|
1 person marked this as a favorite.
|
That's why I said you might have to do it piecemeal, like one school at a time. But I'm looking through the program now and this shit is not hard.
That, and the CDE already does outputs into excel formats. They admit that this is possible in the instructions on their own website.
You can also output it into a .txt format. Portion it off and import that into excel, avoiding the choke. Once it's in excel, this is ridiculously easy to solve. The CDE gives us instructions on how to do it.
thejeff wrote: "build infrastructure" I couldn't resist. :)
Which sort of points to the DoE's privacy "concerns" being a red herring, don't it?
Plus, if enough parents get scared into opting out, the data is useless!
It's win-win!
Orfamay Quest wrote: Irontruth wrote: Orfamay Quest wrote: Irontruth wrote: There are obviously ways to share the information without making it more vulnerable, particularly in this specific case. There have even been documented suggestions that would maintain security on the information, while allowing an investigation to happen. You obviously know a lot more about the DoE's data storage procedures than others in this thread (including myself), if you can verify of your own knowledge that they are "documented suggestions" instead of "wild-assed guesses based upon assumptions about how the database MIGHT look."
You talk about it in this post.
We both know there are ways to format the database to share the required information that doesn't include information like linked names and SSNs. Yes, and neither of us knows if any of those ways are actually feasible given the existing state of the systems. As the old joke has it, "if I wanted to go there, I wouldn't start from here." Or rather, I know that I don't know, which makes my musings "wild-assed guesses."
So, on second run through, the CDE has documented on their own websites how this data could be shared while addressing privacy concerns.
Irontruth wrote: Orfamay Quest wrote: Irontruth wrote: Orfamay Quest wrote: Irontruth wrote: There are obviously ways to share the information without making it more vulnerable, particularly in this specific case. There have even been documented suggestions that would maintain security on the information, while allowing an investigation to happen. You obviously know a lot more about the DoE's data storage procedures than others in this thread (including myself), if you can verify of your own knowledge that they are "documented suggestions" instead of "wild-assed guesses based upon assumptions about how the database MIGHT look."
You talk about it in this post.
We both know there are ways to format the database to share the required information that doesn't include information like linked names and SSNs. Yes, and neither of us knows if any of those ways are actually feasible given the existing state of the systems. As the old joke has it, "if I wanted to go there, I wouldn't start from here." Or rather, I know that I don't know, which makes my musings "wild-assed guesses."
So, on second run through, the CDE has documented on their own websites how this data could be shared while addressing privacy concerns. Link? And why didn't they come up with this years ago?
thejeff wrote: Irontruth wrote: Orfamay Quest wrote: Irontruth wrote: Orfamay Quest wrote: Irontruth wrote: There are obviously ways to share the information without making it more vulnerable, particularly in this specific case. There have even been documented suggestions that would maintain security on the information, while allowing an investigation to happen. You obviously know a lot more about the DoE's data storage procedures than others in this thread (including myself), if you can verify of your own knowledge that they are "documented suggestions" instead of "wild-assed guesses based upon assumptions about how the database MIGHT look."
You talk about it in this post.
We both know there are ways to format the database to share the required information that doesn't include information like linked names and SSNs. Yes, and neither of us knows if any of those ways are actually feasible given the existing state of the systems. As the old joke has it, "if I wanted to go there, I wouldn't start from here." Or rather, I know that I don't know, which makes my musings "wild-assed guesses."
So, on second run through, the CDE has documented on their own websites how this data could be shared while addressing privacy concerns. Link? And why didn't they come up with this years ago? Links... in this post that you replied to.
I'm not joking when I said I downloaded their database program and looked around. I found the "export to excel" options. In both of those links it talks about exporting to text and excel formats. This is a standard and used practice by the CDE.
You don't have to take my word for it, give the manuals and help documents they host on their website a quick search. You can also download the software for yourself. It's actually very, very small. My download took maybe a few seconds and it installed in less than a minute.
Irontruth wrote: thejeff wrote: Irontruth wrote: Orfamay Quest wrote: Irontruth wrote: Orfamay Quest wrote: Irontruth wrote: There are obviously ways to share the information without making it more vulnerable, particularly in this specific case. There have even been documented suggestions that would maintain security on the information, while allowing an investigation to happen. You obviously know a lot more about the DoE's data storage procedures than others in this thread (including myself), if you can verify of your own knowledge that they are "documented suggestions" instead of "wild-assed guesses based upon assumptions about how the database MIGHT look."
You talk about it in this post.
We both know there are ways to format the database to share the required information that doesn't include information like linked names and SSNs. Yes, and neither of us knows if any of those ways are actually feasible given the existing state of the systems. As the old joke has it, "if I wanted to go there, I wouldn't start from here." Or rather, I know that I don't know, which makes my musings "wild-assed guesses."
So, on second run through, the CDE has documented on their own websites how this data could be shared while addressing privacy concerns. Link? And why didn't they come up with this years ago? Links... in this post that you replied to.
I'm not joking when I said I downloaded their database program and looked around. I found the "export to excel" options. In both of those links it talks about exporting to text and excel formats. This is a standard and used practice by the CDE.
You don't have to take my word for it, give the manuals and help documents they host on their website a quick search. You can also download the software for yourself.... Oh, I thought you meant somewhere the CDE specifically proposed a method to get the data needed for the lawsuit without privacy concerns, not a generic manual saying it's possible to export to Excel and your assumption it would be easy to strip identifying information while leaving the data usable.
A process which might, if you had to break it out as you earlier suggested, take hundreds of man-hours and thus be an undue burden.
thejeff wrote: Oh, I thought you meant somewhere the CDE specifically proposed a method to get the data needed for the lawsuit without privacy concerns, not a generic manual saying it's possible to export to Excel and your assumption it would be easy to strip identifying information while leaving the data usable.
A process which might, if you had to break it out as you earlier suggested, take hundreds of man-hours and thus be an undue burden.
I don't know if you can split that hair any smaller.
The CDE has a specified and accepted method of extracting the data from their database. This method is more than capable of removing the identifying data while maintaining the rest of the data.
You: Here's the manual I wrote on changing tires for a 1995 Honda Civic
Me: Can you change the tire for this 1995 Honda Civic?
You: Well... I haven't seen that specific car, so I'm going to have to say no.
They wrote down a method that would work on a publicly accessible website. If they're claiming that they can't do it... that's bullshit, because I can see where they wrote how to do it on their own website.
Did they submit it in a legal document to the court? No. But that's cause they're trying to hide behind bullshit.
Irontruth wrote: thejeff wrote: Oh, I thought you meant somewhere the CDE specifically proposed a method to get the data needed for the lawsuit without privacy concerns, not a generic manual saying it's possible to export to Excel and your assumption it would be easy to strip identifying information while leaving the data usable.
A process which might, if you had to break it out as you earlier suggested, take hundreds of man-hours and thus be an undue burden.
I don't know if you can split that hair any smaller.
The CDE has a specified and accepted method of extracting the data from their database. This method is more than capable of removing the identifying data while maintaining the rest of the data.
You: Here's the manual I wrote on changing tires for a 1995 Honda Civic
Me: Can you change the tire for this 1995 Honda Civic?
You: Well... I haven't seen that specific car, so I'm going to have to say no.
They wrote down a method that would work on a publicly accessible website. If they're claiming that they can't do it... that's b$@@+&!+, because I can see where they wrote how to do it on their own website.
Did they submit it in a legal document to the court? No. But that's cause they're trying to hide behind b*@%$&#~. You can dump any database to a text file. Probably to an Excel sheet as well, at least as csv. Once you've done that you can delete whatever data you want from it.
That doesn't mean that
a) It won't be a monumental task to do so, if only because Excel doesn't gracefully handle huge amounts of data.
b) That the resulting output will be usable. If this was a relational db and the SSNs were unique keys just deleting them would break all the connections. Replacing them with different unique identifiers throughout the db would be a much more complex task and one I'd be much happier doing in a db programmatically than manually on a spreadsheet. Even then data integrity is tricky. I've written some code to make such broad changes in a complex database. It's definitely non-trivial. I don't know the nature of this database or the table structure it's using, so it's hard to say much.
Either of those could easily hit "undue burden".
Looking more closely at those documents, those look to me like end-user report generation instructions, nothing lower level. Some DB programmer has created a front end that lets them access specific reports. Those can be saved as text or opened in Excel, not an actual dump of the database. If a report doesn't already exist that has the necessary information, one would need to be written, if you were going to use that interface. Again, a non-trivial task.
And finally again, the judge's approach is more than sufficient to address privacy concerns. It's overkill, if anything.
I do agree they're trying to hide behind bullshit though. :)
The information is already exported by each school to be collected into the state's central system. The information already exists in smaller files.
CASEMIS isn't a system of terminals on a network accessing a central server, each computer run's it's own copy of the program and stores the information locally. The state collects the information in the form of text and excel files from each school via email.
Recent threads in Off-Topic Discussions
|
