If you have a 2015, 2016, 2017 Subaru Crosstrek and need to pull/replace your window switches – this is a good video on how to do it. You have to unscrew the door panel screw in the inside handle recess and pull the door panel off to access the 2 screws and the wire harness you’ll need to access to remove the trim piece and associated windows controls.
#12 – Obey Gall’s Law: A complex system that works is invariably found to have evolved from a simple system that worked. A complex system designed from scratch never works and cannot be patched up to make it work. You have to start over with a working simple system.
Gall’s Caveat: Working big systems come from working smaller systems – but not every small system can solve/scale to big systems.
Healthcare.gov (the health insurance exchange website linked to the Affordable Care Act) was a disaster by almost all metrics. A report by the Office of Inspector General offers ten key reasons for the disaster, everything from lack of clear leadership, an overly bureaucratic culture, failures of integration, communication, execution, and oversight. The report is thorough, but too vague. Instead, we should have paid attention to Gall’s Law.
In 1975 a pediatrician and systems design theorist John Gall wrote the book Systemantics: How Systems Really Work and How They Fail, in it, he wrote “A complex system that works is invariably found to have evolved from a simple system that worked. A complex system designed from scratch never works and cannot be patched up to make it work. You have to start over with a working simple system.” There’s a tremendous amount of wisdom wrapped up in those lines. Wisdom that explains why lots of great big ambitious projects become huge failures. Healthcare.gov is one, various megaprojects in Dubai, Boston’s Big Dig, Portland’s I-5 bridge, and countless other projects that devolved into massive cost overruns and failure to deliver on major promises.
Instead, extreme programming and the requirement that startups have working prototypes have espoused this simple but functional operating method based on Gall’s observation. In my own work at Intel Labs and other new products, getting a working proof of concept 100% nailed before you move to production assures success is possible while failures happen while things are still cheap.
Which brings us to Jennifer Pahlka’s article where she covers a number of other observations. One of which is Mike Byrne. Byrne built the broadband map for the Federal Communications Commission (FCC) and estimated that most government tech projects could cost 10% of what they do and still provide 85% of the functionality. Apple works this way as well to a certain degree. They do not provide an experience for everyone. They focus the problem, solve it in a particular way, and ignore the critics. They don’t always try to cover all cases nor provide all the features people want – but the cases they do cover are expected, and usually are, rock solid. And that has led them to be the biggest success in tech.
So, instead of designing a system to handle every possible case like was done with Healthcare.gov, it’s ok to leave some cases on the table for manually addressing via phone centers. But the core should work flawlessly.
Gall’s Caveat of Small Systems
There is one caveat Gall observed. This is something I have encountered with younger and recent developers/designers. Some teams/designers rush to a working prototype which is flashy and makes a big splash with some aspects of the key functionality, but quietly and knowingly doesn’t address a number of critical requirements that the design absolutely cannot deal with. They usually hand wave that those will be solved by someone else/some other time – when the reality is that a complete re-write or re-design would be required to handle the requirement. In my experience, that usually happens after they plan on being long gone with the money and awards for the first part. This is probably why we’re see a large number of the Forbes 30 under 30 crowd ending up in jail for fraud.
Gall wrote this caveat: Working big systems come from working smaller systems – but not every small system can scale nor handle all cases it needs to handle. If you need to handle a number of key design requirements, you need to have a working (small) system to solve each of them. Then you combine them.
You must do the design and smaller proven system work for all the things you do need to handle even if you do not deliver those parts today. Do not use simplicity and just building a small core to handle 85% of the cases to avoid the work of ensuring you can handle the remaining 15% of cases. You should always have a design and a plan, even if you don’t need it today (or ever). Maybe that plan is that call centers will handle the remaining 15% – maybe forever. But when things are planned, they can be accounted for, have fixes planned, and not be a huge surprise.
Welcome to dISAppointment! Well, it has been disappointment for anyone wanting to use old retro sound cards or attach 5.25″ floppy drives or MFM/RLL hard drives in modern computers. They all relied on ISA controllers, which have not been supported for over a decade now. But is that about to change?
I recently read about this really clever hacker’s adapter. TheRasteri found he could still access the ISA bus lurking latently in modern computers through the Low Pin Count (LPC) bus which is exposed on the TPM port in many modern motherboards. He created a hardware interface board that connects to the TPM port, exposes an actual ISA slot.
He plugged in a Sound Blaster card via the adapter – and voila! It worked.
More details on his YouTube site, or you can follow the active development thread on Vogons. No word yet if he’s producing any for purchase; but he does want to open source the work and wants to see if he can partner with PCBWay to develop them.
I for one would absolutely buy a few of them.
“That the neural activity of 33 people can predict if millions of others listened to new songs is quite amazing. Nothing close to this accuracy has ever been shown before.”
Have you ever read Isaac Asimov’s short story “Franchise“? The short sci-fi story tells the story that in the future, the United States has converted to an “electronic democracy” where the computer Multivac selects a single person to answer a number of questions. Multivac will then use the answers and other data to determine what the results of an election would be, avoiding the need for an actual election to be held. The story centers around the very average and reluctant Norman Muller of Bloomington, Indiana, who was chosen as the “Voter of the Year” in the 2008 U.S. presidential election and ends with the ironic statement that the US public has “exercised once again their free, untrammeled franchise (voting)”.
Sounds far fetched? Think again.
Researchers in the US have used a comprehensive machine learning technique applied to brain responses and were able to predict hit songs with 97% accuracy. Not only that, but they can do it with as few as testing the song on 33 people.
“By applying machine learning to neurophysiologic data, we could almost perfectly identify hit songs,” said Paul Zak, a professor at Claremont Graduate University and senior author of the study published in Frontiers in Artificial Intelligence. “That the neural activity of 33 people can predict if millions of others listened to new songs is quite amazing. Nothing close to this accuracy has ever been shown before.”
This ability to determine things before they hit markets of scale is being called “Neuroforcasting”.
In the experiment, they equipped participants with off-the-shelf sensors who listened to a set of 24 songs and were asked about their preferences and some demographic data. Researchers used different statistical approaches to assess the predictive accuracy of neurophysiological variables. Linear statistical model identified hit songs at a success rate of 69%. When they applied machine learning to the data they collected, the rate of correctly identified hit songs jumped to 97%. They also applied machine learning to the neural responses to the first minute of the songs. In this case, hits were correctly identified with a success rate of 82%.
This ability to predict hits is called ‘neuroforecasting’ and could even be used instead of recommender systems like you see in Netflix and Pandora.
“If in the future wearable neuroscience technologies, like the ones we used for this study, become commonplace, the right entertainment could be sent to audiences based on their neurophysiology. Instead of being offered hundreds of choices, they might be given just two or three, making it easier and faster for them to choose music that they will enjoy,”
This is one of the best talks on leading and setting expectations on a creative, collaborative production group that I have ever run across. It was given during GDC 2022 by a director at Obsidian Entertainment who clearly came with years of likely painful learning. I loved this quote:
We must bust the myths that equate ‘passion’ with overwork and aggressive ownership, and we must normalize collaborative, sustainable habits.
As she described the work situations and the martyr and maverick worker types, I found myself finally hearing in words many things I experienced but didn’t know how to express. I have personally seen the destruction of many extremely talented teams by just a few divisive elements. They could even be rockstars, but they end up cratering the whole production and driving people away from those teams. Carrie Patel does an amazing job describing how even well intentioned mavericks/martyrs damage the team. This is true to the point that hiring less technically good people that are good teamworkers produces consistently better results than extremely talented people that can’t work with others.
Are mavericks and martyrs inherently bad? No, but they absolutely can be if they are not managed properly. That requires being aware of how their behaviors affect the long-term success of the team.
As it turns out, this is not new learning. Other organizations have learned these lessons long ago.
Special forces teams as well as Mark Cuban point out these same learnings for business and sports teams. People who work more collaboratively and think of the entire team’s impact when making individual choices produce far more productive teams than rockstars/martyrs do. Special forces teams often do not look for the absolute best person – they look for who works in the team. This doesn’t mean you can’t have rockstars or martyr types, but what you want to find and reward is behaviors and an environment that encourages sustainable and truly collaborative teamwork.
The original Tron movie came out in 1982 – and was a huge reason I became a computer programmer and got into real-time graphics (thanks Sid Mead). Before the Tron: Legacy movie came out in 2010, Monolith game studio came out with the game Tron 2.0. I still have my big box version of this game, and right now it’s only $2.09 and the Steam version appears to work on Steamdeck just fine.
I remember many good hours with this game. It was a really good spiritual successor to the original movie, had a great soundtrack, and even had some interesting early graphics effects like real-time glow.
Links:
Covid has changed things. Even in 2023, international travel still muted and people have been doing more domestic travel. That means open US attractions are somewhat overrun. There’s been a lot of bad press that Disney theme parks are terribly overcrowded, dropping the annual pass and switching to a much maligned Genie Pass system (dropping a functional but much mutated Fastpass) and have gigantic wait times (with averages for some rides over 130 minutes).
All of this has caused lawsuits and the ousting of previous Disney president Bob Chapek who is largely seen as badly handling most of these problems. It appears that the new/returning Disney president Bob Iger is restoring annual passes and seen as righting the struggling Disney ship.
Now what about avoiding those 2 hour wait lines? Filming is still allowed in the parks and Youtube channels like Attractions 360, DocumentDisney, WDW and others have created amazing 4k ride videos of just about every ride at every single Disney park – in the world. Many even have multiple copies that include seasonal theming. So now only can visit and ride your favorite ride whenever you want – and each of the other versions in other countries. Though some of those other countries have modified the rides in interesting cultural ways.
Here’s a good example. The Haunted Mansion (a personal favorite) is in 4 different locations: California Disneyland, Florida Disneyworld, Tokyo Disneyland, and Paris Disneyland. I prefer these because there is no commentary or cheesy music. You hear everything including everyone is saying – which is just about the same experience as being there.
Not that one? How about the one in Florida.
Not that? How about the Tokyo Disney Haunted Mansion.
Oh, not that one? Then How about the Paris Haunted Mansion?
While world class buildings aren’t being built in the US at the rate they used to be, they are still being built in Dubai, Saudi Arabia, and China. Buildings such as Atlantis the Royal.
How are they opening these creations? How about paying for high class stars like Beyoncé and Swedish House Mafia to perform in customized shows. They also invite various movie, music, and other stars to be seen there in red carpet like events. They also bring in high-class influencers as well as having outrageous fireworks displays. In short, they create world class see-and-be-seen events like used to happen in New York and other cities.
See a tour of the final project here:
The 90’s Demoscene subculture was famous for building incredible visual demos in astoundingly small executables sizes. Many demo scene gatherings had maximum size requirements – often in just a few hundred, or even a few DOZEN, kilobytes. Figuring out how to get the most amazing tech in the smallest size was one of the great innovation points for these contests.
Once developers had exhausted their technical chops on generating amazing art with miniscule code (using all the tricks in the book they could think of), they quickly found that hand-tuned compression became far too tedious and brittle. So, they started building tools to do this compression for them.
I wrote about a more modern take on this where MattKC tried to fit an entire game into a QR code. Part of his adventure was compressing the executable using an old demoscene tool called Crinkler.
There were others, one of which was called lekktor which was used first on their .kkrieger demo. The story behind the development is a fun read as is an interview he did in 2005.
Apparently it used a form of code coverage as part of the analysis which ran while you ran the application. This had the dubious effect of allowing people to use the down arrow down on menus but not the up arrow – because nobody ever pressed the up arrow when training the compressor.
Links:
Having trouble understanding all the knobs in stable diffusion’s webui interface? This is a great website on github that shows what each feature does and gives some examples and tips on getting the most out of the features.
It covers all the big features like outpainting, inpainting, prompt matrix, AI upscaling, attention, loopback, X/Y plot, textual inversion, resizing, sampling method selection, seed resize, variations, and a whole host of all the other options along with before/after pictures to help you understand the features better.
Links: