OOTS #1071 - The Discussion Thread

[Psyren]

Well, the good news is that if they turn back, they might find Roy's sword, and possibly even the rest of the Order who are trying to catch up with them.

The bad news is that if that crazy giantess reaches the controls, we may very well end up with a dead Bandana and a downed ship.

Either way, well - I think the Dwarven kingdom is screwed.

[Riftwolf]

The engineer I guesOH RIGHT

Guess for future strips: something happens that makes Andi order full speed ahead again... Only to find the ship can't. Turns out engines don't work on the HP system and the mountain beat the break DC.

[Grey_Wolf_c]

Guess for future strips: something happens that makes Andi order full speed ahead again... Only to find the ship can't. Turns out engines don't work on the HP system and the mountain beat the break DC.

More likely: they barely made it pass the last peak. Since then, more air has escaped & several Schroedinger ropes have been cut. I don't think the ship can make it back past the same peak, which means they are trapped in a valley, with a still quite large army of giants in probable pursuit behind them.

Grey Wolf

[factotum]

When you're debugging software, the answer is "no". It can sometimes take many experiments and tests to isolate the problem, which can then be corrected.

I've seen the same thing in the auto shop; While a mechanic may have a reasonably good idea what the issue is, a lot of times it's still a good idea to check the onboard diagnostics before opening 'er up. And in a few cases the symptoms could result from a number of different problems, and it will require some experiments to narrow it down before a proper diagnosis can be made.

None of which is "trial and error". The mechanic or programmer has a good idea of where to find the problem--for instance, if the car is belching blue smoke from the exhaust, there's very little point in checking out the electrical system. "Trial and error" implies that you're so clueless about what the problem might be that you can only fix it by changing random things and seeing if the situation gets worse or better. That works well over the course of millions of years when you're talking evolution, it isn't so hot when you have a problem to solve and you're on the clock.

[Keltest]

None of which is "trial and error". The mechanic or programmer has a good idea of where to find the problem--for instance, if the car is belching blue smoke from the exhaust, there's very little point in checking out the electrical system. "Trial and error" implies that you're so clueless about what the problem might be that you can only fix it by changing random things and seeing if the situation gets worse or better. That works well over the course of millions of years when you're talking evolution, it isn't so hot when you have a problem to solve and you're on the clock.

Trial and error is not synonymous with making random changes. Intelligent and systemic examination of the most likely causes is still trial and error if you don't actually know for sure what the cause is already (most likely through somebody else's trial and error).

[Grey_Wolf_c]

Trial and error is not synonymous with making random changes. Intelligent and systemic examination of the most likely causes is still trial and error if you don't actually know for sure what the cause is already (most likely through somebody else's trial and error).

Trial and error is also not synonymous with divide and conquer. An engineer should start with a rough idea of where the problem is, and by carefully controlled trials, isolate the problem so they can perform a punctual fix. That process is not called trial and error - at least not in my specialty.

"Trying to find your way out of a mountain range by randomly picking the route" is a lot closer to trial and error than to any systematic engineering method, although admittedly it is close to the only guaranteed way out of a maze (as long as she sticks to only ever making left/right turns, which we don't know if she will). However, we call that approach "brute force" and in general it is also a terrible algorithm to fix things, taking large amounts of time and being a last resort solution.

Grey Wolf

[Rogar Demonblud]

I said this, like, strips ago. Just stop already instead of dodging mountains. There's no threat from above or below now they are off the pass. Deal with the giantess; maybe even land, then repair and send for the rest of the party.

As we learned from the storm at the beginning of this book, STOP equals CRASH. And that was before a significant amount of lift was lost. Plus, the ground is where the frost giants (and allied ogres, winter wolves, etc) are. It wouldn't take them long to find you, especially if they use the white dragons to scout.

[Jay R]

I think your idea of what the subplot is is different from mine.

Very much so. I think we're exploring the meaning of mutiny, loyalty, leadership, and envy.

Andi is an antagonist, probably a single-arc antagonist. "Her head gets lopped off as she draws breath to continue to declaim how she's right about everything" would be as much of a resolution to the subplot involving her as most antagonists get.

(I don't think that'll happen because Lesser Confusion only lasts one round, but I won't be at all surprised if something causes her to die abruptly any strip now.)

And then the issues of mutiny, loyalty, leadership, and envy disappear without a resolution.

[arimareiji]

Fwiw, count me among the "Why are some people so d--- good at finding the worst possible time to start teenage melodrama?" contingent. (For better and worse, a reread seems to show Andi's sulk has been brewing for some time.)

Her career going caput?

I believe that qualifies as capital punishment.

(I don't think that'll happen because Lesser Confusion only lasts one round, but I won't be at all surprised if something causes her to die abruptly any strip now.)

Falling overboard after shouting down someone trying to warn her of Frostja's rush would be more darkly-amusing, but decapitation doesn't seem unlikely. As we've seen, Andi is eager to get a head - but looks more likely to lose it.

[Jay R]

Well, if this is how Andi tackles engineering issues, she's also a rotten engineer. "Trial and error" is NOT how you fix a machine, in my experience. That approach implies that you just replace every part at random until you get around to the one that's actually broken. Maybe I'm wrong here, but don't actual engineers know enough about their machines to have a pretty good idea what's wrong before they start working on them?

It occurs to me that trial and error probably works fine- if you're the engineer on a ship that always arrives in the nick of time.

[Doug Lampert]

Trial and error is not synonymous with making random changes. Intelligent and systemic examination of the most likely causes is still trial and error if you don't actually know for sure what the cause is already (most likely through somebody else's trial and error).

Yep, I just dumped a bunch of debug prints in the code, changed the input slightly, and recompiled. Now I'm running to see what happens.
Drat, it worked. I'll need to change the inputs again to regenerate the error.

But now I know that the last change in inputs fixes it, which is something I didn't know before.

What do you call this if not trial and error? I am literally trying different things and seeing if there is an error message and if so what else the debug prints might tell me.

Yes, there's 30 or so years of experience and a Ph.D. in deciding WHERE to make the changes and put the prints and in interpreting what they mean, but it's still trial and error. Calling it anything else is silly.

Trial and error is also not synonymous with divide and conquer. An engineer should start with a rough idea of where the problem is, and by carefully controlled trials, isolate the problem so they can perform a punctual fix. That process is not called trial and error - at least not in my specialty.

"Trying to find your way out of a mountain range by randomly picking the route" is a lot closer to trial and error than to any systematic engineering method, although admittedly it is close to the only guaranteed way out of a maze (as long as she sticks to only ever making left/right turns, which we don't know if she will). However, we call that approach "brute force" and in general it is also a terrible algorithm to fix things, taking large amounts of time and being a last resort solution.

Grey Wolf

Definition of trial and error: a finding out of the best way to reach a desired result or a correct solution by trying out one or more ways or means and by noting and eliminating errors or causes of failure; also : the trying of one thing or another until something succeeds

Are you sure about your definition? Because the one above matches my usage and Keltest's fine. I can be the best engineer in the world, and if I'm trying things and eliminating possible causes of failure based on the outcome of those tests I am doing EXACTLY what is defined above. Nothing requires that the tests be random or uninformed.

[Kish]

The problem with a "trial and error" approach here is that it's on an airship in flight and in a running battle with frost giants. If Andi finds out through trial and error that the airship actually can't fit around the other side of the mountain, "Drat" will not meet the case.

[Keltest]

The problem with a "trial and error" approach here is that it's on an airship in flight and in a running battle with frost giants. If Andi finds out through trial and error that the airship actually can't fit around the other side of the mountain, "Drat" will not meet the case.

Which is certainly an argument against exploring new territory while under attack and on a time sensitive mission, but not an argument against trial and error as a problem solving method.

[Grey_Wolf_c]

Which is certainly an argument against exploring new territory while under attack and on a time sensitive mission, but not an argument against trial and error as a problem solving method.

Oh, sure, given infinite time, money & engineers, you could use trial and error to solve all issues. But this is a flying vehicle. Even when I have had to resort to trail and error because no better method could be applied, I didn't do it in the live process. Trial-and-erroring a, say, production program is a recipe for causing more problems than you already have, which is why development & testing environments exist. Which in a vehicle usually involve a dry-dock.

So, I repeat: I would not get on a vehicle I suspected was fixed by trial-and-error, especially not one that is liable to sustain damage in transit.

Grey Wolf

[Anarion]

Well, it's good that they're turning the ship around. Although honestly, seeing the Mechane take the wrong path is incredibly worrying when it's Julio Scoundrel's ship.

[Keltest]

Oh, sure, given infinite time, money & engineers, you could use trial and error to solve all issues. But this is a flying vehicle. Even when I have had to resort to trail and error because no better method could be applied, I didn't do it in the live process. Trial-and-erroring a, say, production program is a recipe for causing more problems than you already have, which is why development & testing environments exist. Which in a vehicle usually involve a dry-dock.

So, I repeat: I would not get on a vehicle I suspected was fixed by trial-and-error, especially not one that is liable to sustain damage in transit.

Grey Wolf

So youre saying that performing diagnostics and then attempting solutions based on the results of those diagnostics in decreasing probability of success is a bad way to go about fixing things?

[Grey_Wolf_c]

So youre saying that performing diagnostics and then attempting solutions based on the results of those diagnostics in decreasing probability of success is a bad way to go about fixing things?

No, I'm saying that trial and error is much, much more broad a concept than what you are describing. As per the definition quoted earlier: "the trying of one thing or another until something succeeds" does not imply, or necessitate, diagnosis. It is more likely to make things worse than to fix them, because all things being equal, a change to a mostly functioning process is more likely to make it less effective than more.

ETA: it's like saying "percussive maintenance works... as long as you know precisely what hammer to use and which spot to hit, and never miss". Sure, but that is a sub-sub-subset of all methods of approaching a problem that can be described as "percussive maintenance". Equally, while I could agree that very careful diagnosis, followed by a controlled environment testing and identification and isolation of errors can be described as "trial and error", so can "just remove and replace parts of the engine until the engine stops smoking".

ETA2: and I'll add that, given that in the comic it is being compared to trying all the valleys in an uncharted mountain range to look for an exit, I think it is quite clear that Andi meant it more in my general "just try stuff randomly and blindly until it works" than you "carefully isolate the problem" sense.

Grey Wolf

[Keltest]

No, I'm saying that trial and error is much, much more broad a concept than what you are describing. As per the definition quoted earlier: "the trying of one thing or another until something succeeds" does not imply, or necessitate, diagnosis. It is more likely to make things worse than to fix them, because all things being equal, a change to a mostly functioning process is more likely to make it less effective than more.

GW

Yes, "trial and error" does technically include "randomly pulling levers and pushing buttons and seeing what happens." No, that is not the entirety of a trial and error approach and that such a thing technically falls under the umbrella term does not mean every other approach under the term is automatically bad.

I do think its pretty telling though that you immediately conclude that Andi meant that she randomly pushes buttons and pulls levers when she performs maintenance though.

[Grey_Wolf_c]

Yes, "trial and error" does technically include "randomly pulling levers and pushing buttons and seeing what happens." No, that is not the entirety of a trial and error approach and that such a thing technically falls under the umbrella term does not mean every other approach under the term is automatically bad.

I do think its pretty telling though that you immediately conclude that Andi meant that she randomly pushes buttons and pulls levers when she performs maintenance though.

As I edited above: she is suggesting finding an exit from the mountain range by randomly trying every valley they come across, so I think I am in quite solid ground when I say that her definition of trial and error is closer to what I'm using than to yours.

GW

[Keltest]

As I edited above: she is suggesting finding an exit from the mountain range by randomly trying every valley they come across, so I think I am in quite solid ground when I say that her definition of trial and error is closer to what I'm using than to yours.

GW

Except she isn't just randomly picking directions, she is methodically exploring the valley. As someone noted above, that is pretty much the only guaranteed way to navigate a maze who's layout you don't know, so long as you take steps to avoid going the same way multiple times.

[Grey_Wolf_c]

Except she isn't just randomly picking directions, she is methodically exploring the valley. As someone noted above, that is pretty much the only guaranteed way to navigate a maze who's layout you don't know, so long as you take steps to avoid going the same way multiple times.

No, she is randomly picking directions. And that "someone" was me, and I was quite clear we'd need to see her be methodical about the search, which so far she has not been.

It is also a really poor way of escaping a maze in which enemies are behind you, especially if you do know the two ways out, and you decided to NOT go in either direction.

ETA: also, the right hand rule does not work when there is a possibility of loops in the maze, which suffice to say it is not impossible in a mountain range.

Grey Wolf

[Keltest]

No, she is randomly picking directions. And that "someone" was me, and I was quite clear we'd need to see her be methodical about the search, which so far she has not been.

It is also a really poor way of escaping a maze in which enemies are behind you, especially if you do know the two ways out, and you decided to NOT go in either direction.

ETA: also, the right hand rule does not work when there is a possibility of loops in the maze, which suffice to say it is not impossible in a mountain range.

Grey Wolf

Theres only been once fork so far. It is completely arbitrary which way they pick first. And she flat out said how she intends to explore this valley.

I get it, you don't like the character. But theres no need to make up bad things that she's done to justify that dislike, theres plenty of poor onscreen decisions without deciding that she is apparently incapable of any sort of problem solving at all.

[Grey_Wolf_c]

Theres only been once fork so far. It is completely arbitrary which way they pick first. And she flat out said how she intends to explore this valley.

I get it, you don't like the character. But theres no need to make up bad things that she's done to justify that dislike, theres plenty of poor onscreen decisions without deciding that she is apparently incapable of any sort of problem solving at all.

So you have decided to handwave away all my arguments based on what you think I feel about the character, and your assumptions of how said feelings influence my reading of the comic.

OK, we are clearly done here.

GW

[Jasdoif]

More likely: they barely made it pass the last peak. Since then, more air has escaped & several Schroedinger ropes have been cut. I don't think the ship can make it back past the same peak, which means they are trapped in a valley, with a still quite large army of giants in probable pursuit behind them.

The part that's confusing is that Andi told Mateo to go head back to the fork, not the ridge....but they went over the ridge after the fork. Are they going to try going over the ridge again (despite being unable to gain more altitude than they had the first time), or does Mateo have a navigational frame of reference to allow bypassing the ridge?

[SaintRidley]

The part that's confusing is that Andi told Mateo to go head back to the fork, not the ridge....but they went over the ridge after the fork. Are they going to try going over the ridge again (despite being unable to gain more altitude than they had the first time), or does Mateo have a navigational frame of reference to allow bypassing the ridge?

Or is Andi so out of her element and incompetent that she doesn't even understand that what she's asking for is literally impossible?

[Killer Angel]

Well, the good news is that if they turn back, they might find Roy's sword, and possibly even the rest of the Order who are trying to catch up with them.

Yep, I believe we all are hoping they will turn back to the place where the sword has fallen!

[Dr.Zero]

They shouldn't be losing quote anymore, courtesy of the last, working, mending.
I agree that this seems the set for something else instead of managing to retrieve Roy's sword , though.

On the other hand I never correctly foresee the plot twists, so...

Re-edit: thinking twice about it, I suppose that if this was a RL airship they would continue to lose quote as long as they don't compensate dropping something heavy, to compensate for the lost gas.

Tri-edit: substitute "quote" with "height". Thanks to GW for informing me.

The part that's confusing is that Andi told Mateo to go head back to the fork, not the ridge....but they went over the ridge after the fork. Are they going to try going over the ridge again (despite being unable to gain more altitude than they had the first time), or does Mateo have a navigational frame of reference to allow bypassing the ridge?

Now, follow me: if they did drop something really heavy, maybe they managed to gain quo... height.
Roy is known to be so strong that he can tightly and firmly, and with no apparent effort, keep a rope during a storm which cannot be tied, and which two normal persons have a hard time to deal with using all their strengths combined.
Yet, Roy must use his greatsword with two hands.
Therefore his greatsword must be terribly heavy.
So, dropping it, maybe the airship gained enough height that, even taking account of the lost gas, they can pass the ridge.

(Of course it is a joke: more likely the conundrum will be solved next page, when we will learn that moving half speed they can maneuver better and pass it in a slightly lower point or that they dropped some sandbags.... or the bodies of their dead comrades. It seems really unlikely that Mateo didn't point out the problem, if a problem exist, when he had no problem to protest about the height of the ridge the first time).

[Unoriginal]

Theres only been once fork so far. It is completely arbitrary which way they pick first. And she flat out said how she intends to explore this valley.

I get it, you don't like the character. But theres no need to make up bad things that she's done to justify that dislike, theres plenty of poor onscreen decisions without deciding that she is apparently incapable of any sort of problem solving at all.

Actually, she only said "well, this road didn't work, let's go take the other one".

Trial and error can work in certain circumstances, and we know that Andi is capable of some problem-solving, but doing trial and error was never her intention in the first place, she just picked directions and either expected them to be good on first try based on her belief anything is better than what Bandana proposes or just did a random choice at the last second and then pat herself on the back for being awesome, without regards for the actual results aside from "I didn't kill everyone with my decision."

What Andi is doing is like trying sword fighting for the first time in a duel to the death thinking she'll win, then call it trial and error when she get wounded.

[8BitNinja]

Completely unrelated to what everyone else is talking about: Are the dark blue markings tattoos or are they born with them?

[Grey_Wolf_c]

Completely unrelated to what everyone else is talking about: Are the dark blue markings tattoos or are they born with them?

I'd say they are woad war paint.

GW