abhimskywalker

Youtube talk from Kailash Nadh, CTO of Zerodha: largest retail stock broker in India) Talk slides here: https://nadh.in/files/indiafoss-2024-dungbeetle.pdf

"The recent change also means you can run multiple LLaMA ./main processes at the same time, and they'll all share the same memory resources." So this could have a main and multiple sub-worker llm processes possibly collaborating while sharing same memory footprint?

Thanks!

This sounds like a great advice! Will pickup a domain and get started. Would probably look around for hardware/robotics and machine learning domain specific projects to get started and learn :)

This sounds great!

What else would you recommend? Rust?

But if I have to work with hardware, I believe even today (and perhaps for many years to come) C/C++ is the way to go right?

Over last 10 years have used Freemind & (last few years) Xmind to kind of make superfast flow charts in the tools dressed as mindmap tools. So i can see the appeal of knoted here for quick jotting down of thought flows.

But have also enjoyed making nice diagrams in mermaidjs especially the sequence diagrams. Prefer markdown to digram for code maintainability. Was recently seeing a youtube talk from C4 author on composable diagrams and agree to a lot of what was said in the talk. Src: https://www.youtube.com/watch?v=Za1-v4Zkq5E

Keyboard to diagram is much faster than mouse based diagram tools and have found them better for long term use cases. Quick diagraming in excalidraw or figjam doesn't hurt though when brainstorming especially with not very technical folks as they can easily contribute.

Have used sequence diagrams for interactions & state diagram for state machines from mermaid for this. https://mermaid-js.github.io/mermaid/#/ System architecture is probably best in PlantUML as already pointed out in other comments. https://real-world-plantuml.com/ is helpful. But mostly it's a lot of manual effort to keep it updated, but the markdown (in git repo) makes it easier I guess...

=> One thing that worked at times was to have new joinees in team update the docs and the diagram based on their understanding + having this system architecture as part of onboarding sort of ensures it is updated on most occasions. :)

Looks nice :) Have been looking for something as simple as Heroku for a while now. This seems closest as of all the ones I have looked at!

It was delightful to watch him create a Wordle clone with a simple html file coupled with one js file in plain sublime text editor and chrome debugger hacking :D

No complex modules installation or initialisation setup. After having seen the JS complexity grow to from simple jquery stuff to download a whole lot of stuff to just get started creating a simple webpage, I personally found this simplicity absolutely delightful and inspiring. Ended up creating a simple world clone in a couple of hours and quite enjoyed the whole simplicity of it all :D

Thanks!

While I would agree with this to get started, but it's always a good idea to be aware of possibilities beforehand and skim though them once so one knows that the solution exists for the problem being faced, what the solution looks like and where to look for it the quickest. So while i know I will forget the exact implementations of 90% all the cool stuff I can do with vim, but I know the possibilities when I have seen them once somewhere.

I remember feeling combinations of motions with edit commands were like the superpowers when I first saw them. I had to look them up again several times whenever I needed them to get it right. But once I got it, suddenly predicting them to edit text on vim transcended into fun territory for me personally.

Though for vim, I have found to adopt only one or two new things a week at max to improve productivity. Anything more and it used to get overwhelming, or i forgot about it anyways...

* mostly keeping screen brightness at ~30%

* f.lux

* pomodro sessions and walk-breaks

* compulsively execute post-work 1hr evening exercise

* at least 6 hour sleep everyday in darkish room.

* Live with hope that these counter most of the sedentary work-lifestyle damage :)

12ish midnight to 6:30ish am

As a technical co-founder, 5-6 days a week if not 7 at times.

Yep. It's almost there. Used to use excel a lot earlier (on Windows) in data science job. Now have excel on Mac, but google spreadsheets mostly gets the job done on most occasions. (Now I am mostly a programmer/product guy) A few problems I have encountered with google spreadsheet: 1. Sharing some stuff with third parties who exclusively use Windows and just want an excel. 2. Speed of working with all the keyboard shortcuts still seems better in desktop excel as compared to google spreadsheet (but I am happy with the progress of how many of those same shortcuts now work on google spreadsheets).

Could you please elaborate a little on "play your cards right"?

Top priority: How to be Happy.

And also some Kubernetes, Docker in more detail, explore rkt and CoreOS, perhaps also get into details of linux kernel and finish a custom build from http://www.linuxfromscratch.org/ Get more depth into system security

If time permits, would love to learn more about Quantum Computing and explore if I can contribute in any way.

He chooses a topic and does a snapstorm (a bunch of continuous 10 second videos) on those topics. Quite nice focused advice and personal observations on a daily basis. Topics range from how to raise money and reaching out to angels/VCs to other things relevant to everyday business processes like marketing, building teams, metrics to focus on, common mistakes, etc. Justin recommended to follow him once earlier this year I think. Checking out their stories almost daily is one of my prime use cases for Snapchat since then. (I am in India and in late 20's so have very few folks around me on Snapchat to do anything else really...) Also starting the day seeing Justin following up with example on his advice of "Fitness is the first step to greatness" has personally helped me get a lot more motivated towards a better lifestyle :)

Thanks! Was not aware if this is even possible. This looks quite awesome!

Front-end: React.js(if fancy interactive stuff needed), Bootstrap with Flexbox for styling Backend: Django, PostgreSQL, Redis, Elasticsearch, Nginx

Haven't used but wish-list to try next: Docker, GraphQL, and probably try hosting it on google cloud platform

If the use-case fits, even wanna try a Firebase-backend only app as well

Nice. This is quite cool. Just saw the linked xkcd comic by just hovering on the links in some of hackernews comments :)

"If you're going to make an irrational decision, do it out of love, not hate." :) This is awesome advice, specially for seemingly crucial decision points.

Thanks for the references! The idea to follow the texts and readings from one of these courses seem to be a good idea. Will read up on that.

Thanks! Will look into the CFA level 1 material as well. I have some basic hold on accounting concepts, have prepared financial projections for large projects in the past and also can read and understand most parts of company annual reports and financials. Also could you please extrapolate a little on "materials from the exchanges and articles", does it mean routine publications/periodicals and white-papers sort of material? It would be really great if you could just point me to any one of the example source of these.

I am just coming out of previous full-stack dev role. Wanted to spend some time dabbling in a few ideas right now in this space of finance/fintech. So I wanted to develop solid understanding of this field right now first (which has also been a latent personal desire/goal now for a quite some time) as well as experiment with a few ideas by myself first for some time...

My goal was to cover some of the main fundamentals of finance to get a good all-round understanding along with eventually specializing towards practical application based finance... mostly stuff like designing & implementing algorithmic trading strategies, building securities exchanges/marketplaces on my own (maybe in python for now), maybe even design and implement more complex financial instruments & related contracts...

Was looking to get to learn both the theory and real-life implementations of this financial systems & processes. Hence wanted to start with Corporate Finance as it looked the most generic and seemed more like broad revision of stuff i already learned in a lot of econ classes and some management electives I took in college, with some additional knowledge to build good skeletal/context structure of the field and various agents involved, which I can later fill in with and place new concepts i would learn in the two other books. Then next logical step seemed like getting to know ins and out of the financial instruments of more complex nature hence Hull seemed good to start with. (Also the fact that I read first few chapters in college and quite liked them then...) And then Security Analysis seemed like a good way to look at all of the above learnt info in more analytical way specifically from POV of an investor.

Also as suggested in the other comment "Trading and Exchanges" by Harris is definitely going on this list of books to finish for me to get better practical implementation understanding.

As of now quant side of finance seems most interesting, but again haven't explored all other roles/areas fully so can't say for sure as yet where would I like to work. A finance nOOB question... is there something like a full-stack finance engineer? Closest seems to be a quant guy...

Thanks! The "Max Dama on Automated Trading" looks awesome!

"Trading and Exchanges" by Harris also looks very helpful towards actual implementations. Will order it right away. Thanks a lot!

What you want sounds somewhat like WeChat in China...

Ref: http://a16z.com/2015/08/06/wechat-china-mobile-first/

Yes. Mostly soundtracks from some awesome movies I can relate to. Also anime soundtracks are quite awesome. Sometimes the memory connection to some of those great moments from movies/anime/tv series help get me into the zone and give a bit of adrenaline to do get shit done fast.

So consequently there would be a lot of startup failures as well... One more interesting scenario that can result could be if there will be an investor fatigue as well similar to the customer fatigue of investing their time & money in these startups...? As this could potentially make finding initial capital harder for startup teams which could potentially be solving similar problems with better execution strategy but in the same market...