Papers We Love | Hacker News Reader

Papers We Love

paperswelove.org

244 pointswallflower49 comments

In 2017 I went to StrangeLoop and Papers We Love was collocated. Such a great experience. I saw a talk about how type inference in OCAML worked (Hindley–Milner type inference) and I basically didn't understand any of it, but talked briefly to the presenter after.

I left so excited by both conferences and started on pulling on threads around type systems and eventually started a podcast as a result.

The podcast changed over the years, but originally it was just a way for me to investigate things I had heard of at Paper We Love and Strange Loop. As a community, PWL is a nice on-ramp into CS Theory. Thank you Zeeshan Lakhani.

Your podcast is very useful - I have been a subscriber for a couple of years now!

Thank you for doing what you do.

Thanks for listening!

Really enjoy Corecursive. Thanks for all the work you've put into it. A really great resource for learning all about fascinating ideas and personalities in computing.

Thanks!

this!!!

PWLconf vidoes are up here: https://pwlconf.org/#videos

Thank you so much for the work you put into your podcast! I frequently recommend CoRecursive to new and old tech folks so they can have an insider eye on how the industry and especially company culture works. Educational, entertaining, and interesting!

Thank you for recommending it to your friends! That is the highest praise.

Your podcast is the best, Adam! I enjoy every episode of it.

Thanks for listening!

If you want to get into reading some CS papers, I can also recommend these sources:

1. The unfortunately defunct blog "The Morning Paper" that provides a short write-up/explainer of each paper. https://blog.acolyer.org

2. Will Larson's favorite CS papers. https://lethain.com/some-of-my-favorite-technical-papers/

The pandemic has been tough on Papers We Love as many of the local chapter meetups had to shutdown. This had an adverse impact on the Papers We Love Conf (https://pwlconf.org/) as well.

Some chapters (Seattle, Chattanooga, Bangalore) have managed to keep up a semi-regular cadence of Zoom meetups. NYC tried but has a hard time getting speakers (which is a bit surprising tbh.)

If you're interested in running a meetup, being a speaker or volunteering your time hit up @papers_we_love on Twitter.

Are there any good "baby's first CS papers"? I read maybe 1 paper in college and have not thought about anything more than the very practical, day-to-day applications of my degree since then. If I wanted to dip my toes into reading some CS papers are there any good places anyone would recommend to start?

I think the original DynamoDB paper is a good choice - some interesting CS/systems engineering content but relatively accessible.

https://www.allthingsdistributed.com/files/amazon-dynamo-sos...

Notably, the Dynamo paper is _not_ a description of how DynamoDB (the product available from AWS today) works. They are fundamentally different, and it is not possible to implement notable features like conditional updates using the algorithms described in the Dynamo paper.

That’s interesting. I was going to take a stab at reading that first since I frequently use DynamoDB (Amazon product) at work.

If you’ve worked in the systems space, the Flash HTTPD paper is a classic: https://www.usenix.org/legacy/events/usenix99/full_papers/pa...

There's this nice mix of technical and opinion papers: 10 Papers Every Developer Should Read https://michaelfeathers.silvrback.com/10-papers-every-develo...

I have heard good things about https://mitpress.mit.edu/books/ideas-created-future

‘Time, clocks and the ordering of events in a distributed system’ by Leslie Lamport 1973. This is my favorite intro to distributed systems paper. Gives you a new tool to think with, and it’s a straightforward read.

https://lamport.azurewebsites.net/pubs/time-clocks.pdf

“Why Functional Programming Matters” by John Hughes.

https://www.cs.cmu.edu/~crary/819-f09/Strachey67.pdf

Fundamental Concepts In Programming by Christopher Strachey is awesome. It’s a lot less technical because it was based on a series of lecture notes. Christopher Strachey is also an inspiring figure in CS to me as someone who was a perpetual fuck up and a late bloomer.

https://en.wikipedia.org/wiki/Christopher_Strachey https://en.wikipedia.org/wiki/Fundamental_Concepts_in_Progra...

The NYC PWL chapter is pretty awesome. The organizers do a great job in finding speakers and venues. I've made many friends from the group. Darren, Andrew, Sean and David are amazing. So was Zeeshan before he left for his PhD.

A truly amazing team indeed :).

Also wanted to link to the copious amount of PWL videos around (incl. diff chapters, the conference): https://www.youtube.com/c/PapersWeLove/videos. Having run a few internal reading groups at various jobs myself, I know so many who watch the related videos while reading a specific paper (or use the morning paper as cliff notes).

I have organized PWL in Milan at the end of 2019 / start of 2020, then we-know-what happened and put the entire thing on stand-by, because of Zoom fatigue. If anyone is interested, just ping me, though.

Milano: https://github.com/papers-we-love/milano

Any have any tips for getting started on reading a paper? It’s like reading something in a different language.

This might be useful: https://web.stanford.edu/class/ee384m/Handouts/HowtoReadPape...

If you find there is just too much unfamiliar technical language, it’s a good idea to pause and look up a definition. You might need to follow that chain several steps, but that will help you get to an understanding of new terms.

Plan on reading it 4-5 times. The first time it is going to be a whole bunch of new words/concepts -- "a different language" as you put it.

Only read the introduction.

The first reading is to inventory the words, where they are used, and how important they are. Don't try to understand except in the most vague of ways.

The second time you can start to make some sense out of it because hopefully you think you might understand 1/3- 1/2 of the words.

Now look up some definitions. Google might help, otherwise the difficult word(s) might be cited, in which case you need to read the abstract of the relevant paper (only enough to get a definition of what the word might mean).

The third time you read it, you should know 1/2 to 2/3rds of the words and be able to make good connections between them, you start to see the ideas take shape.

The fourth read you fill in the gaps from the outline you made in the third read.

If the paper is well written, the introduction lays out the relevant current thought on the topic, and suggests a way by which some conflict in this thought can be resolved.

The discussion presents that resolution.

This Sean Cribbs talk, The Refreshingly Rewarding Realm of Research Papers, also has a ton of good thoughts on the process of getting started: https://www.youtube.com/watch?v=8eRx5Wo3xYA.

The 3-pass approach mentioned in the Stanford page linked in a sibling comment is a good way to start. It's also worth knowing why you want to read a particular paper. They're usually written for communicating tersely with people "already-in-the-know" so you shouldn't expect to get much out of most papers you come across.

Murat Demirbas a prof at SUNY Buffalo has described his method - http://muratbuffalo.blogspot.com/2013/07/how-i-read-research...

I attended some pwl Kansas City chapter meetings, but the pandemic put the meetings on hold as we started out end of 2019.

Somethings I don't understand. How do they cover the cost for the event?

Does the attendees pay for this and is it enough to cover the cost or is it profitable?

How is meeting people who are not authors of the paper more useful than organizing online?

Companies usually provide the space and will sponsor pizza and drinks, at least in SF. That’s about all the costs beyond the time of the organizers which is offered for free.

Yep. For many other chapters as well.

I went to some Papers We Love meetups here in NYC a few years ago and learned a lot not only from the presenters but from other attendees ranging from database experts to sound engineers. Some very nice, kind people!

Plenty of pizza and beer, too :)

Warmly recommended.

To the last question, just having people put in the work to present and understand a paper, or a theme around a set of papers, can make a huge difference for others to understand the ideas put forth. We've even had events where presenters demonstrated an implementation of the paper or called out interesting references or historical associations that may not have been apparent initially.

Companies love to sponsor the local meetups because they see them as recruiting pools. They will usually provide the space and some $$ for food (pizza)

I have been organizing the Papers We Love Beijing Chapter with a group of engineers. It is a good place to share wonderful ideas that are not necessary useful.

Anyone from Paris ? I’d be glad to take part to sessions

Is there any active pwl group that has regular zoom meetings to discuss papers? Would love to join in.

Seattle does https://paperswelove.org/chapter/seattle/

I'm shocked there's no chapter in my city!

São Paulo has so many universities and computer science groups.

There was initial attempt in 2017: https://github.com/papers-we-love/sao-paulo

Maybe you could get some people together and start one https://github.com/papers-we-love/organizers

Chicago, anyone?

There used to be a Chicago chapter https://paperswelove.org/chapter/chicago/ - would be great to get it restarted

Reply on news.ycombinator.com