The liner, given enough force, could bend if it is thin enough.
This definitely could happen. However, I have tested several low-middle grade liner locks of normal lockbar thickness with intense force (I assure you, it far exceeded the Manix failures here) and there were no problems. Unfortunately, I can't test a huge selection of liner locks merely for financial reasons, but I at least have given strong support (as opposed to proof) to the idea that a merely normal width low end liner lock should not bend under the force a human being can exert.
Furthermore, it's contacting an angled surface, which could push it to the side and thus disengage the lock.
The point of the angled tang and lockbar is precisely that they're
not angled relative to each other. Because the lock has to come off of one side of the handle, it will hit the tang in a weird direction if it's flat, and then we have to worry about it sliding. However, the angle of the tang should put the force directly perpendicular into the lockbar. The important part of this paragraph is that the surfaces aren't angled relative to each other.
There is simply more physical room for error in a linerlock, compared to the tight spaces filled by a lockback's lockbar.
Well, again, no one has any idea as to the tight spaces that might be filled by the lockbar of a lockback--because you can't see them. Should you have a defective one, which, as we can see in this thread, is alarmingly common even in the toughest of lockbacks, you would never know until something happened or you tested it, whereas a sloppily made liner lock could be made obvious by visual inspection (poor lockup, etc).
I think we're mislead, as non knife makers, into thinking the lockback is simpler. It just isn't--I've taken apart many a knife, and no lock (that I've seen) is more complex than the lockback. It far exceeds the axis and variations of liner locks in complexity. And if your intuition is that complexity is inherently more troublesome than simple designs, then you should be worried. Fortunately for your side of the argument, of course, this intuition is patently false. Complex designs are not inherently more prone to failure than simple designs. My only point is, then, that I do think we're mislead to believe that lockbacks are just so much simpler--we need to reevaluate that position.
Looking at benchmade's diagram, we can see that when a lockback has pressure exerted on the spine, it's two flat pieces of metal contacting. The tang depression pulls against the lockbar in a direction it is not designed to move, unlike the linerlock being pushed in a direction it is designed to move.
Again, liner locks aren't being pushed in any direction at all--they're being pushed straight down the line of the lockbar.
They also tend to be more difficult to accidently disengage, have smoother opening due to a lack of ball-detent and have one of the most reliable, longest-running track records of locks used in modern production folders.
I'll grant you the accidental disengagement part of your argument.
I disagree on the smoothness of opening, and I'm a bit troubled that you'd assert that. I'll make the claim that decent liner locks are universally superior in action (all other things being equal) than lockbacks. I can't imagine how someone would come away thinking otherwise, but maybe your liner locks were overly tightened or dirty. As opposed to ball-detents, they're just one more misunderstood link in the puzzle. They work great in my experience, and the knives least likely to open on their own in my collection are definitely the ones with the ball detent. I can easily hand flick lockbacks and axis locks, but my ball detent knives resist a lot.
I'll grant you the longest running track record, but will disagree on the most reliable. The only reason it has the former, of course, is that it's an older design, and this might be valuable for inductive reasoning, but we don't have to use inductive reasoning for our current tests. That a Buck 110 was reliable says little about whether a Manix will be, and even if it were, we don't have to reason by analogy--we can actually test the Manix. Track records on something this testable aren't that valuable.
They are clearly not the most reliable, and I believe this claim is irrefutable. I believe that anyone objectively looking at failure rates will always side with axis locks, which have, essentially, a 0% failure rate, in my experience and I believe this board's experience.
Of course, individual knife design plays a heavy role in the safety of the lock design, so broad comparisons will only mean so much.
Quite true, and I find this criticism one of the most damning of lockbacks. When we're talking liner lock failures, we're usually discussing a 40 dollar CRKT small folder, but here we're talking about a 100+ ultra tough lockback from arguably the most respected lockback maker ever. If Spyderco has difficulty implimenting the lockback using their most precision design, then we may all be in for trouble.
Lock-back folders are quite difficult to make well (especially mid-locks). Geometry is tricky, heat treat critical and small deviations can make big differences. -- Sal Glesser
He does it so we don't have to, then he tells us about it. Sounds perfect to me.
