Software Idea to build databases, have you heard of it and would you buy it?

[podonne]

Hello,

I see this question a lot on this forum, how to setup a database, so I thought I might make some slight tweaks to my existing software and sell a little app that would import the various DRF\BRIS\TSN file formats into the database of your choice, MySQL, Oracle, SQL Server, Access, whatever... It would do all the work, create the entries,races,pps,workout tables, setup the indexes, keys, and referential integrities (except downloading the files themselves) and you could start writing queries right away.

So before I try to do that (and have to pay our humble host to tell you about it) Does this exist? If not, would you buy it? How much?

Curiously yours,
podonne

[njcurveball]

Just my opinion, but if you could package it somehow for "free" and get a kickback from the data companies it could work.

As for selling it as a stand alone, your market is either experienced people who like to do things "their way", or inexperienced people who will not take the time to learn to use the database even if you give them the data on a silver platter.

I work with about 3,000 technical people and my experience is that if someone has an option of taking 40 hours to do something with a familiar tool (say Excel) and has to spend 20 hours to learn a new tool (say Access) that will not only do the same thing in 1 hour, but also offer significant time savings and additional benefits, they will 99% of the time choose the familiar tool and look for people who agree with them that is the best way to do it.

I like your thinking! So good luck with the project, I hope you can find a "market niche" and make some cash. :ThmbUp:

Jim

[Tom]

If it would allow me to choose what factors ( I have several specific ones I need) and perhaps allow me to export...I would be interested.

[Pace Cap'n]

As for selling it as a stand alone, your market is either experienced people who like to do things "their way", or inexperienced people who will not take the time to learn to use the database even if you give them the data on a silver platter.

Jim


There is the name for your new product....Data On A Platter

[plainolebill]

Build the front end yourself with pull down menus for the various factors, date ranges, etc. I think it would be more appealing that way.

[podonne]

Just my opinion, but if you could package it somehow for "free" and get a kickback from the data companies it could work. ...

I like your thinking! So good luck with the project, I hope you can find a "market niche" and make some cash. :ThmbUp:
Jim

Interesting thought. I think some of the software sold through HDW is priced this way. I wonder how open they are about taking on new software. Does any one know if those allow data exporting?

[podonne]

Thanks to everyone for the comments. I've actually gotten a bit more excited about this.

While running errands today I though about "smart tables". If you had a standard representation of the program\result data in a database and a lot of people were using that structure, you could have a community working on new kinds of tables that summarize that data in certain different ways. Since everyone is using the same basic structure, everyone can use the smart tables.

Introduces a nice web 2.0 web community feel.

[Tom Barrister]

Interesting thought. I think some of the software sold through HDW is priced this way. I wonder how open they are about taking on new software. Does any one know if those allow data exporting?

HTR exports to Access, so exporting (for private use) must be allowed.

As far as how open they are about taking on new vendors, the Stealth group uses them, so I'd imagine they'll take on anybody who has a customer base.

[amazingD]

most would want to query by any track, odds, or lots of other things like multiple speeds in one race or lone speed... I would buy it in a second. My software works fine but it is not anything like i need to get more specific on situations.

[raybo]

I assume you're talking about a database template specifically for horse racing.

I have stated before that I was surprised that nobody had developed a free horse racing database template (specifically for Access or possibly MySQL free version). There are literally hundreds of db templates available free on the internet, but as far as I can see, none for horse racing.

I know, for absolute sure, if it were free or priced very, very affordably, you would have many takers, I, for one.

[gopony]

I was planning on waiting until the end of year to post it Open Source.
Still a bit clunky but I get by with it.

It currently requires a web server with PHP connected to a PostgreSQL database (So the market percentage is low). Loads main horse information in one table. PP in another table and Charts in another table.

I currently only use it on Linux systems, and I was not planning on supporting Windows versions.

If you look at HTR and HDW both come out cheaper as far as data is concerned though. But for warehousing a single track and searching your own data it works fine.

Usually a year of data on a single track costs about $100 from TSN and Brisnet.
Currently I use TSN PP's and Brisnet Charts.

There is also a problem with TSN data and Brisnet Charts coming with "/" in some of the fields. I have program for removing it but it's also clunky and would require some work (It's a bash script).

[podonne]

There is also a problem with TSN data and Brisnet Charts coming with "/" in some of the fields. I have program for removing it but it's also clunky and would require some work (It's a bash script).

I believe it is the last 3 or four fields have completely random characters in them. It caused me a hell of alot of problems because I was trying to insert the CSVs without any processing (to save time)

[gopony]

Yeah, I remember that one also.
I don't know if I ever did anything specifically to handle this. But I don't think it caused and many problems as the "\" character. But that is one of the things I would have to retest, since a number of times I was doing this as a very manual operation. My import code would tell me where it stopped and I could look at the line and see the problem.
So I would need to add some more rules to automate the process.

[podonne]

Yeah, I remember that one also.
I don't know if I ever did anything specifically to handle this. But I don't think it caused and many problems as the "\" character. But that is one of the things I would have to retest, since a number of times I was doing this as a very manual operation. My import code would tell me where it stopped and I could look at the line and see the problem.
So I would need to add some more rules to automate the process.

I wonder if they get emails about those fields, what the hell are they for?!? I am putting them into mysql, so I tried to escape them, but with them being random, it still errors on 1 or 2 % of them. I just throw the errored ones out though, too much trouble. If I can't automate something I don't do it.

[hcap]

There are problems with the running positions in the PPs.
The "\" and sometimes-if the horse was eased at some point in the race a "0"
Also blanks and spaces sometimes show up. Go figure.


The TSN fields.

576- 585 1st Call Position(if any)
586- 595 2nd Call Position(if any)
596- 605 Gate Call Position(if any)
606- 615 Stretch Position (if any)
616- 625 Finish Position

I import all 1400+fields in excel 1997-2003 and use string formulas to check those fields and if necessary make corrections. I have not noticed any real performance loss. Excel may not be the best program to use as an importer, but if you call built in functions from VBA, instead of VBA routines to do the dirty work, it does work well.

Checking values in cell BL106 for blanks (ISBLANK), a "O", a space character (code=32), and character "\".

Formula might be...

=IF(BL106=0,99,IF(ISBLANK(BL106),99,IF(OR(BL106="\",CODE(BL106)=32),99,BL106)))

I put a 99 in instead. Pretty much removes horse from contention

[podonne]

I put a 99 in instead. Pretty much removes horse from contention

This is a topic that could use some discussion. My gut would be to either NULL ths field, rather than use 99 (or 92, like in the reference) or set it to field size + 1. Using a high number would throw off certain calculations, average finish, pct finish...

[raybo]

This is a topic that could use some discussion. My gut would be to either NULL ths field, rather than use 99 (or 92, like in the reference) or set it to field size + 1. Using a high number would throw off certain calculations, average finish, pct finish...

Don't know what calculations you are using with the running positions , but, obviously, with no running position actually there you need to decide on something that will not mess up your calculations. The formula hcap showed in the previous post would trap any of the characters that might be there. You just need to decide what to replace the "99" value with.

The same formula hcap shows in a little shorter form would be:

=If(or(BL106=0,ISBLANK(BL106),BL106="\",CODE(BL106)=32),99,BL106)

Just replace the "99" with whatever you decide.

[raybo]

Don't know what calculations you are using with the running positions , but, obviously, with no running position actually there you need to decide on something that will not mess up your calculations. The formula hcap showed in the previous post would trap any of the characters that might be there. You just need to decide what to replace the "99" value with.

The same formula hcap shows in a little shorter form would be:

=If(or(BL106=0,ISBLANK(BL106),BL106="\",CODE(BL106)=32),99,BL106)

Just replace the "99" with whatever you decide.

After trying the formula in Excel, I found that the "CODE(BL106)=32" causes an error, "wrong data type". If a "space" is what you want to test for then, instead of using "CODE(BL106)=32", just use: BL106=" " , this change gets rid of the error and returns "99" for all cell values of: 0, a space, \, or a blank cell.

[hcap]

Yes, if you calculate average finish or average of any running position, removing a '99" from the calculation first might be more representative. I do this with fewer than 4 pace lines showing. Younger horses. One "99" averaged with 3 finishes LESS than 10 would skew the score. I use a "sumif" formula in excel to do this.

On the other hand I leave "99" as is with older horses. Many horses who do receive one 99 may get more than one. Maybe they should be "skewed".

I do use a series of formulas to pick a paceline. And values from that paceline gets more weight in my calculations than the averages I do use.

.................................................. .................................................. .....
Ray, the original formula I posted works. Been using it a number of years. Don't know why your first modification does not. When I have the chance will check it out

[raybo]

Yes, if you calculate average finish or average of any running position, removing a '99" from the calculation first might be more representative. I do this with fewer than 4 pace lines showing. Younger horses. One "99" averaged with 3 finishes LESS than 10 would skew the score. I use a "sumif" formula in excel to do this.

On the other hand I leave "99" as is with older horses. Many horses who do receive one 99 may get more than one. Maybe they should be "skewed".

I do use a series of formulas to pick a paceline. And values from that paceline gets more weight in my calculations than the averages I do use.

.................................................. .................................................. .....
Ray, the original formula I posted works. Been using it a number of years. Don't know why your first modification does not. When I have the chance will check it out

Could be something with Excel 2007. The cell format is set to "general". Should have worked but doesn't (?).

[keith70]

padonne, what did you ever decide to do with this project?