Mae Milano
@mbpmilano
Followers
1,715
Following
611
Media
11
Statuses
350
@PrincetonCS Assistant Professor. I build Programming Languages for Distributed Systems! @mpmilano .bsky.social
Berkeley, CA
Joined November 2018
Don't wanna be here?
Send us removal request.
Explore trending content on Musk Viewer
Oasis
• 497984 Tweets
SB19 GRACES BILLBOARD KOREA
• 283538 Tweets
SB19 GRACES BILLBOARD KOREA
• 283538 Tweets
Starmer
• 249342 Tweets
Ugarte
• 112760 Tweets
PS5値上げ
• 54751 Tweets
オアシス
• 44849 Tweets
コントローラー
• 40730 Tweets
REACT TRAILER JJ
• 34406 Tweets
ニンダイ
• 29380 Tweets
ゲーミングPC
• 23594 Tweets
値上がり
• 19568 Tweets
#日向坂46_四期生ライブ
• 19468 Tweets
スタテン
• 18440 Tweets
Toney
• 15772 Tweets
ゲーム機
• 14060 Tweets
UMA Summer
• 12485 Tweets
Amrabat
• 10355 Tweets
PhD DONE! Just got word my final revisions have been accepted. Thank you everyone who helped me on the way (especially
@AndrewCMyers
, advisor extraordinaire). Now on to a postdoc with
@joe_hellerstein
at
@ucbrise
!
10
3
196
I did a talk about my research! If you like this sort of thing, you should consider coming to
@Princeton
and doing a PhD with me 👀
"Programming Distributed Systems" by Mae Milano
@mbpmilano
from PWLConf 2023 at Strange Loop
0
26
140
4
25
196
Hey internet! I'm on the Computer Science faculty job market this year!! Check out my materials at , coming to a campus near you!
2
38
111
After 7 papers in 7 different conferences, finally breaking my streak with a second PLDI paper!
Type systems that allow fearless concurrency are cool but are either too restrictive or require destructive mutation. Our solution: "A Flexible Type System for Fearless Concurrency" (with
@mbpmilano
and
@JTurcotti
, conditionally accepted to PLDI'22.
3
2
65
7
1
83
POPL: 1, Omicron: 0. The conference was AMAZING, thanks to everyone (especially
@andrewkhirsch
and Anna) for the conversations and the research and see you all soon!!!
0
1
58
Meanwhile at
#pldi2022
... Let's just say not a resounding victory against COVID this time. LOTS of folks testing positive during the conference, minority masking, so... Maybe let's all just take another COVID test before going back to work next week, eh?
POPL: 1, Omicron: 0. The conference was AMAZING, thanks to everyone (especially
@andrewkhirsch
and Anna) for the conversations and the research and see you all soon!!!
0
1
58
3
4
46
Ah, POPL decision day, the one day where the eternal war for control of my timeline is decisively won by "PL Twitter, but kinda sad"
(We too were rejected with lots of "accept / weak accept" reviews and one "weak reject" review. It happens!)
1
0
39
Nevermind POPL is forgiven see y'all at PLDI!!!
(More substantive tweet forthcoming)
Ah, POPL decision day, the one day where the eternal war for control of my timeline is decisively won by "PL Twitter, but kinda sad"
(We too were rejected with lots of "accept / weak accept" reviews and one "weak reject" review. It happens!)
1
0
39
2
0
39
A year and a half after the pandemic cancelled it, I finally got my 2020 PhD graduation 🥰! And I got to carry the class banner! And celebrate with friends I haven't seen in far too long 💜 @ Cornell University
3
0
38
Glad someone's figured out how to manage "the snake fight portion of your PhD thesis defense" during the pandemic. Genuinely no idea where the snake came from. Wish me luck, I guess!
1
0
29
The reception to this has been so kind! I'm glad everyone liked it 💜
Just saw
@mbpmilano
’s talk at
@strangeloop_stl
- incredible tour of distributed systems, and one of those talks that expands your mental model on a topic. I especially loved when she said: “The speed of light is slow - 4 inches per clock cycle! [at 3 Ghz]” 😄
#strangeloop
1
10
46
0
1
29
Hanging out at PL lunch with
@CornellCIS
's CS admitted PhD visit day and it's... a little different than last year. (But it's still great!)
1
0
26
Some folks are asking what we could have done differently, and whether there should be a "permanent" alteration to our conferences. Here's some easy ideas:
- get a "pandemic" sponsor to buy a COVID test per day for every participant. Require them each morning. It's $40.
(1/N)
Meanwhile at
#pldi2022
... Let's just say not a resounding victory against COVID this time. LOTS of folks testing positive during the conference, minority masking, so... Maybe let's all just take another COVID test before going back to work next week, eh?
3
4
46
1
6
26
SNAPL is now officially my favorite conference. Great ideas, really engaged audience, and excellent venue at
@BrownCSDept
. Many thanks to
@ShriramKMurthi
for hosting and organizing!
1
2
25
Good news! My dissertation "Programming Safely with Weak (And Strong) Consistency" has finally posted to both ProQuest and my website! Check it out here: .
...Also, I have a new website (just a port of the old one for now).
1
0
25
I graduated (again)! Also, apparently Instagram's "share to twitter" feature could use some work... here are the photos!
A year and a half after the pandemic cancelled it, I finally got my 2020 PhD graduation 🥰! And I got to carry the class banner! And celebrate with friends I haven't seen in far too long 💜 @ Cornell University
3
0
38
3
0
23
Guys I really wanted to go to FCRC/PLDI this year 😕
Academic PSA: please don’t organize any events in Florida in the future, indefinitely, and expect the list of states to grow in the coming weeks, as legislative sessions wrap. They are not safe for your most vulnerable attendees.
1
16
69
2
3
24
Yay bean friends! We're making a SPLASH in Chicago, I assure you 😎😉
2
2
23
@IanArawjo
I think it depends! Each language you learn comes with an implicit machine model, which colors students' intuitions for how a machine "really" works! Teaching the "next" language is much easier if you can discuss its core semantics in terms of that "intuitive" machine model.
1
1
21
On this
#TransDayOfVisibility
, we celebrate news that, for the first time since the dawn of body scanners, the TSA won't automatically pat down trans junk anymore!! Continuing the theme of "you know I think I'd prefer to be *less* visible, thanks..."
1
1
19
Happy to announce the Derecho paper () has been accepted to TOCS! Also, TOCS appears to be moving again! We'll have a camera ready by the end of January.
0
3
19
Hey! Who do I know who's going to SPLASH in Chicago in a couple weeks? Feeling a touch nervous about re-meeting everyone now that I'm out (and also it's still a pandemic) and I would really appreciate a few friendly faces... 🥺
2
0
19
My advisor, commenting on my dissertation intro: "your text reminds me somehow of Nietzsche—lots of drama". Is this a good thing? I'm gonna go with this is a good thing.
0
0
19
I expected a lot of surprising and interesting things to come up my first time teaching. But what I did not expect was the management overhead! Beside signing up students for this or that and coordinating TAs/assignments, the *actual teaching* load looks pretty small!
2
0
18
Happy to announce I've finally posted the extended MixT (PLDI '18) Tech Report to my academic blogsite! Thanks to
@lindsey
for pointing out that it wasn't online anywhere :)
0
4
18
Wow nothing drives Twitter engagement like the job market... I just managed to spook like multiple people by expressing a little too much enthusiasm for Cornell's new building!
(I did not apply to Cornell. I shall enjoy their building from afar)
2
0
17
I remember a couple of years ago seeing
@siobhcroo
tweet about buying flights for faculty interviews from an airplane and, at the time, I was like "wow I hope I'm never that busy!"
Meanwhile, I've just finished buying flights for February interviews... from somewhere above Ohio
1
0
16
YAY PLDI! We're hard at work on our artifact (thanks
@RiggerManuel
!) which is double-blind this year? I'm used to single-blind artifact evaluations! With the traditional announcement of paper names on twitter well underway, how much are we actually able to say about our papers?
2
1
17
With all the humans sheltering in place, winter is finally returning to Ithaca...
2
0
15
Y'all! I'm helping run the IWACO workshop!! It's got a submission deadline in a little over a week, so if you've got a work-in progress on aliasing, ownership, capabilities, or similar and want some early feedback then send it our way!!
0
4
14
I've finally revamped and redeveloped my academic website, now featuring a blog section! Check it out:
1
2
13
- explicitly move snacks and other hallway-track items outdoors. The "hallway track" will follow.
- expand provided daily tests beyond COVID test to combined COVID+flu test. Provide LAMP instead of antigen tests (these are both expensive! So need a sponsor)
(3/N)
1
2
13
@michael_w_hicks
The killer for me is (and has always been) mathpartir. I really don't know how to write a well-formatted inference rule without it!
1
0
13
I gotta find a way to work this in to literally all my talks :)
"The speed of light is slow! It's only about four inches per clock cycle." --
@mbpmilano
at SNAPL
3
18
112
2
0
13
@XandaSchofield
got
@thompson_laure
and me a cake! We're doctors and now we have the cake to prove it!
1
2
12
- "strongly encourage" conferences to host in locations with large, easily-accessed outdoor areas. Serve lunch and dinner outdoors. (PLDI did this one!)
- mask during sessions, when you are not the speaker.
- provide masks (on day one)
- keep the indoor hallway small
(2/N)
3
2
12
I'm revisiting some of my recent papers and have come upon an example in which Alice and Bob, like all academics, get on planes and travel a lot. Well, I guess that one's failed to keep up with current events.
0
0
11
@TaliaRinger
TEACH ME YOUR WAYS
I mean seriously I'm at the same place and just getting random dudes
1
0
10
Hey! Who do I know at Splash???
0
0
11
If you like our stuff on distributed programming, you're gonna love this. Check it out!
Katara, our system for automatically synthesizing CRDT designs from sequential data types (), is now open-source at ! Deep-dive blog posts into the implementation coming soon!
4
31
161
0
3
11
@lindsey
Guess what -
@joe_hellerstein
is doing it again (this time with me!) Syllabus is still in flux, but keep an eye on Berkeley's CS 294-170: Programming the Cloud. And if you're currently a Berkeley student it's not too late to add!
1
1
10
@AndrewCMyers
I really like Rust's compromise here! Give me something like type classes, give those type classes interface-inheritance, and give me a way to say that "this is a statically-unknown instance of type class T" to replace OOP interfaces.
Decidable inference is a nice-to-have.
0
0
10
UPDATE: We're extending the IWACO deadline until July 21st! Hope to see your exciting submissions :)
Y'all! I'm helping run the IWACO workshop!! It's got a submission deadline in a little over a week, so if you've got a work-in progress on aliasing, ownership, capabilities, or similar and want some early feedback then send it our way!!
0
4
14
0
1
9
I've posted a recording to my website, if anyone wants to watch!
1
0
9
@notypes
@avanhatt
@priyasrikumar_
@davidjustodavid
@koronkebitch
@alpha_convert
@samps
@emilyaherbert
@mwillsey
@ChandrakanaNaN
@cmeik
@HackerF00
@hgoldstein95
Missing y'all! Next year, in person.
0
0
8
- ensure outdoor areas are accessible and shaded areas are easy to find, or at least protected for those who need them.
- have an actual outdoor location, with a projector, for the business meeting and remote-only sessions. (Don't want weather to ruin a paper talk!)
(4/N)
1
1
8
@dedbox
@AndrewCMyers
at
@CornellCIS
/
@cs_cornell
-- look up the Fabric system, I'm a student of his with recent / early work including MixT + Gallifrey. Also
@natefoster
is close! Google + let me know if you have questions.
Also check out UCSC for
@lindsey
@palvaro
too!
2
0
8
I really wanted to make this tweet "Friendship ended with POPL, now PLDI is my best friend" but I don't wanna be mean to POPL 🥺
0
0
8
All of the changes I'm suggesting could be permanent, though some complicate organizing. And nothing changes the feeling of the conference.
Keep thinking about this. Because pandemic or not, it really sucks to assume that you'll just get sick after a conference.
(5/N) N=5.
2
1
8
Excited to be at SNAPL2019! But does anyone know exactly *where* it is? Otherwise I'm just gonna be wandering the CIT looking for an Official Person tomorrow :)
1
0
7
@ShadajL
is an absolute POWERHOUSE, in a few months delivering on a result I've been hoping to see for like three years now! Finally, easy-to-specify, provably-convergent CRDTs are HERE! Check out our paper (from the quoted tweet) at 😎
Super excited that our paper "Synthesizing CRDTs from Sequential Data Types with Verified Lifting" has been conditionally accepted to OOPSLA 2022! 🎉🎉🎉 We explore how program synthesis can be used to generate provably correct CRDTs that mirror sequential data types.
@splashcon
3
19
114
1
1
7
@sydgibs
Never underestimate Facebook marketplace! You'd be shocked at how budget-friendly it can be.
1
0
7
(PS: POPL
@Boston
tried to find literally a dozen warm locations before running out of time and giving in to a known-bookable location in the northeast US in January, so yeah I don't love the location either but like, it wasn't just borne out of a preference for Boston)
2
0
6
1
0
6
@IanArawjo
for example: starting with python, you can explain how things work as "everything is a dictionary." If you then describe pointers as indexes into a hardware-supplied dictionary, they're not so hard to grasp! Not saying this is necessarily a good idea, just came to mind quickly:)
3
0
6
@TaliaRinger
Meanwhile there are several states and countries in which the trans population is, depending on where you draw the line, actively undergoing early-stage genocide right now...
0
0
6
@jmgrosen
At Cornell,
@natefoster
(and maybe
@samps
now?) Sometimes runs a "great works in PL" class, which does what you're looking for: reads old papers and then discusses them in their context. But I'm not sure if it's written up anywhere.
@notypes
would know, he helped start it!
2
1
5
Treee!
0
0
5
0
0
6
I continue to be amazed by the students I get to work with---our line of coordination-averse programming continues, now coming to a databases conference near you!
Our VLDB vision paper dropped today. We describe how the gap between convergent writes and convergent reads in CRDTs is a "foot gun" and suggest monotone queries as a lens. We also propose many ways database researchers can help! If you want to chat about it just DM me!
4
27
174
0
0
6
Hi internet! If you're at the SOSP tutorial day, stop by the Derecho tutorial at 2pm in Waterhouse Ballroom 3. We're building the latest and greatest best-in-class cloud computing framework, and we'd love to have you along for the ride!
0
1
5
TIL that newish Macs STILL ship with a case-insensitive filesystem. How did we discover this? Try cloning a repository with two files whose names differ only in capitalization. Now just run git diff. Enjoy your new game of whack-a-mole! thanks
@EthanCecchetti
0
0
5
@koronkebitch
Ughhhhh I might have to do the same... God I hate pandemics. Even if POPL doesn't officially cancel I'm far from sure that enough people will want to brave the pandemic to make its in-person component viable anyway 😭
0
0
5
@AndrewCMyers
Like 100000x Rust, for the language enthusiasm, non-insane FFI / external linking, ease of syntactic extensions (via procmacro), extra type/concurrency safety, and performance debugging.
But also, I would personally still use C++ and just suffer! (this is a joke. i hope.)
1
0
5
@ShriramKMurthi
@BrownCSDept
@EthanCecchetti
@jseakle
That class forever changed how I write C++! Would take again!
1
0
5
I have a joke about free objects but you'll probably find it too obvious to be particularly interesting.
1
0
5
@lorisdanto
Usually, but at that point they also are going to have to talk to me synchronously about it.
0
0
4
0
0
4
1
0
4
@afd_icl
Tbh the conference introduced a bunch of innovations, so easily integrated into the conference experience that we didn't think of them as mitigations at all! Picking a venue that could have such an outdoor hallway track for example. These could become permanent and I'd be happy!
1
0
4
@dmimno
@XandaSchofield
@maria_antoniak
There's a bunch of settings under "editor" to control this. As far as I know, they only control when the completion dialogue should be displayed; I'm not sure if there's a way to turn it off completely. Things like editor. quickSuggestions, editor. suggestOnTriggerCharacters
0
0
4
Best description I have heard of new Orleans french quarter so far: "you are in a maze of twisting parties, all alike"
0
0
4
We've still got a few hours if you find a typo!!
Mae and Josh's
@PLDI
'22 paper on a new type system for safe concurrent programming is now available.
1
1
14
0
0
4
@joe_hellerstein
@lindsey
@ShadajL
@alvinkcheung
@conor_power23
I'm fairly sure that the approach would... just work for the class of CRDT specifications we already handle, actually! And a traditional metalift approach may well prove effective for the more general synthesis problem. We should chat!
2
0
4
@TaliaRinger
PL PhD
@Cornell
, 2013:
(1) Meet Dexter and
@natefoster
and
@alexandra8silva
(2) spend months on an SMT-based verification project for NetKAT, which fails
(3) merged with
@thompson_laure
's near-complete first year project, to her surprise
(4) submit that to POPL, it gets in
0
0
4
@ilyasergey
Definitely not! But---if the paper is missing something important (to me), I'm more forgiving if that paper is at the page limit. And I feel like every paper I've ever reviewed or written has at least one reviewer clamoring for more. So in practice, maybe it does get penalized???
0
0
4
@lindsey
@Twisol
@joe_hellerstein
Ooo yes can we please turn this thread into overlooked paper suggestions thank you 🙏
1
0
3
@afd_icl
I wanna make it clear that some of PLDI's changes really COULD be permanent! There's a few more changes (snacks outside, outdoor hallway track) that are "in the spirit" of this year's PLDI. Let's not lose track of the good while we argue about whether we'll do anything at all.
0
0
3
@typesanitizer
Yay I'm glad you liked the paper!!! And excited to see a few industry languages in our space too--especially project Verona :) I was actually interning with Verona designer
@ParkyMatthew
waaay back in 2017, before either my language or Verona started... small world!
1
0
3
@davidad
I'm actually a little confused by this modeling! I would have expected unit to be an initial object (co-cartesian, so checks out), but I'm not seeing how we're modelling the commutativity requirement of CRDT writes! I thought cartesian categories are just monoidal?
1
0
3
@acasals
@AndrewCMyers
@joe_hellerstein
@ucbrise
The thesis itself is working its way through proquest, but the talk is here: and you can find the papers that make the thesis posted to my website:
1
1
3