(untitled)

I'm going to be wrapping up my CS degree next year, and we're required to deliver a final team project (from 2 to 6 members where the project's size scales with the number of participants) that we're encouraged to think about from now, probably in order to have us determine which fields we want to work in. I spent the large majority of my college years as a compilers geek, but right now, the field I'm interested in the most is distributed systems, and I want my final project to be my entry to it. Problem being, really have no clue to what a project reflecting that would look like, especially since I spent most of my times on compilers and this change in focus is very recent, I'm not at all experienced or knowledgeable. My questions being, where should I start? What should I think I about? And what would a 'distributed project' look like? You can throw ideas at me, too, because I don't really have a clear image and I need one before next year.

meant to say 'im not at all experienced or knowledgeable in distributed systems' as I've accumulated knowledge on compilers which I have no equivalent on in this area

>>13123 This is right tough, make sure you don't make it to difficult. Try asking a researcher in the area at your establishment. They should be happy to chat. But also keep it simple, this isn't a Masters.

You could do a command and control server. https://github.com/TreeHacks/botnet-hackpack

Make something on top of Blender or POV-Ray render farms. Distrib = farmland.

>>13123 >My questions being, where should I start? What should I think I about? And what would a 'distributed project' look like? You can throw ideas at me, too, because I don't really have a clear image and I need one before next year. Most people publish their final year projects so that's what you should be researching. Especially ones from your own university. Also, you probably figured this out by now, there are no objective standards for higher education. The most important thing is to know who will be evaluating your work and what they like/don't like.

>>13123 > compiler nerd > distributed project Why not a distributed compiler?

You probably need to elaborate on >distributed systems Do you mean something like microservices? Apache Spark/Hadoop clusters? Kubernetes deployments? Jenkins distributed builds? LoRaWAN nodes and gateway? The term is very vague.

Write a domain specific language for real time peer-to-peer state synchronization of text document editing. An example of this that is already in production is Google Docs. When you graduate, you can apply to Google and tell them about your cool project. There are some other companies doing similar things.

You had me thinking of ideas you could use for about 10 minutes before I realized this thread is a year old and this guy prbly graduated already....

>>13628 generally "distributed systems" means knowledge of various different consistency guarantees, working algorithms to guarantee them(such as paxos), etc

>>13123 >where should I start? maybe explore distributed systems to get a sense of where you want to start might i suggest apache nifi, it's pretty nifty otherwise maybe spark, kafka, or inner workings of hdfs