Blog Posts

Here is a series of blog posts related to the project as we were setting it up and doing it, originally published in Jim Downey's Communion of Dreams blog.  We have reproduced them here in chronological order.  They have been slightly edited to leave out extraneous material.  You may click on the title of each one to go to the original post in the Communionblog.

A Little Hint  August 7, 2007

A friend and I got to talking earlier this year about the lack of solid data available on ballistic performance of different common handgun calibers in terms of different barrel lengths and popular ammunition.  If you poke around on any of the forums dedicated to discussion of guns, you’ll quickly discover that a lot of people have opinions on the matter, but there is precious little real solid information out there.  Even the gun manufacturers and ammo makers only have little bits of information, and none of it is uniform or organized in such a way as to allow anyone to compare apples and apples.

Well, from that discussion emerged an idea:  conduct the necessary tests ourselves, compile all the data, then make it freely available to all on a dedicated website.  Sounds like one of those great ideas which no one will ever get around to doing, because of the time and expense involved, right?

Except that I now have sitting here at home a custom designed and built testing platform consisting of a single-shot pistol and 11 different caliber barrels which will mount into the receiver.  Think Thompson/Center Encore pistol in stainless steel with 11 different barrels, all 18" long - each one to be fired using a variety of different ammo, then cut down 1" and fired again, using the same ammo, and that process repeated until you get down to a 1" barrel.  Each firing will be documented using two chronographs, and we’ll standardize conditions as much as possible so we get reliable data.

Oh, I don’t have just the test platform and barrels.  I also have the ammo.  Enough for three rounds of each type at each increment for each caliber.

Now, the shooting and collecting of data still has to be done.  And once all the data is collected this fall, we’ll need to construct a website and put it all online (with pictures documenting the process).  There’s still a lot of work to be done.  But a big first step has been made, and we’re financially committed to this project in a very big way.

I’m excited about this project, and just wanted to share a bit about what I’ve got cooking in my non-writing, non-caregiving, non-bookbinding world.  More bits and pieces about it will probably show up here from time to time, as things progress, and then all of this will be archived on the website for the data when all is said and done.

Another Little Hint  March 6, 2008

In email discussion with a friend the other day, I realized that I had left something dangling which I had meant to address here some time back.  Here’s the relevant email exchange:

I am really glad to hear that you are getting well enuff to get out and blow off some cordite.

Ayup.  That was one of the things that helped keep some shred of sanity for me the last couple of years.  And I am enjoying it even more now that I have a buddy here who shares that interest (and knows more about guns than even I do!)

You may remember that I mentioned a ballistics test that we were planning - here.  Well, everything is now proceeding, and in about ten days we’ll do the first batch of tests.  Three days of actual shooting time - see how far into it we get, and then schedule however many additional sessions we need in the coming weeks.  With over 7,000 rounds of ammo, it’ll take a while to do it all right.  By summer we should have all the data and the website set up - it will be a major source of information for the gun world . . .

In looking back at that post, I see that I had intended to discuss more about the project.  But of course the fall didn’t go the way I planned, mostly because of the increasing demands of caring for Martha Sr.  Oh well.

But now things are in full swing in preparing for the first session of actual testing.  We’ll be doing the shooting on private land, and are deeply into preparing all the logistics necessary (everything from testing equipment to work tables to the chop saw to food & drink) and ironing out our protocols so the data is most valuable.  We’ll start setting everything up next Wednesday, then test for three days, and take everything down on Sunday.  I am expecting that it will take two more such sessions to complete the tests, but I could be wrong - we’ll just have to see how far we get this first time.  It should be an awful lot of fun, as well as a lot of concentrated work - don’t expect me to do a lot of blogging during that time.

Well, anyway, I’m looking forward to doing the testing next week.  Almost a vacation of sorts.

Testing  March 13, 2008

Sorry things have been a bit sparse.  Monday and Tuesday I was busy wrapping up some conservation work for MU Special Collections’ Adopt-A-Book program (and you can see some of my work results on the ‘adopted’ page).  I delivered those Tuesday afternoon, then went and bought a generator.

A generator?  Yup.  And a chronograph, lots of sandwich items, a couple of sawhorses, some plastic drop cloths, some more bore snakes, and various and sundry items.  Because today we’re starting the ballistics testing I have mentioned before.  The last several days have been very busy with running around and getting some of the various items we needed, and then yesterday me and my two buddies with whom I am doing this test went out to the site and started setting things up.  Everything is coming together very nicely, and it is pretty exciting to finally be starting this project we first conceived of over a year ago.

This morning we have a lot more stuff to load up and lug out to the testing site (private land about 20 minutes south of here).  With a little luck, the forecast rain will hold off long enough for us to get set up and settled in (we’re working out of a large cabin tent, which should deal with the weather adequately so long as we’re not facing a deluge).  Once we’re set, rain shouldn’t matter.

So, busy here.  I’ll try and post a bit each morning before going out to do the day’s testing (we’re doing tests through Saturday), but I won’t have a lot of time to do any real research or extensive contemplation.  Not that my usual posts would necessarily make you think I did this anyway, but there you are.

So, later. Whenever.

Did I Mention it went well?  March 14, 2008

From a note I sent a friend last night:

Yes, it rained.  And rained.  And then rained harder.  Surely you can guess how the progression goes?

But we managed quite well.  Thanks to a little advance planning, we had all the materials necessary to cope with the rain (in terms of keeping the wet off of the chronographs, which have to be out between the gun and the target).  And the rest went really well, also.  Today was the steep part of the learning curve, but by the end we were really cranking along.  We should finish up the first three calibers tomorrow (about 1/3 of the total ammo to be tested).  And then get a start on the next batch.  Which we should finish up this weekend.  With a little luck, that’ll be about 2/3 of the data we want to collect.  If so, just one more session in April will be needed to wrap it up, and then the hard work of compiling and constructing the website, all the writing and editing and stuff.

But still, it went *really* well today, and all three of us are happy but tired from the effort.

Did I mention it went well?

It did indeed.  Given that we have something on the order of 7,000 rounds of ammo for testing, if we can wrap it all up with just two testing sessions, that will be remarkable.  Granted, it’s a lot of work, but it is a lot more doable than I had feared.  By the end of the testing yesterday, we were cranking along at about 10 seconds per shot, complete with recording the data, popping out the spent cartridge (the test platform is just one shot, for simplicity and control of mechanical factors), reloading and preparing the next shot.  After each three-shot group of a given type of ammo, we would then run a Bore Snake through the barrel to start with a fresh barrel for the next batch.  Chopping the 1" thick barrels with the abrasive cut-off saw went fine, with only minor glitches, and the process of dressing the end of the barrel was as a result very simple and quick.  We got good data from the chronos, and were already starting to see some interesting results from the testing.  This information is going to be hugely beneficial to the gun community, once we have everything done, sorted, and published online.

Today the weather is supposed to be nice - about sixty and partly sunny.  Tonight storms roll through again, temps drop, and we may have snow mixed with the rain tomorrow morning.  Charming.

Part one done  March 16, 2008

Well, in spite of my optimism in my last post, it took us three full days to get through the first set of ammo/caliber tests, about one-third of the total we’re doing for this extensive research project.

Why?  Well, we kept running into a number of minor problems which would slow us down as we thought through how to resolve them.  Most of these pertained to difficulties with the base platform on which the whole tests are being conducted:  a Thompson/Center Encore pistol.  Don’t get me wrong - the gun is great, and the custom gunsmithing we had done on it to make it more suitable for our testing was fine.  For the most part.  But with any new gun, it takes a little while to get used to it, to learn what quirks and special needs it has.  And in this case, the interchangeable custom 18" barrels also complicated things, once they started getting chopped short.  There were two main problems which cropped up:  difficulty in managing a gun in which the (very) short barrel is essentially hidden by a shroud; and aiming such a beast.

The original design had no sights whatsoever on it - well, when you are aiming an 18" barrel down the range and only need to be able to hit a dinner-plate sized target 15? away, it should be simple, right?  It was.  But as the sight radius dropped lower, it became increasingly difficult to aim the thing.  Particularly since the way the individual barrels mounted into the gun uses a large knurled nut - which would randomly obscure your vision.  Add in the fact that while our cut-off saw was quick and effective for chopping the barrels one inch at a time, it was not a gunsmith-quality tool.  We would ‘dress’ the chopped barrels, to remove slight burrs and whatnot, but even a slight angle on the chop can throw off a bullet.  Basically, after each cut, we would have to learn all over how to point the thing, and hope for the best.

Which led to some hilarious results.  All three of us - with over a century of combined experience in shooting all manner of weapons under a wide variety of conditions - actually managed to clip the vanes of the two chronographs (the bits that stick up like old television antennas).  I got two vanes within about four shots - one on each side!

Our first effort to get around this aiming problem was to install a Weaver rail on the top of the pistol barrel housing (into which the individual barrels mount).  This 4" long U-shaped mount protruded past the front of the barrel, and in theory would give us something to sight down.  In theory.  But just when we would learn how using it related to where the bullets went, the thing would loosen up and shift a bit - and we didn’t have the appropriate loc-tite stuff needed to secure it.

We did finish up testing three full calibers, though:  .380 ACP, 9mm Luger, and .40 S&W.  Altogether, we recorded data for almost 2,000 rounds, including a number of ‘benchmark’ rounds to test the chronographs, as well as a data for several ‘real world’ pistols we own.  And each round had data from two chronographs, specs from the ammo (bullet weight, lot #, manufacturer, brand), ambient temperature, and any additional notes we felt were needed.  A lot of work, not always under the best weather conditions (yesterday it rained all day, and temps were in the mid 30s with a stiff wind).

But we got through it, and already this data would be a priceless contribution to the gun-using world.  And it was a load of fun to do, working with two other people who are more knowledgeable than I am about guns (and I’m no neophyte) and share my joy at coming up with solutions to practical problems.  As my buddy Steve said, there were enough difficulties come up to keep it from ever getting in the least way boring.

Combine the hard work and concentration of conducting the tests for 8-9 hours a day with evenings of good companionship, and I am one tired puppy.  I’m glad that we won’t be conducting the next round of tests until the beginning of April.

But wow - when this is all done, documented, graphed, and put on a website for free access to anyone interested in the result - people are going to go nuts.

Or so I hope.

It was spectacular!  April 4, 2008

Ever see what happens to modern electronics when you shoot them with large caliber handguns?

Wait . . . I’m getting ahead of myself.

* * * * * * *

As noted previously, I’ve been busy the last few days getting things ready to start the next round of ballistics testing.  Round one was about three weeks ago, and in addition to getting a lot of good data about three of the 13 calibers we’re doing the research on, we also learned a great deal about the testing platform and procedures.  Because of that amount of learning, when the three of us got together this week before setting everything up on Wednesday, we decided that we would go ahead and push this weekend to do all the remaining testing.

As a friend said in an email Wednesday night:  “Whoa.”

Yeah, because that is 10 more calibers (eight barrels, since the .38 special and .357 magnum bullets use one barrel, and so does the .44 special and .44 magnum).  And over 5,000 rounds of ammo.

We’re fortunate in that all three of us (me, Jim K and Steve) are in situations where we can set aside our work demands for a time and devote our attention to doing this.  And in looking at the remaining barrel/ammo combinations, it was clear that there would be some efficiency in doing things all at once - a number of the somewhat more unusual calibers have only two or three types of ammo, meaning that we’d be spending more time swapping out the barrels and chopping them than actually shooting and recording data.  So there would be a benefit in getting all those calibers done, then move on to the several calibers where we had a lot of different ammo to test while the other barrels were chopped and prepped (de-burring and modest recrowning to get good consistent results).

And that’s what we did yesterday - dove in head first, in spite of very uncooperative weather (lots of rain and temps in the 40s).  Our set-up keeps us out of the wet (we’re using a fair-sized cabin tent for our work area, with the chronographs outside under a protective tarp) but the damp chill still takes a lot of energy out of you.  The changes we made to the shooting platform - the addition of an inexpensive target laser - meant that you essentially didn’t need to take the time to aim the thing (once we had it zeroed in), all you had to do was control it with the more powerful calibers.

And before we stopped early in the day, we had gotten to the same point with these remaining 10 calibers as we got on the first day of the previous round of testing with just three calibers.

Now, why did we stop early, if things were going so well?

Wasn’t due to the weather.  Not unless you consider a .45 caliber bullet as rain.

What happened was this:  one of us (who shall for now remain nameless, until I can spend more time to write up the saga appropriately) was in the middle of shooting the second most powerful of the calibers we’re testing, and didn’t manage to control the gun completely when he fired the round.  And it went right through both chronographs.  Perfectly.

We use two chronographs, lined up one in front of the other, to be sure we’re getting good data.  He hit the first one right dead center, a little high from the middle.  Like a perfect shot in a movie, hitting the bad guy right between the eyes.  The large bullet punched through the display, destroyed the electronics, and shattered the back of the chrono - then entered the front sensor of the second chrono, exiting out the bottom rear sensor as well.

It was spectacular.  A perfect shot.  I have pix I’ll be posting later.

But it meant we were done for the day.  No chronographs, no way to measure the velocity of the bullets.

But such things are available here, and we’ll pick up a couple more units this morning.  And we’ll be getting the kind which have a remote readout - meaning that it’s just the sensors in the line of fire, the electronics on our shooting bench.  Meaning that we can place some protective armor plates in front of the sensors to prevent this from happening again.

Meaning that we’ll just have to find a new and improved way of screwing up. :)

More when I get the chance.

Damn I'm beat  April 6, 2008

This is going to be fairly brief - I still have a lot to do this morning and only about an hour to do it.

You know how you feel when you first start a new job, with the uncertainty and intense learning curve?  OK, keep that in mind.  Now, you know how you feel after a good workout, doing exercise or a bunch of yardwork?  Add that in.  How about that great feeling of accomplishment, yet exhaustion that comes with finishing a big and difficult project?  Add that, too.  And lastly, if you are an intensely introverted person, used to being alone about 85% of the time, but switch around to being with people constantly for a week?  (Extroverts, I think you’d get a similar feeling by being stranded on a desert island for a week, with no human contact to recharge you.)  That’s how I feel.  All of it together.

In other words:  “Damn, I’m beat.”

The ballistics testing project continues to go very, very well.  With a little luck we should wrap it up today.  We’ve had a variety of minor glitches, each of which has required some good ol’ Yankee Ingenuity to resolve, but nothing major.  The new ‘armour’ protecting the replacement chronographs works - yes, one of us managed to bounce a round right off the top of it (likely would have missed the chronos, but still . . .).  The new chronos themselves are a great improvement over what we had been using.  The laser has had some problems - good thing we ordered in overnight three additional ones on Thursday, as back-ups.  And now that we’ve gotten down to the ‘good parts’ (barrel lengths typical for most defensive handguns), we’re both really excited and really getting pummeled by the effects of the blast of repeated discharges in close quarters.

But we’re coping.  Using the box for the portable generator, combined with additional layers of cardboard, we’ve created a ‘blast shield’ to mitigate the worst of the shock wave.  All the recording process has gotten smoothed out by experience.  The chopping of the barrels continues to be simple.  It’s going really, really well.  We have four more ‘flights’ of ammo to test today (we did six yesterday), going from barrel lengths of 5" down to 2".  When we get to the point where the length of the barrel is less the the housing of the receiver (the pistol), things get tricky, but the aiming laser will help a lot.  Whether we get to doing tests of ‘real world’ guns with the same ammo today is not too critical, since we can do those on another day with a whole lot less equipment (no need for the chop saw, generator, et cetera).  So we should finish the main part of the test today, follow up with the rest later.

Well, gotta run.

Finished. Well, sort of  April 7, 2008

The ballistics testing is finished.  Well, sort of.  As I mentioned yesterday, the primary goal was to get through the barrel-chopping tests, get all the data from that, and to worry about the ‘real world’ guns sometime later.  And we accomplished that.  Early on, as we were cruising along, it looked like we might wrap up really early, perhaps mid-day.

Of course, it couldn’t possibly work out that way.

Just as we were down into the 3" barrel lengths, disaster struck:  the mounting rail for the laser busted loose.  And not just a little bit - the mounting screws stripped out, and there was no way to fix it short of machining in new threads on the receiver.  That’s not the sort of thing you can do out in the field.

Remember I said this, yesterday:

We have four more ‘flights’ of ammo to test today (we did six yesterday), going from barrel lengths of 5" down to 2". When we get to the point where the length of the barrel is less the the housing of the receiver (the pistol), things get tricky, but the aiming laser will help a lot.

Um, oops.

And of course, the thing didn’t fail all at once.  We got to go through that wonderful process of noticing the impending disaster and trying to cope with it, using this or that strategy, each attempt sorta working for a little while before failing.  That took about two hours.  And just when we thought we had it secured, a shot with one of the powerful calibers (I think it was a .45 Colt) sent the thing flying in recoil.

Steve, who probably has the most experience as a shooter of any of us, figured that he might as well see if he could shoot the pistol, sans laser.  And sans any other sort of sight.  Even without the benefit of a barrel, since once we got to that point, the short little bits of barrel were down inside the housing of the Thompson/Center Encore pistol (look down to ‘Number 1842' to get an idea of what just the housing looks like - it’s the part without a barrel or anything attached, off on the left side).

This, needless to say, was going to be a difficult accomplishment, for just a few shots.  Doing it for a couple hundred data points (shots were only considered “good” if we got comparable readings from both chronographs), using always-changing types of ammo and different calibers, almost impossible.

But Steve did it.  We decided that his secret ’super power’ is the ability to shoot amazingly well in really dumb situations.

So, we got it done.  Got all the data we considered critical for the project.  Packed up the stuff we didn’t want stolen, and what needed to go back to Iowa with Jim K, came home.  I’ll get the rest of the set up (tent, tables, et cetera) today.  And Steve and I will go out sometime in the next week or so and do the ‘real world’ testing, using guns on the market with the same ammo as for the test, which will give us some benchmark comparisons to relate to the ‘ideal’ data.

Big project.  The better part of $7,000 worth of ammo, and $5,000 for the custom barrels.  Add in another three or four thousand for incidental stuff (lasers, chronographs, generator, chop saw, and so forth).  Maybe 250 man-hours of labor for the actual testing component.  And we’re still a long ways from being done.  We still need to do all the data entry, design a website, write everything up so that all the information - warts, glitches, and errors - is available freely to anyone who is interested in seeing hard data about what the correlation is between barrel length and ammo performance.  I’m guessing we still have several weeks before we can say that it is done.

But one hell of a big part of it is accomplished.  And that feels really good.

"A website by any other name . . ."  April 8, 2008

One of the major things we still need to do for our ballistics project is to come up with a name for the website/url where all the information and data will be hosted.  It is something we (me, Jim K and Steve) really should have discussed over the course of the long weekend just past, but honestly we were always just too worn out at the end of a long day of testing to be very creative.  And trying to think about it myself right now is problematic, because I am trying to fight a migraine at the same time and only have enough focus for one thing or the other (and sorry, but getting rid of the migraine takes precedence).

So I thought I would throw out the idea here, see if any of my occasional readers would have any thoughts to contribute.  Ideally, the site name/url will be short, easy to remember/type, convey exactly what the project is all about, and available to reserve as a domain.

Here is a list of some of the names we have kicked around previously, to give you some idea where we’re going with this:

    * (et cetera = .etc)
    * ballisticsbytheinch.etc
    * handgunballistics.etc
    * pistolballistics.etc
    * handgunammovsbarreldata.etc
    * pistolammovsbarreldata.etc
    * pistoldata-barrelandammoeffects.etc
    * pistol-barrelvsammodata.etc
    * muzzlemeasure.etc
    * muzzlevelocity.etc
    * muzzledata.etc
    * barrelvelocity.etc

We’d probably prefer to do a .org for the primary domain, if one is available in the name we want, but also buy up the .com and other common variants and have them redirect.  So, that is a consideration.  Of the ones listed above, I like the simpler and shorter ones.


It is done  April 21, 2008

Well, as I mentioned previously, after we did the schedule of ballistic tests using the custom Thompson/Center Encore pistol and had all the “ideal” data relating to barrel length versus bullet speed for a wide variety of ammo and calibers, we still wanted to use the same ammo in a number of “real world” guns - actual handguns from our various collections.  That would give us some head-to-head comparisons to see how they would compare to the “ideal” performance.

Well, yesterday Steve and I had a chance to get out and do this additional testing.  Here’s a message I sent to our third partner in the previous tests:

Thought I would drop you a note, let you know that Steve and I (with another friend tagging along) went out and shot all the “real world” pistols today, using the full run of ammo available.  Lots of good data points on those.  About 6 hours, plus a bit for cleaning up.  I will get copies of the data sheets sent off to you in a day or two.

Mostly, it went smoothly.  The little Berettas in .25 and .32 were a right pain to shoot, and problematic in getting data (we did, but we really had to work for it).  The .380 Walther was OK, the .327 Ruger rough, the big .45 Colt and .44 Mag more pleasant than either of us expected.  We also supplemented with Steyrs in .357 Sig and in .40 S&W, along with the .357 Python, big .357 S&W, .38 Diamondback, .38 S&W 642, and Para Ord .45.  We shot the .357 revolvers with both .38 special and .357 magnums, to have those data points.

Vanes were hit, bullets bounced off the armour plate in front.  Sunburns were earned.  But we got all the data, done done done.  I’ll probably write something up for my blog in the morning, as documentation.  I also took pix today, to go along with the pix from the previous tests.

I heard back from Jim, who said that he knew a number of people were eagerly waiting for the data, and that one fellow in particular who has done a lot of ballistics testing of his own using ballistic gelatin was really looking forward to the comparisons between the “ideal” data and the “real-world” data.  John, he said, expected some real differences but was curious just how much there would be.  My response:

Well, tell him that his expectations will need to be changed.  Here’s some quick head-to-head comparisons:

* .45 ACP (5") - almost no difference, advantage to the Para Ord!

* .40 S&W (4") - marginal difference (less than 50 fps), advantage to the Thompson over Steyr M40

* .357 mag (6") - Significant difference, advantage to the Thompson over .357 S&W (by about 200 fps), more over Python (another 100 fps)

* .38 sp (6") - A little difference, advantage to the Thompson over .357 S&W, more over Python (about 100 fps across the board!)

* .38 sp (4") - Almost no difference, advantage Thompson over Diamondback.

* .38 sp (2") - Significant difference, advantage to S&W 642 - between 100 and 200 fps!

* .357 Sig (4") - almost no difference, advantage to the Thompson over Steyr M357.

I don’t know the barrel length for the rest of your guns, so can’t really say.  Interesting, but not too surprising, that the semi-autos seem to be closer to the Thompson “ideal” than do the revolvers, except with the 642.  Really odd, that.  Oh, wait . . . that could be the difference between the measurement including the chamber and not.  We’ll have to be very careful to note that in the data display, with information about the comparisons.  Hmm.  That would make the revolvers look even worse, since you would effectively be comparing them to a ‘longer’ barrel in the Thompson . . . say between a 3" and 4".  OK, checking that, the data makes more sense.  The 642 falls right there between those, so is fairly comparable, or a little on the underside.  Clear advantage to the semi-autos for power, head-to-head barrel length, then, though with a revolver you get “extra” barrel.


And of course, there are variations between ammos, with some up and some down more than noted.  Once the data is plotted, be interesting to see what the curve comparisons look like.

So, yeah, very interesting!  I do look forward to getting everything entered into the spreadsheet programs and plotted, so that the relationships between one and another are easier to visualize.  But now the testing itself really is done!

Other Resources

BBTI is not the end-all of ballistics testing, just one more component available for the common good.  In addition to extensive discussion about ballistics to be found at many gun forums, here are some other great resources pertaining to ballistics testing you should check out.  (And if you would like to recommend a site to list here, please send an email.)
  • BrassFetcher:  excellent resource, with an emphasis on bullet performance in ballistic gelatin
  • The Box O' Truth:  testing ammo penetration through various barriers
  • Terminal Ballistics Research:  Specializes in the research of cartridge & projectile performance, using hard data gathered from 20 years of hunting game.


We'd like to personally and specifically thank Pat Childs at Fin & Feather in Iowa City, as he not only helped get most of our ammunition and other supplies, he was the brilliant gunsmith who worked with us to make this insane project much more practical.  Without his help all of this would have been much more difficult and perhaps impossible.  Anyone who uses our data owes him a debt of gratitude.

And thanks to our spouses, who were not just tolerant but enthusiastically supportive of this rather nutty project.


This project, and all of its results, is only our fault.  We (well, Jim K, mostly) paid for everything ourselves, and we did not receive any kind of sponsorship or remuneration from anyone.  We did all the work.  We used products we were either familiar with, or because they were what was available, and mentioning them by name does not constitute an endorsement of any kind.  Furthermore, the data is provided purely for entertainment purposes - to better facilitate arguments over what ammo or caliber or gun is "best."  How you use the data is entirely up to you.  And if you think you could do better, feel free to spend the money and do the work and publish your own results.  Or not.  Your choice.