For the upload format, I think JSON is the way to go. (XML is ). I don’t see where the size for that would be an issue. I made a WinTD JSON file from the full SuperNationals II file. The raw JSON file is 7.5Mb, zipping to 600K. And that has, by my guess, about 6 times the data that would be needed for a tournament submission, unless you want to get into the weeds of how a game record actually came about. (WinTD keeps a single game record for each game, so if you change the result, it is reflected for both players. That requires keeping track of which round it is for each player in case of a cross round pairing, and also requires a field to handle the situation where you actually need mismatched results.)
As long as you keep something like the current
WnnnnC
coding for a game, I would think it wouldn’t be all that difficult to implement for any program that can generate the current DBF files. Certainly JSON libraries are easier to come by that DBF ones.
dBase was a good format for the 1980’'s I guess. Does anyone still use Clipper?
Compile-in PHP support for dBase files was dropped years ago, though I think there’s an add-on module for it now, but I don’t think I’ve tried looking for it for either PHP7 or PHP8 lately. PHP8 broke so many things that I’m ready to move on to other tools, and I’m a programming dinosaur.
I’ve spent some time playing with Python and Ruby, I’m just not ‘getting’ either of them yet.
It was not difficult to produce dBase in JavaScript without any modules when I gave it a shot, although JSON or a delimited (please | or something not a comma) format would be even easier.
I find myself using the all-ratings tab-delimited file more than any of the other supplement formats. Given the crickets when I mentioned an all-ratings PDF format test I created, I’m inclined to recommend that Leago not bother with that supplement format in their initial release unless there’s a lot of requests for it once we enter testing.
The biggest limitation with the tab-delimited file is that it has around 1.2 million rows in it, and that blows Excel out of the water.
I’ve written a few programs that built a dBase formatted file, and one that dumps such a file in human-readable form.
Back when I was doing work in the book industry 25 years ago, I served on the BISAC EDI format standards committee, some of the EDI formats were a lot harder to parse manually than dBase files.
Actually, my invoicing program is a home-grown dBase program written in the early 1990’s originally. So far, it ain’t broke and there must be enough legacy apps to keep dBase LLC in business. 30 years ago, the advantages of random access vs sequential was quite important.
Somewhere I have a copy of Clipper as I used that to reprogram the rating software. (I wonder if there is still a demand for COBOL programmers. IIRC, there was a state unemployment system that was still using it just a few years ago).
I understand that COBOL programmers with the right background are still in high demand, I just haven’t felt like that’s contract work I wanted to pursue. Figuring out someone else’s spaghetti code is a lot of work. (I wrote COBOL programs from the mid-70’s until the early 90’s when I switched to doing SQL work, originally in Oracle.)
Whether AI has had an impact on that yet is unclear, I’ve seen some AI-written code that looked pretty good. It is possible that maintaining existing COBOL programs is not as easy for AI tools.
I actually download the all ratings file (as well as the FIDE XML rating file) and immediately copy the parts that I want into DBF files. I have perfectly good (and efficient) code for indexing and searching DBF files. It may not be a good transport format, but for a fixed database task, it’s fine.
…I haven’t messed with FIDE, but I take the all ratings file in its entirety and drop it into a database (Postgres or DB2), at which point indexing is the database’s problem and searching is just SQL.
Does the difference matter between 1 and 2C actually matter all that much now? v1 was a complete mess back when tournaments were submitted on a combination of diskette plus printed xtable when no name (by design) plus blank ID meant someone at the office had to match up a new ID with the name on the cross table. Now there would just be a number of fields that would be blank rather than being filled in based upon the uploaded files. More work for the TD, but not really more work for the office.
Is it still permitted to submit results on paper? A TD (who shall remain nameless) famously did all his local tournaments on paper and refused to go home and enter them into software for submission. I pointed out that he felt he had the unalienable right to make someone at USCF punch his information into a computer rather than doing it himself.
Good luck finding the form on the website, if it is. I tried to find the paper submission form recently as I recall it being the only official source I ever saw of the deadline to submit a rating report after an event, but I was ultimately unable.
It is cunningly hidden on the 'tournament director’s page, which there is a link to at the bottom of the page at www.uschess.org (and many other pages), but hasn’t been updated with the St. Louis address.
The last time I asked, the ratings staff indicated that they still do get a handful of rating reports in the mail every month.
Cunningly hidden? If you’re a US Chess Tournament Director and you don’t have the US Chess TD Page bookmarked, or know where to find it, then to me that speaks more about the TD rather than US Chess trying to cunningly hide it, in the place where it belongs.
This is total madness. If US Chess was formed a bit earlier, I’m guessing it would still be employing Morse code operators to receive telegraphic rating reports. Can someone just cut this stuff off? There shouldn’t be an option for paper or diskettes. The organization in willing to alienate 50% of it’s TDs over safe play but can’t make the easy decision to stop supporting these ridiculous formats. I don’t get it.
While I am fairly certain that Michael’s “cunningly hidden” was at a minimum tongue in cheek, that form is not particularly obvious as the paper submission process on the TD page, nor is it something that is obvious after clicking on “Tournament Rating Reports,” the place that I would expect to find information on, well, Tournament Rating Reports and processes.
I concede that I have not submitted and do not expect to ever submit a TRR on paper, but I find it odd that while Chapter 10, Rule 15A is quite clear about the time to submit a tournament rating report for an online event (three weeks), that there is no corresponding rule giving the seven day number anywhere in the rulebook* for a non-online event.
*At least not in the 7th edition as published or the 2024 updated Chapters 1/2/10/11.
