Advertisements

Showing posts with label Interview. Show all posts
Showing posts with label Interview. Show all posts

Aug 6, 2016

Interview with David Liben-Nowell

David Liben-Nowell has appeared in our blog a few times. He kindly invited me and our local constructors to Carleton College when Matt Ginsberg, creator of Dr. Fill, visited Minnesota in 2013. 

Left to Right: George Barany, David Liben-Nowell, Matt Ginsberg, C.C. & Tom Pepper

Today is David's LA Times debut, but his puzzles have appeared in The New York Times, New York Sun & Games magazine. He was also a contributing constructor to the past four Minnesota Crossword Tournament and created this fantastic puzzle in 2015.

I mentioned the seed entries when I blogged Kevin Christian's puzzle last Saturday, you can read from David's interviews that the seed entries are not always the longest & removing one or two black squares can cause a wholesale change in the grid fill.

I often find myself ditching the second or third seed entries in themeless grids as I just can not find clean fill. Did you have to discard any of your original seeds in today's grid?

Yes ... though actually the grid pattern changed at the same time that I ditched one seed entry.  The original version of this puzzle had four vertical 10s as seeds (5-, 10-, 25-, and 26-Down), but what's now the 20-Down square was black (so what's now the horizontal 15 was a 9+5).  I ended up feeling like the grid was a little too closed in, so I tried removing that black square and refilling from scratch, and, luckily, it worked out when I ditched the fourth 10.  (Aside from three of the four seeds [5-, 25-, and 26-Down], literally only three other entries in the entire grid survived when I refilled -- OLD, LOUTS, and S AND P.)

What were the trouble spots in the filling process? Top middle seems to be thorny enough that you had to add a cheater square there. 

The top middle *was* rough, with multiple entries that I'd rather have avoided.  Actually both the NW and SE corners weren't so easy to fill either:  I worked pretty hard to try to get rid of partials and abbreviations, but what you see was the best fill that I found after a lot of searching.

Even with two cheaters, you still only had five 3-letter entries. Is limiting 3-letter answers an important part of your themeless filling philosophy?

Not explicitly.  If I have 3s, I have 3s.  It's probably a bad habit of mine, but I tend to be willing to accept some cruddier shorter entries if it lets me get livelier longer entries -- I'd almost always take a trade of a couple less-than-ideal 3s and 4s in exchange for an interesting multiword phrase as a 7 or 8.  You can probably see that in the N/NW section:  4-D, 5-D, 16-A, and 19-A were enough for me to accept several lesser short entries.

You've made both themed and themeless grids. What are the major differences in your approach to fill?
 

Honestly, not that much.  Themed or themeless, I still start with the longest fill entries, and try to plug in the most interesting entries (preferably multiword phrases) I can find for those entries -- and then hope I can make everything around it work.

What kind of reference tools do you use for crossword construction & cluing?
 

Lots and lots of websites.  I'll typically look at an online dictionary and thesaurus to try to brainstorm punny clues -- I try to avoid looking at databases of existing clues until after I've come up with something myself -- but otherwise I end up doing a lot of nonauthoritative reading (e.g., Wikipedia) to try to get a grasp on the entry.  I think I clue pretty slowly, but for me it's one of the most interesting parts of construction.  I might spend 30+ minutes of online research into something totally unusable as a clue, but at least I learned something.  Just as an example, the clue that I submitted for 13-Across ["King Harald Bluetooth (namesake of the communications protocol), for one"] was related to a fact that I'd heard at pub trivia once ... Rich was totally right to change it, but that was a fun 20 minutes I spent digging into that fact.  [If you're interested:  http://factually.gizmodo.com/bluetooth-is-named-after-a-medieval-king-who-may-have-h-1671450657]

Can you tell us a bit about background? How did you get into crossword solving and later on construction?

Grad school!  Crossword puzzles turned out to be just the distraction that I needed from research when it wasn't exactly going gangbusters.  I started constructing with a grad school apartmentmate, Ryan O'Donnell, while we were sitting around our apartment's living room; it seemed like a natural extension of solving, and we eventually started to get the hang of it and got a couple of co-constructed puzzles published in the NYT.

Besides crosswords, what else do you do for fun?

A little of this and a little of that.  Reading, relaxing and bantering with friends, playing in random sports leagues.  After I moved to Minnesota for a job a few years back, I started curling -- it's a hoot as long as you don't take yourself too seriously while you're playing.

Jun 24, 2016

Interview with Howard Barkin

For those who are not familiar with the annual American Crossword Puzzle Tournament, Howard Barkin, our constructor today, is the reigning champion. 

You can watch the finals here. Forward to the 1:53 mark if you're too impatient. Howard is on the left. The great Dan Feyer, six-time (consecutive) ACPT champion, is in the middle. On the right is David Plotkin, another super solver.

Howard is also wonderful constructor. Besides L.A. Times, he has been published by the New York Times and BuzzFeed

 
Source: ACPT 2016


How long did it take you to solve this puzzle and what's your average time for a LAT Friday?

I don't solve my own puzzles, so no time on this. On average, and depending on whether there are tricks involved, Friday LAT is solved (online) between 2 1/2 and 4 minutes. On paper, between 3 and 5.

What were the other RING candidates you considered but discarded?

Originally I wanted all theme answers to be fully standalone phrases, with GELATIN MOLD as a ring, but that was not quite specific enough. BOXING ARENA was also a proposed option but again, that is not specifically a ring. So this became BOXING VENUE.

Where were the trouble spots in your filling process? It's not often that a 15-letter theme entry starts on Row 4 rather than the traditional Row 3.


I did want to make a customized, original design and challenge myself a bit. The top-right corner really didn't lead me to the kind of fun fill I wanted, so it's a little bit plain. The KOI / GALEN area was also tricker than expected.

Can you tell us  a bit about your background? How did you get into crossword solving and later on construction?

I got into solving as something to pass 30 minutes on a lunch break of a rather unpleasant job I had at the time. I tried different kinds of puzzles, but crossword solving was the most fun for that time limit. The hobby went from there to a more focused interest.

As far as constructing, I dabbled a little bit in it at first, trying to write (terrible) puzzles on paper and in a spreadsheet. I returned to it in 2013, where I slowly learned the finer points of construction and style; my first LA Times was published that year. I try to construct a little nowadays after the kids are in bed and all the dishes and laundry are done ;).
  
What kind of theme and entries interest you the most and what kind do you try to avoid in your grids?

I tend to construct from my comfort zones. My first puzzle centered around baby items in a CRIB, after my first daughter was born. I like fun little wordplay themes, where letters are in a pattern or changed; something light that you can enjoy with tea or coffee, basically.
 
Which part do you enjoy the most in the construction process: theme development, filling or cluing?

For me, the filling is the most fun; trying to find the best combination of familiarity to meet the intended difficulty of the puzzle. I tend to try to make more accessible puzzles, so as many people as possible can feel good about solving it.

 
Do you use Crossword Compiler to make puzzles also? What kind of reference tools do you use for cluing?

Yes, I use a combination of Crossword Compiler and CrossFire, each of which has its benefits to make construction easier. I also use Xwordinfo's clue finder and Onelook.com for matching specific patterns. But these tools only help to find what has already been done, and how often, or whether or not a particular fill or pattern is viable.
There is no substitute for your own brain.

How many puzzles do you solve every day? Do you prefer solving online or do you print out the puzzles and solve leisurely?

These days, between work and family, there are days in which I solve no puzzles, and some days in which I may binge solve 5-10 puzzles to catch up. For that I solve online for the convenience. But for LA Times puzzles, these are syndicated in a local paper, so I prefer to solve on paper, leisurely with a cup of coffee. Speed-solving is an entirely different thing.

Does constructing puzzles improve your solving time? Also, how does solving puzzles influence the way you construct?

I would say that in general no, it does not improve solving time. What constructing does for solving is give a better idea of what the constructor would do, given certain letters. This helps very much when stuck in an area, to figure out possibilities for letters and break open the tough part.

Besides crosswords, what else do you do for fun?
 
Being daddy to my two daughters (4 and 1 at time of writing)  is the biggest thing. I still occasionally play street hockey at night in our local league - it's pretty popular here in New Jersey. We play by ice hockey rules, but without skates on an indoor court. Just played as a goalie last night and I'm pretty sore today - I'm not getting younger, but it seems the other players are ;).

Feb 17, 2016

Interview with Todd Gross

Todd Gross is a friend I wish I could meet someday. Todd has been incredibly kind and supportive of our blog and my construction efforts from the very start.

Today is our third puzzle from Todd, who made the very first Fireball crossword. Todd also has 13 puzzles published by the New York Times.  His works have also appeared in The Wall Street Journal and The Chronicle of Higher Education. In the past few years, Todd has been helping David Steinberg with the Pre-Shortzian Puzzle Project.

Todd lives in Mesquite, NV
I did not do research. I presume you sifted through quite a few *D*D*D* movies and decided to go with three 15-letter only?

I looked at some lists of movies to find ones with three D's in their titles.  There are surprisingly few of them.  I didn't start out looking for 15-letter movie titles, but finding CROCODILE DUNDEE and INDEPENDENCE DAY, both of which I really wanted to use, I tried to find another 15-letter movie and was lucky to find STAND AND DELIVER.

By the way, did you notice the films are in chronological order?
  
Did you have a low-word count in mind when you designed the grid? We don't often see a themeless-style grid with dedicated 51 theme squares. Where were the trouble spots for you in the filling process?

I wasn't intending to make a low word count puzzle, but the way I construct, I usually start with a lower word count grid and add blocks if I can't find a way to fill the grid cleanly.  With THREED on the bottom row, it was natural to have a stack of 8-letter entries on the other side.  But that wasn't working well so I added the pair of black squares to make the first and last across entries 7-letters long.  And I was able to make that work (Rich didn't ask me to revise my grid and didn't make any changes).

The rest of the grid seems pretty normal to me.  There's even several 3-letter entries in the upper left/lower right corners, which I usually try to avoid.  Having three 15-letter theme answers helps keep the word count low.
 
What's your background? Who introduced you to crossword solving and later on construction?

I discovered puzzles when I was young.  I'm not sure how I was introduced, I just remember seeing them on magazine stands and eventually buying them.  But I didn't like crosswords, they were too hard!  I remember being excited when Dell Pencil Puzzles & Word Games first came out, because it didn't have any crosswords!  Alas, I'm not sure what year that was, mid-late 70's sometime.

I don't have much memory of crosswords specifically until 2008, when I posted a dorky little crossword (5x5) on Ken Jennings's blog site.  A fellow named Bill MacDonald (see http://www.j-archive.com/showplayer.php?player_id=1285) saw my puzzle and suggested I try making actual publishable crosswords.  So I decided to give it a try.  As you know, there's a lot to learn before you can make publishable puzzles.  I was lucky Will accepted my 3rd submission (I might have given up if I'd had 8 or 10 rejections without any acceptances).  Rich accepted a submission at about the same time.  And I've been making (not always publishing) crossword puzzles ever since.
  
What kind of theme and entries interest you the most and what kind do you try to avoid in your grids?

As a constructor, I'm trying to think of original theme ideas.  These puzzles are more interesting to make than themes you've seen many times before.  Original ideas by their nature don't fit into categories well, but I will say I like puzzles with meta answers.  I'm trying to learn to create good ones, I have more to learn.  I'm generally not big on wordplay themes, because they've been done a lot and there are several constructors that are better at making them than me.

As for entries I like and ones I try to avoid, I don't think I'm unusual there.  I really like trivia, so I like entries like RAYEWRY that are interesting but not very well known.  I also like entries that can be figured out even if they aren't well known (you may not know what a DUCHY is, but you've heard of a duchess and a monarchy, so the word makes sense for its definition).  I really don't want to offend people, so I try to stay away from entries that might bother somebody.  I'm also not fond of phrases that feel like words just spliced together...but they do come in handy sometimes (like 13 Down). 

Which part do you enjoy the most in the construction process: theme development, filling or cluing?

I've found all three of those enjoyable at times and frustrating at other times.  Theme development is probably the most fun because it's more creative and less mechanical...but when you try and try and you can't think of enough theme entries for a puzzle, the mechanical process of cluing can be a delight by comparison.  Filling is somewhere in between, with choosing a grid pattern being creative and filling the grid with words more mechanical.  I'm really working to improve my main word list, because all too often I find ugly entries keep being suggested by Crossword Compiler, and I can't find a good fill.  A better list would definitely make filling more enjoyable.

What kind of reference tools do you use for crossword construction and cluing?

I look up pretty much everything online.  I like using Google autocomplete to help me learn about words/phrases/titles I otherwise wouldn't have known about...though this is pretty tedious to do.  Wikipedia helps me with facts than can allow for interesting clues, onelook.com helps me find entries for specific letter patterns.  When I first started, I spent a lot of time looking at existing clues so I can come up with something different.  Now I only look up clues if I can't think of one (it happens more often than I want to admit).  Matt Ginsberg's clue database is particularly good, but I've also used XWordInfo and cruciverb.com to look up clues.
  
What is the best puzzle you've constructed? Fireball #1? I love that puzzle.

If you saw all of my rejections, you'd know I'm not a great judge of which of my puzzles are good.  I'm still surprised by which ones get accepted and which don't.  Personally, my favorite is my first accepted puzzle: the Sunday New York Times crossword titled LET'S PLAY BINGO.  It's not just my first acceptance, it's rather memorable with the bingo card in the middle.  Jim Horne had to rewrite his grid display code to handle an image in the center like that.

I'm also amazed I got to co-author the NYT puzzle that commemorated the crossword's centennial.  Even better, it was with David Steinberg, who of course spearheaded the Pre-Shortzian Puzzle Project, which clearly shows his interest in crossword history.  But we submitted our puzzle before the project had started, and before I knew how appropriate it would be to share a byline with him.

You're in good company liking my Fireball puzzle.  Amy wrote a nice review on her blog.  And Peter Gordon is especially choosy, since Fireball only publishes one puzzle per week.

Besides crosswords, what else do you do for fun?
 
Since I've already mentioned David's blog, let me say I'm really interested in crossword puzzle history, especially learning about different constructors.  We are an interesting, creative, diverse bunch, who have developed this unusual skill, only a very few making a living from it.  In the old days, about the only feedback you'd get was from editors you submitted to.  Much of my research has been posted on David's blog, but there will be more in the future.  Including an interview I recently did by phone with a constructor who's almost 90 now.

Other than that, I lead a pretty quiet life.  I'm actually on disability because I don't handle stress well.  So I spend a lot of time at home leading a low-stress lifestyle, much of it in bed.  I realize that doesn't sound fun to most people, but it helps me a lot.  When I do have energy, I like traveling to different places.  I also like learning things, hence my interest in trivia.  And, of course, crosswords.

Jan 24, 2016

Interview with Fred Piscop

I'm always happy to see Fred Piscop's byline. My first cheat-free Sunday puzzle was made by Fred.  

It's hard to make clean Monday/Tuesday puzzles as Rich and other editors want fill to be very solver-friendly. No obscure entries. It's also hard to make Sunday puzzles. The sheer size alone is daunting. Little dupes (not allowed) happen so often. But it's hardest to make Sunday puzzles with Monday/Tuesday smoothness and cleaness. I tried and failed often.  I have yet to make a Sunday without a partial. But Fred accomplishes this on a regular basis.  

Now I've read Fred's answer regarding his filling philosophy and realized the extremely high standards he set for himself. He forgot Roman numerals. You would not find MCII in his grid either. His #1 priority is always his solvers.

Fred has over 100 puzzles published by the New York Times, not counting the Variety diagramless. He has also been published by the LA Times, the Wall Street Journal, the Crossword Club, Newsday & many other  newspapers and magazines. Fred was the crossword editor for the Washington Post from 2002 to 2008. He is a legend!
 


What are the other theme entries you also considered but discarded for today's "Fare Play"?

To tell you the truth, I don't even remember!  I tossed my notes as soon as Rich Norris approved the theme.

I noticed that you seldom go low on word count and your puzzles are always smooth and clean. No weird three-letter abbreviations, not even one partial in this grid. What's your philosophy when it comes to fill a Sunday grid?

As for word count, I almost always use the maximum count allowable by the editor. The higher the word count, the easier it is to avoid lousy entries.  Solvers as a rule are not concerned with word count, but they are very concerned with junk in the diagram. And besides the partials and weird abbreviations you mentioned, I avoid stuff like: foreign words that have no English usage; uncommon variant spellings; brand names that are not national (such as EDY'S ice cream which is not known in California); sports figures that are either not Hall of Famers or who have been beaten to death in crosswords (like ALOU); "crosswordese" (words that appear practically nowhere else in the word but in crossword puzzles (such as ANOA and ESNE) and so on.  If I'm not happy with an entry in a particular region of the diagram, I try to rework that region.

What's your background? And how did you get into crossword construction?

I graduated from Cornell University in 1970 with a degree in Industrial Engineering, which I didn't do a great deal with.  My early working years were spent at a variety of jobs which I referred to not as a career but as a "history of gainful employment."  Around 1980 I purchased a used Apple ][, began teaching myself computers, and eventually became a PC tech support specialist.

I was always a puzzle person; even as a kid, loved anagrams, word games of all sorts; math brainteasers, and so on. Around 1990 I began trying my hand at constructing crosswords, and found that I had a flair for it. It wasn't long before I was selling crosswords to several major outlets, including the New York Times.  In fact, my first NY Times puzzle, published November 22, 1993, was the first NYT daily puzzle under the editorship of Will Shortz, and the first NYT daily puzzle ever to carry a constructor byline.

In 1995 I got laid off from my tech support job at a defense electronics firm on Long Island. I decided that I would never return to the 9-to-5, and just try constructing crosswords for a living.  I have never looked back.

Which part do you enjoy the most in the construction process: theme development, filling or cluing?

All three have their enjoyable aspects: there's great satisfaction to be had in coming up with a cute theme or a clever clue.  But I'd have to say that filling the diagram is most enjoyable for me.  Filling a diagram is, in effect, solving a puzzle.  I've got to make everything fit, and still maintain the diagram standards I listed above. 

What kind of reference tools do you use for crossword construction & cluing?

The de facto standard computer program in the crossword construction business is Crossword Compiler for Windows (CCW).  Practically every constructor uses it.  Ginsberg's Clue Database is a big help, too.  Online I rely on standard references such as Google and Wikipedia as well as onelook.com, which links to about 1,000 online dictionaries.

You used to be the Crossword Editor for the Washington Post Magazine. How did that job influence the way you make crosswords?

I'd say it's the other way around; the way I made crosswords influenced the way I did my job.  Besides adhering to the aforementioned diagram standards, I worked with constructors to make their theme submissions as sharp and clever as possible, and tried to come up with clues that evoked colorful mental images. 

Submissions went through a three-step process:  coming up with the theme, creating the fill, and writing the clues. There was often a fair amount of back-and-forth between me and the constructor regarding the theme and fill.  Incidentally this is exactly the way I work with Rich Norris now.

During my tenure as editor, I tried never to forget that I was once a beginner myself. So, I tried to be as helpful as possible to new constructors, explaining why certain themes didn't work, certain areas of the grid had to be redone, and so forth. 
 
Besides crosswords, what else do you do for fun?

I'm a solver too!  My puzzle of choice these days is the Guardian (a British paper) cryptic.  In fact my first order of business for the day is to stop at Dunkin' Donuts for a coffee, bagel and the Guardian.

Other than that: I'm a keyboard player and attend several jam sessions every week. And if I'm not playing somewhere, I'm listening to live music and sampling the selection of microbrews that are for sale.  I also do all I can to stay in shape, including running, biking, going to the gym and playing senior (age 60+) softball. And you can often see me around the neighborhood walking my niece's dog Dina.

Jan 23, 2016

Interview with George Barany

Many of our blog regulars are familiar with George Barany, who often entertains us with puzzles from his expansive Barany and Friends group. Some of you are friends with George off the blog.

I first met George when Andrea Carla Michaels visited Minnesota in the summer of 2013. His enthusiasm and passion for crosswords are infectious. George's "Breaking the Code" puzzle for the Chronicle of Higher Education is truly ingenious & innovative.

Today marks George's LA Times debut. He has been published by The New York Times & The Wall Street Journal



I imagine you guys completed the middle quad-stack first, then extended to the top and bottom?

It's an honor to be making my Los Angeles Times crossword construction debut in collaboration with one of my cruciverbal heroes, Martin Ashwood-Smith. As many crossword enthusiasts are aware, MAS has pioneered and championed very wide open grids featuring initially intimidating, but ultimately always fair, triple and quadruple-stacked arrangements.

Over the past two years, MAS and I have developed some novel strategies to facilitate the construction of quad-stack puzzles with interesting answers beyond A_LOT_ON_ONE'S_PLATE, RUSSIAN_ROULETTE, SCARLET_TANAGERS, and A_TEENAGER_IN_LOVE,  among others that may have once been cutting edge, but are now greeted with yawns and no small measure of derision. Far be it for me to give away all our tricks, but suffice it to say that the puzzle you are seeing today is the second one to appear in the mainstream media (the other appeared in the New York Times on September 27, 2014), while more are in the queue or have already appeared on my Barany and Friends website.  And yes, we need to discover the central quads first, and then build our grids around them. 

Where were the trouble spots in your construction? 
 
Based on a review of my notes, e-mail correspondence, and computer files, it seems that the heavy lifting on this puzzle occurred over an intense week-long flurry of activity in mid-July of 2014, involving at least a dozen distinct drafts.  An early concern was CERO and ESTO, both short foreign words, adjacent to each other in the grid, but this was settled (see next paragraph) by creative cluing.  A breakthrough was to discover that ALTMANESQUE (not in any database!) could be run through the grid, and finding ALMOST_THERE to balance it, and then recognizing that ALICE_B could hold together an area below the quad [an earlier version was anchored by SPARE_ROOMS balanced by GIVES_A_HOOT, crossing THE_GRATEFUL_DEAD above the quad and ULTIMATE_FRISBEE below it, held together by O_ROMEO].  Also, we looked at multiple versions that did not include grid-spanning entries above and below the quad.  

Once MAS and I agreed on the fill, there was the usual brainstorming and give-and-take on the clues, which took about a week.  Whenever I received a 3- or 4-paragraph e-mail from MAS, in the middle of the night, that started with the words "In the spirit of friendly debate ..." I knew that whatever plans I had for the next several hours would need to be deferred.  Then, once we heard back from Rich Norris, we still had to make some small fixes to the grid to meet his exacting standards.  Specifically, HAS_A_HOME and RENEW, crossing MEWL, as you see in the published puzzle, were originally HAS_A_HOPE and RENEE, crossing PEEL. 

Tell us a bit about yourself. What's your background? And how did you get into crossword construction?

I was born in Hungary into a family of scientists, grew up in New York City, and have been a member of the University of Minnesota-Twin Cities Department of Chemistry faculty since 1980.  More about me personally and professionally, as well as about my family, can be found here (directly, and following further links). 

I always had an affinity for creating puzzles and games, and started to dabble in crossword construction back in the late 1990's through my personal and professional friendship with Charles Deber, one of the all-time greats.   In the mid-2000's, I became a cyberfriend of, and crossword collaborator with, the brilliant Michael Shteyman.  After my children went off to college, there were large gaps in my discretionary time that had previously been taken up attending their concerts, science fairs, and sporting events, so I decided to try to raise my level of commitment to the art of crossword construction, at least in terms of quantity. 

What kind of theme & fill fascinate you and what kind do you try to avoid in your grids?

What you see today is atypical of my work.  I like themes that skew towards my particular interests in science, math, music, sports, and current events.  I also enjoy creating "tribute" puzzles, which rarely make it into the MSM [a notable exception being this Chronicle of Higher Education puzzle, edited by the amazing Patrick Berry, marking a significant centenary]. Very few things in life compare to the thrill of seeing one's name spelled out inside a crossword grid, so I think that I've been able to make any number of family members, friends, colleagues, and casual acquaintances quite happy.

Which part do you enjoy the most in the construction process: theme development, filling or cluing?

I tend to go for theme density at the expense of "squeaky-clean" fill, and often have to be reined in by more level-headed collaborators.  I do enjoy creating themes that are edgy (within reason), quirky, and/or scholarly, and I'm glad when it's possible to find theme entries that interlock.  The best parts of construction are the social aspects of interacting with my crossword friends, and learning from them.

Elaborating just a bit, theme development is fun and demands much in terms of creativity; filling is mostly mechanical but it can be challenging to do well; and cluing is, relatively speaking, the easiest ... I tend to be a fairly good editor/organizer, and by involving my group of friends, some rather high quality clues emerge. 

What kind of reference tools do you use for crossword construction & cluing? 

When I got serious about crossword construction about a decade or more ago, my A-list collaborators handled grid design and filling, and we jury-rigged spreadsheet software like Excel.  Words were introduced manually, based on searches of the invaluable xwordinfo.com and cruciverb.com databases.  For more broad-ranging searches, we used onelook.com.  

About five years ago, I invested in Crossword Compiler (ccw), which certainly helped our productivity, and also ended the sorts of mechanical errors that slowed down earlier work.  One more "must-have" resource for constructors is the free database established and maintained by Matt Ginsberg.  By now, I have a personalized word/clue list consisting only of entries that have already been vetted by my friends group.  Again, these improve productivity, but there is no substitute for human creativity and ingenuity, coupled with an unsparing  commitment to accuracy and high standards, like avoiding duplications and minimizing "crossword-ese." 

Besides crosswords, what else do you do for fun?

Our motto is, we put the fun into dysfunctional.  My work is fun, I love and am devoted to my family, I have wonderful students (including alumni), and great friends, and I partake in the local sports scene and cultural life.  Even non-glamorous events or aggravations like trips to the dentist or doctor, or getting stuck in an elevator, can be the inspiration for new puzzles.