Sunday, November 15, 2026

Jen-sen Huang

WHATTHEHELLHAPPENED?

I became aware of Nvidia some time back--maybe almost thirty years ago, since I had loved the graphics world ever since the HP 1300A in 1967.   Nvidia was a key partner early for the graphics world that was emerging from E & S efforts and Silicon Graphics (where I spent a cool 36 hours in 1984 as VP of Eng'g before I decided that Jim Clark and I were not going to be compatible).   

Remember the Amiga?  

Amiga, private company then sold to Commodore--NOT Amigo, the ill-fated SOS-based HP computer "so friendly that it followed you home"

Precurser to almost all of the interesting gaming and music graphics work we've all come to know and love, Amiga sold an amazing 5 million units over time (10x what I imagined).  Commodore went bankrupt in 1995, not long after Nvidia was launched, with a graphics chip that itself was hardly noticed by the marketplace.  The next Nvidia chips were dynamite, though, essentially establishing GPU's as a valid CPU alternative for certain applications.  The CUDA environment in 2007 expanded Nvidia's GPU's into more business applications, and they began to "be seen" at more than SIGGRAPH conventions.   

The name NVidia derived from NV = Next Version, and "vidia." Latin for 'envy.'    And envy them is what we've done now for more than thirty years.

One of the three founders of Nvidia was Chirs Malachowsky, whom I got to know modestly well when he joined the Computer History Museum board.  He was by then out of the mainstream at Nvidia.  But when I became Chancellor of Cogswell College, we got front and center with Nvidia, its chips and the tools that supported this work.

And such tools they were (and are).   We were using Maya heavily, along with Renderman, V-Ray, Photoshop, Houdini, Nuke, Mudbox, Mari, Z-Brush and After Effects, to name a few.

And then, at SIGGRAPH '50 (8/23/23), Jenny and I were escorted to the front row with Don Greenberg to hear Jen-sen Huang's incredible talk.   Which AstroVirtual made into a Diorama presentation (ask politely).   And he spent two or three minutes with us (mostly with Don) at the end.


So, naturally, since it was a dynamite speech--in fact, Jenny and I said afterward that it was one of the most inspiring  technical talks we'd ever heard, and we've both been around the block on these things.  And, thanks to an earlier ACM interview with Bill Dally, head of Nvidia research, we'd bought a few shares in the $130 range (up nearly 2000% over the next three years, down a bit this month), and even persuaded family to garner a few shares as well.

And then, Huang started to change his stripes.  He fell in with Trump, not unlike some of the other Silicon Valley multi-billionaires who cannot stomach believing in empathy.   I guess I just assumed that he'd be more like Dave and Bill than like Larry Ellison.    And when he got Trump to cave on the chips to China ban, a problem I spent Washington D.C. time on. with both DoD and DoC when I was HP's corporate engineering director, I lost faith in him.

And the loins?    Apparently this is a problem for rich and powerful men of every political stripe.  Let's see: Trump of course, and the whole Epstein band of bazillionaires that Trump is still doggedly protecting, but . . . Bill Gates, Jeff Bezos, Bil Clinton, FDR, Warren Harding (bad enough that his wife poisoned him), and even those paragons of American presidencies, Thomas Jefferson and George Washington.  Fraternizing with the help isn't that unusual in American business (or that of any other country either).  Even Dave Packard was not immune if we believe Mike Malone's account in Bill 'n Dave.   It isn't always so front-and-center, but it hardly is rare.

But active loins, and even screaming at subordinates (does anyone imagine for a minute that Jobs never raised his voice?), are hardly offenses that cause you to resign a job this powerful, with no obvious replacement, for the most important company on the globe, for whom you've been the driving force from the beginning.   With no clear replacement plan.

So it all seemed surreal, that between Sunday's revelations--just six days ago--that in less than 96 hours Huang would resign in ignominious fashion.   Some have said the employees will beseech hime to return, like Eric Altman and OpenAI, when he will then summarily fire the intemperate Board.  

Don't bet on it.   This time I'm thinking that "Little Marco" got it right.  This set of actions was incredibly treasonous, greed-driven and narcissistic behavior the likes of which the world has seldom seen displayed so nakedly.  Aaron Burr would have been proud.

Such a shame.    Aaron Burr would have been proud.





Sunday, November 1, 2026

Egon Loebner and the Soviets

 Strange and wonderful notes come in from time to time.   Here's one from Chuck Morehouse, one of the stalwart HP Labs folk circa 1970-1990.   This one properly belongs in the HP Blog, but I am also posting it here because Egon represents one of the classic intrapreneurs that HP hosted over the years.

Chuck,

I have a question about long-ago HP that you might know something about.  It is about Egon Loebner.  I have just watched a PBS Compass Points report on a new book on the Havana Syndrome.  I have heard many reports on this but never one that mentions a story about Egon that was circulating anound HP Labs around the time I first joined in the early 1980s.

 

The story was that Egon went to Moscow to investigate rumors of a mysterious illness of some members of the US Embassy.  The suspicion was that it might be caused by some sort of radio or microwave radiation bathing the Embassy.  The story was that he was able to confirm with direct measurements that the Russians were bathing the Embassy with microwaves.

 

Do you have any knowledge of this story?  Might it be true but the story was covered up officially but could still linger in HP Labs among Egon's colleagues?

 

Any thoughts you have on this would be appreciated.

 

Cheers,

Chuck Morehouse


My reply wasn't very satsifactory:

Chuck, I have heard that story several times, but all of them a long time ago.   I didn’t know him very well; Doyle and Hammond did, but they’re both gone.  Binnbaum may know about this???

 

There are lots of ’stories’ but hard to pin them down unless it was someone who really knew.   And if you knew, it was hard to mention sometimes.    Example—the Santa Rosa site was unclear-hardened per Packard’s explicit instructions, but Doug Chance (who ran it, and oversaw construction even for a period) didn’t know that until I told him thirty years later.  I snuck that tidbit into The HP Phenomenon (page 489) in an innocuous paragraph 

 

You might also try John Minck, who has a penchant for wonderful details from HP.  I am including him here.

 

There is apparently an autobiography of Egon’s.   This outlines the first seven chapters https://shoah.deportal.cz/shoah/document/10 

 

That’d be fascinating to find and read.   I think his son may still be living. Gary Loenere, but I couldn’t find him in LinedIn

 

Egon’s Ancestry.com data is at https://www.ancestry.com/family-tree/person/tree/161347171/person/112125226517/facts    It notes three offspring, but no names, which says they were still living when the form was filled out

 

Sorry that I cannot be more helpful

Chuck 

The good news, and this is not atypical, ia that someone else 'had the story' -- in this case, HP super-historinan John Minck, who replied back:


Hi Chuck,(to Morehouse, not me)

I got to know Egon fairly well, he was an LED expert with experience at RCA before HP. I wrote him up in my Narrative History of HP 

and will attach a copy. The incident of the Embassy microwaves is true, because Egon told me about it. I've forgotten which administration was in DC but it was when Packard was appointed as Deputy Secretary of Defense. Egon was recruited to be a Technical Attache in the American Embassy, for three years. The staff there was getting serious medical issues, some more than others, but they finally did some measurements and found that the Embassy was being bombarded by high power microwaves. It turns out that speech audio vibrations of the windows would reflect enough power to detect audio. 

The technical folks then built a separate room inside another room, with those acoustic pyramids, and secret meetings were held there. Plus they covered all internal rooms with aluminum foil to shield the place. 

Egon also told me that the Russians had built a brand new building to be used for the American Embassy, and filled it with hidden microphones, in light fixtures, etc. My memory fades now, but some idea persisits that the Embassy might have taken other buildings where they weren't bugged ahead of time. 

Now that I think about it, the foil would have given strong reflections for the audio recovery? 

If you have other specific questions, just ask.

John


So, how is that for getting answers to questions you never knew to ask?


All the best

Chuck (House)

Wednesday, June 24, 2026

Macro-trends?

 A long-time colleague from Informix, Bernard Golden, and I met yesterday for the first time in a year or so.  "Met" over Zoom, as he now is living in Bedford, England, 40 miles outside London.

Bernard has. a long high-tech career stemming from his design team lead at Informix, where he helped define and create the world's fastest UNIX database for its day--the day being circa 1992.  Most recently, he was several years at VM Ware, as a Field CTO.           https://www.linkedin.com/in/bernardgolden/

Bernard was intrigued by my new appointment to the Irish National University at Maynooth, and was hopeful I was now based in Ireland.  Nope, sorry.  But he wanted to talk about the reasons for that gig, which of course as readers know, was the COVID work we did for InnovaScapes over the past few years.   As the board minutes for many companies note, "Much discussion ensued.

Most engineers, you might have noticed, are focused on How Things Work, in. a pragmatic, down-to-earth manner, not given to reflection or wide speculation.  Bernard is not of that stripe.  Indeed, he shared near the end of our hour that his undergraduate degree was in strategic politics, long before the high-tech bug bit him.

And it was in that context that we veered into a "Macro-Trends" discussion, where he succinctly laid out the three 'drivers' across the globe at the moment that he thinks are affecting all of us.  See what you think of these:

1. Maybe a good way to frame this would be “we’re in the midst of a protracted demographic transition" — a changing of the guard from baby boomers who long for an imagined halcyon past and a new generation impatient for solutions to today’s problems and opportunities. This plays out in politics, media, social movements, and even non-profits.  I vividly recall being on the Computer History Museum Board and newcomers clamoring for 'new blood'.  The guise was 'how about software folk, how about younger folk, how about international folk' but one pointedly said one day "We're done with all you old white guys."   I have to admit, I winced at that jab, but on reflection, I realized that they had a valid point.  Bernard didn't pursue this much further, but I did give voice to groups that I've been in where there seemed to be a very real fear about minorities, immigration, and different religions and dogmas 'taking over.'

2. Most of us, certainly Bernard and me, celebrate the incredible technical progress in recent years.  Think simply about lithium batteries.   EVs are now a respectable 13+% of new car registrations in California, AFTER all the subsidies and tax benefits were removed.  And the new Chinese batteries about to become widely available provide longer range and significantly quicker recharging time.   And your cell phone, hearing aids, entertainment systems on airliners, and all sorts of other things (drones?) all owe their capability to these little marvels, way better than the EverReady batteries and the Energizer Bunny of both our youth and recent years.

In medicines, the recent stunning breakthrough with genetic manipulation for pancreatic cancer markers is just one of many incredible advances for formerly intractable diseases.   And diagnostics?  Two second scans of spit enable spotting COVID, or multiple cancers, or diabetes wtih 98% precision?  Treatments?  Leukemia, still a significant killer, has treaments which have improved the 5-year living expectation from 28% to 71% for the most intracctable form, and for childhood victims, to 88%.   

Not all high-tech, though is greeted with enthusiasm.  Many are terrified about AI and the potential loss of jobs, privacy, and even identity.  It is the Luddite reaction, writ large, across a broad landscape.  What will this mean?  Who can escape it?  You all know the laments and lamentations, and yes, we can aver that new jobs, and great new opportunities will arise, but my Gawd, man, this is DISRUPTIVE, and the future is worse than cloudy.

3. Not nearly so "topical" but perhaps one of the biggest longer-term trends is the shrinkage of the total population, coupled with the very different ethnic balances that will result.  Japan's birth rate has fallen dramatically.   The total fertility rae (average chidren per woman) hit an alltiem low of 1.14 in 2025.  This marks a decent of continuous decline (see the graph below), much faster than official predictions.


Digging a bit more deeply, no country in Western Europe except Switzerland is above 1.7 births per woman (it takes roughly 2.2 to sustain a current population, given youngster fatality rates).  The USA, Canada and the UK are both at 1.6.  China at 1.0 is lower even than Japan's 1.1.  South Korea is at an astonishing 0,7, while other big countries--Australia, Argentina, Brazil, and Canada, for example, are all 1.6 or lower.   Even India is at 2.0..    True, a number of countries are still above 2.2; for the most part, they are African nations, or smaller Asiatic nations.   For data country by country, see   https://data.worldbank.org/indicator/SP.DYN.TFRT.IN

Despite continual Unied Nations estimates that the global population will continue to rise (to 10-11 billion by 2100), it doesn't take a math major to figure out that the world population could shrink if these trends continue, from 8.2 billion people to somewhere around 4 to 5 billion by 2100, if African nations dramatically drop their fecundity rates as did the Asian countries when urbanization and lower infact fatalities became normal.  Under the UN projections, and even if these shrinkages occur, a significantly higher percentage of ethnic and religious groups that don't fit the "proper Western Civilization" model will become the norm.

Bernard commented on our recent move to Sacramento, and also his move to England, with respect to the composition of the populations for either area fifty years ago vs. toady--many societal norms then would seem completely alien now to a 1970's version of Alexis de Tocqueville.   Just as today's version will seem antiquated in another couple of decades.

All in all, the three points were well taken--the world we so recently 'knew' is being DISRUPTED at a rapid pace, and for many it is breathtaking, and for some it is distressingly so.   The trick, we agreed, is to be open to chance, to embrace the changes insofar as possible, and to avoid pity parties wishing for 'the good old days'.

Have you been musing about this too?


Tuesday, June 9, 2026

Big Pharma and Drug Discovery

 The previous two posts have described talks at the FiRe 2026 conference, held in the Qualcomm Institute facility at the University of California at San Diego, one of America's premier science research institutes.  Toward the end of the 2nd day, CEO Mark Anderson of Pattern Computer introduced an hour of Pattern talks by various folk, the first of which was Pattern's work with AI in Pharma, for which Mark's opening (04:00-05:45) was important.    https://vimeo.com/1093395045?share=copy

After Mark's opening, Dr. Ben Brown of Pattern gave a synopsis of LLM and AI evolution and accomplishment, along with difficulties that have been hard to overcome to date.  (05:45-12:10).  It then shifted to a segment that covered the Drug discovery efforts and results for Pattern Computer, with a discussion led by Ian Walker, of IMW Pharma Ltd in Cambridge, England.   See segment 12:10 to 17:27 below:   https://vimeo.com/1093395045?share=copy

The significant slide that outlined Walker's talk is shown here:   


Several key points were made.  First, most of the AI search results to date, for nearly all of Pharma, have yet to yield any significant drug combinations or any disease diagnostics.  Why Pattern?  One is tempted to say, "Simple.  It is due to the multi-dimensional nature of the Pattern Engine", but while true, that is a bit hard to appreciate.  But 27 dimensions (vs. the 3-D world we 'know' or the 4-D world with time included) is vastly hard to appreciate of even conjure, but in such a world, big data combinations of molecular structures for diseases, drugs, and organisms is able to be outlined and researched.   And that has been the basis of the Pattern discoveries that led to the drug combinations listed above.  

Walker acknowledges and honors all of this, but then goes on to outline at length what that means for pharma and the world, which must take into account the lengthy trials and production cycles, followed by the awareness and acceptance and distribution of such solutions.  Clearly a convoluted and complex path to success. 

But it all has to start with efficacy, and for that, the scientific papers are a linchpin, based on the science and the trials.  In this context, the recent announcement of a potent aide for pancreatic cancer (results published in the New England Journal of Medicene, May 31, 2026) occasioned a standing ovation by doctors attending the ASCO meeting.

Each of Pattern's 8 drug trials and 3 biomarker developments deal with similarly difficult cancers--the results will likely be transformative.

Wednesday, June 3, 2026

The Vera Rubin Observatory

 The Future of Innovation in Review, Day 1 of the 2026 conference, opened with a stunning and dramatic talk and movie by Mario Juric, a long-time astronomer at the University of Washington   https://astro.washington.edu/people/mario-juric.   Juric, a Croatian, talked with enthusiasm to the Vera Rubin Observatory in Chile https://rubinobservatory.org/.  That observatory has a web presence stating part of its mission: "Located on a mountaintop in Chile, NSF-DOE Vera C. Rubin Observatory will capture the cosmos in exaquisite detail.  Using the largest camera ever built, Rubin will repeatedly scan the sky for 10 years and create an ultra-wide, ultra-high-definition time-lapse record of our universe" 


The history of how this observatory first came to be, and the major challenges of creating it, is a saga that is fascinating, but I will skip that here, other than to observe that the camera itself was defined to record images with a 3.2 giga-pixel charge-coupled device by far the largest camera yet constructed.  

For FiRe, Juric put forth the radical premise not stated prominently on the Rubin website--the images will all become available for anyone anywhere.   Juric's point was that up until now, astronomers had just a handful of locales where key telescopes exist, and then try to gain user time on those telescopes.  As he noted, this put a premium on places like CalTech (my favorite alma mater) and left folk anywhere else bereft, notably anyone growing up in Croatia where Juric was born and raised.

Just how important might this be?   Well, start with the uniqueness of the images found in the first ten hours of operation: "The NSF-DOE Rubin Observatory has officially begun its journey to map the universe, and it started with a remarkable feat: capturing over 4,000 asteroids, including 2,100 brand new discoveries, in just ten hours of test imaging. "  https://rubinobservatory.org/news/11000-new-asteroids

Add in the charismatic personality of Juric as he described his own perspective: "I’m interested in astronomical ‘Big Data’: developing and applying methods and algorithms that let us use large data sets to answer research questions. Major astronomical surveys of today are routinely collecting hundreds of terabytes of images, creating databases with billions of objects and several billion measurements. Large surveys astronomers are becoming part data scientists. In my research, I go where the data takes me — I’ve worked on topics ranging from asteroids in the Solar System, Galactic structure, to the scale structure of the universe. My current focus is using survey data to understand the structure and evolution of the Milky Way. I also lead the Data Management team for the Large Synoptic Survey Telescope, a project to build the largest sky survey ever undertaken."   LSST, the Legacy Survey of Space and Time, is the name of the ten-year 'scan the entire sky' project at the Rubin Observatory.

Juric illustrated his talk for FiRe with animated clips and movie clips from Rubin.  Many of these an be accessed at the Rubin gallery site: https://rubinobservatory.org/gallery

Tuesday, June 2, 2026

Dave Andermann comments

I spent yesterday 'particupating' in Day 1 of this year's FiRe conference, an annual conflab that is often praised as 'the best futures conference' on the globe, according to the Economist.   I've attended on occasion (I've been a speaker, and I'm certainly an enthusiast, but seldom an actual attendee: the fee is usually a stiff $6K per attendee).  https://www.futureinreview.com/

Pattern Computer was 'born' at this conference in 2015, and I've been privileged to be on Pattern's Advisory board since 2018, which this year earned me a Virtual Seat at the conference.   https://www.patterncomputer.com/about-us/ 

This is quite a privilege in its own right--access to leading-edge talks, if not the side-rooms and chat sessions, and in that connection, there were multiple spectacular talks yesterday, and I anticipate that today and tomorrow wil feature more of the same.

One that I was especially taken by yesterday was by Dave Andermann, CEO of Sovera Inc,, a company spun out of Elon Musk's Starlink and SpaceX endeavors.  Andermann has a pedigree that well suits him for work in this arena, having served as General Counsel for Space X for seven years, and for Lucas Film for 16 years prior to that.   Here's his LinkedIn profile   https://www.linkedin.com/in/davidanderman/details/experience/

His talk focused on How to Build Low-Orbit Satellite Systems for sovereign nations, what that means, and how you go about it--including the Perils of Pauline.  He started from the premise and observation that StarLink was the first global network for the world, despite the size and goals of major network corporations over the years.  He described the ongoing battle for sovereign control by every (major) nation, and what it has taken to counteract that.   Fascinating 150 year old 'history' of government control of information flow, and what that means for geopolitical issues.

The Blue Origins launch pad explosion last week came up, in the context of how constrained are the number of launch sites.   He said, contrary to what I knew, that more than a dozen new launch companies exist, and the 'common' expectation is that in a decade, the world will see 10,000 launches per year, somewhat analogous to how many planes fly worldwide daily.   So launch pads and launch vehicles, and the costs (now that, ala Musk, launch vehicles are 'recoverable') are 'easy issues' to surmount.  

That puts his answers to questions at the end into perspective.  The most telling answer to me was his retort about the brouhaha now surrounding the enormous drag on water for communities where these humongous data centers are being constructed for AI centers.   He pointed out that two things are true in space--solar power is 'free' and 'cold' is free.   So the huge cooling requirement for water to cool these centers is obviated by the coldness of space, and the huge energy demand is answerable by solar energy.

Ergo, why not launch satellite data centers?   Easy to beam data to the earth, compared to the savings in energy and cooling needs demanded by a planet-bound data center.  Oh, and costs?   Why not beam infra-red energy at night to all of the 'dark' solar farms, and render them two to three times more efficient?   Fascinating talk!

This has been in 'the press' but not in a clear way.   Here's one phrase, gleaned from a Google search: "Musk's space and AI ventures are aggressively pursuing off-Earth computing thorugh 'orbital data centers'.  SpaceX has filed an FCC application toluanch up to 1 million satelliees dedicated to running AI workloads in orbit, even when his terrestrial AI firm, xAI, is building massive data centers on the ground."   Well, duh--of course, a both-AND strategy.  

But the pushback to such a novel idea is almost predictable.  Forbes for example couldn't wait to predict a dire future, a bankrupt futre, for Musk if he follows such a ridiculous plan.  Kevin Platt, May 31, 2026 in Forbes: "SpaceX Vow to launch 1 million AI Satellites could spark a Doomsday Dive."   https://www.forbes.com/sites/kevinholdenplatt/2026/05/31/spacex-vow-to-loft-1-million-ai-satellites-could-spark-doomsday-dive/

Andermann closed his talk by describing this possibility, and said: He then went on to say, "Never bet against Elon Musk's crazy ideas".   

Other talks were equally stimulating.   I might post some more stories from that conference.

Sunday, May 31, 2026

And Cleve Moler?

Another stalwart gone.  The founder and long-time leader of MatLab, Cleve Moler, just passed.   https://computerhistory.org/blog/in-memoriam-cleve-moler-1939-2026/

 I first met Cleve when he was a sophomore at Caltech--I think Pete Rux introduced us, but I am not sure about that.   Actually, now that I think about it, it probably was my old high school idol, Karl Pool.

Cleve was one of those 'big men on campus' types, rooming at Ricketts House with an even bigger personality, Brad Efron.  Cleve (pronounced and named for his grandfather who was named Cleveland for the Ohio town) oozed personality, charisma, and in addition, he was obsessed with mathematics.  Brad was, too, but made his way into statistics where his fame now resides.

Cleve's goal was to make mathematics 'available' to the lay public, which explains how he came to create MatLab.  Heres the one sentence about MatLab that I find so wonderful:  MATLAB, the language of engineers and scientists, is a programming and numerical computation environment for algorithm development, data analysis, and visualization.  

Now I grant you that not every person walking up and down the sidewalk in Sacramento (or any other town in America) would be apt to describe such a sentence as truy math 'avaiable' to the lay public, but that is how Cleve saw it.

Searching old records, I found a Caltech newspaper from March 5, 1959 which listed the Ricketts House candidates for office in my sophomore year.  And sure enough, Brad was running for president, and Karl was running for Athletics Director.  Cleve was watching Brad tear up the landscape   https://campuspubs.library.caltech.edu/477/1/1959_03_05_60_20.pdf

I caught up with Cleve in January 2024, for an ACM Archive interview about his career.  It was a solid interview, akin to the one that Thomas Haigh had done twenty years prior for SIAM.   https://history.siam.org/oralhistories/moler.htm.   Tom's was the classic historian interview of a techie; mine was more of a Who and Why as seen from a colleague, characteristic of the 175 interviews I've done for ACM.   Each has value, while providing some different insights.   I have an unedited transcript and Video on request, but they seem unable to attach here (cockpit error, early Sunday morning).

The surprise for me was that Cleve was a Salt Lake City kid, and so were a number of classmates, including Roger Noll and our Nobelist, Kip Thorne.   Cleve played high school basketball with Roger; Roger and Kip were both in my class, and we had some sport talking about our various career paths. Remarkably, Cleve did not know Al Despain, also in my class, who went on at Berkeley to pioneer Chess Games with Dave Patterson, and later founding AcornTechnologies (not to be confused with the British Acorn Computers).   

My first wife was born and raised in Salt Lake City, and my long-time Forest Service partner, Irv McQuarrie led me to his cousin who co-founded Beehive Medical Electronics, with Homer Warner.  Alan Kay's first job while attending U of Utah was at Beehive, where he used my HP1300A XY display for his pioneering PC, the FlexMachine.   Ivan Sutherland was still at NSF, five months from joining the U of Utah to help establish their fabled Graphics program.   Somehow (lament, lament), the HP 1300A never gets mentioned as the first interactive dynamic display that was used at U of Utah's program.  Ah, well.   

Cleve 'found' Jack Dongarra, an intern at Argonne, and persuaded him to join Cleve in a new PhD program at the University of New Mexico.   Jack, later a storied ACM Turing winner, gives Cleve full credit and then some for his keen eye and key leadership for Jack's nascent career.

We've lost another good one, at a mere 86 years.  Au revoir, Cleve.