I understand the production value of it. But as a manager, when someone brings a solution to a simple problem that seems overly complicated, my first question is, what benefit does it bring vs a more simple solution? In his case, he was just writing some unit tests to validate our module. He could not tell me why it was better. I did not see the immediate benefit of TS, and in fact, it caused such a headache on its own that I’m a little suspicious it may have caused the project to nearly fail on its own (the external team we were working with couldn’t easily work with our tests because they didnt know TS).
We were writing a nodejs module in C++ (because n-api and NaN libs are in c++) but you call the functions in node. It probably could have been written in C++ but since you called functions in node and all I had was an intern and a senior dev (the senior dev was architecting the thing) I decided to write the tests at that layer. Was kind of the only choice. I relented to the intern because arguing with him was obnoxious, that was my mistake.
The problem wasn’t so much that the tests did not do what they were supposed to, but as requirements changed, adding to the Test structure became increasingly difficult. I don’t doubt this had to do a lot with the lack of experience of the dev but a little of it seems to be the complexity of typescript itself. I didn’t personally do much diving into the code so I can’t say for certain, it was just a suspicion after looking at it a few times.
So have to laugh at this one. The product I work on at my giant fortune 100 is an in house business intelligence tool. We have a slack support channel that is frequently used. Today a user of our product came in complaining about his table taking too long when he asks for it, usually 30 seconds. After investigation, it turns out that it contains… 1 billion rows.
It doesn’t fail, but it seems like sometimes ssh’ing in causes Apache to crap out because of memory issues. I don’t have much running on this site and I wouldn’t think it would be a problem. I have 1G of RAM. When I ssh into the site on one shell and watch top on another shell I can see that at a couple points a ton of bash processes spawn and then die. And briefly the memory free goes quite low and the “KiB Swap” will drop to “0 free”. After the connection is made it’ll be fine and the numbers will return to pretty much what they are with just the one shell running top.
Is this normal? Is there something I can do other than getting a plan with more RAM? Why does bash start so many processes (hard to tell, but it looks like something like 20 or 30 processes)?
I could use some programmering advice. I am somewhat seriously considered quitting my current job sometime in the next 6 months and trying to start a software business that would develop some professional software tools that there is a glaring need for in my industry. There are a few concepts, ranging in ambition from simple-enough-to-be-a-free-utility (but needs to be wrapped in a friendly and smart UI to make it palatable/learnable and sold to the IT department in secure package before people can use it) to a much more ambitious concept to make a modular architecture for creating semi-custom software solutions for a particular kind of task. The idea is not to become a tech billionaire or rake in a bunch of VC cash, but rather to have a flexibly schedulable job that can, hopefully, eventually, replace a meaningful percentage of the income I’d be walking away from in my current line of work.
To that end, I would want to do as much of the coding work as possible myself, and limit myself to freelancing out the things that require finesse (UI polish, security and network stuff, maybe others?). The big problem with that is that I don’t have any formal software education beyond AP Computer Science or any real experience, other than having spent the last 4 months of nights and weekends working on a prototype of my program. As it stands now, I would say I have a reasonably solid understanding of python fundamentals. I understand HTML/CSS 101, and am working on learning JavaScript now. My current project is putting together a website that will generate illustrated word search puzzles, which I’ll share when it’s functional (hopefully by this weekend). My questions:
Is this completely insane? There’s zero risk of ending up hungry or on the street even if I spend two years working on this and it brings in zero dollars. The big downside is likely torching a fairly guaranteed 1%er+ career path, but I’m OK with that as long as I’m not chasing a complete fantasy.
Any suggestions on what I should focus on learning? My current plan is: (a) finish my toy website with an unnecessary amount of vanilla-JS-powered interactivity, (b) reimplement toy website with an extremely unnecessary framework (thinking Vue.js based on not too much), (c) start trying to build a prototype of the real application. Possibly there needs to be some kind of stepping stone between (b) and (c)? Any resources people have found especially helpful to learning this stuff? I’ve gotten a lot from MDN and javascript.info.
If anyone has any experiences (positive or negative) to share about pursuing something similar, it would be great to hear them.
Is it excessively arrogant to believe that if my business is a complete failure and I end up having irreparably torched my existing career, that it wouldn’t be too hard to line up a 100k-earning job as a software developer somewhere (in Atlanta) as a fallback plan?
A little more than a decade ago (god I’m old), a few POGgers put together a werewolf-oriented site (usgamers.net RIP) that had a ton of amazing functionality. My recollection is a bit hazy, but you had drop-down menus to vote for people when writing posts, full automodding with reveals, night actions were all run automatically through the PM interface, dead people were automatically not allowed to post and got a separate, hidden graveyard thread. It was wonderful. If the PTPB are OK with it, probably (?) you could set up a subforum that has similar (but better!) functionality built in (live vote count in the sidebar??).
You’re not crazy, but it probably won’t work out due to reasons that have nothing to do with your actual software engineering skills.
Skip vanilla JS and just do your UI in either Vue or React (preferably React). No one does vanilla JS dom manipulation nonsense. Use create-react-app.
Everyone just uses MDN for everything. I can’t imagine how many times I’ve googled “mdn splice”.
If your idea doesn’t pan out but you’ve gained a lot of skills in web dev it would be easy to get a 100k+ job in any major city if you can sell yourself a bit.
Wouldn’t bother with the werewolf forum stuff tbh.
I’ll tell you a story of how I spent 5 years programming on a project that I think had pretty good commercial potential. Way back in the late 90’s when Realtor.com was just coming up, I built real estate web software that would do all sorts of stuff that was ahead of the market. Realtors could put up websites that they could customize themselves in a lot of ways, we imported data from the MSL that their customers could search on their site, they could integrate with loan and escrow people to provide tracking for their customers, share documents, basically everything I could think of. We sold some websites for a while and I made a meager living, but imo we failed because sales and marketing was not even close to half-assed, more like 1/10th assed.
Point of the story is, I had a partner who owned 1/3rd and he did some of the work (he was a unix admin - still is, but like upper management unix/security admin). He didn’t quit his regular job, but I did. The venture didn’t lose money - it made a bit of money - but still it was more or less a failure and it left me broke and his career uninterrupted and he’s doing very well.
I think of projects like that all the time and never get much farther than the beginning phase. Kudos to you for doing that much with no experience. However, your overall goal seems very ambitious. I wouldn’t want to discourage you and am generally a “go for it” person, however, but given that you seem to have an already great career, I’m a little confused as to why you want to do this. If you are generally unsatisfied with your career and genuinely enjoy this stuff (I assume you do, or you wouldnt have gotten this far) it seems like a no brainer to just try it. Running a business will give you a lot more challenges than just coding, however - it’s a full time job on its own. Unless you plan on being a 1 man army doing 100 hour weeks…
I am too junior to speak as to whether the experience would be enough to land a job, but I wouldn’t be surprised.
It’s ambitious but not completely insane. I would guess the most likely outcome is that the startup idea never really pans out. But that’s just because most don’t. But the reason it’s not insane is also because of (4).
+1 on learning some framework. I don’t have a strong opinion on which one it ought to be. +1 on MDN also, but you might decide that the “right” framework is the one that seems to have the most useful documentation. You’d have to spend some time evaluating that. I think the nice thing about having a startup project is that the business goals can kind of dictate what you end up having to learn, and there will be plenty to learn. So it’s OK to just focus on figuring out how to get to the next milestone that way.
Speaking as someone who has made a career in software development with no formal credentials, I think it’s possible to do. I think there are definitely some difficulties in getting people who don’t already know you to give you a chance, and basically every job I’ve had I got because I happened to have impressed someone who already worked there through some outside project. So I’m not sure it’s easy, but I think if you develop the skills and can demonstrate them then it should be possible.
Yeah, but you don’t have to quit until you have something close to a product that you can at least run by prospective customers or maybe not even until there’s some income.
