Bevan and Guy founded that site and realized that the UX was perfect for sharing coupons and launched RetailMeNot.com from it, leveraging BMN’s crazy high PageRank to propel RMN to the top of Google. Not to mention the ease of use and social aspects way ahead of its time.
They were able to join a handful of retailer affiliate programs and were probably clearing tens of millions a year in affiliate commissions with no overhead. Then they sold it to WhaleShark Media (funded by Google Ventures among others) and IPO’d it in 2016, cashing in a second time.
Honestly one of the most impressive bootstrap stories you’ll ever find.
No big surprises here. Big outages at big firms seem to always look like this.
The electricity industry has this notion they call a black start, where they bring the grid online from zero. It's a very difficult operation because you can't start a generator straight into an offline grid otherwise it gets hit with full demand and can't keep frequency in the right range, which can damage equipment. So they have to start by disconnecting all the loads and then bringing load blocks back in sequence, trying to roughly balance available generation with the size of the next block. In practice black starts are extremely rare, but they can happen after a cascading failure in which something causes some generating stations to trip, redirecting load onto the remaining stations which can't handle it causing them to trip in turn. The entire grid can collapse in just 60 seconds or so once that kind of runaway failure starts.
This kind of outage is very analogous to a black start. A few servers get into trouble - maybe they're in a crash loop or maybe their caches are cold - the load balancers shift traffic onto the remainder which then can't handle the load either and so the failure cascades. Once a service is in this state sometimes the only remedy is to disconnect all load, bring the servers back to stability, then bring load blocks back online in small chunks, making sure to never hit anything with the full load immediately to ensure any caches or JIT compilers have time to warm up.
It's the sort of thing that requires a bit of preparation because naive ways of doing it like just randomly dropping 90% of requests will yield a broken service for everyone. Ideally your LB can drop traffic by modulus of the session cookie or similar, so users are let back in atomically.
Black starting a service is obviously quite hard to practice, but is probably something that sufficiently well resourced SRE teams should try once a quarter or so on a test cluster. I used to be an SRE at Google and although we regularly started up new clusters and moved traffic around between them, actually starting an isolated service instance entirely from scratch was never something we practiced. I remember in a few cases teams discovered devs had introduced circular dependencies by accident, i.e. starting service X required service X to already exist, or it depended on service Y that in turn depended on X. Sometimes you can only find these things by practicing.
From a former call-center phone operator, I think the 10 1-sec pushes, is just a way to make sure the customer presses the buttons long enough. In our callcenter we often had to ask customer to hold the reset button on their router for 3 seconds. When the router didn't reboot, I would send replacements, until a colleague advised me to tell customers to hold the button for 10 seconds. Apparently when you ask customers to hold the button for 3 seconds, some will actually only hold it for 1 or 2 seconds. Suddenly most customers were able to reset their routers.
I personally operate a fairly advanced closed-loop chlorination system, and this is beyond screwed up.
These systems can easily fail, and they tend to fail in a way that causes them to measure low or zero chlorine and thus overdispense chlorine. More below on why.
First, this means that an operator must always ensure safety, both against excessive chlorine and against insufficient chlorine. In a commercial context, that means that a real human (or maybe a totally different sensor design, e.g. a colorimetric sensor) need to measure the pool chlorine level, preferably with a high quality titration (drop test) every day before opening the pool. They should probably record the result in pen on water resistant paper, but this part is icing on the cake.
Second, hoping that an email gets noticed is ridiculous. If you do not have positive confirmation that the pool is safe to open, don’t open it.
Third, the systems should be sized so that they cannot overdose quickly. The maximum dispensing rate should be such that hours of continuous operation would be required to reach an uncomfortable level, let alone a dangerous level. (Hot tubs are sort of an exception here.)
Fourth, everyone should be using a reasonable amount of CYA in their pool. Even at a lowish level of 30ppm CYA it takes a lot of chlorine to become dangerous unless the operator has also lost control of pH. I realize that some codes are confused about this.
Fifth, chlorine is smelly. Someone should have noticed, even just casually. Admittedly, lots of swimmers may be conditioned to stinky pools, and checking for safety isn’t their job. But the staff should notice.
Sixth, the systems I’ve seen beep when there’s an alert. It’s easy to add flashing lights, too. (In fact, the system I’m familiar with automatically shuts down if it feeds for more than X minutes without reaching the setpoint, and you cannot turn this feature off. You can, however, defeat it by switching the power to the whole controller.)
—- sensors —-
Actual free chlorine sensors intended for pools use an amperometric probe immersed in iodide solution or gel, separated from the pool water by a fancy membrane. Free chlorine oxidizes iodide to iodine, and the sensor measures the iodine concentration. If the membrane fails (or general wear and tear or excessive pressure swings) cause the iodide to be lost, then the chlorine can’t react and the sensor reads zero (or low if there’s tiny bit of iodide left).
Interestingly, these sensors are rather expensive devices that appear to mostly come from France and Germany. The Chinese vendors seem to be rebranding European parts (!), and the major US vendor is very clearly reselling a French company’s product.
There are some vendors who sell a cheap (or absurdly marked up) potentiometric ORP-based closed-loop system. In my considered opinion, these are worse than useless.
You don't need a fancy glass. Your sense of smell makes a huge difference when it comes to food and drink, as anybody with a nasty cold can tell you. You want something that keeps it from diffusing into the air so anything with a wide mouth makes it harder to appreciate.
So: a cognac snifter; sherry glass; a glencairn, which was ""made for whisky""; etc.
But of course it's down to taste.
If the aroma is too heady? Blow air into glass and then inhale the aroma. The burn is gone (important if you're drinking 62% rocket fuel single cask whisky) and you get a more dilute smell.
No need for all this pretension; this glass is that.
I've done biological/ecological surveys (albeit never with fish) and the act of "processing" an animal means identifying it to species, weighing and measuring it, identifying its relative age, sex, and reproductive status, examining for signs of specific disease or injury, etc.
This is done as quickly and with the least impact possible to the animal, though it unavoidably stresses the animal. Infrequently a bird or small mammal will die of a heart attack while being handled (it's predator vs prey from their perspective, or an alien abduction experience at best), so if an animal shows any signs of major physical stress they're released from processing immediately.
The teams performing these surveys will handle hundreds of animals in a season and thousands over their career, and the target animals are often those listed as Threatened or Endangered, where the death of one animal is a Big Deal (from a regulatory perspective, if not from an ecological perspective).
Hi all. Thanks for the interest. I'm going to post an impromptu blog post here just explaining what this is. The rather unfortunate ground stop today made this unexpectedly useful with the pilots here on Orcas just to get a sense of the chatter in our airspace re the situation.
This is an ETTUS B210 attached to a dual 4090 box. It's a polyphase channelizer that takes the entire airband, channelizes it using cuda code on GPU, demodulates it, detects transmissions, throws away data tx and then feeds the tx through a voice model I created for transcription. A separate LLM reads the transcriptions to create a situational awareness picture of the airspace.
I'm interested in open sourcing this and making it available for folks who can run an ETTUS B210 or high quality SDR with enoguh bandwidth and a local GPU machine, so that we can cover more airpsace.
This was born out of an incident we had not too long ago where a pilot around KORS (based at Friday Harbor) lost the engine in his 172, and put it in the water in front of a ferry and was picked up by the ferry. We're a pretty close bunch on Orcas Island and we only heard about it days later via social media. That felt completely unacceptable, and I set out to create a radio that could monitor everything in real-time, all the time, and alert on emergencies. That has extended to transcribing and creating situational awareness for a chunk of airspace.
I'm an IFR pilot. I fly a T206H regularly in IMC around here and I'm a ground instructor. You can email me at mark@defiant.com or find me on x at mmaunder if you'd like to connect. I'm particularly interested in hearing from folks who can run a station on their own hardware and are under bravo or a charlie or even just somewhere interesting re airspace.
Yeah — I don't really see how it can work except on the bass strings (played open).
This is a novel but slightly peculiar take on this problem. But there are lots of other things, right?
Listing for others since I suspect you know: all of this scheming about live-manipulating the basic tuning of an electric guitar dates back to the Parsons-White stringbender (and arguably further, since that was a way to get pedal steel sounds), which has led to stuff like the Duesenberg Multibender, the really extraordinary MayTrem (which keeps all strings in relative tune as the pitch is bent), the amazing Hipshot Doubleshot which lets you tune a guitar or dobro in two separate open tunings and switch between them instantaneously with a lever, etc.
Ed Gerhard added single-string frets to the bass side of his weissenborn (acoustic lap steel) that he can play with the fingers behind his slide or (incredibly) with his thumb.
All of these things are sort of fiddling at the edge of the question of a musical instrument, which is: should you confound their constraints or work within them? It's a bit like the thing about whether a dulcimer should be diatonic, whether you should add just the frets you need, whether it should be chromatic, whether it should have removable frets, etc. etc.
As a sort of hobbyist lap steel player I find all these things fascinating, but I think the ultimate question is — where do you draw the line? There are lots of things I can't play on a six-string lap steel, but should I change the instrument so I can, or should I accept the role the instrument plays, and play the things I can?
Another way to look at it is: why is the Harpejji niche? Why isn't everyone playing the Linnstrument? Why don't all pianists play the ROLI Seaboard?
Ex-MSFT here, sharing my outdated POV based on working on a product that was using Microsoft Exchange and therefore was hooked onto the email delivery pipeline.
Note that the infra for M365, Azure etc is different. They would be running on servers hosted in MSFT DCs but have completely different CI/CD pipelines. M365 services have their own deployment pipelines and it takes on the order of weeks to deploy your changes to production.
Considering the fact that Outlook is down, clearly email doesn’t seem to be working. So email uses something called Microsoft Exchange Server, there used to be an on-premises version of this but MS stopped supporting it to run the cloud version hereby called M365. Email is one big “macro-service”, if you’re running Windows Server, look at the process called “EdgeTransport.exe” this is responsible for sending/receiving email on Windows Servers. There are probably multiple 100s of services supported by 100s of teams that are run while email is routed from sender to receiver via MSFT Exchange servers. My recollection is that even if one of these services goes down or introduces latency in the pipeline, it caused a major incident. When I was working in M365 Security in 2024, this Exchange Server pipeline was called “legacy” because it’s one big .NET Framework service. I don’t know if they upgraded to .NET core but developing on this was as HN comments say, very janky.
The reason Teams doesn’t seem to be down is that Teams services are completely different and are written using .NET Core and have their own separate release pipeline. It was much easier to interact with Teams services. I actually worked closely with the team that was responsible for ingesting Teams messages to the customer’s data store. Their code base was much more modern and easy to understand. Exchange Server not so much.
I don’t know if what I said is accurate but know that Exchange server code was like a fossil that didn’t change as much, over the ~5 years I worked on it from 2019-2024.
According to algolia, there were 57 posts with jane street somewhere in the title or url in the last year, so a bit more than one per week. For a Poisson distribution with rate = 57/52, P(X >= 4) ~= 2.5%. So there should on average be ~5 weeks with 4 or more jane street related posts at that rate.
I created the first online lidar map of the UK (well, England and Wales), as announced here many years ago! (now at https://lidarmap.uk) and delightfully there have been dozens of finds of previously unknown sites of archaeological significance found using it that I'm aware of. Roman villas, forts, henges, all sorts of earthworks. New discoveries are being reported using this and similar tools almost weekly, on forums like https://www.reddit.com/r/LiDARdiscoveryUK/. Britain really is littered with this stuff.
Although it does feel the finds are getting increasingly hard to spot, every now and then you get a how did everyone miss that moment. Perhaps not surprisingly, most of the remaining impressive finds tend to be in woodland, using the Digital Terrain Model data which can peer beneath the forest canopy, like this Mexican one today, or are within newly released coverage which is promptly pounced on by armchair archeologists.
Not sure what my point is here except... to me this story still feels like a function of the amazing recent progress in lidar coverage and data availability (not to mention the incredible technology behind it), rather than an AI one, although I don't doubt that pointing a capable model and a healthy token budget in this direction might well be able to sweep up a few of the remaining discoveries.
This has completely changed my industry (hard rock mining). A lot of the time drillers would be in the middle of nowhere looking for resources and would be sometimes days away from providing results (which also lead to lots of dodgy situations where drillers would know some results before market…) now it’s instant live feed. It’s made revising drilling plans trivial. Very cool
The author, Felix Rosenthal, was a family friend. He had an incredible life, like something out of "The Third Man":
Born on January 12, 1917 in Munich, Felix emigrated from Germany in 1933, shortly after Hitler came to power, and settled with part of his family in his mother's native Florence; after obtaining his Italian high-school degree ("licenza liceale") he went on to study engineering at the Istituto Politecnico in Milan, but when the Fascists' anti-semitic laws of 1938 prohibited Jews from enrolling in public schools and universities, he settled in Paris, again with part of his family. At the outbreak of WWII in September 1939, Felix was in Switzerland and a return to France was impossible; in 1940 he emigrated to Chile before finally settling in Berkeley to join his brother Bernard in 1941, and enrolling in the engineering school of the University of California. From 1943 to 1946 he served in the US Army and became one of "The Ritchie Boys" - GIs of German origin who spoke German natively and were trained in Military Intelligence at Camp Ritchie, Maryland. He was discharged with the rank of Master Sergeant and returned to the University of California to study architecture; after having obtained his MA he became assistant to Eric Mendelsohn, the renowned architect who had recently joined Berkeley's Architecture Department.
We had the original book when I was a kid. My wife gave me a copy (off of AbeBooks or somewhere) a few years ago and I was a bit overcome with emotion at seeing it again.
Specialty cut flower grower here. Backyard scale, but we're profitable and have received some grants so that means we have had a USDA relationship for a couple years. The rep I've talked to and am closest with is skeptical of agrivoltaics due to past dealings with solar companies (not sure what she means). There doesn't seem to be enough reliable public data to make it seem like the win-win-win we want it to be. In their mind, you're adding a non-ag industry to the land and that could have downstream consequences in a generation. Maybe you're scheming to get the land re-zoned for light industrial; I honestly have no idea what they're thinking, but they're averse toward it.
Most flowers in 7b don't want extreme sun. We use shade cloth frequently, so solar coverage would be fine and of course have the energy co-production. Cut flowers aren't harvested mechanically at our scale, so we wouldn't have to worry about tractor access.
Meanwhile, more and more food is produced under cover for a lot of good reasons. Polytunnels, glass houses. You might get some summer shade from solar but what about season extension at the beginning of spring/end of fall? Depending on your crop and location, that could mean 4-8 extra weeks a year. What if you do geothermal energy generation for the tunnel? Now you can grow through the winter and not need the same footprint as solar. That's more appealing to me.
Anyway, I'm forever mad at corn ethanol and we should at least replace all that with solar farms. That's probably the right place to start. But also, again, zoning...
I have a story to share. In 1978 as a preteen, my mother made me take piano lessons. God bless her for trying. Anyway, while waiting for my class in the lobby of the music school, I would often make paper airplanes. One day, an older gentleman came over and said that he wanted to gift me something. What he gave me was his copy of "The great international paper airplane book". And remains one of my cherished books or really possessions. But I'm getting older now and feel that the time has come for me to find a 13-year-old worthy of this gift. But it's quite a leap of faith, for me anyway, to convey such a gift.
OK, I think I see what’s going on with this, in this configuration.
The goal from an attacker’s perspective is to get this to be true: `$ret = crypt($pw, $hash); $ret == $hash` when called by OpenSSH’s sshd. This is how sshd validates that the user logging in knew the account password.
The `$hash` is the “settings” to the crypt call and will normally look like “$6$--gibberish” for users with passwords set, but could look like “*” or “!” for users not intended to be able to login.
The bug is from the intersection of all of these things:
1. musl can return “*” for errors, rather than leave the return value NULL. Other libc have adopted something like this too, but seem to use a different result “*0”.
2. musl can also return “+” for errors, but only does so if $hash starts with “*”. This is intended to eliminate the possibility of an error causing `$ret == $hash` to be true, however it assumes that applications calling `crypt()` will set $pw and $hash to values corresponding to the `$ret == $hash` comparison to be made immediately afterwards.
3. OpenSSH can be convinced to pass the hash of a different user if the hash is “*” or “!” to avoid timing attacks that can enumerate the user list, so it will set $pw and $hash to correspond to different users. $hash will be a real “$6$--gibberish” style hash in this case, and OpenSSH expects `crypt()` to either return NULL or return a hash string. But now there’s two hashes OpenSSH uses. It does: `$ret = crypt($pw, $hash2); $ret == $hash1`;
4. This should normally not be a problem because OpenSSH will pass in valid $pw and $hash2 that should not lead to errors in hashing, so the $ret that musl generates should be a real hash that cannot match the “*” or “!” $hash1 for the login-disabled user. However as a DDoS mitigation, musl will return error if the password is too long when using SHA-2-based hashes (the default, it seems).
5. This error will result in a non-NULL result of “*” or “+”. Since OpenSSH passed in a real hash ($hash2), musl will return “*”, which compares equal to the actual user hash $hash1 for the disabled-login user.
The OpenSSL adjustment ensures a minimum hash length for the user before logging them in, even if crypt() succeeds somehow.
Hi! Co-founder here. Conductor actually kicked off with a HN post many lifetimes ago in July 2025 (https://news.ycombinator.com/item?id=44594584). I owe so much to this community. Thank you all so much!
The whole team is around to help answer any questions. You can reach out to me personally at charlie [at] melty.sh.
> there's no way to say it nicely, I have a big nose.
But there is!
"Érase un hombre a una nariz pegado,
érase una nariz superlativa,
érase una nariz sayón y escriba,
érase un peje espada muy barbado"
(Francisco de Quevedo, 1580-1645)
"There was a nose, which had a man appended.
It was a nose superb, superlative,
it was a nose, a hangman and a scribe,
it was a swordfish, sharp and hairy-ended"
I had the pleasure of riding around once in a Detroit Electric. To call it a tea parlor on wheels would have been an understatement. It came equipped with a flower vase, and instead of a steering wheel (how unladylike!) it came with a tiller that would sit across your lap.
These were cars for very wealthy elite women. But just as big of a deal as it was for women to drive - these cars in particular did not require a servant to start, operate, or service.
Well, not only former Be but former Finder developer here. The OSX Finder was certainly not a cute Finder-like skin, it was in fact huge amounts of the OS9 Finder source code including a complete copy of a modified Metrowerks PowerPlant C++ framework underneath it all. There was even elements of the old MacApp framework there. It was not a skin; it was a source code behemoth with massive amounts of legacy code.
I was working on some Windows Media DRM stuff and would get pinged with the really loopy support emails, most of the time not to have me fix them, just more of a heads-up. One said "WHY THE FUCK ARE YOU ON MY PC EVERY MORNING GET THE FUCK OFF MY COMPUTER". I actually started to reply to this chap to see if I could get some sense, but a VP was nearby and saw my screen. They picked up my keyboard, wrote "FUCK OFF" and hit send.
Had the pleasure of working with Jason Garrett-Glaser (better known as DarkShikari) circa 2011 on a patch for x264 to better utilize more cores (slice-threaded lookahead). At the time he was working with Sony to optimize their PlayStation game streaming service, and this had the potential to lower encoding latency (critical to game streaming) even on same-core systems by spreading the frame over more cores to do more parallel work, and doing more intelligent scheduling and convergence after that tile-based work was reconverged into a frame. At the time we were using servers with 48 physical cores to do transcoding, a bit of a rarity in a single NUMA space back then and there were no instances to test this on so I flamegraph traced a few encodes at very high thread parallel counts and send back the instrumented profiles (from a custom compiled version) showing the core work being done and he used that to optimize the motion estimation code which worked on the first shot to perfectly level out cpu usage and increase encoding speed across ~24 cores (was previously topping out around 8 and running one core hot iirc). I got a look into how exotic and complex this project was, particularly in terms how low level and advanced things like motion estimation, quantizing and perceptual compression were... it was and still is sort of magic how many smart techniques are baked into something that we barely pay attention to. They were, even at that time, working on x265.
His blog, Diary of an x264 developer, was a great read too for anyone wanting to get a vantage point on how crazy that codebase was.
As a proof-of-concept, I've been using extra tokens to formalize the IRS tax forms and worksheets, mostly just to prove this can be done. The idea is that something like this could be adapted into a publicly maintained version of the IRS Direct File program (RIP). I'm up to an estimated 60% coverage of US federal tax situations so far.
One crucial part of my role as my family's de facto IT support was making sure my mom's FreeCell score record (which was just a Windows registry value) made it from each computer to the next across countless upgrades. She hated the new version that came out somewhere along the way because the card decks changed. (Windows 7? Windows 10? I switched to Linux full time in the XP era so my memory is a little fuzzy.) So I made sure to keep a backup of the Win32 version she was used to, and loaded it onto every machine she ever owned. It wasn't quite the 33 years of its lifetime, but probably close to 30.
Mom passed earlier this year, so the FreeCell score record won't ever increment again, but I still made a point to back it up one last time when going through her desktop this summer.
Nuclear strike orders do not require high bandwidth. The orders contain targeting information and arming codes, all of which is pre-programmed on the missiles (the target package is specified with a "pointer" -- the computers on the missiles already have all the details, the order simply identifies which set of details should be used). For ICBM crews it is conceivable that orders could be transmitted on high-bandwidth systems, but for submarines that is not even possible (ELF bands have to be used because of the physical properties of ocean water).
The basic launch sequence is this: a missile crew receives a series of characters and both missileers write the characters down. They check against a codebook and verify that they both heard the same message and that it is a valid launch order. If the message is valid, they unlock the launch keys (from a box secured by two combination locks, each missileer knows the combination for one lock), insert the keys, enter the missile launch unlock code, enter the warhead arming code, turn the keys, and then that's that.
(There are several other steps -- go watch a video of a test launch and listen to the operators read off the checklist, there are quite a few steps. If you are wondering why such information is public, it is meant to reassure the other side that there is no possibility of unauthorized or unintentional launches -- we want our adversaries to know how we control our nuclear forces so they don't panic and preemptively attack us)
Back in my Navy days, we were taught HF is the less susceptible to jamming so is used for critical comms like these in serious situations. My ship performed in one exercise related to a doomsday scenario. We sat all day listening to an HR receiver until we heard these same messages, and we would then record the characters spoken. The message was repeated so we could make sure what we wrote was accurate. We never responded. Instead, we then repeated the same messages on an HF transmitter, presumably as part of some relay.
The most successful push on spelling was when the Chicago Tribune switched to simplified spelling in the 1940s through the 1960s.[1]
They started with about 40 words, and gradually expanded. But after McCormick died, the company gave up, retaining only "tho" for "though". That ended in 1975.[2]
Hi, I'm the creator of Docker and came across this post. It's always fun to see other engineers think about how you designed a system. I haven't been involved in developing Docker in many years, but it's a fun trip down memory lane.
They were able to join a handful of retailer affiliate programs and were probably clearing tens of millions a year in affiliate commissions with no overhead. Then they sold it to WhaleShark Media (funded by Google Ventures among others) and IPO’d it in 2016, cashing in a second time.
Honestly one of the most impressive bootstrap stories you’ll ever find.