Hours of Operation

They screwed up the formatting, but it's clear what they're saying. They quoted you saying you don't even need electricity for archives, and they're telling you that if you don't test your backups, you don't have backups.

You can say it to the C-Suite all you'd like, but Goodluck getting them to budget for anything sometimes. Spinning rust, lto tapes, etc and the software to handle them is generally considered out of budget, at least at most orgs I've worked with (despite the recommendation to practice 3-2-1). Also, you may want to further detail your advice, most storage mediums (save lto tapes) require ongoing power to prevent things like bitrot. It's quite common for hard drives for example to develop bad sectors over time, but algos running on the drives embedded firmware clean this up by repositioning bad sectors to a reserve space. Nvme/flash is even more vulnerable to this and behaves similarly.

Lol, that's an issue with lemmy‘s markdown interpreter.
It looks perfect on pylova. Come see.

@givesomefucks@lemmy.world seems this is what you actually meant?

Getting hard to find WORM drives & media at an affordable price nowadays. For larger orgs like PBS maybe it isn't such a problem, but for us everyday schmoes...

To make your comment more readable, you need to edit your comment to add another line break between what you quoted and your response to it (i.e. a blank line before "Nope.").

"3-2-1" is important when considering your backups. AT LEAST.

3 copies of your data.

2 separate locations.

1 (or more) off-site.

If you want to go a step further(a good idea), add in one more that's immutable(to prevent access to malware).

Do you comprehend actual backup/cyber hygiene?

I don't understand what you're trying to say, and I feel like that's because you don't know how comments work...

important enough to maintain a local copy backup. tis why I tagged #3-2-1%20ignore you only need the machine on when adding to the archive. Then just shit it back down and leave it alone. Nope.
Regular backup tests is part of backup hygiene.
Otherwise you have duds you need to replace.

If it's important enough to pay a monthly fee for archival cloud hosting, it's important enough to maintain a local copy backup.

If it's archival it doesn't even cost electricity, you only need the machine on when adding to the archive. Then just shit it back down and leave it alone.

good grief, be a shame if the market was flooded with incredibly cheap copies of the AI photo to ruin the value for whatever dipshit bought it.

The thread has those aplenty, it is just that you are confidently incorrect so you dont like the answers.

You asked, but didnt like the answer.

I was looking for good answers. Convincing answers. Which I expected to correspond with data volume.

In any case, good answers are defensible, should the occasion arise. When you cannot defend your answers, it indicates a lack of justified confidence despite an expectation that others adopt some kind of blind confidence in your answers.

Is my request unreasonable?

Yes, it is unreasonable. As you have already been told. You asked, but didnt like the answer. Again, you have only a rudimentary understanding of the problem but base everything off your experience and your needs.

It is simple in sqlite (which is purpose-built to be simple and small,) so you assume all other databases are equally simple.

It’s the other way around. I expect a simple DB to be more basic. A more complex DB should be even more featured. If, for example, an Oracle DB cannot easily handle the job that a small and simple home kit can, Oracle should be embarrassed.

and self-centered.

Yikes! What leads you to think this is about me? It’s about databases. I did not invent sqlite. It was an example. You can fuck off with your vitriol.

It is simple in sqlite (which is purpose-built to be simple and small,) so you assume all other databases are equally simple. You then expect library staff to be standing by ready to help with your demands.

Well prepare to be shocked: That expectation is absurdly naive and self-centered. YTA

Someone would need to query the data and encode it to JSON/etc. on the fly. There may not be a way to do this without having one of the programmers directly connect to the DB and run custom queries.

I use sqlite. They are probably using some heavier duty db but for sqlite exporting JSON is trivial so I would be surprised if other DBs did not have a similar mechanism. And to be clear, I said to the library that I prefer JSON but would handle whatever open format they prefer, be it XML or CSV.

To dump everything blindly would be irresponsible. It’s possible the DB could lock up or that incorrect/inappropriate data could be queried.

This does not sound like a realistic problem. I might imagine if they had a DB of all ISBNs, they would obviously have to use a query that limits to their catalog. Apart from that, I don’t see what would be inappropriate. If it’s in their catalog, why hide it? Not sure what you have in mind but I should say it’s not the US where there would be some right wing concern to prevent children from getting sex education type of material, or the Christian right trying to make Darwin’s theories hard to reach.

If you are thinking in terms of sensitive info, like accounts of people and what they borrow, it would be irresponsible if that kind of info were not in a separate table.

Their software is either homebrew or off the shelf. If it’s off the shelf, they may need to put in a support ticket and have their vendor figure it out. It’s highly unlikely there is a government employee or team sitting around and waiting for your request.

I was expecting my request to be ignored, as open data requests often are -- and rarely fufilled in my experience even when they answer. But in the case at hand, they first responded favorably, saying essentially: we can give you some data but your request is vague.. what exactly do you want? I basically replied with “everything”. So they were not opposed to exporting some data, but the volume involved (100s of 1000s of records) seems to be a show-stopper.

I believe that the library has a point about not performing this amount of work on an individual request.

I might agree that it’s a bit much to serve one person. They also said it would take disproportionate resources when they have a whole public to serve. But I was figuring “build it, and they will come”. There would be a first person to make a request. I am a bit disappointed that if it were made available that we could not expect many people to exploit the option to be free from the UIs limitations.

No shit?

Unless you are talking about a library in an urban area where it is heavily funded, your library may have an incredibly limited tech team. It's still possible that they are running all vendor software with a team of mostly contractors or shared with the rest of the region.

Someone would need to query the data and encode it to JSON/etc. on the fly. There may not be a way to do this without having one of the programmers directly connect to the DB and run custom queries. The data is likely spread in several tables, with the bibliographies in a separate large text only section.

To dump everything blindly would be irresponsible. It's possible the DB could lock up or that incorrect/inappropriate data could be queried. They should do a few test queries to determine the size of the data, then break it up into chunks to pull a little at a time and QC/stitch it together.

Their software is either homebrew or off the shelf. If it's off the shelf, they may need to put in a support ticket and have their vendor figure it out. It's highly unlikely there is a government employee or team sitting around and waiting for your request.

It's not that your concerns are invalid, but I believe that the library has a point about not performing this amount of work on an individual request. Have like-minded individuals request the changes you want to the library system. The library is there to serve the public at large, but large projects like this need high level buy in, funding, and hands.

They likely don’t have any technologist on staff capable of direct db exports or any kind of scripting. So they complete these requests manually, which would be excessively burdensome.

good. separating rich people from their money is an honest profession.

Hahahahahahaha

Can't have physical rentals for something that hasnt been physical in 15 years

Yes but the libraries lend them and the people borrow them. Rent has a nasty air of money exchanging hands.

Yes they do.

Should they be

That's a loaded question. Clearly my local community thinks they should.