Join Bridge Winners
All comments by Nicolas Hammond
You are ignoring the author of this comment. Click to temporarily show the comment.
I did say you mentioned his name - Adam - although you were referring originally to another one!

Adam was extremely articulate, but at the same time pointing out that he was an exception, but also pointing out that nearly everyone in the room was the exception!

I don't think you can ask kids in advance what they want, without them know what the event might be like. Got to start somewhere.

The original event appeared to be more set up for ‘us’ to find out what ‘they’ wanted. ‘They’ couldn't figure out why ‘we’ were asking - after all ‘they’ were there. ‘They’ just wanted to have fun. ‘We’ are supposed to have the answers, after all we were all kids once.

ACBL still does something similar. Big shout-out for Bill Gates who makes an effort to show up at these (kid-only) events. He knows that his presence is a big deal for the kids and makes a life-long impression on them (and their parents). He doesn't have to do it, and is very patient with all the photographs and questions. Who else can go back to school and say they met Bill Gates over the summer break?

Also, the younger professionals/well known players (say, those under 26) also have an impact on these events when they talk about the travel, the countries visited. Not all kids will have the opportunity to emulate what they have done, but sharing the stories, showing pictures has a positive impact. Instead of asking them the day before the event, someone should ask them a couple of weeks before so they can do a better ‘presentation’. Kids relate to those closer in age. Some are better than others at the presentation aspect. Perhaps get them to show up with USBF attire, etc.
Jan. 14, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
Peg, I was at the same meeting. (I think I had some of my kids there so I had an interest).

There were more ACBL members/volunteers/staff than there were juniors. There would be 2-3 juniors at a table and 6-7 “old” people at the same table trying to find out why they young ones were they. “We” (the older crowd) out-numbered the junior crowd and so the junior crowd was less interested in attending similar events. They just want to hang out with the peer group which they can only usually at the YNABC or Midnights.

A Junior social, with none (or very few) of us “old people” would be much more fun for them.

There was one very articulate young gentleman (Peg - you mentioned his name) who eloquently described the various issues at the meeting. Getting kids through the door to start with is the biggest hurdle. This is a local issue. By the time they are playing at NABCs, we already have them “hooked”. Finding out why kids drop out, or lose interest, is more important than talking to those that stayed. So the research they were conducting was probably lost because each kid had an individual story, there was rarely a common thread that could be duplicated.

As with all these issues, metrics is very important. Otherwise everyone is guessing and what works in one area is very different than other areas. What I have seen is that it is the local initiative, effort and volunteers that make all the difference. So if you look at the successful programs there is nearly always a tireless, usually well-known, volunteer that has made it all possible. Finding those volunteers, and getting them enthused to start, and then maintain, and then recruit their replacement is the difficult part for ACBL whether it is the full time staff, or the volunteers at the District or Unit level.

Metrics are important. First step is to define what metrics we are looking for, then how we would use them, then collect them. All are difficult.
Jan. 14, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
I will surmise that this motion was for a specific recent BoD member (you can do the research and the position).

The previous rule was 12 months. That has been dropped.

I know that for ACBLscore+ that there were some BoD members interested in volunteering but it was made quite clear to all (by ACBL) that any work would be as a volunteer, and unpaid. My company could not hire (i.e. pay) any BoD members, even if they were technical qualified and the best for the job. In the end, we did not use any BoD members for the work, though the code was available to some for input if volunteered.

The non-favoritism rule is fairly typical for non-profit boards (where the Board has a supervisory role of Management). There is an obvious conflict-of-interest in having Board members also be employees/contractors or potential employees/contractors.

The general rule appears to have been changed for a specific situation; this change can only benefit future board members. Although one could argue the benefits for this specific instance, the general principle of board members being unable to financially or personally benefit from any current or future situations without a 12 month waiting period is sound. I would expect, and hope, that the board would re-implement the existing policy in 12 months time, or at least re-word it so that the intent is 12 months, but that the board can vote to overrule on a specific instance basis, similar to the Section 8 wording for current board members. The latter wording would have been much better this time around.

At the same time that the Board voted to water down working for ACBL after being on the Board, the Board also voted on changes to Section 8 (see minutes) which restricted what Board Members could do while they were on the board. This is all reasonable stuff. This passed unanimously. 8.1 (2) and 8.3 wording is strange - it seemed to be specifically worded for Jay Whipple and Fast Results/Common Game (his name is well known for those two activities so I will mention it), there's also some other board members that do work with results (less well know so I won't mention names) who may be affected - they probably didn't realize that they could be affected. This may cause problems down the road. 8.3 is broad enough that ACBL could shut down Jay's Fast Results (as ACBL offers same), and Common Game (if ACBL offers same) and also affect those board members that offers better results (as ACBL also offers results). Let's hope that management/board applies common sense and doesn't apply this to existing programs; we can only surmise what may happen to future programs that are offered by board members. There are other programs that have been developed by board members that are limited in scope (i.e. unknown to most of you) that are also potentially affected. Again, let's hope common sense prevails.

5.8.6 and 8.1.3 does not work because ACBL does not (at least did not, I recommend changes some time ago) identify what is, and is not, confidential on documents/discussion. Without an explicit information classification system (it can be very basic), a BoD member may not know what is confidential and what is not. This caused huge problems with the ACBLscore+ project with at least two BoD members mishandling information that should be confidential to great detriment and cost to the project.
Jan. 14, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
Ed,

This discussion topic can get very complicated, very quickly. You really needed an experienced lawyer to be able to describe all the different legal situations. IANAL. Through previous work, I'm a little familiar with the policy issues, but I am not an expert.

If the data crosses through a country, as it would for transmission to a military base, then it could be covered. It's complicated.

The old policy was broken. There were issues.

Not only do you have to worry about transmission of data, there is also the issue that the ACBL member database contains PII about EU citizens, irrespective of borders the data may cross.
Jan. 11, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
http://web2.acbl.org/codification/MPBOOK.pdf is dated March 2013.

Tech Files->MP Book for ACBLscore 7.90 is dated Rev 3/2010.

It describes the Masterpoint Calculations (but not assignment or eligibility).

Both versions seems to be out of date.

It doesn't make sense to re-write large parts of ACBLscore just for the sake of re-writing them.

The discrepancy between MPBook and implementation are usually small enough that most people won't notice or care. It is only if you try and implement new code from it that you will see the differences.

The formulae, particularly the implementation for large bracketed KOs, where there are middle brackets that don't have 16 teams, does have an interesting impact. Only really applies in Gatlinburg.

Jan. 11, 2015
Nicolas Hammond edited this comment Jan. 11, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
For example, at the Atlanta NABC, there were rooms that had no WiFi or cell phone service. They were too far underground.

Going forward, ACBL TDs were going to be equipped with Internet tethered phones.

There is a fine line between the current technology in use, and, for example, the technology that WBF uses. We have to be mindful of the price of our tournaments (and clubs) and the technology ability of the TDs. Asking a TD to set up a WiFi hotspot, tethered to a hotel Internet connection may be trivial to a techie, but is complicated to a non-techie to trouble shoot.

We started to see some of that change in Phoenix where ACBL had much more of an IT presence.

Tournaments will start to make more use of technology. It is inevitable. WiFi is part of it.

The thread started out with Masterpoint discussion. Ultimately we need Ping's suggestion - upload game file and recheck. That was the ACBLscore+ design. Locally display the expected masterpoints but verify on the main server before adding to main DB.
Jan. 11, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
As you can probably tell, I have some opinions on the various issues.

If the business decision was purely on technical merits, or perceived costs going forward, I would have no problem with the decision. However, I believe the decision to de-rail ACBLscore+ was made because of the copyright wording (or lack thereof) in the original ACBLscore+ and ACBL league counsel and management trying to cover this up (they were the ones that negotiated the original contract). My lawyer doesn't believe this is an issue. The problem for me is that ACBL league counsel and management tried to cover up the copyright issue. Their attempt to force the renegotiation (not paying invoices etc.) of the original contract, not providing specs etc. has caused delays with the software.

ACBL made little effort to hire programmers during the ACBLscore+ contract. They had used local recruitment agents, but could not find anyone that knew Bridge and Ruby or Ruby on Rails. I told them that they were looking for the wrong criteria. ACBL did not post any job openings in the Bulletin, which would have been the best way to find a developer to take over and maintain ACBLscore+. What they really needed for someone who had a Computer Science (CS) degree and had 2-5 years post-college experience and some knowledge of LAMP (Linux/Apache/mSQL/Perl/Python). RoR (Ruby on Rails) is nothing more than LAMP with the P (Perl/Python/PHP) being equivalent to R (Ruby). Lost everyone on the jargon, but they needed to hire a CS grad with 2-5 years post-college experience and LAMP knowledge. Given that, the new programmer could learn the rest on the job. They also probably needed to hire a similar person with 5-10 years background so they had one senior and one junior programmer. Trying to replace Jim (developer of ACBLscore) with one person is very difficult, almost impossible. A replacement has to know Bridge, movements, Swiss teams, finances, file formats, pair events, scoring, DBs, etc. etc. Irrespective of the software platform, a replacement for Jim is likely to be at least two people. ACBLscore+ was at a point where it did not need a senior architect, most of the remaining work is of the cut and paste (C&P) type, i.e. there is similar code here, just copy it, make some small changes, and integrate it.

If ACBLscore+ work is to continue, then ACBL must look at the various alternatives:

1. Hire programmers to take over the ACBLscore+ work.

To some extent ACBL have made this more difficult.

By May 2014, the ACBLscore+ contract had finished. However, per the terms of the contract, “Upon the request of the ACBL, Nicolas Hammond shall provide up to ten (10) hours per month of his personal services for twelve (12) months following the Completion Date to assist the ACBL in the implementation of ACBLscore+ at ACBL tournaments.” In other words, I had to make myself available to help ACBL with integration work.

When ACBL failed in their attempt to re-negotiate the original ACBL contract, ACBL decided that they did not need my services and “terminated” the contract. There was no cost to ACBL to keep me “on retainer”, unless they decided to use me. Either ACBL were being exceedingly nice to me (for future work I did not need to disclose my contractual requirement to ACBL), or this was done in a fit of pique. I suspect the latter. ACBL had cut off its nose to spite its face.

This decision made hiring, and training, any new employees even more expensive. I've previously written that there is a learning curve for all new hires with any software, not just ACBLscore+, and it is about 3 months.

If you only have a hammer, then every time you see a nail, you will hit it with a hammer. If you have different size nails, eventually you learn that you only need to carry one hammer, the big hammer. You hit any nail with the big hammer. It works. If someone gives you a screw, you will hit it with a hammer, because you don't know anything different. If someone gives you a screwdriver and a screw, you will probably hit the screw with the handle of the screwdriver because you only know how a nail works. If someone shows you how the screw and screwdriver will work, then you might use it, but why - the hammer is so much easier and it does everything. If someone gives you a Philips head screw, then the screwdriver doesn't work, so you hit it with the hammer, or the handle of the screwdriver. If someone gives you a Torx screw that holds a laptop screen to the case, then you will just hit the Torx screw with your big hammer because that's how things work. If someone builds a system with screws, because they know that users have a power screwdriver and it will work better/faster, but you only know about hammers, and have never seen a power screwdriver, then it is hard to see the advantage of screws.

Similarly, software developers have different tools.

Newer software is much easier to develop new code in. I see that ACBL have committed to spend $600K trying to match some of the features of ACBLscore+/Bridgescore+. It makes little sense. They already have the code in ACBLscore+ to start KOs, create an ACBLscore Game File etc. etc.

I'll post on a separate thread how easy it is with new tools to create some of the same functionality. Hopefully it will save the ACBL $600K.

This is one of the reasons why ACBLscore should be replaced with modern technology.

For ACBLscore+, ACBL need to find 1-2 programmers, with some LAMP knowledge, and interest in card games. RoR is easy given a LAMP background.

2. Rent.

I've worked on Bridgescore+ since the end of the ACBLscore+ contract. I've made some enhancements to improving the code, particularly for running events at tournaments and tournament management.

I've made this available for free to Districts (see my previous post for my motivation) and will do so while my marginal cost is close to $0.

After that, one option might be that I just rent its usage (I put it on the Internet) to ACBL for use at its tournaments.

3. Interim usage.

ACBL have written off all of ACBLscore+, however this is an accounting issue, not a technical one. ACBL still have the code.

Are there parts of ACBLscore+ that can be used?

Throwing it all away makes no sense. For example, there is code that helps start a KO, then can create an ACBLscore game file. Why spend $600K re-creating this software when you already have it?
Jan. 11, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
Irrespective of which committee, the items which I expect most people to be interested are the main business questions going forward. Sure, everyone may want a scapegoat for what happened, but not likely to get one.

The main business questions are:

What is the expected cost of ‘finishing’ ACBLscore+ so it can be released? I've addressed some of this at http://bsp.bridgescoreplus.com/?page_id=44

What is the expected cost of enhancing ACBLscore to match the features of ACBLscore+?
ACBLscore+ has some features not in ACBLscore that ACBL are now trying to match.

What is the expected cost of maintaining ACBLscore+?
This is a difficult question because it involves time/money/resources. ACBL have little knowledge of the power of ACBLscore+.

How does this compare to the code of maintaining ACBLscore?
Again difficult question. How hard is it going to be to replace Jim. How many people are needed to replace him. Can they maintain his code?

What is the expected cost of improving being made to ACBLscore+ (e.g. Bridgemates in Swiss events) against the cost of improvements being made to ACBLscore?

What does the future look like with ACBLscore?

What does the future look like with ACBLscore+?

What is the training time for a new TD (does not know ACBLscore or ACBLscore+) to learn a new system?

Will ACBLscore+ save ACBL any money (automates repetitive tasks)?

Does ACBLscore+ improve the player experience?

Can any parts of ACBLscore+ be used?
Jan. 11, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
The Board of Directors (BOD) committee is different.

The first thing I would do is to get the names of the BOD Technology Committee, then add them to the CEO Technology Committee. It would halve the amount of work the BOD Technology Committee would need to do and save the members who are on both committees from going over the same information twice with different groups.

I'd expect that BOD Committee to have a much wider purview, and also not be restricted in the information that they see. For items I listed in the last post on the CEO committee that may get provided to “independent” members, these should all be provided to the BOD committee.

The BOD needs to know what the problems were, how they were caused, what is being done to make sure similar problems do not occur in the future. They also need to review the decision to drop ACBLscore+ and the reasons behind it.

Here's the set of questions I'd be asking if I were on the BOD committee:

1. Did ACBL use outside counsel when creating the original ACBLscore+ contract?
If so, can we see the outside counsel's comments on the ownership/copyright issue.

2. Did ACBL use an outside counsel to review ACBL's various technology contracts after the ACBLscore+ contract was signed?
If so, what was the outside counsel's comments on ownership/copyright?
I'd want to see the original emails from outside counsel to ACBL.

3. When did this review occur?

4. Were the contents of this review reported to the board?

5. Did outside counsel recommend any changes to the original ACBLscore+ contract?

6. Did outside counsel recommend that if changes could not be made to the original ACBLscore+ contract that ACBL might face issues with using ACBLscore+?

7. When was this reported to the board?

8. Did ACBL make it a requirement to the contractor for any future work on ACBLscore+ to be contingent on renegotiating the original ACBLscore+ contract?

9. When did ACBL stop paying the ACBLscore+ outside contractor's monthly invoices?

10. How long did this go on for?

11. Why did ACBL stop making payments? Was the outside contractor told why?

12. What was the impact on the ACBLscore+ contract of not paying invoices?
(ACBL won't be able to answer this one properly - it was major).

13. Did ACBL management report to the board that ACBL had stopped making payments? Did ACBL management give the reasons why?

14. Did ACBL "provide the necessary information for to perform the work“?
(Wording is from the contract).

15. What were the list of specs/other items that ACBL failed to deliver?

16. What was the duration of each delay in providing the specs? What was the cause of each delay? What was the impact of each delay? Why were there delays?

17. At any point during the contract, did ACBL deny access to ACBL employees for Contractor to perform the work?

If I were answering this one for the board, I'd check with all the managers on the ACBLscore+ project, then I'd double check with them to make sure that they were giving me the right answer. Then I'd check for a third time. Then I'd check with the employees.

18. Did ACBL management report the problems from the two March 2013 incidents to the ACBL Board at the July 2013 Board Meeting?

19. What was the impact on the ACBLscore+ project from the March 2013 incidents?
(Not expecting ACBL to be able to answer that one, the impact was substantial).

20. ACBL were supposed to provide 1.5 staff for the ACBLscore+ contract. Who were those people? What work did they do? There should be 3 man years of work, where is it?


Now the important questions:

1. The ACBL counsel that negotiated the original ACBLscore+ contract and subsequent outside counsel have differing views on the importance of the Copyright issue. How does ACBL resolve these differences?

Until this is done, ACBL are not going to do anything with ACBLscore+.

2. To what extent was the ”technical" decision to drop ACBLscore+ based on outside counsel comments on copyright?
Jan. 11, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
I haven't seen a remit for the two Technology committees, but let's assume that they are supposed to look at what happened, what is happening, and what should happen. The past, present, and future.

This thread was on ACBLscore+, not the other ACBL Technology programs, so I will focus on it. We will start with the past and focus on the CEO Technology Committee.

We already knew the outcome of the CEO Technology Committee. 6 of the 10 members were on the committee that decided to drop ACBLscore+.

For the new “independent” members of this committee, it is unclear what information they will be given.

I suspect that the committee will attempt to justify the “technical” decision last year to drop ACBLscore+ so if I were on the committee, I'd want to see the following:

1. ACBLscore+ running on Windows.

I've posted statements on BW about the difference between running in production mode and developer mode. Logging is also an issue that can slow a system down. I'd want to see ACBLscore+ running in production mode, with developer logging turned off. I'd want to see it on Windows XP and Windows 7, the two most common platforms for ACBLscore+. I'd want it configured with the full ACBL member database (310K members). Basically testing a real world platform.

Other than TDs, few people know the full extent of ACBLscore. So I'd want some idea of what I should be testing. I'd want to know what new features/capabilities are in ACBLscore+ so that I can compare.

You've heard me post about the missing specs from ACBL, so I'd want to know what's working, what's not. I'll ignore the reasons why something isn't working for the moment; I just want to know what to expect from all the different components that make up ACBLscore+.

This information is in the various status reports and the project Wiki and some Exhibits.

The project Wiki gives information on how to set this up.

I want to see how it runs in a real world environment. I'd probably want to take it to my local club and a tournament to see how it compares. I might need a TD to help with the tournament financials.

It would help if ACBL can pre-load tournament and club data (easy to do, details on Wiki).

2. ACBLscore+ running on a Mac.

This was an initial requirement, so I'd want to see that ACBLscore+ runs on Windows XP, Windows 7 and Mac.

ACBL have all the code for Windows 7, XP, and Mac (and Linux). The Wiki gives all setup details.

3. Testing

ACBLscore+ comes with built-in testing, so I'd want to see how this works. Easiest example is to take all the game files from a regional, import them to ACBLscore+, re-score, re-rank, re-masterpoint and see the differences between the output of ACBLscore and ACBLscore+. The test code is in ./lib/tasks/admin/test_gfs.rake and the sample game files are in ./examples/gf/regionals.

The ability to have an automated test suite is one of the advantages of ACBLscore+.

We've just seen a recent example of what happens without an automated testing ability.

4. Source code

It is unclear if ACBL will give access to the Committee to the ACBLscore+ source code. If I were asking to evaluate the software, I'd want to see it. I would want to get some idea of the style of the code, how easy it would be to “finish” the code. The majority of the code is Ruby/Ruby on Rails; the masterpoint code is in C. There's about 200K lines of code so a lot to look at. There's other code in C. A quick look at source code can typically give a good consultant an idea of the quality of coding/comments and the expected cost of new code.

For comparison, I'd like to look at the source code for ACBLscore.

The ultimate business decision that is going to be made is it better to continue with ACBLscore or should we work on ACBLscore+.

Knowing the state of both code bases, the practicalities of working with each solution, the ability to hire new staff to work with the code, are crucial for the decision making process.

5. Utilities

ACBLscore+ is not just a single program; it is a collection of different programs. One tool - gfprint - can read game files, output data in XML, then read that XML and create a game file. This tool is very useful for a variety of applications, including fast results, automated testing etc. I'd want to spend some time learning this tool, how it could be used.

6. Design

Source code is one thing; I'd like to understand the design of the underlying system and how it compares to current. This is going to take a little time to research.

If you get this opportunity, start with ./db/schema.rb and the files in ./db/migrate. This will quickly give you an idea of the underlying data structures.

Understanding the design of the software is important, because it comes back to the business decision of what is the best way forward.

Rails can create a flowchart. See ./notes/documentation on how to create documentation.

I'd want to see the erd.pdf, and the source code documentation.

7. IDE

What Integrated Development Environments (IDE) are there for working with RoR?

My experience with ACBL is that there is very little (none?) internal use of modern tools, modern IDEs that can make developing and maintaining code orders of magnitude better and faster than a terminal and basic word editor.


The above is what the CEO committee is likely to focus on.

The bigger question, which everyone probably wants an independent answer to, is what happened to cause ACBL to drop ACBLscore+. I doubt that the CEO Technology Committee will be given the information needed to answer this one correctly.

If the committee is tasked with finding out what went wrong with the ACBLscore+ project, then as an independent person on that committee I'd expect the following:

1. Access to the Monthly Status Reports (MSRs)

They are dull. There are 24 of them. Nothing very interesting in them, normal type status reports for a project of this size. But, you've probably got to read them, or parts of them. No reason not to release these to whoever wants them, except for the last 3-4 reports. These did not have the wide audience of the first 20 or so and were more specific.

2. Access to the Executive Status Reports (ESRs)

There were some months where I provided both a Monthly Status Report, but also an Executive Status Report which went only to ACBL Management.

I've seen prior discussion on if the Technology Committee should be confidential. Absolutely. These Executive Status Reports mention individuals by name.

I am sure that Robert (Hartman, ACBL CEO) will do his best to protect ACBL employees. I would support him in that regard. A long time ago, we both decided that “name and shame” is not going to be helpful to anyone. It was clear some time ago to both of us that there were some serious shortfalls with some ACBL staff. Robert has admitted short-comings with ACBL, and, to his credit, has taken personal responsibility for these problems to protect his staff. Robert did this both at the March 2014 ACBL Board Meeting, and also at the Phoenix ACBL Board of Governors Meeting. One of the things that everyone must agree on, is what is the purpose of the Committee. Name and shame should not be its purpose.

Releasing these Executive Status Reports, which I intended to be confidential, to outsiders is something that should be taken with great care. But, this is an ACBL issue.

The Executive Status Reports list the various problems that were occurring in more detail.

3. Private letters.

There were some private letters that I sent to Robert.

These would be even more confidential than the ESRs.

The first letter was regarding two separate (ACBL-caused) incidents in March 2013 that had serious consequences for the ACBLscore+ work. That letter doesn't mention names, but one can be deduced from the context of the problem.

I would expect Robert to have reported these incidents to the ACBL board at its next meeting in July 2013, probably without names. I assume the ACBL board are aware; releasing the issue/names to a wider public won't help though I'm sure everyone is now curious.

There were other letters, one in January 2014 comes to mind. I expect Robert to have informed the Board; but releasing them to the CEO Technology Committee is probably unlikely.

4. Emails

There are obviously a large number of emails back/forth. There is going to be little interesting in them. Most of the relevant or important emails ended up on the Wiki.

5. Wiki

We set up an on-line Wiki for the ACBLscore+ project. This was the main tracking mechanism for the project. After the contract was over, we took it down. A copy of all data was delivered to ACBL on DVD at the end of the contract. It should be trivial to set up the Wiki. This Wiki would be very useful to the Technology Committee to review as it will greatly simplify their work.

I've referred to missing specs, and other issues.

If I were an independent person on the committee, I'd like to see what they are, what the scope of the problems were.

On the Wiki, go to the top level items marked “ACBL Integration”, “ACBL Decisions”, “Board Decisions”. These track the various issues, along with dates.

The Wiki was open to all developers, and ACBL, there is little/nothing overly confidential on it.

6. Internal status reports

This was the most high profile project within ACBL.

I will assume that there were internal status reports generated on the work that ACBL was supposed to do as part of the ACBLscore+ work. I'd like to see them.

7. Talk to TDs.

ACBLscore+, or at least parts of it, have been used at different tournaments in the past 2 years. I'd want to talk to the TDs at those tournaments to get their feedback. Nothing can be better than unbiased comments from real world usage.

I've probably done more tournaments in D7 than elsewhere so I'd want to talk to the D7 TDs and also the TCs for D7 tournaments that have run ACBLscore+/Bridgescore+.

I'd also want to talk to the DICs. There's probably 4-5 of them. Ask what they think of ACBLscore+/Bridgescore+ when it has been used.

8. Did contractor do his job?

I'd like to see the original scope of work. It's in the contract but ACBL may not want to release. They can redact confidential items from the contract and provide a copy. Exhibits A, B, C, F, G can be provided. (D is financial, E is the copyright/ownership). Exhibits B and G are the main Exhibits because they lists the phases. Nothing particularly confidential in them, so copies of B and G at a minimum.

The work for ACBLscore+ was stored on the Internet using Github. ACBL should have been downloading a copy at least one a week, but certainly once a month, as part of due diligence to make sure that they always had a copy. (I never cut off access to ACBL to this site, even when they were not paying their bills). I'd want to see a copy of each week's code download to make sure the contractor was delivering.

Given the make-up of the committee, the expected report on ACBLscore+ can be summed up:

1. We (ACBL) made mistakes. We acknowledge them, we are working hard to rectify them and make sure that they don't happen again.
Jan. 11, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
I've received the threat. ACBL league counsel can verify. My lawyer reviewed it, did not seem to be anything for me to worry about. I assume that it was reported to the BOD in Providence. Not going to post here.
Jan. 11, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
A learning curve was expected.

Better way of looking at it:

A new CD, who knows nothing about ACBLscore or ACBLscore+, which is quicker for them to learn?

The original UI has evolved from 30 years ago as new features were added and new functionality put in. I doubt you would design the current UI given today's tools.
Jan. 11, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
The old policy was broken. There were many issues.

This policy stuff was something that I started working with ACBL during the start of the ACBLscore+ contract, so about 3 years ago. My issue was that I did not want my company liable for transmission of data which could result in a privacy lawsuit. The new ACBLscore+ was supposed to have one database that contained all the ACBL members (about 310K names, with about 170K active members). No more downloading local Unit files, or merging local data, or all the other problems that club managers have with their data. Lots more tools to help with getting data off the system.

Let me try and give broad details:

The ACBL member database contains various information about you, including phone, email, ACBL membership data. It is transferred from ACBL to various places.

The database contains information on European Union (EU) members.

The database crosses state lines, and country lines.

The database can cross over into EU, e.g. ACBL club at an army base in Germany.

The database contains information on minors (different definition in different states).

The database has to be protected under different EU and US law.

The laws on protection of information on Personally Identifiable Information (PII), particular on minors, can result in severe financial penalties.

It is hard to find someone knowledgeable about all of the above, sufficiently knowledgeable to be able to give clear and concise legal direction given the complexities of minors, EU, US, Canada, Mexico, Bermuda law.
Jan. 11, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
As a Unit Officer, you can request member data. The rules of use are (from MyACBL->Member Rosters).

“Rules
The list or ACBLscore Player Information file requested from ACBL is not to be used for personal or commercial use. It may be used solely for ACBL Districts, Units and Clubs to promote ACBL affiliated bridge activities.

These lists are not to be shared with any third parties. No information from these lists may be published on a web site, bulletin board or on similar resources without the express written permission of the ACBL member whose name is identified.

Violation of these rules may subject the violating party to discipline or sanctions under ACBL Rules and Regulations”

A strict interpretation of this rule is that you cannot create a local directory using the ACBL member database without getting written permission from each member.

An even sillier interpretation is that you can't print results from your Sectional on the Unit web site where you list the player's name and city/state because the correlation from your ACBL player number to your city/state is from data in the file you need to download.
Jan. 11, 2015
Nicolas Hammond edited this comment Jan. 15, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
The testing of ACBLscore is difficult. It is all manual testing, there are no automated tests.

The masterpoint code has been through lots of iterations. This latest iteration (Jan 1, 2015) is Masterpoint Revision 13 in Jim's tabulation of masterpoint revisions. In other words, he's been through at least 13 major masterpoint revisions with ACBLscore.

I feel for Jim. Previously releases would come out; if there were bugs they would be quietly fixed within the ACBL family and players would be none the wiser. With the Internet, his bugs are unfortunately more public. Jim is very proud, and rightly so, of ACBLscore. When I visited ACBL, Jim was always one of the first people there, the time between 6am->8am was usually the most productive for me because I could sit with Jim and go over various issues. I am sure that he is mortified that there were problems. The fact that they were fixed quickly is a testament to Jim.

There are so many variations with Masterpoints that it is impossible to test each and every situation. Also, it is very slow with ACBLscore to create a new test scenario. In other words, if it is 43 clicks (or whatever the number is) to create a new event that is going to be tested, then it takes an incredible amount of time to create the different scenarios for manual testing.

It is very easy for anyone to make the recommendation that ACBL needs a test department.

For those that have been involved with testing, test departments, it is very hard to find long-term employees.

What ACBL needs is someone with solid Test Management credentials, but I doubt they will find, or hire, this person any time soon. The concept of white box, black box, grey box, TDD, test coverage are all alien.

Also, if you look at the variety of ACBL products/services, there is a very broad range of services from ACBLscore to TourneyTrax to interactive web site. They each involve a different set of testing criteria.

The Masterpoint code within ACBLscore is some of most complex code. The formula has been tweaked at least 13 times. There are 3 similar, but different, software “modules” (different *.PAS files, not ‘modules’ in the classic sense) that do masterpoint code. For example, the masterpoint chart code is different than the scoring code, so you may have to make the changes to two different places.

Trying to replace Jim is difficult. He's been the only programmer for 30+ years on ACBLscore. ACBL have been unable to replace him. Jim is probably past 65, if that is the modern retirement age, but keeps working.

The structure of ACBLscore makes automated testing difficult.

There are some things that ACBL could do.

For example, autohotkey allows for automation of repetitive key strokes/data input. We used it in the ACBLscore+ contract to automatically generate the output from ACBLscore of the various movement files. I believe that some people use it for automated results generation from ACBLscore. Using autohotkey would allow for rapid creation of game files and speed it testing.

Another feature within ACBLscore is the DOS Macro function. Properly implemented this can speed up a lot of repetitive tasks within ACBLscore. Few people use it. The most prolific ACBL TD user retired recently so many of his scripts may have retired with him.

As part of the ACBLscore+ contract, we developed a software tool, ‘gfprint’ or simply ‘gf’ that takes a game file, and converts to XML. It can do the reverse, take the XML and create a game file. This is a very powerful tool for large scale processing of game files. Very useful for testing. We used it extensively. ACBL seem to have a NIH mentality so I do not know if they use it or not.

Setting up any of these test environments would take time. You then have to implement a process to make sure that a software release goes through a testing cycle. If you are the developer, or the test engineer, then you need to make sure that each time there is a software release that all the tests are updated. If there is a ‘bug’ you have to figure out if it is in the test code, or the in the production code. ACBL does not have that development mentality. What is done has worked for some time. Occasionally there will be blips, as happened recently.

Masterpoints are complicated. During the ACBLscore+ contract, we asked for the Masterpoint Assignment and Eligibility specs. There was over a 6 month delay in trying to get someone at ACBL to start working on providing these specs. In the end, they never did deliver. I still think it would be very useful for players to have all of the masterpoint calculations, assignments and eligibility rules in one place.

There is the MPBOOK. This defines some of the calculation rules. But ACBLscore implements what it wants and doesn't follow the rules in the MPBOOK. As no-one has ever complained, what should probably happen is that the MPBOOK should be updated to reflect the actual MP rules that are implemented in ACBLscore, and not what the BOD passed. Those that have read the MPBOOK can see the places where the rules are not followed, and quite frankly, how many players really care to check? Making ACBLscore follow the rules in MPBOOK would be too much work at this point. Much easier to update the MPBOOK to reflect reality.

I also don't think players realize (or care) about the difficulty of calculating the masterpoints. How often do you as a player ever check the calculations behind the masterpoints assigned to you from an event?

As part of testing for ACBLscore+, we took the data from an ACBL tournament, ran it through ACBLscore+, re-scored, re-ranked, re-masterpointed using ACBLscore+'s implementation of masterpoints. We rarely have any discrepancies in the re-scoring, or the re-ranking, but lots of discrepancies with masterpoints. This is either a bug in ACBLscore, a bug in ACBLscore+, or incorrect configuration by the TDs. For ACBLscore+ we automate a lot of the calculations that must be done manually by a TD when calculating MPs for an event.

What we found for the tournament was that 10% of game files have masterpoint errors in them. Usually it was a TD mis-configuration, because it is very easy to mis-configure ACBLscore for a “complex” event for masterpoint calculations/awards. In some cases the errors was small, e.g. less than 0.1 masterpoints, but in some cases it was large (2+ MPs), or the wrong depth calculation which means either more people got masterpoints than they should, or far fewer.

The current ACBLscore process for masterpoint calculation/assignment/eligibility is manual. Always has been. There are requirements on ACBL TDs to manually check the results for MPs by using the masterpoint chart functionality but not all do as it is a very time-consuming process.

I agree with the comments about Doug Grove. He is one of the best. Another one that is retiring soon. He has the best Tournament cash tracking software (Excel spreadsheet) that I have seen. This should be used by more TDs. I hope that this does not get lost before Doug retires.


Jan. 11, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
There is no equivalent testing in Turbo Pascal.

All of ACBLscore testing is manual.

Rails has a great testing framework. Some of the ACBLscore+ work was Test Driven Development (TDD) where you write the tests first, then write the code. We also added lots of additional testing, e.g. take all game files from a regional/sectional, import to ACBLscore+, re-score, re-rank, re-masterpoint, compare to ACBLscore.

The “Personal Web Server” is a different topic. This is one of ACBL's way of trying to dismiss ACBLscore+. Any attempt to use a local service means the software won't work. You are correct to rant.
Jan. 11, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
During the ACBLscore+ work I went to many clubs, of various sizes, in different parts of the country. There are 3,000+ ACBL clubs. The majority are small. Typically they rent space from a community hall, church basement etc. Quite a few did not have Internet access. I don't see this changing in a while.

For many tournaments, the rate for the hotel WiFi can be prohibitive. This is the rate the convention center charges, not the (free) WiFi you get in your room. Often TDs are taking the results on a stick back to their room so that they can upload to Fast Results etc.

I don't see in the near (3 years) future this changing. It is still a reasonable requirement that the Bridge Scoring Program (BSP) does not require Internet access.
Jan. 11, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
http://www.acbl.org/tournaments_page/nabcs/past-nabcs/nabc-casebooks/

Last one posted is Memphis 2012.
Jan. 2, 2015
You are ignoring the author of this comment. Click to temporarily show the comment.
Jeff was fantastic. I don't think people realized how hard he and his team worked behind the scenes at NABCs. He took little/no personal credit for the work. He was always very protective of his staff, and would not tolerate abuse by players. Most players probably don't realize who he is/was. He often handed out Bulletins when you got to the top of stairs at NABCs. Jeff negotiated most of the NABC contracts - I know some of you have complained but he was always on the look-out for the best deal for everyone including ACBL. These are sometimes negotiated 5-10+ years in advance.

He will be sorely missed by ACBL. Condolences to his family. He was one of the best.
Dec. 31, 2014
You are ignoring the author of this comment. Click to temporarily show the comment.
Robert, Tony, Ken, Bruce, Merlin, Russ. Surprised no-one pointed it out before. I don't know who else was on the committee that looked at ACBLscore+ - I wasn't involved or asked.

Tony is new; he's a consultant ACBL brought on-board in March 2014. Never met him. Talked to him in March, seems like a sharp guy. Tony is not a bridge player, but don't hold that against him.

Bruce is ACBL IT Manager (apologies if I got his title wrong), if it's a working committee he belongs on it; if it's a supervisory committee, then they are really supervising his department. He is not a bridge player. He joined ACBL about 2+ years ago.

Ken is a TD, works v. hard, he does not have a tech background. He was i/c of roll-out of TourneyTrax (not the developer/consultant on the project, but the ACBL face to the project). He got added to the ACBLscore+ project towards the end.

Merlin created a pre-cursor to ACBLscore (yes, that long ago!). He does work on results in D20. He was on the ACBLscore+ email list and Wiki etc. but not involved with it.
Dec. 31, 2014
Nicolas Hammond edited this comment Jan. 1, 2015
.

Bottom Home Top