Snowflake - founded by Marcin Zukowski, his academic grandson via Peter Boncz.
HyperDB - Actually not columnar at all... but Martin's student, Peter Boncz, works closely with Thomas Neumann, the main originator of HyperDB.
Vectorwise was created by Peter, with involvement of Sandor Heman and Marcin Zukowski, and limited involvement of Martin himself and others at the CWI DB Architecture research group.
DuckDB - also not columnar IIANM, but the work of Martin's research group.
Arrow compute, ClickHouse etc. - inspired by Martin and others' work on column stores.
Martin was an inspiring group leader even after his retirement... he was always so upbeat and optimistic. I'll also remember him for his flowery and poetic use of metaphors and parables in papers and titles.
* MonetDB Solutions: https://www.monetdbsolutions.com/ is the commercial spin-off of MonetDB. Martin chaired the company hands-on after retiring, almost to his last day. They do integration, training, deployment, professional support... interestingly, it is fully FOSS, and commercial clients who ask for features get those features implemented as FOSS.
Kersten led amazing research and MonetDB was really ground-breaking in many ways (columnar, shared memory interaction with client languages, etc.). A sad day.
Sad to hear of this. MonetDB really is/was pioneering and interesting work. I hope development on it continues to progress. There's some quality practical research and engineering in there.
Cause of death seems to occupy a strange place in our society.
Almost everyone is eager to know that detail, but can't pinpoint why.
On the other hand, especially when we weren't close to the deceased, we know that it's information we probably don't need. And that people close to the deceased may find it painful to have publicly discussed.
It’s the same reason why people watch accident compilations on YouTube. It helps you understand why others died, so that you may be able to reduce the chance of getting into the same situation. Moderate morbid curiosity is healthy in the sense that helps you understand the world without needing first-hand experience which may actually kill you.
> It helps you understand why others died, so that you may be able to reduce the chance of getting into the same situation
I'd like to clarify this isn't universally true. It's worth noting that in cases of suicide, specifics aren't published, because publishing specifics increases suicides. (I'm not saying this was suicide).
But in certain cases, I think you're right. High level information about causes of death in mountaineering can be useful for other mountaineers.
And finally, in other cases? I don't think it's relevant.
If people were really this rational they would just look at actuarial tables, not individual deaths. People are morbidly curious, but I think it is just to see whether he brought it upon himself(drug overdose, suicide, heart disease in a fat person, drunk driving, etc), or whether it is something they should mourn for (cancer, heart disease in a non-fat person, killed by drunk driver, etc).
Is it just me or do a lot of famous software people meet their end by cancer?
It reminds me of something from a bit of mystic literature. It was offered that serious practicioners of "concentration meditation" tend to get the crab.
And software developers are nothing if not concentrated.
Makes you think.
What is this "attention" thing anyway? A field? A ray? A wishing machine?
There are many causes of death, and globally the distribution of cause of death is quite wide.
But software developers are generally removed from a large portion on it. They’re not working with hazardous things, typically can avoid low-income related consequences (including living in dangerous areas), and have a higher than average intellect.
When you remove most of the common causes of death, cancer becomes proportionally more common. It’s like rolling 100 dice, tossing out anything less than 5, and noticing a lot of 6s.
*> Is it just me or do a lot of famous software people meet their end by cancer?
If you live long enough cancer is one of the biggest concerns, and people who don't have particularly dangerous lives generally live long enough for this to be an issue.
Heart disease is the other top dog on the list. In many places this is still the top killer over-all, but it tends to take people earlier (obviously talking averages here, either can strike at practically any age) and is more deadly among those with lesser access to good health care (the non-famous) so the better-off may experience both, survive the “heart problems” cut, and get hit by cancer later.
When you're famous and smart, you're likely to live to a ripe old age.
Cancer is an endgame disease, like gout being a disease of nobility. If you don't get killed off by violence or diet first, eventually you live long enough where you begin to seeing the result of running a copy through a copy machine a few billion times, which is cancer.
Our cells take damage every day, eventually that damage is just enough for some cells to lose control of how and to what extent they replicate, soon some of their friends join and soon before you know it they want their own veinous supply.
RIP Martin and thank you for your contribution. The world of data analytics honors you!