Browsing through the small subset of nodes above, I see that while the end of AG is documented sufficiently, the beginnings really are not. The attempts start out by claiming that AG was a file-sharing utility; this is only partially true, and definitely not the reason I signed on - I would like this view to be reconsidered by at least a few readers. Having worked at AudioGalaxy from humble start to humbler finish (well, nearly), I should be able to provide a bit more of an insider's view on at least a part of what it was like. Keep in mind however that I was only a web coder monkey (HTML and graphics manipulation, mainly), part-time snooty music reviewer, part time band representative, talent scout, and software/hardware reviewer - this last position was mine alone, as everything else was shared between all the part-timers there. As such, my picture is far from complete, but nevertheless may prove enlightening. Some liberties may have been taken with my recollection of things to make the node flow (such as the precise dates of my job interview, which I do not recall, nor the exact text of the ad), and I can only disclose the names of the people already in the public domain for the sake of privacy. Some personal stuff is kept vague to prevent this from being overly GTKY nodish.

Late Summer, 1998

I was a Computer Science student at UT Austin, junior going on post-senior ennui. I saw a few plain paper announcements around the labs, but didn't think I could work and study at the same time. A few days later, when the cash was at a lower ebb than usual, I changed my mind. The ad mentioned reviewing music, MP3 technology, and an online business model - back then, this was all stuff of legends to me and I couldn't resist. I made the call and scheduled an appointment a few days later.

I had met my future bossman and PHP/mySQL-god inspiration David McArthur (System Administrator, main coder, co-owner and host of the physical location of AudioGalaxy for a while) at Insomnia cafe, on the main drag, directly opposite the campus. Later the cafe had magically morphed into the identical yet somehow blander Metro - same decor, same theme, same spot, but the lame name did wonders to ruin the atmosphere. We walked inside, greeted by a blast of multiple exposed AC units (it was one of those neo techno cafes, lots of chrome and steel and exposed piping), doing their best to ward off the Austin summer heat. It was midday, so fairly quiet - only the caffeine junkies from the previous night were sitting blearily over their books, not really seeing them. Grabbing a coffee we ventured to a couch - me dressed in tentative optimistic interview wear, him in hawaiian shorts and an ill-fitting tee; an enthusiastic teddy bear geek, around 30ish.

So, what exactly are you guys up to, I asked. This was a bad move, as it launched a 20 minute monologue on technologies, PHP (which must have been around version 3-ish then, young and buggy but full of glee), Apache web servers, MP3s and the injustice of corporate music (the RIAA, today on everyone's mind, was as yet ignorant of this whole MP3 thing, hard as that may be to believe - back then the worry was that Fraunhoffer would make something of their claim that the MP3 codec was theirs and theirs alone). I listened entranced, and thought that sounded...well, pretty damn cool. To be a part of a potential record label attempting to make record labels obsolete - enlightened by college radio, I was quite well aware of the pervasive problem of manufactured music - that was intense stuff. At first I only signed on to write reviews, but there were definite hints that being involved with website design and creation was going to be nearly required. Dynamic webpages were not in quite as common use yet, either. Every page was done by hand, in those days. Whee!

What we did

We started out completely freelance - I think there was a colocated server somewhere in another state, with us contributing dribs and drabs of music writeups, and thinking how cool it would be if we singlehandedly took on the disposessed, impoverished and wretched artists, and made ourselves their own gateway to the world. A kind of vanity press for the musicians, with a review, a small media sample and the ability to run off CDs cheap whenever required. For now we scoured the web, fanzines and newspapers for amateur band contacts, tried to reach them by email, phone or fax, waited on the results while at the same time negotiating our involvement with a second band, creating new webpages for a third band, listening to music samples for fourth and fifth and scribbling notes for a review for both at once.

While we waited for our vision to solidify, we listened to a lot of bad music.

Uninspired, redundant, bland, off-key, cheesy or downright laughable - you name it, we had it (and it's probably still there). Amongst the chaff there were a few good names that made our front page repeatedly and made us seen, for which no praise can be enough. Groundbreaking one-man IDM projects like Cerebral out of Delaware, esoteric and fun Russian Rock'n'Roll like Olga Arefieva & ковчег Band from Moscow, youthful and energetic metal acts like Sophistes from Poland or Power Symphony from Italy. These were a pleasure to review and reaffirmed our goals whenever the latest angry artist called to complain that our review "didn't comprehend his/her vision" or we cringed to yet another experimental act consisting of drugged out kittens being dragged across a sandy slope in midwinter Sweden.

There were a handful of us doing this while the masters above us (Michael Merhej primarily) worried about server loads, dynamic pages, logos (no logo in the first day, and a true sign of Internet newbie-hood - our page had a black background and yellow/red on the rest of the site), site layout and how to present our slowly growing content. Somewhere in the back, something that would one day be the AudioGalaxy Satellite (but for now called simply the AGent) germinated quietly, like a digital time bomb.

My workload grew by bits and pieces. As we got more content, we started defining genres, and what a debacle that was. Some parties were opposed to the concept, claiming that some music could not be classified easily; the tendency to lump more and more material into simple electronic or experimental did not help matters. Eventually a set of genres was agreed upon, good intentions succumbing to web usability - users needed starting points and clear definitions to find the music, and that was that. Eventually everyone found their genre niche and settled into it (I took Metal, Folk and occasional Electronica since it was the most plentiful), developing their departmental presentations and fleshing out their own content. The process of solidifying the genres took about a month.

By this time I was all over the MP3 phenomenon, and HTML was pretty much instinctive. In the process of establishing artists' websites and presence (sometimes only linking to their preexisting pages, sometimes doing up a full-blown presentation with pictures, blurbs, music and logos), I had to learn to rip & encode music. At the time, the best tools were DOS tools, and all-in-ones existed only in buggy, incomplete forms. There was a lot of enthusiastic amateur talent out there however, and my software library grew. From command line rippers that had you specify all sorts of cryptic options and crashed your computer as often as not, to encoders that were slower than 50% realtime (takes 2 hours to encode a 1 hour album), through a whole slew of frontends that were simply visual interfaces done up for command line tools, eventually the solid all-in-ones (rips and encodes in one step) emerged and formed their own dotcoms. When MP3s began to have built-in tags everyone thought that was an awesome idea, but the effort required to keep naming songs soon grew too much - so whole new tools were made to incorporate the naming process into the encoding. Around this time the giant free formed out of nowhere, and had thousands of users submitting hundreds of thousands titles, songs and artists. This of course required a whole new slew of CDDB-equipped MP3 players and all-in-ones.

Technologies grew at an astonishing pace all around us, and we were right in the middle of it all - it was a very exciting time. I found my own niche as a gatherer and reviewer of MP3 software, each new day bringing a new entry (or five) to AudioGalaxy's software repository. I set up a few independent developers with webspace on AudioGalaxy's servers and helped spread the word about others. About this time WinAmp 1.0 came out, distinguishing itself only by its relatively small memory footprint and economic use of screen real estate - it was also quite stable. At the time it had quite a bit of competition such as FreeAmp, K-Jofol, QCD or 4 MP3 - over time, most of these fell to the wayside as their developers either got bored or moved to other projects. Browsing the software archive (it still exists) is like a bit of MP3 history, the sheer numbers telling of the infectious enthusiasm that spawned prodigious feats of code in a very short amount of time.

We had moved into our 6th street offices by now, having done a short stint in David's new house; ultimately the location proved to be too far out of the way, as most of us were still students. As we were all spending our meager cash reserves on broadband (no 56k for us! It was a valid business expense!), we all had direct hookups to the site from home, and rarely did we need to range out. Once the offices opened, that all changed; there's something about an office that makes you feel obligated to come in, like all of a sudden the business gains legitimacy so you should behave more real-world-like. And so we did, some more than others - this was probably the start of the shift in attitudes, as management got their offices upstairs, and us content creators shared one big room downstairs. We also hired a couple of coders to increase the speed on the AGent.

AudioGalaxy v2.0

The site got a facelift around this time. The AGent interface was still in beta, but it was usable beta, and we wanted to present a slick new face to the world for its upcoming launch. Decisions were still being made democratically (more or less), so we all got to see the plans for the new site, argue about the logo ($5000 to create that triple-circley logo, would you believe that? Branding is an insane business - that year's designs all included a lot of swooshes, multicolored letters and three word slogans - at least we avoided that!), and consider the layout in light of our content. I think we spent a few weeks just deciding the exact shade of websafe backgrounds (you had to worry about those things in those days - a significant population percentage was still surfing in 256 color, and sites had to optimize graphics for 56k modems) for each of the genre subpages. We all graduated to "Department Editors"; unfortunately this did not include a raise (oddly enough).

After settling in to the new site and continuing my duties as reviewer and software guy, word started passing around that some company called Diamond (didn't they make lousy video cards?) wanted to make a hardware MP3 player, using memory cards, of all things. There was much scoffing, mainly having to do with the price of memory cards - by now all of us had made multiple CDs full of MP3s and of course the inevitable idea we hit upon (a CD MP3 player) would come to fruition only a few months later - but in the meantime, scoffing was much in evidence. After a few more of these appeared and showed no signs of bombing, I realized it was my solemn duty to review them as well, and another phase started. A hardware section grew slowly next to the software.

Hints of the End

Things progressed, as they do. The artists continued to be added, the songs reviewed, the departments (now crystallized) had their monthly Editor Updates, their weekly changeover of features, and constant automated New Sounds (bands that just signed up). Somewhere in there an automatic sign-up utility was made, and a backend to the site was created for bands to create their own material on AudioGalaxy's machines. We did less design and more listening and reviewing, as the balance tilted, and even began settling into something approaching routine. At this time we got a "PR Manager" to coordinate matters, which wasn't met with much popularity at first, as we've always done everything ourselves. Eventually the need was realized, or perhaps impressed upon all, and we were disconnected from another process of creating the artist's AudioGalaxy presence. We had lost some writers earlier on, and it seemed that a dedicated PR person would make up for that lack, allowing the remaining writers to focus solely on their task; and perhaps that's what happened. The proof, for or against, I leave as an exercise to the reader...

The AGent was becoming less beta and more real, and soon afterwards was renamed The Satellite and added to the site's offerings. At first, the only part that was obvious was that it was a p2p program that would allow you to get our site material faster from other users - not until the "Music Search" subsite was revealed that it became fully obvious that it could be used to obtain ANY other material from other users. I was skeptical at first, especially as speeds were slow, but as more and more users signed on, and the web interface solidified until you could obtain the song rating, the song's download speed, all titles near to what you were looking for - then click on a single button in your browser and have your tiny Satellite icon do the rest...well, I was convinced.

The End's Beam-It service started, was applaused wildly (insert your legal CD into your computer, fire up Beam-It, have the disk scanned for validity, and then be able to listen to that CD anywhere with Internet access via immediate streaming) by just about everyone as a means to have music on-the-go anywhere, and was shot down by the RIAA faster than a liberal at an NRA convention. Perhaps it should have been a warning sign since we afterall did a very similar thing...just skipping the CD validation step.

About this time I started growing more separate from the company (now company in truth, as M. Merhej now signed all of his emails as "CEO", and inevitably delayed paychecks until supplicants made their way upstairs and humbly requested them - ugh), as schoolwork set in and graduation closed in. At the time I thought it was my separation from the office (I increasingly worked from home as time grew short) that resulted in my inevitable downsizing (December 2000), but really it was a host of factors that combined to do the company in. Unresolved payments from advertising firms (as they went under themselves), huge bills for the insane amounts of bandwidth we used, stirrings from music labels (the "P2P is piracy" argument hasn't been made very strong yet, and labels took care of their own complaints) demanding that filtration be put into the Satellite to leave their music out (this was done, but it was, and still is, impossible to bypass human ingenuity with automatic filters) decreasing our traffic further, desperate attempts to save the company by bundling it with spyware and receiving comission for it (try searching for VX2 and Satellite on Google), all marks of floundering. I found out later, when I dropped by to get my belongings, that AudioGalaxy now consisted of about 5 people (David, Michael, our PR Rep, a coder and a single reviewer), down from about 15 at its most effective. It was easy to see the end from here.

AudioGalaxy Ethics

I couldn't really say that I thought much about the ethics of the situation; I definitely cannot speak for anyone other than myself. As a poor college student I bought my share of the music I liked, borrowed music from AudioGalaxy's stash of artists, or used the usenet to look for more esoteric offerings, as it was the only source I could get the stuff I like from (death metal and evil folk music aren't exactly easily found, nor played on the radio *gasp*) - but here was this amazing tool that would hook me up to like-minded (or not) people and enable me to expand my musical knowledge - although the more esoteric stuff was still rare. I guess I still hoped that most of our users came for the good stuff on the site, not for the search.

So just like usenet or "borrowing" music, this was a temporary way to find out what I liked prior to the inevitable purchase (or dismissal). I don't think I ever fooled myself into thinking that what I was doing was legal, but considering the lack of any other alternative (while reading reviews helps, I have bought plenty a bad album based on someone else's say-so) I saw no better way to reach the music goodness that I was after. However, I do know that I have never felt that the MP3 is really the song - having worked with the earliest encoders (anyone remember XAudio? *shudder*) I simply knew that the quality wasn't there, and anything worth listening was worth buying just for the quality of sound. It didn't help that the maximum quality was 128kb/sec, and some users even experimented with less. Or maybe that was just what I told myself (again, exercise for the reader...)

Some Speculation

In retrospect it's fairly easy to see what happened. The Satellite was a very large assumption to base a company on; I think we all knew that sooner or later someone would notice the 80,000 hits per day that AudioGalaxy was getting, and investigate the reasons for it. The fact that it was even more centralized than Napster only gave the eventual lawsuit a bigger target. If instead, AudioGalaxy had gone ahead to become a new way of distributing music legally (as Valve is now doing with Steam for gaming), it might have become something new and interesting, although the road there would undoubtedly have been far more arduous and not very rewarding. By choosing the quick way to make a buck during the dotcom boom, AudioGalaxy doomed itself to the fate suffered by many others.

Oh, it still exists, in a way. Some of my reviews still appear, some independent artists still survive. For the most part though, it's just another castoff from the brief time of dotcoms, empty and somehow sad.

Michael Merhej has dropped off the radar; googling him dates his last beeps somewhere in early/mid 2001. AudioGalaxy's departmental pages still hold their last monthly editor letter; the metal page still has traces of my presence (speaking of copyrights, I drew that red/gray masterpiece all by myself - not sure why it's still being used for someone else, tsk), and the software pages still exist, the homepages of code experiments long defunct telling another sad tale.

Some sources:
I was very glad to find this to corroborate some of my vague half-rememberings. While I barely remember Kennon (He signed on later, Michael and Tom did all the early work), his account fills in a few holes and echoes eerily of my own account. His AG visitor count is off by a factor of 100, though. Perhaps it is fitting that a coder wrote for kuro5hin, and a reviewer for E2?

Some facts and dates for those who like that sort of thing:

My last entry at AG, all blithe and unknowing *sniff*: