I have few things I do differently with some of the test mechanics. I thought that I would offer them out to see if anyone else found them useful.

Extended Tests
I don't like a test that just tells me how long it will take but doesn't really reflect the difficulty. If there is no difficulty in it then just take the threshold and divide it by the skill rating or purchase the successes. What is the point half the time? I especially don't like the idea on hacking rolls where the longer you are at it the more alarms you will set off.

I try to always include a threshold in my extended test rolls. If you fail the roll twice or more then you have to make a willpower roll or give up. You become frustrated too much. The fun thing is too alter the threshold depending on what they do. If they try and do too many things then it goes way up. They get nothing done that day. If they come up with a good plan or the player researches it then it goes down.

Static Ratings for Equipment
I tend to think that the reason for the variety of success and failure in dice pools is because of the wild differences in focus that a person can have from moment to moment. But a device doesn't have attention span problems. It should provide a flat bonus or penalty whenever possible instead of altering dice pools.

For example, if you are facing a piece of equipment with a rating then that rating should be your threshold and not rolled by the GM. Then if you have a piece of equipment that helps against the other equipment then it's rating should lower the threshold back down. For example, a rating 3 maglock passkey should lower the maglock rating 4 to a threshold of 1. Or if the passkeyrating was greater then or equal to then it is exactly the device for the job and no roll.

I do the same thing with a lot of hacking rolls. I use one program to lower the rating of their programs or defenses. Exploit cancels out Firewall. Etc.

Do the same for ID checks. If your ID is of a higher rating then the reader then no roll by anyone. If your ID is less then the reader then start making rolls.

Use a threshold for range instead of a dice penalty but use the scope level to cancel that threshold out.

Cool, I like the first one. Agree with your reasons.
I'm kind of iffy on the second. I like the game mechanics of opposed rolls often enough, especially for things like technological boosts - I like the idea, for example, that a smartlink adjusts the die pool (offering a potential aid) rather than having a flat bonus or penalty. Mostly for the game effects it causes during play then for any realism considerations.

As you saw in my custom hacking rules, I've included an "Extensive" test that replaces most "extended" test. An Extensive test is just like any other die roll, except that a player can invest more time in it to add bonus dice to the test. In my campaigns, every day that you devote to a test adds a few dice... so it could take a week to find some information if you just Data Search for it, for example. But a highly skilled person with cool programs could of found the same information right away, as a standard test.

I also like your first one. In fact's, it's the only one in your tons of houserules that I might consider stealing.

For extended tests, you could also simply stop when a Glitch occurs.

Bye
Thanee

If you have an Extended Tests that you want to answer "IF" instead of "WHEN" then just use the rule that's in the book. Limit the rolls. From experience I highly recommend the standard maximum of 4 rolls. Or 3 if you want to be a hardass. Simple, fast (no looking up Skill value to figure out how many rolls you get....a variation that doesn't even give that great of results in play anyway).

Cetiah I like the idea of the Extensive test. I would combine the two. I was thinking of even having my extended test cover healing to bring it more in line with SR3. Basically your damage penalty acts as a threshold for your tests. Modify this by your living conditions.

I could see some equipment giving a bonus and some removing threshold. The advantage to removing a threshold is that is means you are getting guarenteed successes equal to the rating instead of maybe getting successes. Of course this doesn't change the fact that you still have to suceed on the roll but it makes the roll easier.

I agree with stopping some rolls when you glitch. You could also reduce your successes by that much. I think that it depends on what kind of roll it is. If you are hacking then once you glitch you should probably jack out fast. If you are building something then you might have set yourself back a few intervals. If you are healing then you don't want to limit rolls. Now a critical glitch could change things alot on any of thos rolls.

I use the "only skill+1 rolls" rule to cover difficulty (depending on the test, some test are really open ended).

I jsut follwo this from the book, p. 58

"Th ough it may seem that characters are guaranteed of success
in an Extended Test over time, that is not always the case.
Th e character may have a limited timeframe in which to accomplish
the task, so she may run out of time before she fi nishes the
job. Th e gamemaster can also limit the number of rolls under
the assumption that if the character can’t fi nish it with a certain
amount of eff ort, she simply doesn’t have the skills to complete
it. A good limit is to allow a maximum number of rolls equal
to the character’s dice pool (so a character rolling 6 dice has 6
attempts to get it done). A character can also fail an Extended
Test by glitching (see below)."