Ask HN: What's your earliest web development story? | Hacker News Reader

Ask HN: What's your earliest web development story?

9 pointssmoyer6 comments

This post was inspired by a small comment thread on the "May 2014 Freelancer / Seeking Freelancer" post. In response to a comment I made about an early web adopter [2], rueven [3] responded to my comment with the following message [4]:

    The story is that a friend of mine at MIT came back to
    the student newspaper and said, "I just went to a lecture
    given by this guy Tim Berners-Lee, describing this thing
    he's calling the World-Wide Web. We should get on this!"

    So we set up a Web server with some of the newspaper's
    archives, and then e-mailed TimBL and said, "Hey, we set
    up a Web site. What do you think?" His response was
    something like, "That's great! I'll put it on my list of
    all of the Web sites in the world." And that's how my
    career got started; turns out that we were one of the
    first 100 sites in the world.

I love hearing stories like this and am especially interested in those who created web sites and applications before there was anything but the HTTP and HTML protocols. In those early days, I know we did everything ourselves ... are there any other early web adopters who are willing to share their stories? I hope reuven will expand on his early experiences and I'll add my story to the comments below. I look forward to reading other stories about our shared history!

[1] https://news.ycombinator.com/item?id=7679422

[2] https://news.ycombinator.com/item?id=7681366

[3] https://news.ycombinator.com/user?id=reuven

[4] https://news.ycombinator.com/item?id=7681874

Hey, just noticed this thread!

There is a lot to say, and I"m not sure where to begin. But I was (as I wrote in my original story) one of the editors of the MIT student newspaper. We had a lot of fun, and some great computers. And every so often, someone would come by and ask to search through our archives.

Now, our "great computers" were Atex systems for typesetting and layout, based (I think) on PDPs of some sort. So to search through the archives meant loading a "disk pack," which is a laughably large hard disk that you had to put into the machine or take off of it. Just searching through the 100+ years of the newspaper's archives was a huge task.

At some point, we figured out how to get the information from Atex onto a more modern computer. And I set up the index, for my undergrad thesis, to be accessible on the Internet (not the Web). But the interface was clunky, and while it was better than before, the search was far from perfect.

So it was really great timing when Jeremy Hylton (now of Python/Google) went to that talk. We turned off my thesis software, turned on a Web site, and set up a simple search to work on our Web site. The thing is, while this was a great feature, there was one missing component -- users. So we actually took out large "house ads" (i.e., advertisements that the newspaper takes out itself, rather than selling to advertisers) telling people how to get a browser, what a URL was, how to enter the URL into a browser... all so that they could search the newspaper archives.

This was in early 1993, and I was still living in Boston at the time. A few months later, I went to a computer show, in which someone was showing off Web authoring tools. He asked who had a Web site, and I was the only one to raise my hand. I think that it just started at that point to become known among the technological elites, but the "elite" part meant that you had ever browsed the Web, not that you created sites yourself.

Another quick anecdote: When we set up the newspaper's Web site, we wanted people to be able to contact us. HTML forms had just been invented, as had the CGI protocol. So I hacked together a little program that took the contents of a form and sent it, via e-mail, to our administrators. I called it form-mail, and it was written in Perl, and all was great... until a few years later, when someone named "Matt" took it, removed my name, added his, took out all of the security features I had put in, and then made it famous as part of "Matt's Script Archive," which was the bane of Perl programmers' existence for years.

I seem to have lost the e-mail exchange that I had with him at the time, but when I told him that you cannot just remove someone else's name from GPL'ed code, he claimed to be scared, and then never did anything about it. Which is just as well, given that he so sullied the name of my program that I'm happy to give him the credit for all subsequent versions.

One last point: There wasn't any CGI.pm, or other Perl module to do Web stuff, back then. So I wrote my own code to decode data from POST-submitted forms. If you ever see the Perl code "for $pair in @pairs", then it's almost certainly copied from the code that I wrote in that first form-mail program, oh-so-many years ago.

Fun stuff, indeed!

Thanks for expanding your story ... I wish this thread had become popular enough to contain another couple dozen stories (histories). Sometimes posting here feels like buying a lottery ticket.

We were building distance learning systems and the classroom controllers were low powered PC hardware running DOS. These systems were connected to the system controller by Lantastic networking (we had no TCP/IP stack on the classroom controllers) and yet we needed both network monitoring (without SNMP) and status information related to the classroom hardware (tied to the classroom controller via RS-485) as well as the classroom's state (was it being used, which devices were powered on, etc.)

We ended up writing a device driver for the DOS machines that could message other Lantastic nodes with the system's current status. Once we had the status stored back at the system controller, we needed a way to allow administrators, operators and teachers to easily see the overall status ... my initial choice was to provide the data via a web server since the system controller had a TCP/IP stack and was connected to "the outside world".

In 1995, what do you use to write a web server that has no static content? Tcl/Tk was one of the first scripting languages to include the ability to open a "server socket" and it provided lots of libraries that made it easy to template text (in this case text/html). We ended up with one summary page that displayed the status of the entire system and a detail page for each classroom. There were no graphics, system status was shown via text and colored backgrounds using in-line styles, and the server automatically redirected you to the index if you asked for a non-existent page.

Sadly, the system never made it to production as we found we also needed a web-based method for scheduling classes, provisioning users, etc. This system was written in (pre-Java) ColdFusion and the monitoring functions were migrated into that system.

In 1995, I worked on getting a large piece of enterprise software to send output to browsers in HTML format. The product took off pretty quickly and is still being used today by big companies. HTML had only very basic markup (no CSS yet) and the only method for talking to a web server was CGI programs. But you could do tables, and those were adequate for generating business reports. Java 1.0 had just been released.

This was long before the web had become a popular culture phenomenon: most people outside of academia didn't know what the internet was, and search engines were in their infancy (Alta Vista was founded that year). Windows 3.1 did not have a built-in TCP/IP stack, so you had to be a geek just to figure out how to connect your home machine to an ISP via a dial-up modem (Windows 95 made it much easier) and install a browser. However, the web was starting to catch on in the business world as companies started putting useful information on the web and their intranets. It was an exciting time to be working with that technology since the potential was largely untapped.

Ahh, the good old days. In 1997/98 I created a website for my hometown with photos, cinema schedule and the like. I "designed" it Coral Draw und Adobe GoLive. It was a time when you were hip if you used tables for the layout. And if you used Server Side Includes, you were quite the hacker.

My first real web development project was an appointment scheduler. In 2000 a potential client wanted one but we couldn't license one from the only company that offered that kind of software at the time in Germany. Thus I taught myself PHP and MySQL and within a week I had a working prototype. Users could reserve appointments and admins could manage them.

The deal with the potential client fell through, though, but I continued working on the code. It was hard going because back in the day, there were no PHP frameworks (that I knew of) and no CSS frameworks.

It's been moderately successful and I still have happy customers from that time. I only wish I had done early on what I did two years ago: offer it as a hosted service (or SaaS, as it is now called).

In 1994 or 95 my elementary school worked with Academic Resource Center (ARC) which was a local elective based school that would host different schools each day for special courses. This was a "gifted" thing.

One class was to make a website, which I didn't know the definition of at the time (being about 8 or 9 and living in a home without a computer).

Hardware:

* Macintosh LCs

* Macintosh 512k

Software:

* Lynx on the classic Macs

* Netscape on the LCs

* Pico for text editing

* Pine for email

* Fetch for uploading

We learned basic HTML which was enough for me to keep pushing toward development once my family finally bought a computer.

When I realized that I wasn't going back to college to finish because my business was doing so well I thought back to ARC and hunted down one of my old teachers to try to make a donation. I couldn't find the one who taught HTML so I reached out to the lady that taught us Logo Writer, but ARC had long been shut down so no dice.

Reply on news.ycombinator.com