hi. I can use shred to securely delete a file, if really usefull on ext4. But how can I remove traces of filename in filesystem metadata ? Because extundelete picks up easily deleted file names.
@solsTiCe I'd recommend physically destroying the drive. shred will overwrite the file's last known location on disk, but if you've edited the file or copied it around, downloaded it, or had it delivered through email, there may be traces of it (or of temporary files with the same data) in all sorts of locations.
@solsTiCe There might be no point to any of this, because the NSA already has a copy. Strong crypto is generally the recommended way of keeping things safe that you really want to say hidden, in this less than perfect of worlds.
@Caleb since you've worked on ISP I got a question about an investigation I'm doing: are there any cost associated with the raw transmission of data through ISP's? My POV is that ISP has fixed costs and that their users download 10MiB, 10GiB or 10TB doesn't make operations more costly (ignoring for a moment if they have the installed capacity to handle the load).
@FaheemMitha there was an extension back in the day we were looking to migrate to indexeddb.
changing the bad single quote characters fixed it but yes that may have been the case. since I modified them in atom which I have set to use LF so it probably removed it.
@FaheemMitha The core POSIX functionality is the same, mostly (GNU command line parsing is funky: ls . -l for example). Then there are extensions on both sides.
I would like to use find and rsync to copy any directory that contains at least one "wav" file. The copying should copy the whole folder if a "wav" file is present. I have this so far:
find . -name "*.wav" | rsync -armR --include="*/" --include="*.wav" --exclude="*" /users/lukew/desktop/merged...
I'm trying to tidy-up my photos which are, for various historic reasons, scattered all over my system. To enable me to make a start on this task, I've been trying to use the command line to construct a list of all directories that contain one or more jpg files. I'm certain that I don't have to b...
@Gilles His original question was slightly different as he wanted to recursively copy all directories with the files, however he has since changed his mind and now wants to delete all directories that DON'T contain the files.
@Braiam Oh absolutely there is cost there. Except for the rare case of when we provided point to point tunnels between locations on dedicated hardware so traffic started and ended with us every byte cost us something to move.
ISP's have to pay for their upstream link(s) too, and pricing for that is more and more directly linked to actual data moved the higher up you get in the link chain.
Consumers pay for a link at a potential speed, ISP's pay for the amount of data they actually move and spread that cost out across their users.
@Caleb In terms of the big picture, does moving internet traffic around really cost a lot? In terms of energy? Or is it just people trying to make a (big) profit on their sunk investments in hardware?
Also the higher you go up the chain the more everything runs at capacity. The last mile link houses may be fast but you count on it not being saturated all the time. An ISP's link up tends to stay pretty full, possibly even saturating during peak. Top level pipelines stay pretty full and earn money based on usage not potential speed.
It doesn't cost energy to actually move (fibre is very low power) but infustructure costs to maintain and constantly needs upgrading. The only way you ever make profits is by having lots of customers. It costs too much to operate to make any money at all one a small user base.
If you're trying to ask if fair use quotas are reasonable or just a scam to earn more money the answer is that they are absolutely necessary — ISPs would never stay afloat without those kind of caps in place: moving the data costs too much.
@Caleb WOAH there, I was under the impression that price went by installed capacity upstream not by actual bytes moved, since if bytes/$ wast too costly I just connected to another upstream.
@Caleb Ok, but those cost may actually go down with more users, no?
Also, isn't hardware more efficient nowadays? I know that they may be more expensive and require more maintenance cost (specialized personnel) but moving the bits be cheaper overall?
cheaper in the way of potential-capacity/bucks-spent
Because data requirements in general are always increasing. My company handles something like 600 PB of data but last year it was only expected to be around 150PB
It's probably going to be several hundred ZB by next year
@Jesse_b I think home users tend to use more data if it's avaliable, but there's a natural cutoff. I now use my computer as a TV, more or less. Which I would never have thought of doing till recently. But I've got to wonder where all this need is coming from.
@FaheemMitha Yea, I've heard the statistic that netflix accounts for like 70-75% of all available bandwidth in north america, because generally streaming data will use as much bandwidth as is available to it.
I guess to be fair that just means they should upgrade their infrastructure to support the services they sell but I think it's likely not the top level ISP's fault. It's likely the end user ISPs that are oversubscribing
oversubscribing is probably not the right word but overselling I guess
not always. Generally in markets where there is competition, services are fine. In the US and Canada anyway however most areas only have one end-user ISP
and in those areas the service is generally barely acceptable
I would like to use find and rsync to copy any directory that contains at least one "wav" file. The copying should copy the whole folder if a "wav" file is present. I have this so far:
find . -name "*.wav" | rsync -armR --include="*/" --include="*.wav" --exclude="*" /users/lukew/desktop/merged...
