The C part of that looks pretty damn easy, the only possibly tricky part would be knowing exactly what the result looks like for the code with the memory management error (the exam doesn't seem to ask for that bit of info). But spotting the error when you are told it is there is fairly easy. I remember the introductory C course i [occasionally ] attended was much more coding centered in the class and final evaluation. From that exam it looks like a puff class on the coding side, more of an introduction to programming.

Of coure i never said an exam, i was talking about a project/task done within the language. Exams introduce and tend to require stuff outside the actual coding where the prof is looking for an echo back of what they said in the course, in the particular terminology the prof used.

The "Big O notation" is a good example of something that doesn't have a lot to do directly with coding. It is more a communication with others issue because a person can identify and understand a problem in their own "notation".

Although i noticed that in this particular exam it does explain the terminology of "double-linked list".

I'm not sure a final project is a fair test of anything relevant to this conversation. Final projects require far more than a couple of hours of work--so you'd spend the time on the project both learning the language and doing the project. Fine for an extended test, but not one that's supposed to resolve in seconds or minutes.

Closed book final exams are better at testing what you actually know at that moment.

How quickly you can sling code is much less important than how you approach a problem.

~J

Dude, not everyone uses an IDE. Last time I checked there isn't a reference manual built into TextEdit, nor should there be.

~J, doing a stunningly bad job of staying out of this one.

If we want to view the written skill level not as 'actual skill' but as a sort of 'effort taken to learn' and admitting the fact that the 'actual skill levels might be highly different' this does produce something like the current SR4 results.

Assuming we rate skills at 'stat plus skill' (and not just skill level) then rewrite the 'average skill table' then things might make more sense.

So a 'skill level' (actuall stat plus skill) is what is being measure not the skilli in isolation (as the skill in isolation dosen't measure anything.

So you might have (for shooting stuff) the following skill levels:

4 raw recruit
5 most police officers
6 regular army
8 highly trained or veteran
10 highly trained and veteran
12 olympic class

This might help reduce the problem of being 'accidentally good' at something because your charater has a high linked stat. However, we run into the problem of granularity here. Someone with a stat of 6 can't only know a 'little bit' about the subject. They are either defaulting (and doing a pretty suck job at it) or they have a skill level of one and are kicking ass already.

There's an easy fix that doesn't mess with the rules here. The GM can limit or tailor any "given" information (meaning stuff he just tells you without needing a skill roll) based on your skill rating.

Example 1: While planning an infiltration you get ahold of a copy of the invoice for the installation of a model H63 Novatech security terminal.
Skill 1 guy will need to do a little research to find out what tools he'll need to bypass it.
Skill 6 guy knows it's late model voice and retinal scanner. It's in-set into the wall and is a real bitch to crack open. When properly accessed, the connected door stays open for exactly 7 seconds. Blocking the door as it shuts sends an alert to central control, but only if it remains blocked for more than 2 seconds.

In the end, they both have equal competency in executing the task, but Skill 6 guy shows himself to the knowledgeable veteran thief. Skill 1 guy might have to do it twice after the door shuts before the whole group walks through, depending on the extent of his research, that is.


Or for harsher GMs:
Example 2: During a gun fight you run out of ammo. You grab a tricked-out assault rifle off the corpse of one of your enemies.
Skill 1 guy wants to take out a far-away enemy. He takes aim through the scope and misses.
Skill 6 guy realizes that the dead guy has a very different build from his own. The sights will be horribly off, so he switches to full-auto and fills the air with bullets instead...


This method gives tangible benefits to skill-boys without screwing up the basic mechanics of the skill roll (actually bypassing the security terminal/shootin' folk). The second example shouldn't be the norm. It's just one of those things that might happen every now and then. Poor guy loses an action, but he won't make the same mistake twice, assuming he lives to take another action.
My favorite part: The bonus info given to high-skill characters sets a precedent "You'll never be the best of the best without a high skill". Skill 6 guy is constantly reminded he's cool. Skill 1 guy has at least some incentive to raise skills. Don't get me wrong, it's still a trade off. Higher stats will make you better overall, but higher skills makes you elite.
The more you, the GM, hate rampant stat-buffing, the more serious the bonus or lack of info will be.

Make sure your players know what you're up to, and they'll probably come around to your way of thinking.

The above is the way we always do it isn´t it? There are after all skills that you never roll, especially knowledge skills. Skills have always represented how much you know about a subject, and how well versed you are in it. Some of you are in such disagreement over the programming example it has become a bad example. Pick another one. Is it still unimportant wether you have gained your proficiancy through raw aptitude or trained skill?

Lets say we take electronics as an example. We have this moron janitor on one hand, with logic 1 and electronics 4 (he´s been working hard to get that stat). On the other hand we have some player´s street sam character with high stats all over, and 4 in logic, but who has neglected his electronics skill, so that´s still 1. Now, in a contest with moderate difficulty we will be faced with a stupid guy doing what he does every day and a smart guy doing something likely for the first time. I wouldn´t feel bad giving the janitor some kind of edge in that contest.

(That would not be arbitrary then, but situational. Faced with a problem without immediate suloution the moron janitor would be completely lost. Then he would be the one getting a penalty)

Off course I might just not be smart enough to see that superior, innate, logical capacity can take you past the greatest scholars in any given field.

I have come to a conclusion regarding SR4's karma system. Even though the SR4 system recommends giving less karma per run by neglecting the "threat" karma from previous editions and they have made the costs of improving attributes go up as well they have made up for it in another area. They have made that one point slightly more valuable. Because there are attribute caps and the cost of attributes have increased the average runner will not max out them all. In most groups I played with the average character has maxed out the attributes that apply to him. Look at your SR3 characters and honestly how many of your mages have at least the willpower maxed out, and probably the intelligence too, and even the charisma if they conjure or plan on getting into astral combat. How many sami's have their Body, and strength maxed out? They probably also have the quickness and intelligence maxed out too for the reactions effects. They probably even have the willpower maxed out for the combat pool and to resist mana spells. In SR3 everything depended on your attributes. So far SR4 doesn't look nearly so extreme. Also, the need to have your skills maxed out wasn't so severe. Think of the hardship of learning a new skill, especially if it was a melee combat skill. If you have the skill at 1 through 3 you would probably get your ass kicked. Now a little natural talent is evident in the attribute+skill development. All in all I think the changes aren't that bad. Even if you have to work a little harder to earn that one point increase in a skill or attribute I think it will be worth more.

Defaulting to attributes is appropriate. It's possible, but overwhelmingly overshadowed by skill (even in small quantities).

~J

Defaulting to attributes is like someone who's got a very very basic idea on how to use a pistol just going 'frag it' and doing what he thinks he should, because not everything makes sense together yet.

Skill 1: He knows the steps, but he's having trouble with the order, aiming, just in general.

Default to attribute: He just points and pulls the trigger, doesn't try thinking about sighting, all that. He's probably got all sorts of 'errors' in the technique, but he's got a better chance then confusing himself trying to remember the proper technique.

Microsoft makes BIOS software?

Blackkie, it's funny how you read my post yet seemed to avoid the examples you complain I'm lacking. Twice now I've specifically addressed your problem.

But the ego police was cute.

Neither of which actually deals with my question...but sure i'll comment on these....

"Why do you shut your eyes?"

"So that the room will be empty."

Do you have any evidence that HoloLISP is interpreted rather than compiled, or any other reason to believe that your point #1 is at all relevant? Perhaps you're trying to say a different architecture instead, in which case the compiled code might indeed be unusuable—but it's still running on your machine, not the server. Perhaps you mean a different networking protocol?

I must say, that last post has either exposed that I need much more coffee today or that you have much less clue what you're talking about than previously surmised.

~J

As Hacking is done with Hacking+Program - is there even the possibility to default?

If you don't have a Hacking skill, you can still just let the program run by itself (the script kiddies, as stated before). If you don't have the program, you can't do the action, period. I suppose a generous GM would allow someone to come up with a program on the fly if you had an hour of time and a Software skill, but you certainly can't crack a device without an Exploit program or command a device without either a Command program or legal access (which doesn't need a Command program). Note that this is one of the main advantages of being a Technomancer, as you can use Threading to whip up Complex Forms that you don't already have.

Also, a program is only necessary for actions that require a program. Doing specific actions on specific devices that do not require programs uses a Hacking + Logic roll as usual, and thus can be defaulted. Hardware, Software, and Electronic Warfare are three skills that CANNOT be defaulted according to the skill list.