Decking, the SR3R way Options

[Kagetenshi]

Analyze Icon should only be relatively noninterceptable if you assume that the icons are making some sort of direct connection—that you're directly scanning the other person, rather than going through the host to do so. I don't believe that to be the case.

~J

[Eyeless Blond]

[nezumi]

In regards to letting the deckers more easily sit and wait, I'd recommend the following changes:

1) The Null Operation test is unusual (once per hour or somesuch)
2) Global tally naturally goes up and down on its own (since tally is no longer connected with a decker, we're only worried about global tally) - if something happens in second 1, then again in second 3, but then stops for ten or twenty minutes, the system figures the decker has most likely moved on, or it was just a bug, and the tally goes down a few notches.
3) The "random background tally" naturally rises and falls as random stuff goes on. So the decker might log in while the background tally is 13, but as the crisis passes thirty minutes later, the background tally drops to 2.

The second one especially I think is critical. The tally MUST go down after periods of inactivity, or systems would just keep accruing tallies until they rebooted and started again from scratch. How long it takes for the tally to drop again should depend on the nature of the system and what our security sheafs finally look like.

[Platinum]

I really don't like the global tally having such a really strong effect on the decker.

I think once a minute there should be a roll of 2d6, on a 2 it goes down by 1 point ... on 12 it goes up by 1 point, but a decker's tally still needs to be tracked, and that is the major mechanic here.

The global tally is a modifier... Hey ... what about a -modifier... there are times that the system is just at peace.... there are no threats, and so it can actually start as a negative.

[nezumi]

I agree, the global tally should never account for more than 10 points under normal operations (plot-wise, there are reasons it should go up, for instance, during a crash, and the "global" tally will also go up when the decker is close to another troublesome decker or program, whether aware of it or not.)

[Moon-Hawk]

If the tally is going to decay (and I agree with this) then it should be set at a rate such that if an "average" hacker is simply performing null operations (at whatever interval you decide you like) the tally should break even and not go anywhere.
Thus, a below-average hacker doing nothing or an average hacker doing stuff will tend to increase their tally, but an above average hacker performing nothing but null operations will have their tally slowly decrease.

It just seems natural to me that that's the level of paranoia average systems would be tuned to. There's going to be all kinds of modifiers, but IMO that's where you should set your "break-even" point when everything involved is typical.

[nezumi]

I can agree to that, as long as the hard-core hackers can let it decrease.

[Kagetenshi]

Hackers/crackers can't accrue any tally, as they're at least 20 years in the past (assuming an earliest game date of 2049).

That said, if we're going to do something like that, it'd probably be best to pick either host colour or host difficulty, tie them to a decker skill level, and then work the numbers so that that skill level has their tally stay pretty constant while performing nullops on that host, lower skill levels have it rise, and higher skill levels have it lower slowly.

~J

[Eyeless Blond]

Hm. Okay, how about this rule for Null Ops then. Note this should probably be in a separate section from other System operations (and thus referred to directly in the Table of Contents), rather than in-lined with the rest of 'em as seen in the SR3 book, to emphasize its separateness and importance:

(Edit: due to comments)
Null Operation: When a decker is illegally logged in to a host, he is constantly vying with the host's security system to avoid detection, even when he is not doing anything especially important. If the period of time is sufficiently long, say ten minutes or more, the GM may call for a "null operation" test to reflect this. Treat this as a modified System Operation, with the decker making a test against the Control subsystem and using his Masking persona chip as a utility. The decker receives an extra 2 dice on this test for every extra initiative die he has beyond the first, and the system receives an extra 2 dice for every rank it Security Code is above Blue (+2 dice for Green, +4 dice for Orange, +6 dice for Red). If the decker wins, he reduces the host's Security Tally by 1d6/2 for every two net successes he earns; if the host wins, it increases its Tally by 1d6/2 for every two net successes (note that the host does not simply add each success to the decker's tally; that particular rule is suspended for this Operation). The Rule of Six does not apply on the security tally dice.

Of course, the decker himself is not the only thing generating (or eliminating) Security Tally. Each time the GM calls for a Null Operation, he also rolls 2d6-7. This number, whether positive or negative, is added to the Tally, to reflect the random factors affecting tally over which the decker has no control.

If the Security Test raises the decker's security tally and triggers a response from the host, the gamemaster should activate the response as he sees fit. perhaps after a percentage interval of the decker's period of inactivity.

Note that the period of time represented by a Null Operation is kept deliberately vague; a single test should suffice for any period between ten minutes and an hour. If the decker is going to remain in a system for extremely long periods of time, the GM may choose to call for multiple Null Operation tests, possibly one for every hour or few hours the decker spends connected to a host. In this case each net success by the decker or system affect the tally by 1d6/2, and the GM rolls 3d6-11 for the random tally generation.


How's that? I'm mainly trying to keep the whole idea simple, while factoring in both the decker's contribution to the Tally as well as the random. Of course all this needs massive playtesting, but what do you think of the idea, and the numbers?

[Link]

On Null operations, the standard rules use 1 test and adjust it based on the time spent idle rather than 10 minute intervals. You dislike tables in play, I prefer less dice rolls too ;) Could you incorporate this into your revision? Also, what's the dice bonus due to initiative bonus about - just interested in the reasoning?

Also here's the Null ops rule if anyone wants a look.

[ Spoiler ]

Null Operation
Test; Control
Utility: Deception
Action; Complex
The gamemaster may require a decker to perform one or
more Null Operations whenever the decker is waiting for some-
thing to happen, whether it is an event on the Matrix, the end of
an ongoing operation, or something else that Involves hanging
around in cyberspace without making System Tests. The
gamemaster may also call for a Null Operation if a decker is
doing anything that requires actions but not System Tests, such
as maintaining an Edit Slave. The gamemaster may secretly per-
form these operations on behalf of the decker, If he so desires.
Use the host's base Security Value for the Success Contest if
the decker is inactive on the host for less than I 0 seconds. If the
period of inactivity is less than a minute but more than I 0 sec-
onds, apply a +1 modifier to the Security Value, If the period is
less than an hour but more than a minute, apply a -t-2 modifier. If
the period is less than I 2 hours but more than I hour, apply a +4
modifier: apply an additional +1 modifier for every additional I 2
hours. The gamemaster may set an upper limit on the inactive
period, depending on the decker's ability to avoid falling asleep
in the event of such implausibly long times.
Selena begins a download that will complete in I 2
seconds. She has nothing else she wants to do. so she
waits. The gamernaster requires a Null Operation and adds
+1 to the host's Security Value,
On a different run, Selena is waiting for a ground team
to get through a locked door (don't you just hate those
manual doors with no computer overrides on the lock?).
TTie gamemaster determines that the maglock penetration
task will take 7 minutes. He also performs a secret Null
Operation for Selena and applies a +2 modifier to the
host's Security Value for the Success Contest, The test
yields several successes that push the security tally past its
next trigger step, which triggers a truly unpleasant I C pro-
gram. The gamemaster decides to have the I C show up 3
minutes into the waiting period.
If the Security Test raises the decker's security tally and
triggers a response from the host, the gamemaster should acti-
vate the response as he sees fit. perhaps after a percentage
interval of the decker's period of inactivity.

[Eyeless Blond]

Yes, on page 218. Okay, I went ahead and edited the entry to make the time units more flexible. What I intended was to completely remove all Null Ops tests if the wait time was less than ~10 minutes, to help cut down on dice rolls. But you're right; the period between 10 minutes and an hour would've had too many die rolls. So, fixed.

As to init boosters providing bonuses, well mechanically it was to balance the bonus based on the host's security level. Intuitively, it was meant to toss a bone to people who have init boosters, as they should be able to react faster to what the host is doing. Flavor-wise, if you can cover your tracks twice as quickly then the host should have less reason to suspect something it going on.

[Eyeless Blond]

I'm not even sure where we are here. There's been a lot of ideas in the past few pages, and very little reaction to those ideas. Anyone care to respond?

[nezumi]

I've been overall supportive of most of the ideas posted (and I've given long posts as to why). That said, astral space and decking are the two threads I haven't been taking notes on as of late as I catch up on everything else.

[Eyeless Blond]

Wow, has it really been over two weeks since we've had a meaningful post here? We're making a lot of progress; let's keep up the pace!

[Link]

A quick question regarding Null Operation (either version). Is it for situations where the decker is constantly vying with the host to remain anonymous or where the decker has their attention elsewhere such as communicating with team-mates? Or both/neither?

[Eyeless Blond]

I think both/either. The relevant text from the original rule is:

"The gamemaster may require a decker to perform one or
more Null Operations whenever the decker is waiting for some-
thing to happen, whether it is an event on the Matrix, the end of
an ongoing operation, or something else that Involves hanging
around in cyberspace without making System Tests. The
gamemaster may also call for a Null Operation if a decker is
doing anything that requires actions but not System Tests, such
as maintaining an Edit Slave."

[nezumi]

Alright, I've got a character joining my game as a decker, so I have good reason to push this along. Here is my updated list of quandries. I'm running out of space here, so we need to start shelving some issues!! Please tell me which issues we are definitely NOT including, or what has been solved and what the solution is.

Agreed upon rules:

[ Spoiler ]

Decking
A) Combined functions (see decking thread)
B) Validate and invalidate Account works only on main user database. All other Validate functions can be used anywhere
C) Link utilities are removed, only HW is required for a wireless link
D) Sleaze has been removed. Masking alone determined Detection Factor.
E) Combat utilities use decker's skill with program rating determining power. Killjoy and Black Hammer determine damage based on the host
F) Guardian is now gone, replaced by appropriate security hardware
G) Hacking pool can be exchanged for Detection Factor
H) Tally is kept system wide, brought up to the tally of the highest individual on the host
I) Detection modes are removed

Suggested rules:
A) Drop response increase in favor of mental-initiative boosting cyber (like a VCR)
1. Should be capped by deck
2. Include 'mental initiative' cyberware/stat
3. Have multiple ratings for datajacks
B) Redirect Datatrail success adds +1 to the Location cycle (making it take longer, not harder)
C) IC are just frame agents loaded with appropriate programs
1. All IC get the features of Probe IC (adding to the value of System Tests)
D) Persona attributes can be redistributed with a complex action while in the system
1. Optional rule: you can go above the MPCP limit by paying 2 MPCP pts per attribute point
E) Drop memory requirements (like SR4)
F) Agent rules are kinder, allowing a decker to have multiple agents. Drones are agents with bodies (SR4)
G) AR allows for people to do simple things without physically disengaging from the real world (SR4)
H) (this area free)
I) All electronic devices (including cyber) can be connected to the matrix and thereby hacked (SR4)
1. ACIFS of 2*rating for all areas
J) Everyone should be able to use AR and therefore the matrix, making the computer skill much more common
K) IC should have to spend time searching rather than immediately appear next to the decker after the tally has increased
L) The decker should be able to directly deceive the IC, not JUST the host
M) A legitimate user who commits an illegal operation may be dumped with no tests
N) Prices should be included for non-illegal decks in the main manual
O) Nezumi's matrix metaphor - host as a physical facility with stationary guards, searching guards, security tally based by approximate location:
1. IC reacts instantly, but it takes time to FIND the decker based on local activity and rating
2. A search is a Analyze Icon cycle on all local icons until a suspicious one is found
- IC can attack the wrong icon if it looks to be suspicious
- Evade detection puts the decker farther away and increases time for the search
3. Security tally is applied to the lowest level possible, to the user (if legitimate), region, host, system, grid in that order
- A system may decide to try and isolate a decker by disallowing traffic from a high-tally area
4. IC serve as the host's eyes. If IC has a visual lock on the decker, other IC can find him immediately and the user may be dumped
5. Stationary IC will analyze every icon that passes by and will automatically attack suspicious icons, regardless as to their tally
6. Increases in alert levels are system-wide, and possibly network wide
P) Hacking pool should have more basis on skill or attributes
1. (Int + Wil + MPCP/2)/3
2. (Int + Wil + Computer skill)/3
Q) Create a 'false positive' background count for security tally
1. 1d6-1 for >50 users, 2d6-2 50-500, 3d6-3 >500
-For ease of calculation, drop the -1/-2/-3 and just start the sheaf higher
2. Is tally based on the host, the network, the PLTG, or the most popular host connected to it?
3. Should background tally be linked to the rating of the host? Red 4d6, Orange 3d6, etc.?
4. Tally decreases by 6 for blue, 4 for green, 2 for orange, 1 for red every twenty minutes, + die roll again
R) Tunneled Connection operation - Simple Action every turn to maintain a connection through the chokepoints, which allows for your Masking, makes Encephalon useful
1. What if the deck allowed for certain 'sustained' actions?
2. In complex network diagrams, decker has to tunnel through intermediate hosts to hide his trail
3. Security tally in inhereted downwards, RTG->LTG->Host->Subhost
-RTGs don't pass tally, LTG's pass tally but don't receive tally, PLTG's share tally between all hosts
4. Tunneling doesn't take actions, but imposes initiative penalties based on the rating of the chokepoint
S) Deck construction... Too complex? Becoming unnecessary?
1. Allow decks to sustain certain actions like tunnel firewall
2. Decks bottleneck mental initiative
3. Drop memory costs
4. MPCP limits number of programs and processes running
T) Tally accumulates too fast to allow for matrix overwatch
1. Operations involving Index subsystem, Analyze utility don't accumulate tally
2. Add an operation that lowers your tally?
3. Tally decreases over time, assuming no further illegal operations (may increase more though, due to null ops and background tally)
- Tally decreases by 6 for blue, 4 for green, 2 for orange, 1 for red every twenty minutes
4. Increase space between attacks on the sheaf
- This helps matrix overwatch, but makes datasteals too easy
5. Null Ops. may DECREASE the security tally
U) Evasion or IC rating should be the TN for cybercombat tests, IC's attack program for damage


Problems:
1. Detection factor is too important, but never above 7 at chargen. It should be higher.
- Due to the exponential growth of TN difficulties, a DF of 16 is largely unbeatable. Is 7-16 a wide enough range?
2. Multiple systems for deckers, one quicker for interaction with non-decking PCs, one more expansive, both equivalent except for speed and detail
3. Determine if programming tools is too easy

[nezumi]

To give my opinion on what I just wrote...

A) (Response increase cyber)
I've been considering this. I think I like it. Yes, yes and eh.

B) (Redirect datatrail makes traces longer, not harder) - Yes
C) (IC are frame agents) - Yes
D) (CA to redistribute persona attributes) - Eh... No, too complex without much benefit
E) (Drop memory requirements) - I don't care. They could do with simplification though.
F) (Nicer agent rules) - Yes, I like this, and I like that drones are agents. Simplicity and compatibility rule
G) (AR allows for basic matrix functions) - Yes
I) (All electronic devices (including cyber) are matrix compatible) - Not all, but most should be
J) (Everyone should have access to AR, computer skill better) - No, not at this point. It would be alright for a SOTA65 supplement, but right now it would have too much of an impact
K) (IC spends time searching) - Yes, but we need a specific rule on that, perhaps based off evasion?
L) (Decker can directly deceive IC, not just host) - Yes, this needs a mechanic
M) (Legitimate users can be dumped with tally of 1) - I think this is a requirement, really
N) (Prices for legal equipment) - Yes
O) Nezumi's matrix metaphor - Yes, of course, I'll be testing it soon, actually (although I'll drop the tally being broken to such low levels as regions for now)
P) (Hacking pool) - I like #2. Computer skill should be the important bit.
Q) (Background tally count) - Yes yes yes! The question is what mechanic. I feel like the host rating should be a critical part of that, as well as population, but that quickly gets too complex. I definitely support 4 (decrease tally and reroll)
R) (Tunneling) - Not huge on it, but I could go either way. Perhaps a good reason to make the decker enter with the group.
S) (Deck construction) - 1, 2 and maybe 4.
T) (Tally accumulation) - 1, 3. I think 5 is accounted for by 3. I don't like 4, it makes datasteals too easy.
U) (Use evasion for cybercombat) - Yes, good thinking.

[Eyeless Blond]

Forgot about: (EDIT: proposal "V" changed to "H", as suggested by nezumi)
H) Extending the "Tally is kept system wide, brought up to the tally of the highest individual on the host" rule, both to mesh better with the original rules and to be more sensible:
H1) Tally inheritance, Part I: Tally accumulated on an LTG is passed to all PLTGs/hosts "underneath" it. If the tally increases on the LTG, then the same amount is added to every host/PLTG that uses that LTG to hook up to the Matrix, but if a host or PLTG increase their own security tally, the LTG's tally is unaffected.
H2) Tally inheritance, Part II: RTG tally wouldn't be passed to the lower-level LTGs/PLTG/hosts, however; this is to create an actual mechanical difference between RTGs and LTGs, where now the difference is entirely symbolic. Linked RTGs also don't pass tally to each other.
H3) Tally inheritance, Part III: PLTGs share security tally. A PLTG has a single security tally, which is increased or decreased by each host in it, as if they were all one big superhost with a single value for tally.
H4) All of Q) actually goes under here. The main reason Q exists is for flavor: without "false-positive" tally then all tally would have to be the result of real intruders, and it would be insane to have systems ever remain online even after a single point of tally is accumulated.


Impressions:

Huh, and here I was under the impression that all of the following were already accepted by everyone: A, B, C/F (though we do need specific Agent/IC/drone construction rules, and I really don't think C1 is a good idea), K, N.

D) Eh, I wouldn't have a problem with this, though Nexumi is right it's mostly complexity without added benefit. Maybe if more attributes were important than just Masking.

E) Eh, in my games it never matters, as I usually end up getting more active/storage/offline storage memory than I can ever use, just to have one less number to worry about.

G/J) No AR-for-everyone, but something like a DNI-enabled pocket secretary, or anything other significantly-advanced DNI-enabled electronics should have some sort of HUD, rather like the car's virtual dashboard, which can be accessed without having to activate an RAS override. So no AR, but DNI-enabled tortise mode.

I) I like rating+4 myself. Remember that TNs increase exponentially in difficulty; adding to the TN results in multiplying the difficulty; doubling the TN actually squares the difficulty.

L) Er, this already has a mechanic. It's Evade Detection, classified as a combat maneuver, the decker rolling Evasion against the IC rating-decker's Cloak utility, the IC rolling its Rating against the decker's Evasion-IC's Lock-on(I think?) utility. If the decker wins he evades the IC for a number of turns equal to his net successes. I think this one's a keeper.

M) Auto-dump I think should apply only to users without Masking; *everyone* is trying to fake being a legitimate user; the ones with Masking just can do so without being under the host's control.

O) The Matrix being run as a pseudo-office building? *shrug* I can get behind that, though what I'd rather do is have sculpted systems be possible, but have the decker be able (and willing) to bypass the metaphors when it's not convenient to the game. More on this later; I've got a few ideas that I want to flesh out.

P-2) Absolutely agree that (Int+Wil+Comp Skill)/3 is the best.
HOWEVER, if we do this, we might want to consider making computer skill a little harder to come by. Maybe require that getting a new rank in computer skill requires 3 times as many points invested in computer-related knowledge skills, with 2 skills at equal or greater level? Something to add a little extra karma cost to upgrading computer skill, since each rank in the skill would add 1.333 dice to any decking test.

Q-2) The original idea was that all tally, including the randomly-generated stuff, was inherited, according to the rules above (see H)
Q-3) No, tally should depend on the size of the host, the number of users, etc. Essentially, the number of times that innocent transactions can be mis-identified as hostile acts. Higher security hosts already have shorter trigger steps, so the higher the security level the more IC/alerts will be pre-activated due to the same default background tally.
Q-4) No, random tally fluctuations will be handled by Null Ops when the decker is online, and a simple reroll when the decker is offline. Besides, constant decrease would mean all host tallies would go to zero over time.

R-1) Rather than the deck, this sort of thing should be the realm of cyberware. Specifically the encephalon, which according to the flavor text is supposed to supplement your thought processes; this rule would actually allow the 'ware to do that by taking over sustaining a currently active thought process.
R-4) We could actually do both, and in fact if we do the first we should do the second, simply to prevent large corps from making strings of 20 hosts to completely shut down outside incursion of their PLTGs (doing so with R 1-4 enacted would mean that noone could actually use such a daisy-chained host, making it less likely to exist.)

S) Yes, deck construction need to be rethought.

T) 5 can accomplish 2 and 3, as well as Q-4, all in one test. If things are too hot the decker tries to "lay low" for awhile and do what he can to lower the tally. See here for details. I also like T-1, though that might be a bit much.
T-4) I don't think this will be necessary. So long as K and/or O are implemented, having even several pieces of active IC wandering around won't be a big problem.

U) I came up with this, so obviously I like it.

[Eyeless Blond]

Oh, I just had a quick brainstorm:

W) "Backdoor" utility metaphor, to eliminate all that silly system sculpting when it's not useful--namely during Overwatch missions.
W1) The "normal" metaphors that scripted hosts have work fine for the vast majority of users, and most of the time for deckers as well.
W2) For security deckers, however, a more utilitarian approach is needed. Security deckers need to be watching the security feeds of the actual building, not ordering a hamburger or dancing with a brunette icon or whatever the sculpted equivalents are.
W2.1) So, such deckers have access to a special "utility metaphor," The utility metaphor almost invariably looks like a wireframe digital representation of the building the decker is patrolling, and gives them easier access to slaved systems and other security operations.
W2.2) Intruding deckers can gain access to this metaphor as well. All they need to do is find a "backdoor" into the metaphor with a Locate Resource operation, and can be accessed with an "Logon to utility metaphor" operation (essentially the same thing as the "Logon" operation). It can be exited just as simply with a "Graceful Logoff" operation. Intruding deckers use this mode for overwatch missions usually.
W2.3) Mechanically the utility metaphor just changes around a few of the ACIFS ratings, like so:
Access- Unchanged.
Control- Null Operations: -1TN; the metaphor is built specifically for security deckers and other long-term inactive occupants, and so such occumpants are not observed as closely. Otherwise unchanged.
Index- To Locate any Slave system: -1TN To Locate anything else: +2TN. Note that this does make the decker somewhat more vulnerable to being attacked by IC, as he's less likely to notice them around. I feel this is a good thing, as there may even be cases where a decker in a real pinch would have to log back into the "real" sculpted metaphor to locate and destroy/hide from a bunch of IC.
Files- +4 to all TNs. This metaphor is not meant for storing or manipulating file information
Slave- -2 to all TNs. The primary purpose of this metaphor is for security deckers to be able to monitor several slaved security systems at once.
W2.4) (Maybe) In anticipation of deckers swapping between metaphors as needed, maybe Index, Files and Slave systems should all be raised a point by default?


Okay I've looked over O) again, and I like it a lot more, especially the stuff about IC.

O-1,2) Yes, absolutely, though change the word "reacts" to "activates": "Patrolling IC activates instantly when a specific security tally value it reached, but it takes time to FIND the decker based on local activity and rating."
O-3) Counter-proposal: V1-4. Note that "Region" is exceptionally vague in a place where geography is entirely subjective btw; note how some metaphors put an entire host inside a single dance hall, while others include whole buildings.
O-4) Absolutely, though I vote that a user should only be automatically dumped if he doesn't have Masking.
O-5) Hm, I'm not sure about "passing by"; that's especially vague in a world with subjective distance/geography/iconography. How about this instead:
"Guard IC, unlike patrolling IC, are always active, and due to the resource drain they cause are fairly rare. They actively guard one or a number of specific resources (specific files, slaved systems, etc) and scan each icon that passes by and will automatically activate when noticing suspicious icons, regardless as to their tally." This is where things like Scramble IC come in, as they are clearly not the patrolling-IC type.
O-6) No, I kinda think that alert levels should stay on the originating host.


Wow, three posts to the same thread in one evening. I wonder if anyone will actually read all that? :P

[nezumi]

Because we're running out of letters, I'm moving V under H (but otherwise keeping it more or less the same), although Q. is still independent. I think it's separate enough to be approached as its own issue.

A I thought was still under discussion, especially since we hadn't gotten the details down. I do agree that the general idea seems to have been well received though.

B I can agree as being settled and I'll put it where it belongs.

C I can agree as being settled, since no one disagreed. I'll drop suggestion 1.

K and N haven't really been discussed and they need specific rules. K especially. How long does it take for IC to find its target? Rating + Analyze vs. TN of the decker's Evasion?

I've rearranged some letters on my file (you won't see the changes until I post it again though, so don't worry about it).

L - You evade them for only a short period of time, and that's evading, not deceiving. I have players who want to show a "fake ID" to the IC and let their detection factor get to work again, which sort of makes sense, given the metaphor.

P - If we try to make computer skill more expensive, it might be better to break it into multiple skills. So Hacking pool is based off cybercombat, a new skill that you use for well, cybercombat. Computer still applies to all other non-combat tests. Debatably, hacking pool no longer applies to non-combat tests, in which case the average ACIFS ratings should be decreased. But that might be making things far too complicated.

Your editing changes I've made, since they're good. The exception is the 'stationary IC' one which I've shortened slightly. Keep in mind, I generally don't bother keeping the entire justification for something, only a line, maybe two, since that would quickly get too long to comfortably navigate.

[Eyeless Blond]

[nezumi]

[Eyeless Blond]

[nezumi]

Ay! You and your huge posts! I'll read this from bottom up...