Ask HN: Long time tech folks, what's your story?

35 years.<p>Tech support, programmer, systems administrator, occasional supervisor.<p>DBA, sysadm, programmer.<p>How has it not changed? The first customer site I went to on my own had a minicomputer with a storage component consisting of a 25-MB drive and an 8&quot; floppy that held 1.x MB. Shortly before I was planning to head to the airport, there was a &quot;spang&quot; and the system panicked. The belt drive of the 25 MB drive had fallen off the spindle. That unit was the size of crib (without legs), and who knows what it cost? My first semi-real programming was in MV&#x2F;Eclipse assembler.<p>I don&#x27;t know how I&#x27;d change tech today.<p>The negatives seem to me those that have been around a long time. Fifty-five years ago, I heard my father grumbling about a professional association that had given Union Carbide an award, when everyone knew how miserably it treated its technical staff. Union Carbide then, IBM, etc. now. Positives? You can start developing on amazingly cheap hardware with freely downloadable software, you can deploy applications to amazing cheap VMs. The resources are amazing.<p>Advice? Go back to your old code (last week&#x27;s, last month&#x27;s), see what you did right, what you did wrong, and what you wished you had documented. Learn different sorts of languages. Use version control.<p>My two cents, anyway.

How has tech changed since you started?<p>Technology is more mainstream and affordable. The concepts are the same but renamed and reimplemented. Before the internet the airlines were able to conduct transactions worldwide and in real time with green screen technology. The only real innovation the last 20 years is the widespread adoption of smartphones with LTE and now 5G. The entire world now is on the internet(s) and communicating with each other.<p>If you could change tech today, what would you do?<p>I wish more of the tech infrastructure we take for granted is in the hands of nonprofit foundations whose goal is the common good.<p>What are some negatives and positives of the current tech landscape?<p>Big tech has captured market share and is behaving like every other big corporation and has little regard for their customers or their employees or their communities.<p>What advice do you have to give to others?<p>Think for yourself. Master your craft.

I have been programming since I was 10, so 30 years. I feel truly blessed to get to write code, nothing makes me feel more like a little kid.<p>Early as a kid on it was AppleBasic, then QuickBasic, then TurboPascal in high school. I loved the computer lab from my magnet school where we would one up each other. I built an American football simulator using player stays, several games, and a simple 3D engine. I started college in 1999. We did C and C++ for the most part. Then I had about a 10 year stint doing embedded or Win32 development. Then I switched gears and now focus exclusively on building search engines.<p>The world has changed a ton:<p>- certain books that existed in the 90s and 2000s are replaced by the internet.<p>- because of open source and the internet, and less reliability on mainstream tech press, it’s easier for more obscure tech stacks to blossom. Nineties and Early 2000s were dominated by C, C++, Java...<p>- in the 90s and 2000s almost all consumer facing native software dev would be Windows. Some web app dev existed, done in Perl, Java, and gradually more PHP<p>- remote work was not a thing. It was rare to have VPN access. Teams were heavily collocated. Outsourcing was brand new and just starting.<p>- there were times where <i>it was hard to get a job as a software dev</i> if you can believe it! After the dot com bust in particular.<p>- version control was relatively new and primitive, with tools like visual source safe and CVS. Gradually Subversions became a cool thing in the late 2000s<p>- the posix command line was not the universal dev environment we assume nowadays. Instead it was often Visual Studio and a pile of bat files :)<p>- software was not well understood by management, and treated akin to how machine learning is treated today: something mystical but the magic fairy dust we can sprinkle on everything<p>About the current tech landscape? It’s very broad and diverse to say much about it. There’s much more of a long tail of low quality and low cost software development now. There’s many more jobs where coding is just part of the job. There’s few places that can do software engineering really well, and these places have increasing social power.<p>On the plus side, the number of options and tools we have is exciting. We never have to worry about being out of interesting work. We have good middle to upper middle class careers. In some places software and the tech industry is having a positive impact<p>If I could change anything about today it would be to increase tech and data literacy across all disciplines.<p>Advice for others: don’t just learn the surface stuff. Get deep to how things really work. Be able to write and speak well. Work really hard on your empathy for customers and colleagues. Find and work on the hardest problems. Pick a niche. “software development” is so broad it now describes a significant percentage of the economy. This is why I focus now on search engines.

I&#x27;ve been doing electronics and computers since 1980. I&#x27;ve actual built computers from scratch (you know, read the data sheets, chose clock generators, RAM and EPROM chips), add 7400 series TTL to select peripherals.<p>I&#x27;ve built a lot of things, including a detector to help align building-to-building Infrared laser links, and a box that connected to a phone line, and a number of voltages... you&#x27;d call the number, and it told you (via a speech chip) what the voltages are, twice, then hung up.<p>I&#x27;ve been an electronics technician, repaired Magnetic Amplifiers (a cool old technology from the 1960s), all the way up to 480v, 1000 AMP, 3 phase SCR packs. Oh.. and Cesium Beam Atomic Clocks. ;-)<p>I&#x27;ve programmed 6502, Z80, 6809, 8088 in Assembler, Basic, Fortran, Pascal, PL&#x2F;N, Turbo Pascal, Delphi, Visual Basic, Python, C, C++.<p>It is my opinion that VB6 &#x2F; Borland Delphi was the peak of user accessible programming.<p>I&#x27;ve used CP&#x2F;M, MS-DOS, Windows, Multilink, Unix, Linux, RSTS&#x2F;E, VAX&#x2F;VMS.<p>I&#x27;ve networked with RS-232, DSL, T1, Cable Modems, Arcnet, ThinNet, 10,100,1000 Mb Ethernet, and Wifi.<p>I&#x27;ve used BBSs, AOL, IRC, FTP, Gopher, WWW, Usenet, RSS.<p>I&#x27;ve seen almost everything once... ask me anything.<p>[Edit] To answer the questions at the top<p>1. 40+ years<p>2. User, Programmer, ISP Sysadmin&#x2F;Sole Phone Support 200+ Users, Electronics Technician&#x2F;Engineer, Developer&#x2F;Field Support for an Inspection System, Manager of IS for a Marketing Company 15 years, and I&#x27;ve made gears for 5 years.<p>3. I&#x27;m wondering how I&#x27;m going to get a job... they&#x27;ve always fallen into my lap... I&#x27;ve never had to look before.<p>4. Computers are thousands of times faster, with millions of times the storage, and yet cheaper. You don&#x27;t have to write our own Basic I&#x2F;O System in order to get CP&#x2F;M to boot... stuff just works. Persistent Internet is a thing... it used to be dial-up to CBBS, FidoNet, Usenet.<p>GIT is one of the greatest inventions of mankind. I used to make a ZIP file of my source code before I attempted changes... I still have stacks of floppies filled with almost identical ZIP files. GIT is about the only thing that has really improved in the software side of things.<p>5. I&#x27;d go back to the old way of social networking, blogs and replies, all powered by RSS.<p>6. Our current crop of Operating Systems (Linux, Windows, Mac) are all unsuitable for a world of persistent internet. They all default to the wrong thing... trusting programs the user runs with every resource the user is allowed to access. This has resulted in the chaos that is forced updates, virii, hacks every day. The opposite is to trust nothing, and only supply a process with the I&#x2F;O and CPU time the user wishes to give it.. we&#x27;re still a few years out. I&#x27;ve been making this prediction since about 2000.<p>7. If you&#x27;ve got an idea for something, just do it... don&#x27;t talk about it or share it until AFTER it is done. If you talk first, your brain will get bored with it, before you&#x27;ve even started.<p>8. Atomic clocks are amazing things, once you get them fixed... you can put a pair of them into the X&#x2F;Y channels of a scope, and the circle will just sit there, not going out of phase, for hours.