Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

RIP Usenix ATC

bcantrill.dtrace.org

Post reply on HN
Print view View original

21–30 of 40 posts

Re: RIP Usenix ATC

#21

Post by linguae »

Ar-Curunir wrote: post #20
khuey wrote: post #3

IMO the implication that the "in-person conference" nature of the forum is what led to its demise rather than its capture by academics/abandonment by industry practitioners is incorrect. I think a better summary of the cause is that most academic research is not that relevant to industry but academics need the prestige of forums like Usenix ATC in a way that industry does not which naturally leads to squeezing indust…

I don't think academics view ATC as particularly highly prestigious. Just within systems there are many other conferences which are either more, or just as, prestigious as ATC: OSDI, SOSP, NSDI, SIGMOD, VLDB, SIGCOMM, EuroSys. Especially the first three are viewed to be much higher prestige than ATC, at least from my perspective.

I used to be a grad student in a file systems lab at UC Santa Cruz over a decade ago. It was considered a dream to publish in OSDI or SOSP; our lab’s most successful paper was the OSDI 2006 paper on Ceph. We also thought very highly of VLDB, though most of our work was outside the interests of the database community. We regularly published in USENIX FAST and made occasional appearances at USENIX ATC.

Re: RIP Usenix ATC

#22

Post by bcantrill »

Ar-Curunir wrote: post #20
khuey wrote: post #3

IMO the implication that the "in-person conference" nature of the forum is what led to its demise rather than its capture by academics/abandonment by industry practitioners is incorrect. I think a better summary of the cause is that most academic research is not that relevant to industry but academics need the prestige of forums like Usenix ATC in a way that industry does not which naturally leads to squeezing indust…

I don't think academics view ATC as particularly highly prestigious. Just within systems there are many other conferences which are either more, or just as, prestigious as ATC: OSDI, SOSP, NSDI, SIGMOD, VLDB, SIGCOMM, EuroSys. Especially the first three are viewed to be much higher prestige than ATC, at least from my perspective.

I really don't care about how prestigious academics view (or viewed) ATC, but for whatever it's worth when our paper was accepted in 2004, was just under 13%. That is bluntly too oversubscribed for the practitioner (whose job is to ship systems, not win paper popularity contests -- or adjudicate those popularity contests by serving on program committees). In some ways, if ATC wasn't thought prestigious, it's even worse -- it means that practitioners lost our conference for nothing at all.

A final note: given that OSDI is on your list of prestigious conferences, the way the program committee was conducted in 2010 (which I outlined in my ATC keynote[0]) should be particularly galling.

[0] https://speakerdeck.com/bcantrill/a-wardrobe-for-the-emperor...

Re: RIP Usenix ATC

#23

Post by bcantrill »

Ar-Curunir wrote: post #19
mlyle wrote: post #15

Earlier quoted context omitted.

I think what they're saying is this: USENIX ATC at its best was always focused on operations and real use of Unix. It was the Unix User 's Group. But academic CS has a shortage of real publishing opportunities, causing less immediately practical things to flood USENIX ATC and smother it.

Academic CS has no shortage of prestigious or even good publishing venues. Just within systems and networking, there is SOSP, OSDI, NSDI, EuroSys, PODC, DISC, FAST, SIGCOMM, CIDR. If you extend to databases there's also VLDB, SIGMOD, etc. These are all Tier-1/Tier-1.5 venues. Academic CS didn't take over ATC because of a shortage of venues.

You made this comment twice, so I guess I'll reply to it twice, but the reality is that the acceptance rates of ATC was very, very low (again, under 13% in 2004). Low acceptance rates actually do indicate a shortage of good publishing venues (certainly relative to the number of submissions), but it would be interesting to look at the ATC acceptance rate over time; if it was much (much) higher in its final years, it would be easier to accept your assertion.

Re: RIP Usenix ATC

#24

Post by raddan »

CalChris wrote: post #11

Similarly, Stanford's EE380 has slipped from these mortal coils. Brian Cantrill spoke there as well. BITD, EE380 seemed like the Senate of Silicon Valley. Startups would announce at EE380. But then it went from relevance to thinly attended and taped for YouTube to archives. This wasn't academic capture. There's a simpler explanation. No one goes to talks anymore.

It would be seriously disappointing if the world we are moving to is one where people don’t care about in person talks anymore. Not only do I love in-person conferences, I would probably stop caring about publishing papers if it was disconnected from the social parts. The social parts of PLDI, SPLASH, and yes, ATC (where I was fortunate enough to win a best paper on my one and only submission there) have been major highlights of my career, and more than once have led to friendships, interesting projects, and even employment. I was part of the team that helped make PLDI virtual during the pandemic in 2020, and as well as it went, I found it so painful to attend that I basically did not attend.

I understand that I am in the minority here when it comes to remote work, etc, but if opportunities to learn from and socialize in person with colleagues go away, then I will too. Woodworkers don’t seem so enamored with virtual this and that, so I would probably jump ship and do something else.

Re: RIP Usenix ATC

#26

Post by romesmoke »

I perceive a negative colouring of the "academic" notion in this post. And I totally get some of the pathologies implied: publish-or-perish is a harsh environment, lots of snobbery.

With all respect, however, this industry/academia dualism looks to me at best as a false and at worst as a harmful dichotomy. I mean, there are similarly cliché pathologies in whatever one chooses to overload "industry" with. There are successful academic projects that end up in industry (RISC-V) and successful industrial open-source (?) projects loved by academics (Rust). At the end of the day, everyone wants to build something. Tools and methods are needed. Some of them may not exist yet. The act of building one's tools and methods (or something completely new with existing tools and methods) is research regardless from where it happens or how it is labelled.

Why is this not enough? What extra benefits go with differentiating myself as either an academic or a practitioner? Honest question!

POV: CompEng PhD looking for industrial vacancies.

Re: RIP Usenix ATC

#27

Post by mmooss »

Building relationships and community is still essential, and it does require some in-person events. A central annual conference is time-efficient. How will these people get together?

Re: RIP Usenix ATC

#28

Post by dadrian »

I think people are glossing over the fact that Usenix ATC wasn’t even a good academic conference. If you’re submitting high-quality academic work to Usenix, you’re sending it to one of the higher prestige Usenix conferences in a specific subfield, such as Usenix Security or NSDI.

Re: RIP Usenix ATC

#29

Post by AdieuToLogic »

gnat wrote: post #14

Computers and the Internet started with academics. Distributed systems were theoretical before they were practical. The first sites of the Internet were universities and the first non-academic site was a milestone. Harvard and Dartmouth and MIT all contributed to computer design and programming advances. UNIX is a product of a quasi-academic research lab. The Center of Gravity shifted in the late 90s/early 2000s, I r…

> Computers and the Internet started with academics.

Only if one categorizes academics as a research arm of military funded research goals. Not that there is anything wrong with that, but let's be intellectually honest.

> UNIX is a product of a quasi-academic research lab.

A wholly owned subsidiary of the commercial entity AT&T as far as the UNIX trademark is concerned. The BSD family would have something to say about Unix (note case difference) however.

> The Center of Gravity shifted in the late 90s/early 2000s, I reckon. Real world needed distributed systems ...

Distributed systems and the networking protocols now assumed (TCP and UDP) was a battle fought before the late 90's. See Novell[0] for an alternative many espoused at the time and Sun's "The Network is the Computer" mantra from the mid 80's.

> Not saying USENIX didn’t have factions. I just question the word “capture” about academics, who invented all this stuff before there was money in it.

There was always "money in it." Academic funding is, by definition, justifying the allocation of money for the advancement of knowledge often pitched as relevant to the benefactor. Again, not that there is anything wrong with this.

0 - https://en.wikipedia.org/wiki/Novell

Re: RIP Usenix ATC

#30

Post by khuey »

romesmoke wrote: post #26

I perceive a negative colouring of the "academic" notion in this post. And I totally get some of the pathologies implied: publish-or-perish is a harsh environment, lots of snobbery. With all respect, however, this industry/academia dualism looks to me at best as a false and at worst as a harmful dichotomy. I mean, there are similarly cliché pathologies in whatever one chooses to overload "industry" with. There are su…

> successful industrial open-source (?) projects loved by academics (Rust)

Is Rust loved by academics? And much more importantly in my mind, was it even recognized by academics before it became an industrial success?

The very first published Rust paper that I'm aware of appeared in the "ACM SIGAda Ada Letters" (https://dl.acm.org/doi/10.1145/2692956.2663188). Today that paper is the most cited paper to ever appear in that journal (which has a 4 decade history) and it's not even close. It also has more citations that all but three papers that appeared at PLDI that same year, for comparison. That certainly doesn't suggest to me that it was recognized at the time. This was also published only 6 months before Rust shipped 1.0. It wasn't that early.

My (third-hand and now-decade-old, so take it for the very little it's worth) recollection is that academic forums weren't interested in Rust because nothing in it is particularly novel in a PL theory sense (a point Graydon himself made from the very beginning, see slides 6 and 7 at http://venge.net/graydon/talks/intro-talk-2.pdf). But it did package those ideas into something that was practically usable for industry and at this point the results speak for themselves.

IMO this is a good example of why a lot of "practitioners" in industry wouldn't bother trying to publish anything in a forum dominated by academics.

Post reply on HN