GuildPortal Dev Blog

Updates from Aaron Lewis, GuildPortal Code Monkey

Posted 7/23/2005 4:21 AM by Aaron Lewis. 153063 reads. Share:

We've recently finished our latest hardware upgrade -- this one affecting every bit of hardware we've got running over at RackSpace's data centers in Texas.  New database server (monster thing), web, and firewall upgrades are done and the migration of code and data took a total of about 6 hours.  The DNS stuff took a bit longer for those living in parts of the net where updating of name servers takes longer than others, but overall it was a smooth move to the new boxes and data center.

We can't express how grateful we are to the members of GuildPortal who are spreading the word in their games, directly leading to more people signing their guilds up for hosting here.  This is especially important to us in the new games that come out, when they pull in people who have played no previous MMOG and hence, have not heard of us.  World of Warcraft is a prime example of this.

That said, though the traffic is increasing, it's not necessarily leading to a proportionate increase in the subscribers compared to the sites being used.  In essence, our costs rose dramatically with the upgrades (check out RackSpace's prices for an idea -- they're one of the best, but their pricing reflects it), but our ability to cover the costs hasn't necessarily risen in proportion. 

That's OK!  As long as the service pays for itself (and for food, etc.), it's all good!

But now a shadow looms on the horizon.  Our old friend Sharbel at GuildWizard needed more time to focus on other business, and sold GuildWizard to IGE.  Well, actually, he sold it to OGaming, which is owned by IGE (info: 1, 2, 3).

Typically, when you have a new product (it's not new, but its ownership is), you have a couple options when you want to generate some traffic in an already-saturated marketplace.  1)  You make it SO good that nothing else compares or 2) You offer it for free for a limited period of time.  With IGE/GuildWizard, it looks like they've gone the route of #2.  It's clearly a market share grab, and while it might sound a bit repugnant to some, it's standard fare and commonly practiced.

Now, what's scary?  The best product doesn't always necessarily win.  If you have a major financial backer (like IGE), you can hire a huge staff of programmers and throw them at it.  The product itself doesn't have to be good enough, initially, to pay for its own development, because your other product lines (selling in-game items, currency, and accounts in IGE's case) can cover the costs for as long as necessary.  Offer it for free, and you're a double threat.

Add to that the other heavily-trafficked sites owned, such as OGaming.  Leverage the traffic there (by means of ad banners directing people to the GuildWizard service), and now you're getting people who've never heard of your competitor (GuildPortal).  If they're not the comparison shopper types, they may never hear of or visit any other product.  Grab a big enough market share, and you may never actually need to improve the product at all.  One day, turn off the free offering, start collecting the data that guild owners are entering (like quest or item information, including drop locations and frequency), and you've got a guild hosting product that will not only stand on its own as a revenue generator, but it could also act as a data source for your game information/news sites (OGaming), and as a pretty decent reference when your staff of item/currency farmers (IGE staff) are looking for good spots in the games to monopolize spawns in order to sell items for real cash.

It all sounds a bit like conspiracy theory, and maybe it us.  But it also sounds like good business, if "good" means "revenue generating" and nothing else.

Why is it scary to GuildPortal?  Well, if they've got unlimited resources to throw at grabbing market share, then we may have something to worry about.  As I said before, the new upgrades to the service are extremely costly, and we don't have other marketing channels that can fund GP -- it's been self-funding since the day it went live.  It had to pay its own bills, and it had to do so by being a good product.  MMOG players are some of the most technical, skilled, and demanding customers I could ever imagine.  They've led us to creating a product that meets their needs, and we're grateful beyond words for that.  Hopefully -- in this case and in spite of a plethora of examples to the contrary in every market -- the best product will win.

Anyway, this isn't a press release or official statement from or Axiom Shift, LLC.  It's just my blog.   

Posted 12/22/2004 10:53 AM by Aaron Lewis. 27715 reads. Share:

Looks like there won't be any escape from Microsoft this year, as far as the server platform and development tools are concerned.

Every once in a while, I pay a visit to the “other” operating system, Linux.  I've tried different distributions: SUSE, Redhat and Mandrake.  A couple of days ago, I tried out SUSE and Redhat again, using Microsoft Virtual PC 2004.  They don't support running Linux in Virtual PC, but it works and it's actually pretty nice.  You don't need a dedicated box just to play around with alternatives.

The installation for SUSE was problem-free and much faster than Red Hat.  The end result, after about an hour (I didn't allocate a lot of RAM to the instance), was a very nice install of SUSE with KDE up and running with no problems.  Red Hat, on the other hand, didn't like the emulator's chosen video card.  I should have done research before-hand -- turns out Virtual PC supports 8, 16 and 32-bit color, but the auto-detect in Red Hat defaulted to 24-bit, which I accepted.  Not really Red Hat's fault, but with an installation that took three times as long as SUSE, it was irritating.

On both distributions, I started by installing MySQL.  Relatively painless with the RPMs being readily available.  I didn't allow the MySQL packages to be installed with the core operating systems, since I wanted the latest version.  Configuration was a different matter, however, as both distributions defaulted to not allowing the local machine to login to SQL.   I found instructions for setting up access on a website (not MySQL's), but for SUSE, the package to install the Query Analyzer (or whatever it's called for MySQL) simply didn't work.  It had dependencies that returned no matches on the web.  I tried Google, RPMFind, and some other RPM search engine I've forgotten the name of.

Red Hat's RPMs for MySQL worked great, so in Red Hat I could actually run queries with a graphical tool that made life a little easier.  In SUSE, it was command-line queries.  Fine. 

Apache was a pretty clean install, but for the latest version's RPM (2.0), I had to grab it from an RPM search engine since it wasn't linked anywhere on Apache's own site.

Next was MONO.  This is the Linux version of .NET/C# that Novell is currently running.  Used to be Ximian.  Based upon my experience with attempting to install it, however, I'm not sure Novell taking control of MONO was a good thing.  Let me just get right to it -- the RPMs for MONO are fragmented all to heck and back.  I don't mean the files are corrupt, I mean they have dependencies that aren't readily available on the net, and certainly aren't available on either the MONO home page or RPMFind. 

The result of my stubborness to get it all installed and functioning was at least a combined total of 20 hours of hunting down dependencies.  I understand that Linux/Unix folks see themselves as the “elite,” and as far as I'm concerned, they certainly are entitled to that distinction if they're willing to go through the serious pain involved in installing MONO and MonoDevelop (an available IDE for MONO/C#).  However, I eventually got MONO running -- even to the point where I was connecting to MySQL and SQL Server back-ends to throw up some data on a web page.  I never got the MonoDevelop IDE up and running, however, on either Linux distro.

I really had my hopes up.  It's been a year since I took a look at Linux, and I'd hoped that the plethora of open-source developers would have put some real effort into standardizing installation and distribution of software.  Heck, I'd have been happy if they'd tested what they had on a clean OS install, to make sure that all the dependencies for a package would be included in the package, or at least readily available.

Why am I looking for an alternative to Microsoft products?  We pay outragous fees for SQL Server 2000 and Windows Server 2003 every month, they outsource jobs overseas, and I don't like being tied to a single vendor.  However, 20+ hours of effort resulting in next to nothing means I'll be sending more and more money to Redmond.  Using Windows Server 2003, Microsoft .NET, and SQL Server 2000, I can deploy GuildPortal to a new environment in under 3 hours.  That's including installation of the operating system, software, and application deployment.

I'm not interested in recompiling the Linux Kernel or even seeing its source code.  I'm not impressed with the lack of up-to-date documentation or the complete disregard given to installation and distribution.  Finally, given the elitist attitude typical of the “community,“ I'm sure it's all entirely my fault for not being technical enough.

I'll be paying another visit to Linux next year, right before 2006.  Hopefully, “Open Source” will have nailed down “Usable Software” by then.  2004, much like 2002 and 1999, was supposed to be “The Year of the Penguin,” and it was -- the development tools, environments, and distrubitions for Linux really do resemble flightless birds.