What project should I finish?

Discussion in 'App Development' started by kingquant, Aug 2, 2020 at 4:51 AM.

  1. kingquant

    kingquant

    I have a few infrastructure apps that I have started but not finished (ie 90-95%+ complete - I get bored easily)! Such as...

    1. Market and transaction data database (rdbms support ms sql server, oracle, mysql, firebird, postgres..)
    2. REST API and Market Data Downloader for above rdbms
    3. Trading Decision Support System (ie issues and automatically journals trading advisory instruction based on your trading algorithm/strategy building an extra level of discipline and rigor into your trading)
    4. Visual Trade Order Management software - an agile trade management tool designed to help coordinate and visualize trades, manage trades-in-progress, maximize trade flow and reduce risk using coloured building blocks.
    5. Trading Dashboard
    6. Trade book-keeping app for the above transaction database

    I was looking to integrate all the above into one app but want to get stuff finished now. Which ones if any of the above would be the most useful to get out first for others to use? I can add a bit more info if needed but have no documentation yet (one of the things that need finishing! ).
     
    PlatformFX likes this.
  2. fan27

    fan27

    Unless you have someone ready to pay you money for one of your completed projects, you should finish what you will get the most benefit from. Unless I am contracted to build a specific solution, my rule is to always build software that I need. That way, it will always be a worth while en-devour as I have solved my problem even if no one else is interested in the software.
     
    guru, d08, qlai and 1 other person like this.
  3. DaveV

    DaveV

    None. Advertise your services on a site such as Upwork and build software for which you are getting paid. As fan27 alluded, unless you are writing trading software that is actually being used by either yourself or someone else, it is almost impossible to come up a with truly useful system.
     
    qlai and fan27 like this.
  4. kingquant

    kingquant

    I agree it is difficult to come up with useful software unless it is being used by others hence the question. They are all working prototypes and as such are useful for my own trading but no doubt would be made a lot more useful if used by others. Therefore was looking to find out if I should spend any more time on any of them in order to get them "production" ready for others to use. I also agree if it is useful software then another gauge is whether people are willing to pay for it but am not interested in getting paid for it by forum members or being "employed" by someone on Upwork to produce a bespoke solution for their own individual use and possibly of no use to anyone else including me. In terms of return on investment or benefit they are trading infrastructure tools and only indirectly help to generate alpha which is what traders should be focusing their time on and not for the most part side projects such as these. As a computer/math geek I enjoy solving problems, research and having fun with trading - professionally I have experience of the investment banking sector including trading system development/implementation. The work to date has been beneficial for my trading as the software has proved to be useful for me but no doubt could be made better.....
     
  5. traider

    traider

    All those pieces are overkill even for a small fund.
    For example 1. and 2. is already taken care of by quantconnect
     
  6. kingquant

    kingquant

    Probably overkill for most retail traders but for professional investment managers, small funds and even professional independent traders lower end (retail level) systems do not have the industrial strength necessary. I would not accept free data (or even data from data vendors) from any source without due diligence and processing. Even small funds should be using multiple data sources to reduce risk and either roll their own bespoke infrastructure or purchase systems offering the right set of products for their stage of development to enable them to grow with minimal expense.
     
    Last edited: Aug 4, 2020 at 8:38 AM
  7. When you work on a personal project, and there is no one to make you finish what you started, everything easily falls apart. If you don't accomplish your task, no one will know about it. Life will take its course, you will still have the same job, and no one will treat you worse. You'll know one thing and you'll perceive yourself differently. Maybe you'll start to think you're not a disciplined enough person to complete a task. You will think that some people are born with the gift of completing things, but you aren't one of them.
    But the fact is, you're able to complete whatever you're doing. I think the biggest problem most people have is that they just choose the wrong strategy.
    Personally, I don't think you should complete your projects. You've learned from those projects, gained invaluable experience. And now you can start a new project with this knowledge that will be better than others and you will want to finish it 100%. Personally, when I can't finish something, I use two tips that always help me.
    -Make a concrete plan. Define its phases and the time required to implement each. By realistically evaluating your abilities, you will maintain internal balance as you solve the task at hand.
    -Moving forward to the goal gradually. The pace of work slows down when the enthusiasm with which you start a new business comes to an end. A calendar with marked end dates for each stage will help you use your working time more intelligently. By checking with it, you can adjust the rhythm of work.
     
    DaveV likes this.
  8. kingquant

    kingquant

    Agreed the last mile is the hardest and I was trying to get a way to prioritise the order to finish the software by seeing whether it would be useful for others. I always finish the stuff I start and it is all fully working software as is - just needs tidying up in places. Decided therefore to integrate everything into a single application and will have it done by the end of this month.