When the US adopted the Beretta Model 92 as the new M9 standard issue handgun in 1985, it was not without controversy. In particular, there was a scandal of cracked and broken M9 slides shortly after procurement began. The military did not budge on awarding the M9 contract to Beretta, but in 1987 they did open a new procurement competition for an M10 pistol, which would supplement the M9. The Army needed 142,000 more handguns, and they were willing to buy something different if something could outperform the M9.
Two series of trials resulted. In the first one, four companies submitted designs – S&W, Tanfoglio (TZ-99), Ruger (P85), and Browning (BDA). The S&W entry was an improved version of their XM9 design; a variation of the Model 459. This was a double-stack, SA/DA pistol with an aluminum slide and steel frame. It had larger than normal sights, an ambidextrous safety/decocking lever, and phosphate finish. It was also supplied with an extended 20-round magazine for the trials. A total of 40 pistols were made, and they performed well enough to get into the second round.
In that second round, the contenders were reduced to S&W, Ruger, and the Beretta M9 as a control. The Beretta performed best, and as a result the M10 designation was scrapped and the Army simply filled the rest of its handgun needs with M9 Berettas.
I think Ian should have gone over the entire M11 procurement situation, as well. That was a strangely bizarre set of circumstances.
I still don’t quite grasp the entire point of the M10 program, except to shut up the idjit class who were dead set on buying anything other than Beretta. That being the case…? Why did they bother spending the money?
And, the M11 itself? What was that, a sop to shut up Sig-Sauer over their complaint about missing out on the M9? Did it make any sense to procure the Sig-Sauer P228, when there was the perfectly serviceable little Compact Beretta 92?
“(…)scandal of cracked and broken M9 slides shortly after procurement began.(…)procurement situation(…)bizarre set of circumstances(…)”
I am wondering about timeline of nearby events: detection of cracks, finding reason of said cracks, launch of search for another automatic pistol.
Did they actually find why said cracks appeared or planned to deploy next automatic pistol disregarding why and hoped that it will not crack (allowing declaration of SUCCESS)?
The slide cracking appears after 4000-5000 rounds of US Army 9mm ammo. I was in the US Air Force when the problem appeared, about 12-18 months after M9 deliveries to the Air Force began. The Air Force blamed the US Army supplied cartridges, which are loaded to higher chamber pressures than NATO standard.
The earliest problems were noted by the Navy’s SEAL teams, with slide cracking at the locking lug cuts developing at 2000-2500 rounds.
In that case, the reason was the SEALs using IMI “Uzi carbine” high-pressure 9 x 19mm (technically +P+), which they preferred for its penetration on protective vests. They were using it in the H&K MP5 series weapons, and apparently didn’t read the back of the box which stated DO NOT USE IN HANDGUNS.
Beretta responded with the enlarged hammer-pin head and the cut for it on the underside of the slide in the M92FS. The idea being that if the slide fractured, the hammer-pin head would prevent the rear half being projected into the shooter’s face like the bolt of a Ross rifle.
In fact, slide fracturing of this sort has occurred in every pistol with the Beretta, or more correctly Walther P.38 type breech-locking system, including Walther P.38s and every previous iteration of the Beretta M951/M92/M93R and Taurus PT92/99 series. Taurus initially beefed up the slide on the PT99 to counteract this, then finally gave up on the M92 design concept and went back to the Colt-Browning type lock on the Taurus 908, or more exactly the SiG-Sauer version based on the Petter-designed MAS P35S.
While sold as a “Compact” PT99 variant with an 8-round single-stack magazine, field-stripping shows that the PT 908 and all PT 900 series Taurus pistols are in fact essentially Smith & Wesson or SiG-type systems. The PT99-style slide shape is purely cosmetic.
The moral is both RTFM, and that the P.38 type locking system has limitations which cannot be flaunted with impunity. Frankly, the older and more difficult to machine Luger-type toggle lock, inherited from the original Maxim machine gun and before that the Merrill carbine of the American Civil War, was and is considerably better at handling high pressures than the P.38 system.
clear ether
eon
Mind that, in the XM9 program endurance test, only the 92F and the H&K P7 reached the 7000 rounds threshold with all three guns tested intact. All the Browning system pistols (included the control 1911) had cracked frames and/or slides in at least one of the samples.
Still the M45A1, tested for endurance at the adoption, cracked the slide before 15.000 rounds fired in all four samples tested.
So, it depends on how you do it. And it’s better to not use rounds that develop peak pressure in excess of proofloads.
“(…)SEAL(…)DO NOT USE(…)RTFM(…)”
This means problem was of organizational nature rather than technical.
Introducing new (beefier) automatic pistol might alleviate that, but I presume would lead to new problem: wait can I use this ammunition in my automatic pistol or not?
“(…)slide fracturing of this sort has occurred in every pistol with the Beretta(…)Walther(…)”
Which did not prevent either from being produced for several years, see
http://modernfirearms.net/en/handguns/handguns-en/germany-semi-automatic-pistols/walther-p5-eng/
“(…)P.38 type locking system has limitations which cannot be flaunted with impunity(…)”
I agree with that, but that is true for all locking system.
Thank you, eon. I didn’t know the SEALS had cracking problems earlier. May not have made the Air Force circuit before the CATM folks started seeing problems using US Army 9mm.
@ Daweo;
While it’s perfectly true that all “locking systems” have limits, the P.38 lockup seems to be the weakest and the most dependent on not just good but superb metallurgy.
In descending order of “not poking you in the eye with the slide”, the lineup goes;
1. Maxim toggle-lock (Parabellum)
2. Colt-Browning lock (in all iterations)
3. Steyr rotating barrel lock (M1912, Obregon, etc.)
4. Wedge-lock (Mauser C/96, Nambu)
5. Walther lock (P.38, Beretta m951 and etc.)
The H & K “roller-lock”, as seen on the P9S, the CZ 52, and various rifles is not a “lock” but essentially a delayed-blowback system.
On a personal note, I was amazed to see that it didn’t fail on the CZ 52 when fired with original WP-era Czech-milspec 7.62 x 25mm FMJ, which was loaded to near- U.S. .30 Carbine ballistics for use in the Sa 24 and Sa 26 SMGs.
I have however seen CZ 52s with cracked slides at the roller recesses, not unlike M9s and P.38s.
cheers
eon
@ eon
I maintain that it depends on how you implement it.
Walther lock simply requires a pair of recesses to be cut in the slide for the lock to work. How much metal to put around those recesses is a technical and aesthetical choice. Also the current iteration of the Browning lock has the lock acting on a weak point of the slide (the ejection port).
Mind that, in the 92 series, Beretta weakened the slide in respect to the M1951, infact the 92 recesses are cut all the way through the slide, while the M1951 ones aren’t.
Why they did that? Because they determined the slide was plenty robust that way, and what they needed was a beefier locking block.
And, at that time, they were right. It’s pretty ironic the 92 series to be known for “slide cracking” as, in the ’70s and ’80s, those slides were arguably more durable than those of all the Browning locked competitors. Those slides lasted, at minimum, for over 20.000 rounds of NATO spec ammo and, at that time, a pistol was not supposed to last for over 20.000 rounds.
Today there are Browning designs that are more durable than Beretta ’70s design (or, better, than Beratta’s ’70s modification of their ’50s design), but today a Walther lock as, or more, durable than those can be designed as well.
For a classification of actions, and leaving out pure blowbacks,
1) delayed blowbacks in general (they distribute the recoil force, from the slide to the frame, over an extended period of time, so reducing the peak force).
2) toggle locked actions (once the toggle is correctly sized, there is very little the recoil can do to them).
3) wedge lock (having a bolt that’s not a “U” section hollow piece of metal helps)
4) rotating bolt (see point “1”, but that only works to lessen the impact force of the slide on the frame in their case).
5) tilting short recoil actions with a bolt/slide external to the barrel (Walther, Browning…)
There had been two separate issues.
First the Seals had two cases of cracked slides (one ended with a slight mouth injury) not on the 92F/M9, but on the 92SB they received from Beretta way before the XM9 trials.
It had been determined than one of the two pistols had fired over 30,000 rounds and the other over 10,000 at that point (for a comparison, in the XM9 trials was required a service life on average not inferior to 5,000 rounds), and it couldn’t be determined what kind of rounds had been used (there had been voices of handloads used on at least one of them). Beretta refused to take responsibility for both incidents.
The cracked slides on the Army M9 happened on 12 samples used for endurance tests and, even if the requirement of “a service life on average not inferior to 5,000 rounds” was amply respected, with samples firing over 30,000 rounds without problems, the fact that one sample cracked the slide under 5,000 rounds fired, another under 7,000, and that, if the slide cracked, there was nothing preventing it to hit the shooter, made so that the delivery of M9 pistols to the Army was stopped, waiting for the cause to be found and the remedy implemented. The slide cracked, on the 12 samples tested, from the worst to the best, at: 4,908, 6,007, 7,806, 17,408, 21,264, 21,486, 21,942, 23,310, 24,656, 27,684, 30,083 and 30,545 rounds fired
The Picatinny Arsenal determined the cause was the “tellurium” used in the steel alloy of the slide (and you can still hear people talking of “tellurium contaminated steel”), and the Army required the alloy to be changed. Beretta also modified the hammer’s hub, so that a cracked slide could no more be projected out of the weapon.
However, the Beretta technicians were not persuaded. Mind that they were selling “92” pistols to military and police forces all over the world, and had problems only with the M9 for the US Army.
Analyzing the results of the endurance tests, they noticed two different propellants had been used on the test rounds. The breakages at low number of rounds not only happened only with one of the two, but only with a single batch of rounds made by Federal Cartridge (FC026). If that single batch was not involved, the worst result was the one of the pistol that cracked the slide at 23,310 rounds.
At that point, Beretta tested the rounds and then sued the US Government for the delays, the unnecessary modification required to the guns and so on, producing the results of four independent labs that certified the rounds produced peak pressure in excess of proofloads. The lawsuit was settled with the Government paying Beretta furter $ 10m.
Mind that all ot that happened in 1988, AFTER the 1987 Department of Defense Appropriations Act, where the Congress decided that: “During the current fiscal year [1987], the Department of Defense shall conduct a new competition for 9mm handguns, with procurement starting in fiscal year 1988 in parallel with the current contract.” The XM10 program was due only to S&W being disgruntled for having been kicked out of the XM9 program after having failed the endurance test (infact technically, it didn’t fail the test, because, on average, the pistols lasted over 5,000 rounds fired, but still was, by far, the worst performer of the three pistols that should have been allowed to compete on price).
Well, Kirk, there were some idiots who still toted the M1911 as the greatest pistol invented since the turn of the century. If they couldn’t force the M1911 back into service, they’d get something else like it because they wanted everything to be ALL AMERICAN EVEN IF IT KILLS THE QUARTERMASTER. Yes, I’m sure this is wrong.
Kirk:
Given the chauvinism of American firearms procurement, the choice of a Beretta pistol was unusual. It may have won the reliability tests, but the Ordnance Board can tweak them to give whatever result they desire. They made sure the M14 and M60 were selected after all.
The odd thing is that the Smith & Wesson was a perfectly good pistol. So was the Beretta. In the past, the American pistol would have won automatically. I wonder what happened?
“(…)Ordnance Board(…)”
According to https://www.americanrifleman.org/content/the-beretta-m9-25-years-of-service/
Air Force determined that the Model 92S-1 was superior to all other entries and recommended that it replace the M1911A1.
The M60 did not win any competition before adoption, and the battle rifle trial was tweaked in favour of the M14 by allowing Springfield to use custom prepared rifles for the Arctic test but, once there, it effectively performed better than the FAL. They tweaked the conditions of the test, not the result.
In the XM9 trials, the S&W, in the endurance test, did break a frame before having fired 5000 rounds (the 92F and the H&K P7 were the only pistols that reached the 7000 rounds threshold with all three guns intact), and shown a reliability that was less than 1/4 that of the Beretta (434 vs 1750 MRBF). After the trials were over, the Army simply didn’t want anything else than the M9, yet the Congress still forced them to redo the competition, hoping for a different result.
Dogwalker:
So the Ordnance Board made sure the M60 was adopted by not running a competition (because the MAG would have won), and they made sure the M14 was adopted by, your word, “tweaking” the trial. How does that contradict anything I wrote?
My point was that the Ordnance Board can make sure any American designed weapon is selected. That is why I wondered why the Beretta got through? There is nothing much wrong with the Smith & Wesson, and if the Ordnance Board had wanted it to win, it would have.
The point is that it’s easier to tweak the conditions of the test than the result. If, once the test is performed, a gun performs poorly anyway, then it’s far more difficult to “adjust” the result.
And, to tweak the conditions of the test, you have to know that the conditions have to be tweaked in first place. But the Army, before the decisive 1984 XM9 tests, had the results of the 1981 tests. There the S&W had been the most reliable gun, the third more durable (and the threshold had been lowered from 10.000 rounds to 7.000 in the 1984 tests) and only failed the mud test. None of the guns of the 1981 trials had been deemed to be acceptable but, had there been a winner, it would have been the S&W (the only other gun that failed in only one department was the H&K P7, that failed on reliability). So, once solved their problems with mud (and they did) the guys at S&W could think they were in a optimal position.
What they couldn’t foresee was that, between 1981 and 1984, Beretta and SIG would have improved their reliability so much (S&W doubled its own, but Beretta and SIG improved theirs ten times) and that one of the S&W pistols would have failed the durability test before reaching 5000 rounds, while in 1981 the worst one did break at 9500 rounds. Without that single accident, S&W would have competed on price with the other two.
Dogwalker:
Your comments are true, but they miss the point. The S&W was good enough to be selected, and as you say, the Ordnance Board can tweak the rules to get any result they want. What I find interesting is that they “allowed” an Italian design to win. I may be a cynic, but I do not believe that Beretta’s design won over an American design just because it happened to be better.
JohnK:
They can’t really get the result they want. They can favor a participant, yes, but if that one wastes the help, then it ends there. Had the M14 failed the arctic test, despite the help, the result couldn’t have been fixed, and yet, despite that result, the Army didn’t express a preference. The Marines gave a slight preference to the M14.
The Ordnance Board couldn’t undo the S&W breaking a frame under 5000 rounds fired, nor they knew in advance S&W needed help in that test, since not even S&W knew their gun was going to fail that.
They could have accepted S&W anyway, like they did with SIG after it failed the dry mud test, but it would have been much more difficult to say that the endurance test was “not so important”, and the less than stellar performance in reliability would have remained. To choose the S&W after the results of the reliability and endurance test would have been like plainly stating that the program was rigged.
Dogwalker:
I am quite sure that the Ordnance Board could “tweak” any result it wanted. They wanted the M14, and they got it. If it failed any tests they would have devised new tests which it would not have failed. All of which makes me wonder how the Beretta came to beat the Smith & Wesson? Did Ordnance have a sudden attack of fairness? I suppose there has to be a first time for everything.
The tests are so politicized and nonsensical that just about anything is possible. The rumor at the time was that Beretta got the nod because of the powers-that-were needing the cruise missile basing rights in Italy… Was that true? No damn idea.
Personally, I suspect that there was a lot of “Hey, we need to be seen to be fair to our NATO allies on these things…”, and since everyone was buying F-16s…
From my perspective as an armorer, I think the wrong pistol won. The right pistol was never even in the competition, because it did not fit the criteria. Supporting the freakin’ M9 was a pain in the ass, due mainly to all the damn parts. The damn pistols were really difficult to clean with just a field strip, and the friggin’ officers all thought they were smarter than everyone else in the world, so they’d detail-strip the goddamn things and then lose parts. Their very favorite part to lose was that little spring under right handgrip that pushed the transfer bar up. That goddamn thing…
I’ve really bad memories of one of my senior brigade officers giving his M9 a “proper cleaning” before going north into Iraq back in 2003. He lost that goddamn little spring, and since they’d taken all our Small Arms Repair Parts away from us years before during the Clinton Administration, nobody had spares. I’d used to make sure we always had about five of the damn things for every headquarters company and at least one out in the company arms rooms. Because we’d never gotten filled with the wartime SARP allocation, there I was in the middle of Kuwait with a senior officer whose pistol was NMC… Cue an hour and a half of frantic driving in to find the KBR small arms repair guys I knew down in Doha, and I got him his damn spring so he could shoot someone if he needed to, going north…
The M9 was a decent pistol, but if you’d let me make the choice? We’d have never procured it. Too big for what the caliber was, too many parts, too hard to maintain. If I’d had the choice? I’d have gone with a Glock 19 for mass issue…
Kirk:
The political aspect makes sense. I feel there had to be some sort of quid pro quo for the decision, the Ordnance Board does not choose foreign designs just because they win some steenkin’ contest.
I had not known that the M9 was so fiddly to clean, that is interesting knowledge.
“Fiddly to clean” isn’t really what I meant to convey… The pistols mostly worked well with basic field-stripping and cleaning. Where the problems came in was with the sheer number of parts and the complexity when someone chose to go well past the authorized-for-operator cleaning…
Which was tempting to do, because the damn things were magnets for dust, and the only way to really get them “clean” for mindless military satisfaction was to take them entirely apart… Which the operators weren’t supposed to be doing in the first damn place.
Unless you got yourself into a situation where the pistol got douched in something really fine, got wet, and then dried out…? You’d be fine with just a normal field-strip. Moon dust, water, then getting dried without a visit to a dunk tank of some solvent, or having someone use a can of brake cleaner? All those little parts and crevices would get bound up in what appeared to be concrete. Which rendered all of those fussy little parts immobile, and the pistol generally dysfunctional.
I used to shoot the M9 competitively, and had one for daily carry. My experience was that they didn’t do well without constant deep cleaning and that the one I carried in deep concealment was horribly prone to sweat-induced corrosion problems in all those small parts.
If I remember the numbers right, there are like 72 parts for the Beretta, and something like 35 for the Glock. The majority of the “excess parts” on the Beretta design are small springs, plungers, and other fussy little things. The Glock does not need detail stripping for cleaning, and I’ve seen examples that had unGodly amounts of ammo put through them without either cleaning or maintenance. The first one I shot? Owner bought it as a plinker, couldn’t ever figure out how it came apart, and just fired it and fired it and fired it. His estimate, based on how much cheap 9mm ammo he’d purchased since he got the gun? Somewhere around 10,000 rounds. He just periodically douched it with WD-40 or some other solvent, and kept shooting it. When I borrowed it from him, that goddamn pistol literally had sludge formed from dirt, unburnt powder, and brass shavings oozing out of it when you drew the slide back. It was filthy to a degree that boggled the mind; I spent roughly six hours just getting most of that crap out of it after I borrowed it, and that was using a borrowed dunk tank from my gunsmith friend. While I borrowed it, I shot one of my best IPSC matches ever, and while I was cleaning it, I contemplated the fact that the reason I had to borrow it was that my damn carry Beretta had locked up on me due to condensation-caused internal corrosion. At that point in my life, I apparently had unusually acidic sweat, and the result was that anything steel carried near my person rusted in short order.
There were reasons I abandoned the M9 platform, and they had mostly to do with its size, complexity, and the utter lack of experienced corrosion resistance in my personally-owned examples. I know a lot of people love them, but… I had a lot of grief with mine, and zero grief with my Glocks. Which, I must acknowledge, I had dismissed as “combat tupperware” for a very long time without actually doing anything to really evaluate them. After I did, and realized that they “just worked” for me and my purposes? I was a convert. Still am.
Hi Kirk:
I like the term “mindless military satisfaction”. That has been responsible for a lot of ills.
I had not known the 92 was such a beast. Thanks for the info. I also agree about the Glock. I didn’t like the look of it when I first read about it in the 80s, but I think you are right about it. I think it might have been a bit too late on the scene for the M9 trials, and anyway the US did not plan to base cruise missiles in Austria, so that was a non-starter.
I agree that it is rude to detail strip some guns. I had a Browning Auto 5 which just worked, as they do. Beyond cleaning the barrel I left well alone. I leave detail stripping guns like that to Ian.
I used to use it on the skeet ground at RAF Lakenheath, where the Rod & Gun Club were very welcoming. Happy memories.
The Glock suffered from being so innovative that it broke everyone’s expectations for what a “pistol” ought to be. I know I rejected it out of hand for years simply because “polymer ain’t guns”, and it felt cheap and tacky to me the few times I handled one.
Took desperation and needing to borrow a pistol to run a second stage after my personal Beretta defecated in the bed due to “issues”. After a string with that borrowed Glock, I was a convert.
Personally, given the power to have made the decision, I’d have happily adopted the Glock “system” for the US military. The mentality behind its design is better than the half-ass “Make everyone an expert on the weapon” ideal behind most US pistol procurements, which always winds up getting shortchanged in reality due to budget and time constraints. The US keeps adopting pistols that are really better suited for experts than utter tyros, which was exemplified by the M9.
I mean, seriously… Look at the number of options there are with the safety system and trigger pull. It’s OK if you’ve got the time to train and make the budget sacrifice to get the ammo out to every shooter, but let’s get real: That ain’t happening. I’ve watched so many M9 shooters screw the pooch on the range for qualification that it’s not even funny, and the majority of the problems stem from that damn safety and trigger system. Add in panic, unfamiliarity with the weapon, and all the rest of the complications when you’ve finally put your staff officer or medic into a real-life last-ditch self-defense situation? Yeesh. Some of them might be better off just throwing the entire pistol at their opponent and running for it…
Glock, on the other hand? Properly carried, with a round in the chamber in a safe holster, you draw one and pull the trigger; that’s it. Much less problem with regards to figuring out the safety/trigger, with a continually consistent trigger pull. The single/double-action trigger on the Beretta was a constant issue, especially for people with weak hand strength. Add in fear of the pistol itself, and… Yeah. It just wasn’t a good choice for the role they wanted it for. Something I think they should have tested for, after thinking things through about just what it would be expected to do. Reality is, you’re not putting these things into the hands of “trained experts” that have extensive trigger time and familiarity with handguns; you’re putting them into the hands of people who’re minimally trained, whose background doesn’t include pistols, and expecting them to make what is actually a fairly complex ergonomic system work consistently and reliably under high stress.
The Beretta was not suited for that role, at all. As a pistol for someone like a military policeman, or other “individual with training”, it worked. As a lowest-common-denominator self-defense weapon for the usual run of person we gave them to? It was a nightmare. Too complex, too big, and too hard to train effectively with the limited resources allocated to the task.
Ian: I think you meant to write steel SLIDE and aluminum FRAME.
M
Gun Jesus is only human after all.
This pistol reminds me of a Star Super B with the exception of the magazine and takedown.
Was there any influence or was it just coincidental?
The Star Super B and the XM10 are influenced by John M. Browning’s designs. The takedown is pure JMB, as is the pivoting barrel.
Except that the military had to buy a smaller frame 9mm almost immediately as the Beretta is too bulky for concealed carry for C I D personnel and people with hands too small for the Beretta frame and grips. I read that the Sig P228 was adopted for the secondary roles without the combat conditions involved in the original testing.
“(…)C I D personnel(…)”
What is this?
Criminal Investigation Department.
CID is the Criminal Investigation Division of the US Army. CID usually works in civilian clothes. CID handles serious crimes outside the usual military police jurisdiction, such as rape, human trafficking, fraud, drug sales, and espionage.
Thank you for the very useful explanation of the initials I used in my earlier comments. There are aspects of those personnel involved in Criminal Investigation that I had not been aware of earlier. I learned two things 1) Do not expect everyone to understand sets of initials and 2) how much specialized investigation CID personnel deal with.
The SIG P226 went through the same testing regime as the Beretta M92. The Beretta product won based on the bid numbers, not because it was a better gun or more thoroughly tested. They both met the standards. When the time came for the M11, the P228 was stipulated to have met the standards because it was derived from the P226.
They wound up being issued to people who needed something more easily carried than the gawdawful huge M9, like aviation and some other fields besides just CID. Some of them wound up substituted for the M9 in “normal units”, but that was often a result of someone having made some sort of “drug deal” or outright accident.
The vagaries of the MTOE system are manifold and unfathomable to mere mortals. Swear to God, I’ve no idea at all how they arrived at some of the “basis of issue” decisions they made over the years.
Serving from 1986 to 2019 I used the Beretta a lot and saw it used a lot. I shot the hell out of mine in training and on combat deployments and so did other Soldiers I knew. It is so commonplace to read about how it had slide cracking problems and sand problems with certain magazines and how it is so bulky and the de-cocker is poorly placed and on and on with the whining about how it sucks. Truly I do not know if it was the best pistol selected at the time of the tests but we never saw one fail. My buddies and I were fine with it.
Scott:
Given the information that the Berettas with cracked slides were shooting +P+ that is hardly surprising. But the locking system does mean they need quite wide slides, so are not suitable for concealed carry, which is not what they were bought for.
“(…)contenders were reduced to S&W, Ruger, and the Beretta M9 as a control(…)”
There is photo suggesting there was another one
https://www.thedrive.com/the-war-zone/32663/hot-dog-gun-shows-how-someone-really-felt-about-the-armys-failed-xm10-pistol-competition
🙂