The plans for the Grading System

[Roger de Coverly]

Clubs not getting first names or surnames correct at input stage is an issue.

Correct identification of players new to a club, is and always has been a key issue for the reliability and accuracy of a grading system. It's made easier by having online access, so club captains and secretaries, league controllers and graders can identify whether a player has played previously and reuse that code. One of the weaknesses of the current system, which was major in its early years, fifteen or twenty years ago, was assigning codes rapidly for new players.

The method of assigning codes to new players is something that would have to be addressed. One League Management system with which I am familiar as a user, employs the ECF grading code as a unique identifier. This works very well, apart from new players who have to be given a temporary code prior to one being allocated by the ECF.

[Roger de Coverly]

Paolo's API idea:

API is a jargon term understood by system developers.

In English what does it mean for how systems would operate?

I'm familiar as a user with two League Management Systems. Both work in broadly the same manner. As a match captain, you log in, select the fixture just played, select the names of your team and their opponents from club lists and key in and confirm the results. It takes a couple of minutes usually. Grading data is only supplied to the ECF at the end of the six monthly periods, either as a direct output or by the grader running a data pasting exercise.

How does an API change this? Would it be that the match results would immediately, or with a short delay, be transmitted to the ECF grading server? That doesn't imply that the ECF would do anything immediate with them, other than show them as results on grading histories.

[Michael Flatt]

Regarding API's this wiki page might be helpful: https://en.wikipedia.org/wiki/Applicati ... _interface

The benefit of using them was touched upon earlier where it was mentioned that the exchange of data between an LMS and the ECF grading system could be automated to dispense with the need to submit emails to the ECF grading administrator, Richard Haddrell, who reviews each and every submission before uploading them to the grading system.

Hence, the discussion on the quality of input data: player's names, grading code, etc.
In place of the grading administrator's role some form of error reporting procedure might be introduced to allow investigation and correction of anomalies to an individual's details or results. It is a role that could be shared rather than rely on Richard to do everything.

If players want live grades some emphasis needs to placed on simplifying and automating the submission process.

[Paolo Casaschi]

Paolo's API idea:

API is a jargon term understood by system developers.

In English what does it mean for how systems would operate?

Introducing an API usually does not fundamentally change what you can do but defines a standardized way for different systems to do that in an automated way; more important, allows any new/existing system to be able to perform that function if compliant to the API.

In the case in hand, the easiest way to implement the communication between the new ECF league management system (LMS) and the grading server (GS) is to define a communication protocol of some sort, without telling anyone how it works. This has the major advantage that you do not need a lot of upfront thinking for that communication protocol, you can modify things as you go along and you do not need a lot of formal documentation. The drawback is that you end with only one ECF LMS that supports that direct communication and you need to spend a lot of cash for developing a full new system.

When defining an API, you spend a lot more time upfront defining the communication protocol between the LMS and the GS, then you produce detailed documentation and possibly sample code using the API. The advantage is that, once the GS supports that, any existing/new LMS can be modified to support it and you do not need a proprietary ECF LMS to ensure this communication is flawless.

Using such an API, whether the LMS pushes data to the GS every day or only every six months (or whether the GS pulls data from registered LMS instances) this is the "what to do" part of the action and remains to be decided. The API allow you to solve the issue of "how to do" such a what.

PS: nothing stops the ECF to both develope a new ECF LMS and at the same time publish the communication APIs for other LMS to use as well. You have to wonder though what's the ECF is trying to achieve, what are the priorities and what are the available resources.

[NickFaulks]

You have to wonder though what's the ECF is trying to achieve

There has been much comment here from people who clearly know what they're talking about, but how can any start be made until this rather basic point is clarified?

My impression is that the ECF wants to move to monthly lists ( probably to be applauded ) and have set out their specifications in the belief that this is the best or only way to do it. I read the feeling of this forum to be that there are other ways.

The statement does include "The tender should include: 1. Any variations in the specification you feel are appropriate" but it is not clear how radical these variations can be before they are thrown out.

[Michael Flatt]

The invitation to tender is restricted to creating an LMS which only duplicates what is already in existence and of which nobody has expressed any dissatisfaction with regards functionality or reliability.

The discussion has revealed the problem is on the ECF side of the interface. The ECF grading system is reliant on an obsolete database programming language, FoxPro, that is no longer commercially available nor supported by its manufacturer, Microsoft.

To boot, the ECF have a single hard working individual keeping the system up and running. Every offer of assistance appears to be spurned. The idea of setting up a technical working party to develop the ECF Grading System doesn't seem to fit in with the ECF approach of doing things.

[Carl Hibbard]

The database side technology wise is the first issue you need something you can even host online before you can add a xml web service to receive the data.

[Mick Norris]

nothing stops the ECF to both develope a new ECF LMS and at the same time publish the communication APIs for other LMS to use as well. You have to wonder though what's the ECF is trying to achieve, what are the priorities and what are the available resources.

I don't think the ECF want to try and stop leagues using the other LMS (and they wouldn't succeed anyway), but the main issue would be those leagues that don't use any LMS at the moment - it may be that the ECF think that developing their own LMS will encourage those leagues to use it, but from what I am reading here, the better solution may be for the ECF to work with the existing LMS and then publish a list of LMS that are compatible with ECF grading requirements

Of course, it is possible that the ECF have a problem with some of the people that run the LMS and favour other system developers, which would be a pity

[NickFaulks]

The invitation to tender is restricted to creating an LMS which only duplicates what is already in existence and of which nobody has expressed any dissatisfaction with regards functionality or reliability.

The discussion has revealed the problem is on the ECF side of the interface. The ECF grading system is reliant on an obsolete database programming language, FoxPro, that is no longer commercially available nor supported by its manufacturer, Microsoft.

To boot, the ECF have a single hard working individual keeping the system up and running. Every offer of assistance appears to be spurned. The idea of setting up a technical working party to develop the ECF Grading System doesn't seem to fit in with the ECF approach of doing things.

A depressing analysis, although I can see how you reach it. There are some sensible people on the ECF side, so let's hope there will be a bit of two-way communication.

[Carl Hibbard]

The discussion has revealed the problem is on the ECF side of the interface.

No we are all guessing based on what the technology used to be so far more information is needed before looking too gloomy.

[Angus French]

You have to wonder though what's the ECF is trying to achieve

There has been much comment here from people who clearly know what they're talking about, but how can any start be made until this rather basic point is clarified?

My impression is that the ECF wants to move to monthly lists ( probably to be applauded ) and have set out their specifications in the belief that this is the best or only way to do it. I read the feeling of this forum to be that there are other ways.

The statement does include "The tender should include: 1. Any variations in the specification you feel are appropriate" but it is not clear how radical these variations can be before they are thrown out.

Well quite.

Surely the first thing to do is to establish the need (or not) for monthly publication of grades. Then, assuming a need is determined, you would work out how - in a way that’s cost-effective - to implement change and this may or may not involve the creation of a new, free those who want it, LMS system.

I thought, from reading the Home Director's election address and from listening to him at the ECF AGM last October, that there was going to be consultation and investigation. I was expecting leagues and local graders – those who would be most affected and who, collectively, probably do more work than anyone else to get grades published - would be consulted. I was expecting them to be asked whether they thought monthly publication of grades would be beneficial and whether they would be alright with submitting results monthly.

[Mick Norris]

I thought, from reading the Home Director's election address and from listening to him at the ECF AGM last October, that there was going to be consultation and investigation. I was expecting leagues and local graders – those who would be most affected and who, collectively, probably do more work than anyone else to get grades published - would be consulted. I was expecting them to be asked whether they thought monthly publication of grades would be beneficial and whether they would be alright with submitting results monthly.

Do we know if the consultation has taken place?

[Paolo Casaschi]

The statement does include "The tender should include: 1. Any variations in the specification you feel are appropriate" but it is not clear how radical these variations can be before they are thrown out.

True. But the current tender asks for someone to develop an ECF proprietary LMS that the ECF is willing to pay for in full. Most likely the developers of existing LMS systems will respond. How likely are they to reply: "no thanks, you do not need to pay me for the whole LMS software, please define APIs towards the grading server and I'll adapt my system to work with those and anyone else should be allowed to as well."

[Carl Hibbard]

The statement does include "The tender should include: 1. Any variations in the specification you feel are appropriate" but it is not clear how radical these variations can be before they are thrown out.

True. But the current tender asks for someone to develop an ECF proprietary LMS that the ECF is willing to pay for in full. Most likely the developers of existing LMS systems will respond. How likely are they to reply: "no thanks, you do not need to pay me for the whole LMS software, please define APIs towards the grading server and I'll adapt my system to work with those and anyone else should be allowed to as well."

The grading server however is not the grading database it is just a MySQL copy (heavily modified...) of whatever the master database is now.

[Angus French]

I thought, from reading the Home Director's election address and from listening to him at the ECF AGM last October, that there was going to be consultation and investigation. I was expecting leagues and local graders – those who would be most affected and who, collectively, probably do more work than anyone else to get grades published - would be consulted. I was expecting them to be asked whether they thought monthly publication of grades would be beneficial and whether they would be alright with submitting results monthly.

Do we know if the consultation has taken place?

In his email to local graders (quoted in the first post of this thread), Brian Valentine wrote "the only benefit I can give is to let you know what is going on". I think this implies there's not been any consultation. I'm a local grader and I've not been consulted.