4 ms·

This is slightly off topic and relatively minor, but man does it represent the piss poor state of Tech writers and their articles. "I was curious how often thi

This is slightly off topic and relatively minor, but man does it represent the piss poor state of Tech writers and their articles. "I was curious how often this go-to one liner was used. I scraped every reported notification to the California attorney general, a requirement under state law in the event of a breach or security lapse, stitched them together, and converted it into machine-readable text. About one-third of all 285 data breach notifications had some variation of the line." Don't bother providing a link to either you data source or your code, the ability for someone to independently verify the validity of this claim and its results isn't important, we'll just "trust" you.

Vinnl 8y ago
When tech writers do that it's one thing, but this is unfortunately also the status quo among publicly funded researchers...
Ensorceled 8y ago
I don't know of any journalism source that would do better than what you just quoted. They explained their source and their methodology. If you think they are lying it's not extremely hard to prove it, but why are you assuming they are lying?
- resonanttoe 8y ago
  It's not up to me as a reader to prove whether they are lying or not. As a writer you either have integrity and assume your audience is knowledgeable and skeptical of unverifiable claims or you can exploit the fact that they aren't. There is a clear line between good writing and not. I'm not assuming they're lying but they've given me no reason to trust that they aren't. But lets be clear, they explained nothing and hand waved around what they did. In an article that is indirectly about bugs in software, I want to be sure that the "simple" methodology doesn't have bugs. This is easily shown by the author by just providing the scripts and methods they used. Choosing not to is lazy. Here's some simple questions that that "sourced" passage doesn't answer. 1. Where/how exactly did you get the info from the CA Attorney General? I searched and found this https://oag.ca.gov/privacy/databreach/list https://oag.ca.gov/privacy/databreach/list which extends back to 2012. Did you just download and parse the PDFs because some have two notifications listed, some have one. Some could probably have more. Are they counted as one entity or multiple? 2. "Stitched them together" Well that can just mean anything. Considering that PDF Parsing is still seen to be painful to most (https://stackoverflow.com/questions/22675690/if-identifying-text-structure-in-pdf-documents-is-so-difficult-how-do-pdf-reade/ https://stackoverflow.com/questions/22675690/if-identifying-...) there are errors that can be introduced here. 3. How did you handle string repetitions in the same document? If I put "We take your privacy seriously" at the start and the end, is that n = 1 or n = 2? 4. Assuming that 285 (of 1559 line entries) of the ones you parsed are single line entries (That sometimes contain multiple breaches and notifications) the author stopped at the 20th of January 2018 for some reason.... Why then? I could go on and yes these are pretty pedantic questions. None of which are design to accusatory or suggest anything untoward about the author and what they wrote. But you know what, all of this is answerable if they post their source and their source code. I can take a look and get a better picture of the full-arsed or half-arsed effort that went in to this article. (Sometimes half-arsed is okay, sometimes it's not) If a tech website can't "do better than what [I]sic just quoted", then I circle back to the initial statement. "Man does it represent the piss poor state of Tech writers and their articles."
- TeMPOraL 8y ago
  If they cared about informing the reader, they would link to the source, so that interested readers could explore the topic further. It's as simple as that.
JustSomeNobody 8y ago
Why are you attempting to equate tech writers with journalistic integrity. TechCrunch is basically a blog.
- TeMPOraL 8y ago
  Well, regular bloggers have higher standards than journalists because they actually link to their sources. News sites do their damnest to avoid linking or even mentioning things by name if the article isn't explicitly trying to promote them.

Vinnl 8y ago

When tech writers do that it's one thing, but this is unfortunately also the status quo among publicly funded researchers...

Ensorceled 8y ago

I don't know of any journalism source that would do better than what you just quoted. They explained their source and their methodology. If you think they are lying it's not extremely hard to prove it, but why are you assuming they are lying?

resonanttoe 8y ago
It's not up to me as a reader to prove whether they are lying or not. As a writer you either have integrity and assume your audience is knowledgeable and skeptical of unverifiable claims or you can exploit the fact that they aren't. There is a clear line between good writing and not. I'm not assuming they're lying but they've given me no reason to trust that they aren't. But lets be clear, they explained nothing and hand waved around what they did. In an article that is indirectly about bugs in software, I want to be sure that the "simple" methodology doesn't have bugs. This is easily shown by the author by just providing the scripts and methods they used. Choosing not to is lazy. Here's some simple questions that that "sourced" passage doesn't answer. 1. Where/how exactly did you get the info from the CA Attorney General? I searched and found this https://oag.ca.gov/privacy/databreach/list https://oag.ca.gov/privacy/databreach/list which extends back to 2012. Did you just download and parse the PDFs because some have two notifications listed, some have one. Some could probably have more. Are they counted as one entity or multiple? 2. "Stitched them together" Well that can just mean anything. Considering that PDF Parsing is still seen to be painful to most (https://stackoverflow.com/questions/22675690/if-identifying-text-structure-in-pdf-documents-is-so-difficult-how-do-pdf-reade/ https://stackoverflow.com/questions/22675690/if-identifying-...) there are errors that can be introduced here. 3. How did you handle string repetitions in the same document? If I put "We take your privacy seriously" at the start and the end, is that n = 1 or n = 2? 4. Assuming that 285 (of 1559 line entries) of the ones you parsed are single line entries (That sometimes contain multiple breaches and notifications) the author stopped at the 20th of January 2018 for some reason.... Why then? I could go on and yes these are pretty pedantic questions. None of which are design to accusatory or suggest anything untoward about the author and what they wrote. But you know what, all of this is answerable if they post their source and their source code. I can take a look and get a better picture of the full-arsed or half-arsed effort that went in to this article. (Sometimes half-arsed is okay, sometimes it's not) If a tech website can't "do better than what [I]sic just quoted", then I circle back to the initial statement. "Man does it represent the piss poor state of Tech writers and their articles."
TeMPOraL 8y ago
If they cared about informing the reader, they would link to the source, so that interested readers could explore the topic further. It's as simple as that.

JustSomeNobody 8y ago

Why are you attempting to equate tech writers with journalistic integrity. TechCrunch is basically a blog.

TeMPOraL 8y ago
Well, regular bloggers have higher standards than journalists because they actually link to their sources. News sites do their damnest to avoid linking or even mentioning things by name if the article isn't explicitly trying to promote them.