status of high alert if that is the protocol. Instead they will continue to do what they do. Trust in their systems for the most part. Hosts will have Patrol IC in the system doing their thing, and maybe a spider or two on hand to manage things. You will eventually lose against the host, and that is what they rely upon. I have never had a hack not alert a high rated host (Rated 6+) at some point. It happens. And it will generally always happen. You have too many actions that are performed for them to all go without a hitch. When it happens, that is an awesome day (or you are hacking a low rated host), but do not rely upon it. Anything challenging will eventually pick you out of the background noise of the system.