The FIRST Alliance - A whole new way to play

[Dave Flowerday]

Originally Posted by GGCO

What features would The FIRST Alliance need for users to switch over?

I'll tell you what TBA has that you don't have yet: a proven record. What are your plans to keep this site going? How is it being paid for? Is it sustainable if it starts getting heavy use? There would be nothing more frustrating to me to develop a cool mobile app using your API and then have it disappear from underneath me because you get sick of paying the server bill or something. Or if you change the API in some way that breaks my app.

Originally Posted by GGCO

we've been getting 5,000 api hits per day

You've quoted that number a few times now. Let me give you an idea what you'll be up against if your service actually grows. The private FRC Spyder API server that we run (which provides similar data, obviously) receives 5,000 hits per minute during some events. We received 8 million hits during the 7 weeks of regionals. We can handle that kind of traffic with a modest server because we tightly control the client and make extensive use of HTTP headers to only send data when it's absolutely needed. Even then, it's a server instance that costs money. We used to run on a free App Engine instance, but we quickly outgrew that (especially after they changed their pricing model last fall).

Basically, what I would want to know is: how do you plan to survive? If you had all the traffic of FRC Spyder plus a few dozen other custom apps, and you have high school kids hitting your server who aren't respectful of the amount of traffic they're generating, how do you plan to handle it?

Although we don't use TBA for FRC Spyder, they've been around long enough to prove to me that they can deliver on their promises. Right now your service looks like a hobby project, and I would constantly have to wonder when you'd get bored of it and shut it down.

[Siri]

Originally Posted by jwfoss

I disagree, i'd like to look at a teams performance over time. It can show you their consistency and if they are trending up or down as well.

I too use previous years' data for many purposes. It's quite handy to have it all in one place.

As to other reasons I'm not switching over, one (not necessarily the only/biggest) is simply integration and habit. You see that little TBA-> beside your team number in your user title? Convenient, no? I probably use frclinks.com/tba/nnnn almost as much as I use ChiefDelphi. It's amazingly convenient. (And maybe I'm just an old fogey at heart.)


All that said, I'm loving the OPR/DPR history charts. Is there anyway the events could be displayed in chronological/reverse chronological order? For instance, 357's is (reading down): DC Regional (week 5), MAR Championship (week 7), Hatboro-Horsham District (week 1), Chestnut Hill District (week 2). I guess you're displaying them in FIRST's "Regional Event" page order now? Maybe it's just me; that order really bugs me.

[GGCO]

Originally Posted by Dave Flowerday

Right now your service looks like a hobby project, and I would constantly have to wonder when you'd get bored of it and shut it down.

Thanks for the great feedback/insight, Dave. When you say TFA looks like a hobby project, what do you mean by that? Is it the fact that we're just the new guys on the block? Or is it something in the site's design or functionality that makes you say that?

[connor.worley]

Originally Posted by bduddy

While I understand where you're coming from, lately FIRST data has felt a little like this...

I see where you're coming from, but teams are developing products, not standards. Competition is good and improves overall quality.

[artdutra04]

Originally Posted by connor.worley

I see where you're coming from, but teams are developing products, not standards. Competition is good and improves overall quality.

The only reason competition is good in the sake of capitalism is because people are motivated to make money. When people want to make more money, in theory they will deliver a better product/service to gain more market share.

When people are working on open-source or community projects for the benefit of society/larger group of people, it's better that they work together on a unified approach, otherwise a large part of their resources and man-hours of work are simply spent on duplicating what already exists.

[GGCO]

Originally Posted by artdutra04

The only reason competition is good in the sake of capitalism is because people are motivated to make money. When people want to make more money, in theory they will deliver a better product/service to gain more market share.

When people are working on open-source or community projects for the benefit of society/larger group of people, it's better that they work together on a unified approach, otherwise a large part of their resources and man-hours of work are simply spent on duplicating what already exists.

The economics of this are a bit sketchy. Check your premises as to why people work and compete. Also, TFA isn't really an open source project or project for the community in the sense that you're thinking it is.

[Chris is me]

Originally Posted by GGCO

Also, TFA isn't really an open source project or project for the community in the sense that you're thinking it is.

That's part of his point... he's arguing against the merit of having multiple competing standards, when there is already one that is both well established and open source.

[Basel A]

Originally Posted by GGCO

Currently the only things we are missing are stats from previous years (which I don't think many people use anyways) and videos (which are a priority right now).

Frankly, those are pretty much the only things that get me to TBA these days. Even if you replicate their database, TBA remains the easier option for me. For current year results, I go to Spyder. Advanced stats, I do myself. What's left? I don't know.

That's what would bring me to TFA: original ideas, give me something useful that I've never seen before. Or, at the very least, collaborate instead of bragging about your great success. People other than you two working on this project isn't going to dilute the mission; it will only make it stronger. You're not exactly being welcoming to the FIRST community here.

[GGCO]

Originally Posted by Basel A

People other than you two working on this project isn't going to dilute the mission; it will only make it stronger. You're not exactly being welcoming to the FIRST community here.

That's really not a fair statement to make. If you go back to my previous posts where I outline exactly why we're not collaborating or making this open source you'll hopefully see why.

If anything we're being extremely welcoming to the FIRST community by providing rapid changes, support, and fixes as soon as we can. It's because of our small size and commitment that we can do this. Right now that's something we're just not willing to give up.

[GGCO]

Originally Posted by Chris is me

That's part of his point... he's arguing against the merit of having multiple competing standards, when there is already one that is both well established and open source.

Competition is always good. Why settle for an inferior standard? Also, don't confuse standard and product.

[plnyyanks]

Originally Posted by GGCO

Maybe a good question to ask is why? Currently the only things we are missing are stats from previous years (which I don't think many people use anyways) and videos (which are a priority right now).

The quotes below basically sum up my answer to your question. I use TBA's historical data all the time - whether its to look at a team's recent trends, to find a particularly awesome match from many years ago, or to just gather ideas and learn from teams of the past. I'm sure that, in the last two years, the majority of teams looked back to match video from 2007 and 2006 respectively to gain insight on the current year's game. That's the real beauty of TBA's huge archives.

Take a step back and look at the track record/competing standard argument from a different point of view. TBA has been around for a really, really long time (I don't even know when it was founded, but I do know that it's long before my time). They're established, I know I can rely on them and their server for anything I have my website throw at them. Now, TFA enters the scene. To me, it just seems a little bit redundant - why spend so much of your time writing a similar website to an already established one? Because face it, there are a ton of similarities between TFA and TBA. What if you, instead of spending lots of time writing portions of code that do exactly the same thing, you spent that time improving the existing codebase? Your site has a lot of cool features, and some ideas that I like, but I feel there's too much overlap between the two.

But hey, feel free to build an awesome site and convert me. You'll just have to establish your own niche.

Originally Posted by jwfoss

I disagree, i'd like to look at a teams performance over time. It can show you their consistency and if they are trending up or down as well.

Originally Posted by Dave Flowerday

I'll tell you what TBA has that you don't have yet: a proven record.

Originally Posted by Chris is me

That's part of his point... he's arguing against the merit of having multiple competing standards, when there is already one that is both well established and open source.

I haven't been able to keep up with this thread, and don't have the time to read all 100 posts, so I apologize if this has already been asked.

Will The FIRST Alliance have records for offseasons as well?

[mobilegamer999]

Originally Posted by SuperNerd256

I haven't been able to keep up with this thread, and don't have the time to read all 100 posts, so I apologize if this has already been asked.

Will The FIRST Alliance have records for offseasons as well?

Its a possibility, me and grant will talk about this one tonight.

[mobilegamer999]

Wow, this thread exploded a bit overnight. I'm going to lay out certain responses as I see them from my point of view.

Why did we make our own site?
There were quite a few reasons why we chose to make our own website. One reason was for the enjoyment/practice it takes to plan, design, create, debug, launch a website. It was something I had never done before and I felt like it would be a good project, and post-launch, if it got a good rep and people were using it, I would have no issue continuing to pay for the hosting. The second reason is that we wanted a site where anybody and everybody in the FIRST community could submit an idea and have it seriously considered as a new feature to the website, because, in my personal opinion, TBA is a website that is there and does not change unless the developers want to add/remove something, regardless of if that is their model or not, that is the feeling that TBA gave me personally.

Why not just build upon TBA's codebase?
The main reason is that its not THAT simple, TBA would not let some random outsider start modifying their code and pushing it live to the website. Even if they let us help them with their code, every time more people are added to a project, the slower changes become and the less often changes can be made.

Open Source?
I can almost guarantee that we will not open-source the website. Our biggest reason for this is that we made the API public and extremely easy to get started with so that there was no need to open-source the website. While we wont be open-sourcing the website, we WILL be providing very simple/easy to read code example for how to do just about every feature on the website in many common languages (C#, Java, Ruby, PHP, etc.)

How do we know the site is going to stay up?
I am going to say that as long as people are still visiting/using the site, I will continue to keep it up. You do not need to trust me on that, but that is what I'm sticking to.

How do we know that the API is going to stay compatible?
This is a very simple one. All versions of the API will be kept and stay accessible. The second version of the API will be requested at api/api.v2.json.php and so on and so forth. This is the easiest/cleanest way to implement a multi-version API.

Why use TFA over TBA?
There is no reason that you have to, its simply a matter of preference.

[Dave Flowerday]

Originally Posted by GGCO

When you say TFA looks like a hobby project, what do you mean by that?

I mean that it looks like you're paying for it out of your own pocket, or you're running it on a free service somewhere which likely won't handle the load when it becomes non-trivial. Anything that costs money to sustain and doesn't have a planned way to at least recoup operating costs qualifies as a hobby to me. End-user sites, like TBA and CD, can use advertising if needed to help defray the cost of hosting. You can't do that when your primary product is an API, unless you charge people to access the API which you have not indicated that you are doing. So, it comes back to my question of how are you going to pay for it when it gets busy enough that you need to run multiple servers and load balancers and such. Are you prepared to foot a $100+ bill per month yourself? If so, for how long?

Originally Posted by GGCO

The economics of this are a bit sketchy.

Originally Posted by mobilegamer999

if it got a good rep and people were using it, I would have no issue continuing to pay for the hosting.

Originally Posted by mobilegamer999

I am going to say that as long as people are still visiting/using the site, I will continue to keep it up.

These quotes should be major red flags for anyone considering developing against your API. You're admitting that it costs money to run, you're paying for it yourself, and basically you'll keep running it "if it gets a good rep and people are using it". That's not much of a guarantee to someone considering investing 100s of hours in developing an app against this API.