AIDA Results Register

[trux]

Jome has developed an excellent tool for keeping competition results - the AIDA Results Register at http://jome.homelinux.net/results/ I am afraid only few people here on DB can appreciate how much time and work is behind the development and data maintenance. I would like to thank Jome for that work and time spent on it.

However, I think it is time that the system evolutes and gets enforced to all national and international competitions. For it to happen, there needs to be done a lot more, though, and I am afraid that's no more in the possibilities of a single volunteer. I believe it should become a collective Open Source project, where more software developers from the freediving community can participate.

I do not know the interface that is available to competition organizers, so do not know how convenient and flexible it is, but I find pity that many competition organizers go on inventing their own systems, which are most of the time the source of many problems. I saw it this weekend on a completion here in France. If instead of it people focused on improving Jome's system, we could soon have an amazing tool saving a lot of troubles to all, and providing excellent service to organizers, judges, competitors, and the public.

As I wrote, I do not know how the current organizer interface works, but there are certainly many features and improvements that can be still added. Local language versions are also needed, because organizers in many countries will unfortunately refuse (or be unable) using software with interface in English.

As for extensions or additions, the following comes on the mind: offline version for running the competition and collecting results even without access to the Internet, and uploading all the data automatically later as soon as connected to the web (I do not know - perhaps it is already included). Then synchronization with automated top-time mp3 announcements (again diverse language versions may be needed); perhaps an interface for loading videos, photos, dive profiles would be interesting too; interface for exchanging data with FreediveCentral (or other similar systems already in use); interface for commentators feeding them with statistical data about the competitors in warm-up or just competing; interface to some common billboard systems used in bigger swimming pools; some additions or improvements in the user interface can be done too; etc. etc. - people will certainly come with many ideas, but a community of developers will be needed for implementing them.

Also, the website with the database is apparently hosted on Jome's home PC, so the reliability is limited. Right now it is inaccessible. It would be nice if it were hosted on AIDA's server , which (I suppose) is at a professional hosting service with 24/24h maintenance, and proper backup routes. Additionally a backup server would be useful too, of course. That could be on Jome's home PC or at another hosting provider, but should be in a different location (and preferably different continent) than AIDA's own servers. The backup server should then synchronize with the main server regularly. All this would require a lot of work, so again, IT professionals from the freediving community could help.

Jome, AIDA officials, freedivers, what do you think? Would you agree turning the AIDA Results Register into an Open Source project for example at sourceforge.org?

 

[chrismar]

Open Source projects often end up being more work than closed source for the original developer, mostly because of increased organisational and standardisation problems. AIDA needs a proper results register and Jome's is as good a place to start as any, but let's not get carried away. I'm happy to assist if Jome want's help, but a clear indication from AIDA on the rankings is more important right now.

 

[jome]

Hi,

I've had similar thoughts as well. The current system does ok and works as a "proof of concept", however it's screaming for a re-write. When I started it I had no experience of php and now...Well, I have a little more, but still not an expert at it.

Also the SQL structure, while it works, is far from optimal.

I'm all for the idea of open source in the sense of the code being public, but as chris suggests, I think we'd get much better results with a close group of motivated people at the core of the project at least.

If we we're to start such a project I'd suggest basically using the SQL dump of the results and divers from the old, but not recycling much else.

But on short term, I would really appreciate any volunteers who could help with keeping the results in sync. I haven't even started with 2009 yet because there's incomplete data in 2008 - it's a monstrous activity currently to update the results for one person, but with say 3-4 it would be an ok task.

In short term, even if majority of organizers would use some kind of standard excel-template for submitting results, I could at least semi-automate it. I've drafted a proposal of that, but not yet put it up. Any organizer out there who'd like to pilot this?

I hate making excuses, but currently I have very little computer time at home, because we have two little girls (2 months of age) running the house at the moment. It's a very rare occurence that I get to type with 2 hands out side of work

 

[ericvrp]

Hi,

During the last few months I have helped Simo with entering a few competitions and it is indeed time consuming. But I think this has little to do with the software. It is more because AIDA allows organizers to deliver the results in a non-structured form. Names of athletes are for instance not spellchecked against the current AIDA member database. What Simo has created now (the excel template) would go a good way towards improving this. Having a simple online system might work even better.

Simo, please correct me if I'm wrong, but the reason it's on your home PC is because the AIDA server is a windows machine that doesn't have a LAMP (linux, apache, mysql, php) software bundle (currently) that is used for the current ranking system.

I am all in favor of turning the current system into an open source project or starting with fresh code and the current database dump.

 

[jome]

Yes eric has been helping and a huge thanks for that. Also correct that aida webserver cannot currently host the project and in any case i think it would be wise to keep them separate for technical reason. Although it would be a good idea to merge athlete, judge, student, instructor and other aisa "career" info into one system but for now lets limit it to athlete profiles.

One other idea would be to offer some kind of public api so if someone wants to create their own look and feel, let them but within limits.

But the first step would be to get a server, some actions are ongoing so please hold your breath!
Posted via Mobile Device

 

[trux]

This project is definitely out of possibilities of a single person. So, a bigger number of people for running the project is needed for different tasks (software development, database maintenance, data feeding, data import scripts, translations of localized versions, documentation, support for organizers, server administration). You will need to coordinate the people somehow anyway, so Chrismar's argument about Open Source not being suited for it, is void. Making it a close application means less chance to get comments, suggestions, bug fixes, translations, and other participations from people who simply peek in and offer their advice or help, but do not have time for permanent participation. Or from those who download the source code, adjust it for own needs (perhaps even in a completely different sport), and then propose the modifications back. Being an Open Source project does not mean that it cannot be coordinated in the exactly same way as a closed source project, with well defined authorizations and privileges for individual participants. It sure is a lot of work, but the current state at most competitions is rather disappointing. Once the system works at least partially, it will save great deal of time for all parties involved.

And data feeding, except of writing scripts for the automated or semi-automated import of data from diverse sources, should not be done by the developers, because it simply takes too much time and energy that is needed for the development.

So I really think this should be started as an independent Open Source project, with a dedicated website for the project management, where volunteers can easily register and participate, and where public can submit suggestions too. A project at SourceForge comes in mind as the first, because it already has all the needed infrastructure. Then, AIDA-International should help with promoting the project website, sending information about it to all national branches, and to all past competition organizers. Only small part of those people come here to DB, know about the register, and have so chance to get involved.

Personally, I do not develop in PHP, although have some experience with it, but I have great experience with other programming languages, including some web based ones. I am rather busy to become intensively involved, but if the project were open, I'd be happy to participate when time allows. I think the best would be if the project coordinator came from the inner AIDA circle, knowing personally AIDA officials as well as many organizers and judges, and in the same time being an IT professional. I do not know though if such person is available for the project managment.

 

[trux]

BTW, big part of AIDA competiton results is stored at FreediveCentral in a consistent form, where the participants are selected from the database of registered memebrs, hence avoding so duplicate records with spelling errors at individual freedivers. Also the format is consistent. Do you, Jome, already have an interface loading results from FreediveCentral automatically?

 

[trux]

Also: at APNEA.cz, I have an automated script parsing several external websites - ApneaMania, AIDA register, DeeperBlue, and FreediveCentral and compares names or aliases of APNEA.cz members with user profiles at those other websites. It then associates the profiles when matches are found. It also allows me handling partial matches (i.e. some surnames in some profiles are often abbreviated or used in a different form than in others), some misspellings, or swapped surnames and family names (very frequent at FreediveCentral, for example). It means I have a rather huge database of freediver profiles, where at many of them I have links to several their profiles, with additional info. This could greatly help with cross linking AIDA register user profiles with those at FreediveCentral, and also with completing the profiles with additional information about the competitors (age, nationality, residence, photo, PB, etc).

 

[cal]

Hi All

I strongly vote for an open source approach. I think many not familiar with open source software developement overestimate the interesst and therefore the coordination work arrising from beeing open source. For a smaller (as in the size of the interrest-group, not in code), niche-project such as this as one, the comunity will be tinier as one might expect. It would however allow people to peek into the code and improve it as the need arrises.

I don't have that much time to get involved as a "core developper". But I would definitively have had a look into the code if it was available at the moment. Without beeing able to review the work, there ist no chance of even thinking about contributing.

Sourceforge is surely a very good place to publish the current work and add a little documentation on how the project is thought of beeing conducted and what's the way to proceed on. An independent hosting would benice as well as jomes PC is currently not reachable This is where I probably could help a bit.

Have fun,

Michi

 

[jome]

Hi All


as jomes PC is currently not reachable This is where I probably could help a bit.

Michi

Again? For the love of...Yes, this I would really like to get rid of

 

[trux]

As for the hosting, I can also eventually propose a server - I have two dedicated FreeBSD servers in the USA, so I could place it together with some my other websites. MySQL and PHP are available, of course too. But perhaps someone can offer and independent server, which would be even better, especially with the possible future need for heavier use on mind.

 

[jome]

I guess if we really start to do this, we could divide it into 2 main areas (as currently)

-public interface for viewing the results

-admin interface for submitting and editing the results

I think the current public one is "ok" for now with minor bug fixes and facelifts, more urgent work is needed on the admin side. More specificly, there should be an simple, intuitive and safe way for organizers to submit results (currently it's none of these).

If we narrow down the "problem" to that, it really shouldn't be a huge task for a small group of motivated people.

Personally I'd be most comfortable in PHP in web programming, but if there's a good reason to select another one, I'm open to suggestions. I think Eric suggested python once?

 

[trux]

Jome, I think there is much more to do if you really want to start enforcing it at competitions, or if at least you want that organizers start using it voluntarily. First of all, translated localized versions - although incomprehensible to most of us, and despite English being taught in schools in most countries, there are still people unable to use interfaces written in English (or even entire nations - like for example the French). Then, an off-line version is needed, where the data can be collected, and stored in real-time without access to the Internet, and then uploaded to the server later. Also the standard web-interface must offer a fall-back option to off-line mode without losing data. It should be used directly at the competition as the main live result register, and not just to convert the data later and send them to AIDA.

However, I think that the purpose of the interface should be not only data collection, but a very complete competition event manager software - including registering to competitions (similar as Freedive Central), scheduling of top times, planning of tasks for all involved organizers, judges, and safeties, and even perhaps payment management with an interface for online payment of registration and other fees (protests, anti-doping, and perhaps even meals and accommodation). It should simply offer all what a competition organizer needs for preparing a good competition - including check lists, rules, tips and hints, detailed information for commentators, prinitng of badges, schedules, line-ups, etc. Only when the software is complete, organizers will stop inventing their own systems, having then each results in a completely different form, or not delivering them at all.

It must be also able of handling different exceptions, like protests, penalties, disqualifications, or other modification applied later, etc. There are many other ideas coming to the mind, and that's also the reason I think it should become a public Open Source project, so that people can add their own ideas or extensions.

 

[trux]

As for the programming language, Java would be probably the best choice, because it would allow coding both the web interface as well as handling the off-line use. And it would run on any platform - from Windows, over Mac, to Linux or other OS's without the need of installation of anything. In contrary, if you want to develop it in other languages, and especially if you keep on mind that you need also temporarily storing the data without Internet access, then you would need installing locally the the application compiled for the specific platform, or language interpreter, database engine, or other tools. With Java it should be doable all without any installation at all (except of loading the application by calling a web page with a Java-enabled browser).

The disadvantage of Java is that it is a language used more by professionals than amateurs, so the number of available participants would be smaller.

 

[ericvrp]

Perhaps Google Gears would be nice

 

[cal]

As for the programming language, Java would be probably the best choice, because it would allow coding both the web interface as well as handling the off-line use.

In my opinion the frontends (Web, Stanallone/Offline clients) and the Backend don't need to be written in the same language. I would rather well define the interface to access the backend. This way it would remain flexible enough for future extensions and replacements.

About the functionality: I can see, you would like to have a tool that can do anything for you. But If you try to build a cathedral in one shot you most likely will fail. Catherdrals usually end up with a big building place and are never finished nor usable when they once shuold get finished. Make small modules, one by one and fit them together. One more thing that's easier with well definded interfaces.

just my 2 cents.

have fun,

Michi

 

[trux]

Yes, that's not a bad idea. Although Gears would require installation on the local PC for the users of the admin interface (organizers), it is probably not such a hassle as it might be with other programming/database environments. I have no personal experience with Gears, but it looks promissing.

 

[trux]

Michi, yes, the back-end can be in a different language, of course. As for cathedrals, I think you do not need to worry about a huge collapsing application. What you need is a simple system, but well though for extensibility and scalability, so that the additional features can be added easily. If you start building a system without taking all possible future extensions in view, you will find yourself soon very limited, and in need of re-writing or hacking the application to add new functions.

 

[ericvrp]

I've never worked with Gears myself either but it's presentation feels good.

@Cal: I'm all for the one-step-at-a-time approach. Coming from a great test-driven (XP, eXtreme Programming. Yes it exists! ;-) ) project myself. Let's start small.

 

[cal]

If you start building a system without taking all possible future extensions in view, you will find yourself soon very limited, and in need of re-writing or hacking the application to add new functions.

Of course you need to choose a clever design (which requires some serious thinking . I just wanted to point out that I dont think it's wise to start building every feature imaginable from the begining. Keep in mind that it may be comming i future, but first keep yourself to implementing the basics which are really needed.

Michi