It’s 1997 and you want to build a website
thehistoryoftheweb.com
thehistoryoftheweb.com
This article is missing two important ways to learn HTML back in the day:
1) Read the source. Websites were simple. There was no JavaScript or CSS. Everything was right there in one file. If you saw a site you liked, just take a copy of the source and tweak it locally to understand how it works.
2) GUI tools. There was Frontpage and HoTMetaL and some others. These were actually pretty decent because there wasn't so much you could do with HTML. You could drag'n'drop your content, then clean up manually if you wanted.
Here's a live example of a 1995 website: http://www.os2ezine.com/v1n1/
Look at the source. It's easy to understand the basics of HTML just by comparing the source and output. Deeply nested divs and tables were not a thing yet.
Yeah here we are, Netscape Composer, released in 1997 - https://en.wikipedia.org/wiki/Netscape_Composer
I think I built my first web page with that... LOL. Long time ago now :)
Soon the even bigger revelation came: It's all just plain text!
Also: I don't remember anyone using books. The web was much smaller back than, but if it has had anything in abundance it was resources for learning how to make websites.
An artifact of its time. If the Web was designed today it would certainly not have viewable source.
Almost 30 years have passed and this website still works and renders perfectly in modern browsers! Will we be able to say the same about current websites 20 years from now?
Maybe some browser specific features but only very few websites use those.
- familiar font on grey background is quite readable
- blue underlined link tell you what to click on
- link turning violet to indicate you've visited it already? Genius!
Isn't it funny how these early sites managed to be completely functional and useful in their plain and unpretentious way without pulling down gigabytes of cruft designed primarily to make it look like you aren't looking at a website. Given the choice between the 2022 and 1993 presentations of IMDB, I'd take the earlier one in a heartbeat.
Fired up opera, which still had cached websites in the tabs - most of them long gone. Funny to see 2006-era Facebook in the flesh.
Tried to fire up a some news websites, and ended up with BSOD rather quickly. Chewed right through the paltry memory.
edit: I guess it is possible to browse some modern websites by disabling javascript, media downloads, and what not. But it's probably going to be a limited experience as far as functionality goes.
I personally think it could be getting better (at least in the realm of Government websites)
The UK Gov Design System for example:
https://design-system.service.gov.uk/
Clean, accessible and functional first.
That's why I included right-click-protection scripts on all my pages! Couldn't have my friends steal my cool gifs (that I'd shamelessly stolen somewhere else) or copy my cool menu hover effect or whatever!
Of course all this was futile, but we weren't knowledgeable enough to know better, heh. At some point, I also had a script that would "encrypt" my HTML. Looking back, it was just base64 encoding the source, and then through JS converting that string to html and inserting into the body dom element.
telnet your.host.domain 80
GET / HTTP/1.0
and copy and paste from the scrollback buffer of the terminal. Chances were that the code was handcrafted and easily readable.That would work unless you were in the vast unlucky majority with Windows and no telnet or no access to a Linux / Unix server.
Anyway, I did it plenty of times. Then wget and curl.
And if you want to be really specific, Hyperterminal could to telnet.
Not mentioning the GUI tools was a pretty poor omission: Frontpage was everywhere back then. I personally hand-coded but had to use it in my first professional role in '98.
Even the books were wrong. I had (it might still be in a box at my parents) a 1st Ed. copy of HTML: The Definitive Guide, later followed by other O'Reilly books: CSS Def. Guide and one I wore to death as it was the MDN of the day: the DHTML Def. Guide.
That site is awesome; I had to look and yep, allcaps!
Also I once found this humble getting started guide for FPE, miraculously still online with screenshots: http://www.iwaynet.net/support/Web/FrontPage/fpe_create.html
I knew one dude who ran a local ISP and connected to it using a 14.4 baud modem (he might have ran the isp through his house). He also had a small web server where I can host my website.
I had all sorts of fancy stuff like marquees and blinking tags. I think I even used a table to lay out certain graphics and text. Those were good times.
Yes, but also in the following years, js and css were simple too. I learned js and css (well html as well I guess, but that took like... a day) initally by reading website source, before minification and obfuscation, webpack, SASS, etc.
It's a beautiful thing to use a web application where all the internals are visible and tweakable. These days I try to develop with as little obfuscation as possible, but it sure is going against the grain. See https://dev.to/open-wc/developing-without-a-build-1-introduc... and Snowpack.
Yes! This was the big one. I got to college in late ‘95 and almost immediately someone in my dorm showed me how to both code HTML tags and look at page source in Netscape. That began an iterative feedback loop of learning how to make things based on what I saw already out there — and looking at personal home pages (at my school or not) was a great creative playground. By ‘97 I was hand-coding pages for small businesses in the school incubator, several departments, and my own page. I don’t remember getting a hold of many books for a couple more years. And the only GUI tool I remember in use around then was FrontPage, but it spit out so much redundant code and I never took to it.
Webmonkey & Castro's book = everything you need.
https://www.math.columbia.edu/~bayer/coffee/
One wondered then as much as ever about things one read on the web. For example, Kopi Luwak was a legendary coffee fished out of paradoxurus marsupial feces. All references on the web could be traced to a single source, yet apparently this is true.
This really helped me power up my HTML skills. I initially hosted my site on Geocities, and later my university's home directory based hosting. I first created documents by hand using a basic text editor for Macintosh (often: view source on another site to copy something nice over), but I later found GUI tools like Frontpage with some basic templating features built in.
Once I figured out what the GUI tools were doing, I could more easily understand how to create documents manually or - eventually - generate them dynamically from perl cgi.
It's easy to look to the past with rose colored glasses, and while the technologies then were much simpler, they were also terribly hard to discover and much more limited. I certainly don't think my path to learning web development (at the time I would have said: learning how to be a webmaster) was straightforward!
I was mesmerized. I knew absolutely nothing about ssh, http, html etc. But I knew I had to spend the night figuring it out. And have a website up and running the next morning.
So I ran home and started reading and trying things out. As I couldn't really see much use in all that html gibberish, I just used full size images for every page and interlinked them via the <map> element. Hurray! Next morning my website was online! Mostly built by just using Photoshop :)
I have not thought about the <map> element for decades now. But it seems it is still supported! That's what I love about HTML.
By the middle of the nineties I had to write-protect everything and leave only an "incoming" directory with write access, because some jerk deleted everything just for fun.
By the end of the nineties I was fed up with everyone trying to use my "incoming" directory for storing warez/pr0n so I gave up (yes I tried setting it write-only but what's the point ?). Well, it was fine while it lasted, the community of people on the internet definitely changed in a few years, and it didn't get any better in the following decades.
Likewise, IRC was also created by a fellow finn back in 1988, Jarkko "Wiz" Oikarinen, at the University of Oulu.
i haven't heard anything about it since. it was just a hobby. and it wasn't very useful anyways because it only supported AT harddisks and the license didn't allow commercial use.
It did spread very quickly, as the issue it solves was quite obvious and there weren't a lot of competing solutions. So it's pretty possible that that was the actual tech used.
How many BOFHs generated keys for everyone, mailed them out, and disabled telnetd / rlogind?
To actually host the page, ~10 old me didn't know what to do. But a relative in Denmark had suddenly become the "webmaster" for her local government, and gotten to learn some stuff.
Remotely teaching me ftp or ssh was out of the question, but I got tipped about Tripod/Lycos. They had an online interface where I could upload a file! And then later she showed me how to apply for dot.tk and have my very own domain. Very cool 8)
Was a bit dormant, though. What more could I do than adding some texts and under construction gifs?
Then the Harry Potter craze hit. Lots of people made cool websites, and I wanted to join! To up the game, me and a friend made stuff in FrontPage. Still manually uploading one and one file through Tripod's web interface. I could do so much more than the little HTML I had managed to do in Notepad! Then I later got to spend a day shadowing a professional web developer, and she gave me the CD for the previous version of Dreamweaver! First I used the wysiwyg mode, but then discovered the editor had auto completion! Suddenly I could write HTML properly.
As many others, my Harry Potter page was a "school" where people could do stuff and earn house points. But not knowing programming, everything was static and I had to manually edit and reupload files. Luckily only like two friends used the page hehe. But other HP pages were dynamic! How? I had to learn PHP!
-
As an aside, it's curious how many of those professionals I interacted with the first years were women. Companies/govagencies felt they needed a presence, but didn't care much, and just gave the task to someone else already doing communication/admin work. And they learned on the job and suddenly became the first experts here.
I think this is more or less an argument in favor of tools that are helpful to the developer, like feature rich IDEs with parameter hints and autocomplete/suggestions, text editors with plugins that give you a similar experience or even a little bit of AI help.
Some people have the attitude that you might want to learn a language/framework with just a text editor to know the internals, but I think that if you want results fast and focus on solving problems (getting the site up and running) quickly, then tools like these are invaluable!
> I have not thought about the <map> element for decades now. But it seems it
> is still supported! That's what I love about HTML.
Good thing you had no hard dependency on <blink>! blink {
text-decoration: blink;
animation: blink 1s steps(5, start) infinite;
-webkit-animation: blink 1s steps(5, start) infinite;
}
@keyframes blink { to { visibility: hidden } }
@-webkit-keyframes blink { to { visibility: hidden } }Back in 1995, we paid equivalent of almost $4000 for a PC with 133 MHz Pentium processor - don't remember the RAM or HDD, but it came with Windows 95. Since we lived somewhat rural back then, libraries were woefully outdated on the computer/IT department - both as far as books and actual PCs go.
The one bookstore we had, didn't really carry much of those things either. Most of my learning came through various PC magazine tutorials, and ordering books via said magazines. Or more realistically - trying to find out what other people worked or dabbled with IT in my little town, and trying to learn through them. Everything was very, very DIY and bootstrapped.
And just for the sake of nostalgia - one of the local PC-enthusiasts had parents that worked at the municipality, and managed to argue that we (the local PC "community") needed a space of our own. So we got this run-down part of an older school building, rent-free, for maybe 7 years.
As long as we paid the utility for that part, and any internet connection, we could pretty much do as pleased. You had probably a dozen or so guys pretty much living there, playing various game over LAN. Doom, Quake, Age of empires, and what not.
Another part of guys were mostly tinkering with websites, software, or what not. Still vividly remember all the Borland boxes laying around. And of course, the more free spirited guys "Buy what? No, let me show you our warez".
But you know what - we made money, real and good money, from just churning out websites. There were really no beauty standards, little (to no) functionality, and hosting was usually done via some local dude that run a bunch of servers out of his shop.
A PC was actually not necessary getting started. I wrote my first Web-site (called "homepage" in those days) in 1994 on my Atari ST I had purchased in 1989. I uploaded it via FTP to my university Web space and checked it via telnetting to the text-based browser Lynx[1] on the university's remote host. Later I used Lynx directly on my Atari ST. This was all done via a 1,200 baud modem.
For example, the Atari ST launched at June 1985 for the price of $800. Adjusted to today's prices (via https://www.bls.gov/data/inflation_calculator.htm), that would be ~$2200, which is a lot for most people.
So I think the point still stands. The barrier to entry was acquiring the device in order to be able to do anything related to it.
However, I would admit, that an Atari ST and a 1,200 baud modem was far from being an ideal setup for getting started with the Internet. I therefore went most of the time to the university's computer lab, where I had my first contact with Unix. But what should you do when the lab was closed ...
It still works. All I used at the time were server side includes and a bit of PERL that was in a /cgi-bin/ . Sure there are some dead links (to some real media files hosted by a third party long-since defunkt) but mostly it just works as SSI is still supported in most web servers and fastCGI exists.
I'm installing it on a laptop tomorrow morning to show the documentary maker. Interestingly they were interviewing me about one thing, but the site shows another thing... that we were one of the first music acts to allow the full download of an album online. This was in early 1998 when we put the entire catalogue online as media files to listen to (first as Real Audio files, later as mp3s, before also embracing Napster and even making live concerts available for free via that).
It's good stuff, early web sites were fun and exciting.
The best feature of this website which was for a record label, is that the bands didn't know what to do with a website so instead I gave control to fans... the whole site was user-generated content as far back as 1996. Things like guitar tabs, lyrics, mishearings, stories, fan artwork... the official record label site was made by fans and just became an archive of everything that the fans and label together cared about. In 1996 this was a link on every page "email if you want to add your content"... it worked.
I verified with the band (Belle & Sebastian) that it was legit and they wanted me to be involved, but beyond that I'll find out tomorrow.
From the prelim questions, I understand that their angle is looking at the early web and the music industry and that this is one of two they're initially researching. They're interested in Belle & Sebastian due to the 1999 Brit Award win https://en.wikipedia.org/wiki/Brit_Awards_1999#Belle_&_Sebas... and how websites connected fans to bands. They're also interested in Wilco as they believed that to be the first instance of a band publishing a whole album on their website (but B&S did that before, but I thought we were late to the game so someone else must've been first). It sounded early stage though, so I figure their research will inform and help evolve their ideas.
Major deja-vu 20 years later throwing a Golang executable up on a server; really felt like I was back where I started.
I remember using this site primarily to learn about HTML: https://web.archive.org/web/19990224220115/http://www.htmlgo...
I also used Netscape's HTML tool, played around with HotMetalPro. Here is a link that somewhat is a reference to 1997: https://www.tf.uni-kiel.de/matwis/amat/html_en/kap_2/backbon...
Edit correction. Can't find the printout. Maybe it's still at home? I still got my basic and turbo pascal binder and some weird shit I can't quite remember.
It also had a French version. I remember it as being a very good resource.
A couple of weeks later I was hitching a ride with someone and we started talking about what we do. He said that "I'm teaching people this new thing called HTML". Then he reached back into the back seat and pulled out a booklet. He had a whole box of them. Apparently he had just taken a whole box of them from the university and was using this as his course material.
I never told him I wrote it. Actually, I was a bit proud that someone could make a living using teaching materials I'd written and I didn't want to embarrass him. (And since the university had printed boxes of them it wasn't like there was any shortage. My main mission was to get people to use WWW so I didn't mind).
I'd probably cringe if I read it today :-).
(However the thing I'm most proud of from that period was to ensure that policies on student access to publishing materials on their homepages without needing to get the content explicitly approved were adopted across norwegian universities. There was a meeting between the universities to figure out what policies to adopt. I didn't have any real authority, but I managed to convince everyone that what students put on their homepages is none of our business as long as they don't break the law or hurt themselves or someone else.
During that period I also learned that one of the most efficient ways to kill bad policies is to follow them to the letter and generate so much work for other people they need to rethink the policies. :-))
The big money maker was having Frontpage Extensions enabled on the web server - anyone serious wanted it, and few people provided it. I also wrote my own (awful PHP3 based) payment integrations, allowing automated billing. Quite often, the business ran itself.
Ten years later I reluctantly taught my Photoshop and Illustrator students how to use the built-in web export tools and was legitimately shocked by what they created. They totally got alt text and titles and things, too.
So, lesson learned; now there are some projects for which I really lean into the $app's built-in web export tools. Usually fast, easy, and appropriate to the project, rather than whatever happens when the output is compared against the entire web...
Static sites, created using Frontpage Express (later, full fat Frontpage, then later again Macromedia Dreamweaver and Macromedia Fireworks). Blink, Marquee etc tags being used etc. Got to have a visitor counter and guest book too.
I remember creating sites for various things, hosted on a whole load of different places over the years (.tk brought back memories). I also remember figuring out that I had to upload images and why my image links wouldn't work.
Fun times.
That also made me a firm believer in open-source.
My middle school computer club was basically a book group for raw HTML files. We started messing around to see what we could get it to do, but it was mostly feeling around in the dark. I didn't learn about documentation until getting into QBASIC a few years later.
Step 3: pray there’s no crc errors after hours of downloading
By 1997, I was working in a consultancy doing custom sites and early Internet-style applications, which was back then insanely lucrative. Like, we pitched $moviecompany a web site for $1,000,000, and got the deal (it involved DB integration and more active behavior, etc, all of which was novel at the time).
I maintained one of those corporate sites…here's a library of the IBM homepages for the five years I was the webmaster: https://epcostello.net/library/ibmcom/
I'd just finished art school and needed to put together a portfolio website. I thought I was so clever... I made the front page contain an illustration of mine dead center on the screen of a very obese man in a jock strap. You had to click his belly to enter the site.
Anyway, that site got me a "Web designer" job at a company in Richardson, Texas, named Micografx. They were a competitor to Adobe at the time and made some pretty cool software including Picture Publisher (a very robust Windows-only photo editor). They had an entire suite of graphical tools. They also made software called Create a Card, where you could create digital postcards on your computer and email them to people.
My first job was to work on a team to convert that software from CD-ROM to 100% online. It was an incredible learning experience and launched my career into the exciting Web design world on the cusp of the dot-com boom.
I partially owe my career to Netscape Navigator 4.01, which was released in 1997.
I really miss that version of the web sometimes.
Maybe it was some Microsoft tool. But since back then I mainly used notepad to edit my html it's not really that important :) it's only too bad that none of my websites survived (never hosted anything of importance back then).
The biggest challenge (if you did not work at the ISP) was keeping it on 24x7, and registering your DNS. It sure seems like DNS is a lot easier now than before. You had to call NetSol, and there was a multi-step fax-back and forth process to prove who you were to register your DNS. Sometimes it took days. You needed an act of congress to change things afterwards too, and be sure you did not mess up your NS server information. As for keeping it on 24x7, I think our pipes now are better than ever. T1 down? Yes, why yes it is, there is something wrong with the frame relay again. Or someone at the telco changed my dlci. Or the crazy person with the backhoe cut it. Or the router decided to freak for no reason. Or our custom X.25 connector went bad or lost a solder connection.
Good times... :)
If Joe Burns, PhD reads this then THANK YOU! You had a profound impact on the 14-17 year old me.
Wayback Machine link for those interested: https://web.archive.org/web/19970529013015/http://htmlgoodie...
I suppose I was lucky to be a teenager in the mid-90's with access to a computer with dial up internet. Untold hours using the View Source option in the browser and then trying to make my own stuff using Notepad (Windows). Hundreds of lines of black/white HTML snuffing out syntax errors. Trying to make images used bootleg versions of Photoshop. I recall when I started to see Javascript get popular in webpages and just being blown away people could write that stuff on their own.
Time flies.
You learn NOTHING from copy pasting stackoverflow. I always look and retype or actually figure out what’s happening. Otherwise I won’t be able to solve problems, just put bandaids on them.
It was fun work in that everything was so new, it was almost an entirely new industry. Today many people claim to know how to do everything (even though that's pretty foolish), but then you were learning on the fly. When I started at the beginning of the personal computer era (80's) it was a similar kind of wild experimentation era.
To be perfectly honest, despite some obvious drawbacks, I miss the simplicity of HTML. And I really dislike how much my every step is tracked online ( and how much it slows everything down.. my online bank when allowed to go nuts loads up for full 30 seconds in chromium; naturally it does not help that it does not want to cooperate when it is not allowed to do that ).
I guess I miss the old days.
I put up a website for a lawyer friend in EU zone recently in pure HTML ( so super simple - nothing fancy, but she did not need that much ). I did not hear any complaints, but then I sold it as a privacy consideration ( you can tell your customers you don't do any analytics, you don't pollute their device with random cookies ).
HTML Goodies (and the DHTML/JS sister site, I forget the branding), also mentioned, was a great next step to expand my knowledge.
But initially AOL actually had some good HTML learning resources. All caps for the tags. I opted for black backgrounds and white or blue text ... :) Developed in Notepad initially.
Somewhere I think I still have printed off copies of some of those old resources I printed so I could look at them offline.
He did make some mis-predictions (Linux turned out pretty good in the end), but I will probably re-read it some time soon, and I expect it to have aged well.
But there were definitely applications for building larger websites, including ColdFusion[1] from Allaire and, infamously, Broadvision[2].
[1] https://en.wikipedia.org/wiki/Adobe_ColdFusion [2] https://en.wikipedia.org/wiki/Broadvision
It also worked well enough at work... we had a "brochure ware" type page for the first few years. It wasn't until 2000 that someone with Graphics skills took over.
[1] https://web.archive.org/web/19971013164938/http://www.tsrcom...
I used Frontpage to create my website and learned HTML by viewing the source code. I then used the command line Windows ftp to upload the web site to a 10MB server I got access to for free from the academic network of the primary school. Later I stole a bunch of MIDI-s, animated GIF-s and Javascript code from other websites.
I remember I was pulling my hair out because the images wouldn't transfer correctly. I didn't know I had to switch to binary mode.
Soon after, we used to print the entire catalog of registered domain names, just to see what was taken.
The ORA Nutshell books were a great series for learning many technologies around the turn of the millennium.
Also Netscape Navigator had an option to create HTML sites.
Not as bad as Frontpage, mind you. At least initially.
Hard disagree on this...
But to the point, back then web pages were simple enough for it to be trivial to learn by viewing the source, especially if you already had experience with code. You didn't even have to be studying or working in CS specifically, BASIC and Logo were taught in many grade schools.
1998 launching the first European website of a major worldwide photography company as a projectmanager, webdesigner, contentwriter, fotographer, "developer" one man team.
was paid 120 Schilling per hour (about 9 EUR).
fun times before the invention of separation of labor on the CyberDataHighway
This entire thread brings back a lot of memories.
One of the biggest changes is that "view source" is no longer helpful to learn HTML from scratch.
i was among the first in the class to run an http server by my self including cgi and server side includes in perl (setting up a server wasn't part of the class) which earned me a job as assistant sysadmin.
then i got invitet by a classmate to a student club to help them build their website, which i did using spinner/roxen which did serverside includes on steroids by making it look and feel like html itself.
that allowed me to create fancy dynamic websites with generated graphics among other things
Interaction - mainly scripts - were provided by matts script archive.
Yup. Download a web site you liked the look of, see how they did it and build your own based on that ;)
Give them a book on programming instead, much more likely to be useful and appreciated.
TikTok sure, insta sure, not Facebook. Carrd is popular for making websites among Twitter users, but that may not be relevant depending on how you define kid.
... just wait a decade or two; it will get much easier.
Seriously, I built a very simply static website around 1997 or 1998 with Adobe Page Mill. It was pretty easy.