Open sourcing is stoopid

Discussion in 'App Development' started by Aquarians, Jan 10, 2021 at 4:38 AM.

  1. Well perhaps it made sense in the days of Linus Torvalds and later SourceForge which at peak popularity hosted some 200,000 projects. Today GitHub hosts 100,000,000 projects, so your identity was diluted a further 500 times! Think of that for inflation, in the days of SourceForge your identity if nothing else was "one in 200,000" now it's "one in those 500 buckets of those one in 200,000". Fuck this shit!

    Having any sort of valuable work open sourced is totally pointless today because:
    1) You only expose to competition what tiny edge you may have. And don't think of competition as "some other developer like me", think the countless faceless corporations who don't even send an automated reply that you failed their automated programming tests that take hard months to train for and pit a 20 years senior in the field on the same level as a college grad.
    2) Apart possibly getting copied and stolen, noone cares. It won't make absolutely no difference when you apply to the countless faceless corporations, you still have to take the same darn automated leetcode tests like stupid college grads who likely had more time and energy to put into this crap. Afterall they only have one thing to do.

    So whatever you code, if it has any sort of value, keep it to yourself.
     
    931, stochastix and zenlot like this.
  2. Ouch, that sounded depressed and frustrated. I hope you are OK.

     
    d08 likes this.
  3. Also, OpenAI has learned code from GitHub and can now create scripts based on natural language input. I think this will curate the programming industry or force the bottom tier to improve. This has been happening with photography, too, with phones capable of better and better snapshots.
     
  4. VicBee

    VicBee

    I'd say that has happened to many creative and less creative industries, thanks to the internet. You need a logo? Jingle? Translation? Standard legal document? It's all now available on the web for a 10th of what is used to cost to get these done.
     
    931 and george_the_second like this.
  5. I'm fine, just ranting :p
     
  6. Simples

    Simples

    On the contrary. Closed source code can never be taken over by someone else, at least you need to give it away at some point. Never has the bar been so low to find code, glue it together and produce something tangible in shortest amount of time. Or never has so much code been provided for free, to learn and improve your skills on. The simple truth is you never had "an edge" producing and testing code. It's either your own business, hobby or you have managers and everyone breathing down your neck about the slightest things they can't figure out themselves. Your business and salary are kind of edges, while your edge in hobbies are just to have fun! The main problem with programmers is that sacrosanct worship of programming code, which is devalued by everyone else. Just look at every passing fad, and realize it's all BS, though your long experience from it may not be.

    If a company values 20-somethings working overtime for free: Either you need to find where that puts you in that picture, maybe grind less, or find a worthy pursuit where your skills are not covered by inexperience.
     
    stochastix likes this.
  7. I'll probably open source a tiny part of my strategy backtester, just to have something to show to the corporate clueless if necessary.

    gervais.jpg

    Anything beyond toy level / demonstrative purposes would be a net loss for myself.
     
    Trader200K likes this.
  8. stochastix

    stochastix

    as an open-source pioneer , I concur . I made 34k on the VA Linux ipo shares i was given as a gift due to my work on Debian and other things. now I do not recommend anyone code for free or employment now. I only code for myself now
     
    931 and Aquarians like this.
  9. d08

    d08

    What makes your backtester unique or interesting? Is it the fastest, can it smartly reuse data from previous backtests to save time? Does it process data not found elsewhere?

    Yeah, I don't want to open-source mine either because it's in a way unique but also incomplete. So while for me it's of great value, I doubt it would get attention from more than a dozen people.

    In the end open-source value is in distribution of labor. It's a massive undertaking unless you just execute from terminal and use basic serial backtesting methods.
     
    DiceAreCast likes this.
  10. Covers options, delta-hedging, pricing model customization... Ida know. Point is the customer (end user) doesn't care how stuff is built, only about results: profitable and reliable trading strategies.

    So those whose money I'm after won't download my source code and build it themselves then spend years researching something that might work (most probably not). They will either buy something that works NOW or move along, bottom line the code is irrelevant to them.

    The code can only be relevant to my competition. And don't wanna make it the least easier to them.