|
Chalks posted:The proposed legislation is absurd. It has the potential to require you to retroactively redact the email address of a customer stored in a PDF file inside a database, and every copy of that information, including offline tape backups. And you'll need any third party provider, regardless of their location, to be able to do the same. I feel that maybe it's a good thing companies would be forced to consider things like that when designing their bespoke enterprise data sprawl. 4lokos basilisk fucked around with this message at 15:35 on Jul 28, 2017 |
# ? Jul 28, 2017 15:27 |
|
|
# ? May 30, 2024 12:10 |
Subjunctive posted:I thought you were saying that techbros need to read the law. are lawyers techbros now? have you heard of context
|
|
# ? Jul 28, 2017 16:01 |
Chalks posted:The proposed legislation is absurd. It has the potential to require you to retroactively redact the email address of a customer stored in a PDF file inside a database, and every copy of that information, including offline tape backups. And you'll need any third party provider, regardless of their location, to be able to do the same. "sure people regularly get their lives utterly wrecked by our collected information getting into the hands of the wrong people, but have you considered that complying with this regulation would be really hard if we don't actually address the root cause of the problem in our industry's complete disregard for securing or even keeping track of where that sort of information is located?"
|
|
# ? Jul 28, 2017 16:11 |
|
Penisface posted:
Well, at the moment the legislation isn't finalised so fingers crossed they can come up with wording that means it's actually feasible. Right now my company isn't worrying about it because we can't plan for something that's as ill defined as this is. The backups issue is a huge one. Offline tape backups are, by definition, read only. Intentionally - it's a fundamental part of the backup system that you can't accidentally go and destroy them with a poorly written script. So what are you meant to do? Run a delete script on all of your historical backups? Presumably without taking a backup first. Not to mention the fact that you're meant to be doing this automatically at the request of random members of the public when ever they feel like it. We have separate databases per customer, with incremental backups taken between periodical full backups. These backups are stored on azure and mirrored on S3. Those storage accounts are themselves backed up with microsoft/amazon's backup facilities. Our customers keep PDF copies of things like customer invoices in these databases along with huge amounts of other data, such copies of emails which may contain zipped attachments containing personal details. The astronomical clusterfuck that it would be to implement a system to automatically delete personal information from a system like this is impossible to imagine. If we were writing something from scratch with this requirement in mind then perhaps we could come up with something but there's no way you could retrofit something like this. Shifty Pony posted:"sure people regularly get their lives utterly wrecked by our collected information getting into the hands of the wrong people, but have you considered that complying with this regulation would be really hard if we don't actually address the root cause of the problem in our industry's complete disregard for securing or even keeping track of where that sort of information is located?" We know exactly where the customer data is stored and it's completely secure. Having the data form part of an encrypted incremental backup in cold storage is one of the biggest problems. I mean for fucks sake, if we had a bunch of unencrpyed data sitting around without any backups, it'd be trivial to delete stuff from it!! Maybe that's how we should have designed it from the beginning because then we could easily comply with this dumbass proposed law. And you know what sort of "collected information" we're talking about here? poo poo like an invoice from a plumber that's got your name at the top of it. We're not exactly talking about big data here or invasive analytics, just basic poo poo that small businesses have in order to run their perfectly ordinary business. Chalks fucked around with this message at 16:19 on Jul 28, 2017 |
# ? Jul 28, 2017 16:13 |
|
Oh fun fact: the wifi chip in macs is a bcm43xx so it's vulnerable to broadpwn
|
# ? Jul 28, 2017 16:34 |
|
Chalks posted:And you know what sort of "collected information" we're talking about here? poo poo like an invoice from a plumber that's got your name at the top of it. We're not exactly talking about big data here or invasive analytics, just basic poo poo that small businesses have in order to run their perfectly ordinary business. but anyway it's not really a security fuckup
|
# ? Jul 28, 2017 16:53 |
|
cinci zoo sniper posted:have you heard of context yeah, that was dumb, sorry
|
# ? Jul 28, 2017 16:55 |
|
i'm loving GDPR because now we're actually considering a data retention policy that isn't "never delete anything and let someone else worry about it in a few years"
|
# ? Jul 28, 2017 16:58 |
Subjunctive posted:yeah, that was dumb, sorry tbh i almost fell for it but was distracted with shitposting elsewhere anyways, some secfucks: quote:Ruben Santamarta, a security researcher for IOActive, has found various vulnerabilities in nuclear radiation monitoring equipment from three vendors, who when contacted by the researcher, declined to fix the reported flaws, each for various reasons. and the good old "hackers can turn your car wash into a bomb". as expected from internet of poo poo, it's shipped with web server that has default admin password of 12345, and no one cares to fix
|
|
# ? Jul 28, 2017 17:00 |
|
spankmeister posted:Oh fun fact: the wifi chip in macs is a bcm43xx so it's vulnerable to broadpwn Wasn't that addressed by the security patch just issued within the last week or so for iOS and macOS?
|
# ? Jul 28, 2017 17:06 |
|
Proteus Jones posted:Wasn't that addressed by the security patch just issued within the last week or so for iOS and macOS? it was, but i think it's interesting because people focus on mobile devices but it's a lot of aiport devices as well
|
# ? Jul 28, 2017 17:08 |
|
Not to mention that the codebase is probably shared between a LOT of broadcom wifi chips so things like routers, tv's, IoT devices etc are likely vulnerable.
|
# ? Jul 28, 2017 17:09 |
|
spankmeister posted:it was, but i think it's interesting because people focus on mobile devices but it's a lot of aiport devices as well It is interesting that the same vulnerability could be leveraged across phones, laptops and desktops. That's an intersection you don't see a lot of.
|
# ? Jul 28, 2017 17:12 |
|
Chalks posted:The proposed legislation is absurd. It has the potential to require you to retroactively redact the email address of a customer stored in a PDF file inside a database, and every copy of that information, including offline tape backups. And you'll need any third party provider, regardless of their location, to be able to do the same. i mean, i will admit that i may have undersold it when i said that the "transition may be difficult", but the entire point would be that your personal information is not theirs to put in a pdf inside a database somewhere the fact that the easy solution in all instances is to opt to be conservative with storing personal information is precisely the kind of outcome which i find highly desirable
|
# ? Jul 28, 2017 17:14 |
|
My work deals with a LOT of personal data which means we get regular government audits and all that stuff already. I believe the current law/policy is that we can't keep any personal data longer than 2 years after our last dealing with them. It's not hard to just run a scheduled delete for that on all live systems. For backups, within 2 years things change so much that 2yo backups are completely useless to us anyway, so no reason to keep them.
|
# ? Jul 28, 2017 18:10 |
|
Chalks posted:Well, at the moment the legislation isn't finalised so fingers crossed they can come up with wording that means it's actually feasible. Right now my company isn't worrying about it because we can't plan for something that's as ill defined as this is. This entire post is just a good example WHY the law is a good idea. I'm hoping it has a loophole for read-only backups where you can have a secure vault for them where they're never accessed without multiple staff members present. But no, the law's designed to make it the easier option to not hang onto personal data you don't need... which is good for anyone who doesn't like their data being revealed to the public by some ten-year-old unpatched system getting compromised.
|
# ? Jul 28, 2017 18:12 |
|
more broadly sweden has had requirements which require the basic awareness of what is personal information and where it is since 1998. not as comprehensive at all, but one can e.g. request a copy of all information an organization has attached to one personally, and one can demand corrections in that information be made. i am aware that it has been a pretty complex administrative task, but it really is the kind of stuff where it is good for real people that companies *do* take care i especially like that the "request copy" thing has been expanded into being able to do a full transfer of the information between platforms (which i assume amounts to all services being required to have an equivalent of google takeout, though possibly even google takeout would have to get deeper)
|
# ? Jul 28, 2017 18:17 |
|
Avenging_Mikon posted:Go away, Nintendo Kid nice meltdown, guy who is literally too dumb to know google's primary business is ads
|
# ? Jul 28, 2017 18:34 |
|
fishmech posted:nice meltdown, guy who is literally too dumb to know google's primary business is ads lol if you don't
|
# ? Jul 28, 2017 18:47 |
|
https://github.com/xoreaxeaxeax/sandsifterquote:Lastly, a so-called ‘halt and catch fire’ instruction was discovered on an as-yet unnamed x86 processor. This instruction, executed in ring 3 from an unprivileged process, appears to lock the processor entirely. To rule out kernel bugs, the instruction was tested against three Linux kernels and two Windows kernels, yielding the same results. Kernel debugging with serial I/O and interrupt hooks appeared to corroborate the results. At the time of this paper’s publishing, the vendor has not been provided sufficient time to respond to the issue. The details of the instruction and the processors affected will be enumerated when responsible disclosure is complete, and an updated version of this whitepaper will be released. Such instructions pose a critical security risk, as they allow unprivileged users to mount denial of service attacks against shared systems.
|
# ? Jul 28, 2017 19:18 |
|
endlessmonotony posted:This entire post is just a good example WHY the law is a good idea. An exemption for backup data is basically required IMO. You say "secure vault with multiple staff members etc" but seriously, even just a tape backup containing incremental database backups - how do you write an automated process for removing contextual data from that? You say "hanging onto data you don't need" but in reality we're talking about data you do need up until the point that you're told to remove it. It's not like you could have a policy of never storing unnecessary data and be unaffected by this legislation as a result.
|
# ? Jul 28, 2017 19:32 |
|
Chalks posted:An exemption for backup data is basically required IMO. You say "secure vault with multiple staff members etc" but seriously, even just a tape backup containing incremental database backups - how do you write an automated process for removing contextual data from that? you store and backup the personal data with a process separated out for that purpose or you bloody well trawl the full backup when needed to prune out the personal info. you make it sound like you have some natural right to handle peoples personal information in whatever way is convenient to you, and, turns out; you don't. if you actually cannot control the data fully i personally will be quite pleased when you go out of business.
|
# ? Jul 28, 2017 19:40 |
|
so if I gchat your email address to someone, can you require google to excise it from their records? delete all the mail that copied you from gmail?
|
# ? Jul 28, 2017 19:41 |
|
Cybernetic Vermin posted:you store and backup the personal data with a process separated out for that purpose In a customer management system there's little else besides personal data. Contact details, correspondence, support tickets, order details... that's basically all there is. And it's not just a case of trawling through backups - in the case of an incremental backup you can't even access the data without restoring the increment against its full backup. So you do that and remove the data.. then what? Instead of the 200mb incremental you started with you've got a 200gb full database. What are you meant to do with that? So we outlaw the use of tape backups, we outlaw the use of incremental backups and we outlaw the use of offline data storage...? It's perfectly possible to have reasonable data protection legislation without putting a completely unreasonable burden on backup processes. It would be silly to pretend there are only two options, this or literally no control over how personal information is stored and handled.
|
# ? Jul 28, 2017 19:49 |
|
Could I create phony payroll records with -REDACTED- as the payee and when I'm audited say "well we scrubbed their personal info" as a money laundering mechanism asking for a friend
|
# ? Jul 28, 2017 19:53 |
|
this is a super hackercore interface
|
# ? Jul 28, 2017 19:55 |
|
wow, it's almost like the idea of "right to be forgotten" is stupid as hell and has made sense never in the history of the world as some sort of all-reaching order
|
# ? Jul 28, 2017 19:57 |
|
It's like that stupid cookie law all over again. Sure am glad I have to read "hey we use cookies for some stuff and if you don't like it feel free to gently caress off" on every website I visit, very valuable legislation.
|
# ? Jul 28, 2017 20:09 |
|
wyoak posted:Could I create phony payroll records with -REDACTED- as the payee and when I'm audited say "well we scrubbed their personal info" as a money laundering mechanism no when it comes to financial data like that you're allowed to keep it which makes the whole thing loving pointless
|
# ? Jul 28, 2017 20:15 |
|
I'm glad storing people's data is a huge pain in the rear end an you all sound like losers who are mad that it's your job to figure out how to do it.
|
# ? Jul 28, 2017 20:15 |
|
Subjunctive posted:so if I gchat your email address to someone, can you require google to excise it from their records? delete all the mail that copied you from gmail? wyoak posted:Could I create phony payroll records with -REDACTED- as the payee and when I'm audited say "well we scrubbed their personal info" as a money laundering mechanism fishmech posted:wow, it's almost like the idea of "right to be forgotten" is stupid as hell and has made sense never in the history of the world as some sort of all-reaching order
|
# ? Jul 28, 2017 20:15 |
|
Subjunctive posted:so if I gchat your email address to someone, can you require google to excise it from their records? delete all the mail that copied you from gmail? yes, and if they don't comply immediately they get nationalized
|
# ? Jul 28, 2017 20:16 |
|
Bonfire Lit posted:no, because in this case google benefits from the safe harbor provisions of 2000/31/EC articles 12-14. this is called out in GDPR article 2(4) thanks!
|
# ? Jul 28, 2017 20:16 |
|
Bonfire Lit posted:it's almost like nobody has looked at the directive but everyone loves arzying fishmech is just keeping on brand with his whole "privacy never actually existed" schtick
|
# ? Jul 28, 2017 20:17 |
|
Bonfire Lit posted:no, because you're still allowed to store data you need to keep to fulfill for legal obligations. which keeping accurate payroll is.
|
# ? Jul 28, 2017 20:19 |
|
Maximum Leader posted:no when it comes to financial data like that you're allowed to keep it which makes the whole thing loving pointless lol ate all the Oreos posted:fishmech is just keeping on brand with his whole "privacy never actually existed" schtick hey guess what: that's true. but also you can't actually make people forget anything, or keep them from talking in private. which is why this is really dumb
|
# ? Jul 28, 2017 20:22 |
|
oh no think of the poor companies. the evil lawmakers are fuckingw ith the free market again
|
# ? Jul 28, 2017 20:22 |
|
wyoak posted:how many breaches are from datasets unrelated to financial / legal obligations the vast majority of them? like the ones that get the loudest coverage have to do with finance/legal stuff (and get free credit monitoring!!!!) but the ones with the largest sheer magnitude of users affected tend to just be account dumps and stuff like that, and those seem to happen way more often (because companies think security only applies when you're handling money, if at all)
|
# ? Jul 28, 2017 20:23 |
|
wyoak posted:how many breaches are from datasets unrelated to financial / legal obligations fishmech posted:but also you can't actually make people forget anything, or keep them from talking in private. which is why this is really dumb
|
# ? Jul 28, 2017 20:27 |
|
|
# ? May 30, 2024 12:10 |
|
ate all the Oreos posted:the vast majority of them? like the ones that get the loudest coverage have to do with finance/legal stuff (and get free credit monitoring!!!!) but the ones with the largest sheer magnitude of users affected tend to just be account dumps and stuff like that, and those seem to happen way more often (because companies think security only applies when you're handling money, if at all) At its face regulation like this seems like something that big companies will skirt because of their well-paid legal departments while smaller companies could get hit really hard
|
# ? Jul 28, 2017 20:41 |