Seems bad for Snowflake? Iceberg is a big part of Snowflake's data lake offering, and I assumed it was a Snowflake-originated OSS project until this announcement (all Snowflake products have snow related names).
Yesterday we announced Polaris specifically so (1) customers don't get locked into a catalog; (2) people know Snowflake works with AWS, Azure, Confluent, etc.
Iceberg is not Snowflake originated, it was built by folks at Netflix - the same folks who built Tabular.
But yes, this is definitely bad for Snowflake, Databricks can position itself as a very strong competitor with this move and moving more towards Iceberg.
So this makes it sound like a positive thing for Iceberg as a format. Others seem to be suggesting that DataBricks will be working to undermine the format in favor of their older Delta Lake format but that seems overly cynical.
My guess is Data Bricks saw the popularity of Iceberg and realized that they were starting to look a little irrelevant still trying to promote their competing Delta Lake format. Have an Iceberg lakehouse? Well then Databricks just didn't seem very relevant to you. With their purchase of Tabular they're given some legitimacy when they start marketing their products as being iceberg compatible. This doesn't signal to me that Iceberg is going to be harmed in the near or medium term.
Having worked with a team using Snowpark, there are a couple things that bother me about it as a platform. For example, it only supported Python 3.8 until 3.9/10 recently entered preview mode. It feels a bit like a rushed project designed to compete with Databricks/Spark at the bullet point level, but not quite at the same quality level.
But that's fine! It has only existed for around a year in public preview, and appears to be improving quickly. My issue was with how aggressively Snowflake sales tried to push it as a production-ready ML platform. Whenever I asked questions about version control/CI, model versioning/ops, package managers, etc. the sales engineers and data scientists consistently oversold the product.
Yeah it's definitely not ready for modelling. It's pretty rocking for ETL though, and much easier to test and abstract than regular SQL. Granted it's a PySpark clone but our data is already in Snowflake.
The top part of Joe Eskenazi’s earlier article https://missionlocal.org/2023/04/bob-lee-crazy-bob-mobilecoi... was good (where he reported evidence that the Bob Lee killing was not a robbery), but I agree with Garry Tan that the bottom part of the article (where he makes a broader commentary on whether San Francisco is “safe” unrelated to Bob Lee) is gaslighting and reductionist. It makes many claims and implications that are questionable: e.g., that rampant property crime does not make you unsafe (despite the fact that many thieves are armed), that festering drug addiction does not make you unsafe (which may have been a contributing factor to the car not stopping for the victim), that the problem with crime is “feelings” rather than real risk, and that those who are concerned about crime must have come from sheltered “suburbs”. And there’s no mention of anti-Asian robberies that got the previous DA recalled.
Eskenazi is a well-connected journalist, but he is also arrogant and often presents only one side of issues. For example, virtually nothing that he wrote in this article (anonymously sourced from disgruntled politicians) about the magnet school Lowell High school ended up being true (magnet schools do not violate state code as claimed, and the school did return to test-based admission which he claimed would not happen) https://missionlocal.org/2022/02/lowells-old-merit-based-adm.... So while his reporting is mostly good, you have to be aware of his bias.
I think you are trying to pigeonhole Eskenazi's argument into the standard "progressives don't care about crime" punditry that's popular on the right. I would suggest re-reading the column with a more open mind. His argument is that feeling safe is as important, if not more important from a policymaking perspective, as empirical measures of safety ("real risk") such as violent crime rate. He is in fact arguing the exact opposite of what you are characterizing him as arguing ("rampant property crime does not make you unsafe", "festering drug addiction does not make you unsafe"), and seems to have advised politicians to ignore these issues at their own peril.
The problem is that just like violent crime rates don't fully explain feelings of safety, things that make one feel unsafe don't fully explain all violent crime. Since Bob Lee's murder did not seem to be a result of either drug-induced psychosis or a mugging gone wrong, Joe made the correct call that the murder was likely unrelated to either of those issues.
That all being said, it appears that you have issues with him based on unrelated reporting on an issue you seem to care deeply about. A good of a time as any to examine any potential biases you might have when receiving new information so you don't accidentally embarrass yourself on Twitter!
> I think you are trying to pigeonhole Eskenazi's argument into the standard "progressives don't care about crime" punditry that's popular on the right. I would suggest re-reading the column with a more open mind. His argument is that feeling safe is as important, if not more important from a policymaking perspective, as empirical measures of safety ("real risk") such as violent crime rate.
From the article:
> But the city’s violent crime rate is at a near-historic low, and is lower than most mid-to-large-sized cities.
> Lee’s death, however, was packaged in the media and on social media into a highlight reel of recent San Francisco misfortunes and crimes: large groups of young people brawling at Stonestown; the abrupt closure of the mid-market Whole Foods, leaving San Franciscans just eight other Whole Foods within city limits; the severe beating of former fire commissioner Don Carmignani in the Marina District, allegedly by belligerent homeless people — it all adds up to a feeling of a city coming undone.
> This manner of coverage, however, does not capture the actual lived experience of the vast majority of San Franciscans.
I'm reading his article again.. where is his argument that feeling safe is important?
He states that violent crime is low, and that newspapers shouldn't be cherry-picking and sensationalizing how bad it is, and that he knows this is not the actual experience for the vast majority of San Franciscans.
> His argument is that feeling safe is as important, if not more important from a policymaking perspective
I’ll set aside his smear that opponents of crime come from suburbs and his strawman that those who oppose crime only want “more cops, stiffer sentences and a return to the Gov. Reagan-era incarceration of the mentally ill”.
His main argument is that by “objective” measures, San Francisco is safe and any increase in danger is only subjective “feelings”, but that “feelings” still affect tourism and politics. The first part (which is the same party line we have seen in the SF Chronicle) is the gaslighting. I want to distinguish between psychological feelings, and risk which can be real but not fully measurable. I hypothesize that for many residents, actual risk of injury, not just feelings, has increased over the past 5 years. There are many mechanisms by which risk may not show up in the citywide violence statistics. Crimes may shift from one neighborhood to another (e.g. to touristy places) or from one demographic to another (e.g. against Asians) while staying steady citywide. Armed robbers may primarily target wealthier people who give them what they want, but victims who have less to lose and resist are more likely to be attacked. Underage thieves may injure you through reckless driving instead of attacks. With fewer pedestrians commuting after COVID-19, a street that is more dangerous may get fewer violent incidents. Or victims may underreport crimes. The point is, when there are so many changes in behavior among commuters, thieves, and addicts, I don’t believe a couple citywide metrics give the whole picture.
> you have issues with him based on unrelated reporting on an issue you seem to care deeply about
That’s just the most egregious example. I learned to read Eskenazi’s articles skeptically because I often read him pushing one lazy narrative but not getting the details right or not getting the other side of the argument. Another example is this article on Proposition 22 https://missionlocal.org/2020/09/prop-22-chronicle-uber-lyft.... In it, he claims that “Airbnb and its ilk skirted paying hotel taxes for years… And they kept their money”, which is not true; Airbnb settled with the city to pay all the 15% hotel back taxes which exceeded their own 6% revenue during that time. And notice how overtly one-sided that article is; it makes no attempt to get the other side of the delivery issue and whether paying drivers who are are waiting at home with their app open makes any business sense for Uber or for taxi dispatch companies for that matter. Another example I recall is the reporting on HubHaus https://missionlocal.org/2019/08/san-francisco-rental-platfo... which claimed without evidence that the room share company was “exacerbating the already onerous cost of housing” and showed very little interest in what it would take to actually follow the definition of family, and whether the definition of family itself is what is exacerbating the housing crisis. He is good at ferreting out a certain kind of bureaucratic corruption (e.g. DBI), but he turns a blind eye to other kinds of corrupt rules that benefit incumbents that politicians like Arron Peskin (coincidentally one of his favorite sources) specialize in. In other words, he’s biased, and you often get only one side from his articles.
I know in Twitter everyone is inflammatory, however a CEO should exercise restraint and show leadership, and hold judgement until facts are clear, in this case he clearly failed.
Of course, I know better. He's too cowardly to come out and say what he's feeling, but just take a look at his likes from today. Apparently he thinks that journalist should be jailed for embarrassing him. https://twitter.com/garrytan/likes
This is a prime example of the growing stated perception outside of tech bubble, "tech bros killing each other and blaming the homeless". It didn't take long for people to start posting that.
Yes, street crime is itself a taking of rights and freedom of the people living and working in a place, and it is more important to contain it than to maximize the criminals' rights.
But this shows that the reputation of the tech industry is pretty much underwater, and premature postings like that don't help.
> It's easy for a childless bohemian to have no problem with needles in parks, but for those of us raising future citizens, it's not fun.
> Your entire statement to me seems driven by emotion
> Sorry, but the whole comment reeks of luxury beliefs.
C'mon man. You don't know the person you are responding to and included multiple personal attacks in your response. There's a way to make your argument without making the person you are responding to your own personal hate-object.
Reminiscent of the 606 shelving system Vitsoe produces. Given the section dedicated to the desk's "Field Rail," I assume TE is going to release a couple more projects using it as part of some modular system. A brand with similar cult status, Snow Peak, revealed their own modular office concept[0] around a year ago, so I'm guessing that Teenage Engineering is planning similar things.
Lundia shelving system was invented in Sweden in 1946 but is made in Finland and New Zealand at least. It is odd in that it's light, adaptable, modular, durable and timeless, all at the same time. (Usually durable things are heavy.) This is because it is actually designed with mechanics in mind and the materials are good: solid wood for the compression members, thin steel for tension members and solid wood shelves for the bending strength. The shelves are movable without tools.
Also if you're in Finland, there's a huge aftermarket so it's easy to buy and sell Lundia "ladders" and shelves etc whenever you're moving or refactoring your home. They last basically forever and you can mix and match old and new parts. And it's a mover's delight as everything disassembles into easy to carry parts.
It's a complete antithesis both to the old heavy furniture and the Ikea cardboard throwaway things. All the furniture chains are making bad adaptations of the idea. Ditch those, get the original.
I was thinking Vitsoe as well, but without the deep thought. I have a single set of Vitsoe shelves that are excellent and have made it through two countries and counting. I liked - but returned - the TE OP-1, it was a luxury toy and the whole company seems to pander to the likes of Wallpaper*
To a certain extent, isn't this all a moot point? The sanctioned ETH addresses associated with the Ronin hack didn't use the immutable TC contract, they used the proxy contract[0][1] which can be updated from the Governance address (which is controlled by the Tornado Cash DAO). It appears that the transactions happened after the Treasury Dept. published the sanctions list.
To use your metaphor, a person who it would be a crime to sell anything to went to a gun vending machine to buy a gun. Given the nature of gun vending machines, it would be difficult, but not impossible for the vending machine company to prevent this scenario. However, in terms of assessing whether a legal violation occurred, it doesn't matter that the item the person bought was a gun, if the gun was then used to commit a crime, or if the item was instead a rubber band or a brown paper bag, or if the gun vending machine company designed the gun/band/bag or produced the gun/band/bag.
* DHTs are "invented" at Berkeley/MIT. Online storage is too expensive for most end users, so the target market is helping large companies index data in their filesystems.
* A bunch of content distribution protocols get built on top of the concept, the most popular being Bittorrent. Torrent protocols run into issues with the free rider problem, resulting in slow downloads. Meanwhile, storage on the web is becoming cheaper and cheaper.
* To solve the free rider problem, IPFS was created. A brilliant incentive structure was created so that asset hosting no longer relied on the benevolence torrent seeders. Meanwhile, companies like Cloudflare, Mega, Google and Amazon make online storage essentially free.
* Cloudflare starts hosting IPFS assets for free. Decentralized storage still exists, but is still slow in comparison and, in IPFS's case, is more expensive than free, so people only use IPFS addresses to download assets from large, centralized services.
* IPFS is mostly used as a distributed network of hashed addresses that can be used to look up data in a large company's filesystem.
BitTorrent was published in the same year as the original DHT protocols, and they kinda had nothing to do with each other until four years later (according to Wikipedia the Mainline DHT was added to the mainline client in November 2005).
