Register
View RSS Feed

WoW-GPS Dev Blog

WoW-GPS 2.0 - A Path Forward

Rate this Entry
So, now that you know a little bit more about what I've got in mind, I'd like to get some feedback on how to move this along, and what that will look like.

Obviously, a project overhaul of this magnitude isn't without its share of obstacles. The 3 biggest for WoW-GPS 2.0, I believe are: Resources, Features, and Scope.

Resources
I had previously outlined SOME of the resource limitations here: http://stormspire.net/blogs/kathroma...api-first.html. Unfortunately, those are only a few of the hoops I expect I'll have to jump through. Let's start with application framework.

WoW-GPS 2.0 will be built on Rails - of that I'm 99.9% certain. The ONLY way this project gains traction and builds momentum is if it's fun, and for me, right now, that means Rails. The biggest problem here is that my shared host (the one where I've proposed that all the AH data will be stored, and accessed via API) doesn't support Rails. To be fair, I've read that I can TECHNICALLY run a Rails application with them, but it would need to use v 2.x something, which is 2 FULL versions behind current, and really not worth the hassle. This means that the public-facing application, the one you access when visiting www.wow-gps.net, would be running on a separate server from the data itself.
This isn't a big deal, since I've got a relatively unused linode available that I run Rails apps on for clients, but it's not exactly ideal, so it's definitely something I'll have to keep an eye on.

So, essentially, I'd have portions of the new setup running on 3 separate environments. The application would run on my linode, the data would reside on my shared host, and any number of worker processes would be running through Heroku. I definitely expect some challenges getting all those moving parts working together nicely, but for the cost (essentially free, based on server costs already covered by current clients), it worth trying at least.

Features
WoW-GPS 2.0 has a lot of planned features, and some of what I've mentioned hasn't even been all that specific yet. I'm sure there are also a number of features that users would like to see that I haven't thought of yet. Either way, with such an extensive list, it will be extremely difficult to even get off the ground without first prioritizing all those features and requests. Not only will I need to determine what needs to get built, but I also need to figure out what needs to get built FIRST, so that I can start getting some Beta-quality functionality into people's hands for testing.

I think the first step will be to compile my own list of planned features and then open it up to user requests, so keep an eye out for that as probably the next piece of communication. Once that's complete, I'll need some way to prioritize everything. I'll probably use some combination of my own discretion and user feedback (up-voting, raw ranking, etc.) to figure out where to start.

Scope
At this point, probably the biggest hurdle, potentially hanging like a Sword of Damocles - ready to derail the entire project, is the sheer size of it all. It feels like it's going to take FOREVER to finish - which branches into a couple deeper issues: 1) Can I do this by myself? and 2) Is it WORTH the amount of time required?

1) Can I do this by myself?
The short answer here is "Yes, I probably can". The long answer is that it will obviously go a lot faster if I can find help. I've always considered the Open Source route with WoW-GPS, it really seemed to be a good fit with the module nature of version 1.0 - I just never really had the time to figure out exactly how it could work. With 2.0, there doesn't seem to be as natural a separation of components, but it's also quite a bit more work, overall, so the added value of bringing on more team members seems more reasonable. So, while open-sourcing the project doesn't seem to make as much sense anymore, I'm seriously considering bringing on additional contributors. One of the next steps for that will be to see if there are any Rails, PHP, or Front-End developers out there who are interested in being a part of the project.

2) Is it Worth it?
Part of the reason for investing so much time into what might be considered a "hobby" of a project is for the experience in developing a commercial-grade web application. That said, commercial-grade web applications generally also revolve around the notion of commerce, or profit, and I find myself wondering if the "experience" is enough. Don't get me wrong, this is a project that I enjoy and plan on working on regardless of financial viability, but I have to be realistic - if there's a choice between paying clients and supporting a hobby, WoW-GPS could get shelved for an extended period of time while things settle down.

So, I've wondered about a for-profit model - a natural inclination for any goblin, I'm sure you'll agree. And obviously, there are so many different options for what that might look like. Donations, monthly subscriptions, different "tiers", etc. It's something, though, that I'm not crazy enough to implement without feedback, so I'd like to send out some sort of survey to gauge the level of interest for some of those options.

Obviously, any sort of financial aspect will complicate matters for #1 - regarding team members or external contributors. Ultimately, I'd like to weigh the level of interest between the 2 to see which might hold more traction moving forward and go from there. If there are a few people interested in contributing to development and no financial interest, I'll probably avoid a profit model, altogether. If there's viable interest in both, I guess I'll have some other decisions to make...


So, next steps for WoW-GPS are really about data gathering. I'm going to have to get that feature list together and throw up a few surveys, so keep an eye out for that.

As always, any comments or feedback would be greatly appreciated, so fire away!
Categories
Uncategorized

Comments

  1. Ord's Avatar
    Regarding features, I'd love to see some outlines of what you think GPS 2.0 can do. I think we can all make a huge list of "every possible thing that could be any degree of awesome" and from there, trim the fat and focus on what will actually matter or not. I also love the idea of collaboration and community input.

    As for scope, my personal opinion is that you should keep GPS something that's fun for you, and trying to adapt a paid-model (besides voluntary donations) could be the one thing that turns it from a fun hobby to a stressful commitment. A modular approach sounds fantastic whereas any community member (or group thereof) can develop modules of their own and share them with others, or customize or "perfect" existing modules. This might not be to difficult if you are planning on developing a full-featured GPS API. There could be public API calls, and there could be internal ones only used for developing 'modules.'

    I've always been a strong supporter of open-source, community-run projects, but that's not to say that you don't deserve compensation for your time and effort. Would site ads be enough revenue to cover server costs? What about donations and a voluntary ad-free subscription model? There are many different ways of approaching it, I just fear that one that relies on incoming money could be what kills your drive to pursue it to completion at all.

    Lastly, given how interested I am in what you have planned for GPS, I would be more than willing to lend whatever assistance I can. I don't have any frint-end web dev experience or rails, but I am fairly decent at looking at code and understanding what it's doing: something that I think could be helpful for initial beta testing/debugging. I do have experience with SQL and relational databases including MySQL from a reporting/data-analytics perspective and if anything that I do know how to do well becomes in-scope for the project I gladly volunteer in-advance for helping however I can.
  2. Kathroman's Avatar
    Quote Originally Posted by Ord
    Regarding features, I'd love to see some outlines of what you think GPS 2.0 can do. I think we can all make a huge list of "every possible thing that could be any degree of awesome" and from there, trim the fat and focus on what will actually matter or not. I also love the idea of collaboration and community input.

    As for scope, my personal opinion is that you should keep GPS something that's fun for you, and trying to adapt a paid-model (besides voluntary donations) could be the one thing that turns it from a fun hobby to a stressful commitment. A modular approach sounds fantastic whereas any community member (or group thereof) can develop modules of their own and share them with others, or customize or "perfect" existing modules. This might not be to difficult if you are planning on developing a full-featured GPS API. There could be public API calls, and there could be internal ones only used for developing 'modules.'

    I've always been a strong supporter of open-source, community-run projects, but that's not to say that you don't deserve compensation for your time and effort. Would site ads be enough revenue to cover server costs? What about donations and a voluntary ad-free subscription model? There are many different ways of approaching it, I just fear that one that relies on incoming money could be what kills your drive to pursue it to completion at all.

    Lastly, given how interested I am in what you have planned for GPS, I would be more than willing to lend whatever assistance I can. I don't have any frint-end web dev experience or rails, but I am fairly decent at looking at code and understanding what it's doing: something that I think could be helpful for initial beta testing/debugging. I do have experience with SQL and relational databases including MySQL from a reporting/data-analytics perspective and if anything that I do know how to do well becomes in-scope for the project I gladly volunteer in-advance for helping however I can.
    @Ord - thanks again for the comments.

    That's definitely what I had in mind for the features: just brainstorm everything possible, then go back and organize/strip it down. It'll probably need a couple passes, too, since it's likely to start off being pretty disconnected until some clear structure starts to take form.

    As for the $$ - the server costs are essentially already covered. What remains to be seen is whether the current server setup can handle the project. If not, then server costs will definitely need to increase in order for the project to continue/progress. We'll cross that bridge when we come to it. It's really just the time investment - I don't want to end up resenting the work or getting pulled too far away from it that it's difficult to get back into.

    Ultimately, my perspective on this changes dramatically if this ends up as more of a collaborative project. Me having to take some time off in order to focus on my own, paying clients isn't nearly a big deal if there's someone else, or even a team of people to continue working on it. So, the response I end up getting from potential team members will play a big role in this, I expect.

    I'll reiterate this - I don't really think WoW-GPS 2.0 will work well as an Open Source project. The "modular" nature of v1 is realistically going away in exchange for the more connected and versatile Account setup. With v1, it was theoretically possible for someone to develop an entire module on their own, submit it to me, and have me drop the files on the server and add a link on the front page. 2.0 really won't work like that, however. Any functionality for individual markets, for example, will need to also integrate back into everything else. That said, I expect there's still plenty of opportunity for Team-based development, and I'm excited to see what comes of that.

    I'll definitely keep you in mind, though - Rails developer or not, the project can always use support from enthusiastic people, and the DB situation is FAR from resolved
  3. Kathroman's Avatar