Is there a reason cipher /w fills up temporary files instead of writing to the raw disk?

The Windows command cipher /w can be used to securely wipe a hard drive’s free space without the use of third-party tools or GUIs:

enter image description here

cipher works by creating a folder called EFSTMPWP on the root of the target drive; inside this folder, it successively fills three temporary files with zeroes, ones, and random numbers respectively, one after the other, to the size of the empty space left on the drive.

By the time a file has taken up all of the drive’s empty space, it’s effectively forced the file system to overwrite all data held in its free space with the file’s newly-written data, rendering any data previously held there permanently irrecoverable.

enter image description here

I understand Windows doesn’t typically grant users access to the raw drive like Linux does, but I’m confused as to why Windows’ own utilities would choose to write data indirectly to files in this way and be subjected to disk I/O bottlenecks in the process, instead of simply writing to the raw disk itself.

Is there a particular reason it wipes disks using files instead of writing to the raw disk, or was this likely just a design oversight?

Create a PDF with specific page ranges of multiple PDF files?

I have found many related questions in this forum. But, none addresses my issue. Please double-check before marking it as duplicate.

Suppose I have two PDF files.

  1. first.pdf having 10 pages.

  2. second.pdf having 20 pages.

I want to create a new PDF file, where I need pages – 2,5,6,9 from first.pdf and pages 6,7,15,19 from second.pdf.

How to do it from command line?

CentOS 7: installed devtoolset-7, but GCC 7.8 is using GCC 4.8 installation’s old C++ include files

I’m trying to compile and install Pistache on a CentOS 7 machine, following their instructions.

This is a package that requires C++ 11, so since the Development Tools package on CentOS 7 installs an old version of gcc (4.8.3) that doesn’t support C++ 11 features, I’ve installed devtoolset-7 and enabled it with scl enable devtoolset-7 bash. This appears to be functioning properly, as gcc --version displays gcc (GCC) 7.3.1 20180303 (Red Hat 7.3.1-5).

However, when I try to make the Pistache sources, I get this compilation error:

/root/compiler_installation_script/pistache/src/server/ error: no matching function for call to ‘regex_replace(const string&, std::regex&, const char [2])’
     SegmentTreeNode::multiple_slash, "/");
 /root/compiler_installation_script/pistache/src/server/ note: candidates are:
 In file included from /usr/include/c++/4.8.2/regex:62:0,

Unless I’m misunderstanding, this seems to indicate that my compiler is using outdated C++ include headers that are compatible with gcc 4.8.3, resulting in the error. How can I fix this?

Note: I’m sure the issue isn’t with Pistache source itself, as I’ve compiled this successfully on other machines, including a fresh CentOS 7 machine yesterday.

Delete old windows / program files from second drive?

My PC has a bunch of extra drives. Most of them contain old copies of Windows and Program Files. My PC is called “PC”, and my admin user is called “Tim”. I’ve assigned ownership of these second drives to the user “PCTim”, and also given the user “PCTim” full control on the drives. I then try to delete either the Windows or Program Files folders on these drives, and get the message.

“You require permission from PCTim to make changes to this folder.”

The current owner of these folders is listed as “Tim (PCTim)”. The effective permissions for these folders lists this user has every right.

The absurd “answers” on the microsoft community of course offer things like “Use Disk Cleanup” (which is a blatantly uneducated answer) or “just reformat the drive” (formatting avoids the problem rather than addressing it).

What are the correct steps to delete these old files?

How to store files in Microsoft Outlook?

What’s the best way to store files in Microsoft Outlook?
Please don’t ask why. That won’t help.

I’d like to have them arranged in some kind of hierarchy. Most of the files will probably be document-type, but not all.

It would be good to be able to:

  1. Search the content of the documents, just like you can with email.
  2. Associate tags/labels with the documents.
  3. Link to them from tasks

How to move files to hdfs in hadoop using unix command

I have 2 directories “dft” and “hdfs” in the unpacked Hadoop folder. I am trying to copy the file(StartUnit.txt) in dft into hdfs.
The corresponding command and exception are as shown below

313159@PC213710 ~/hadoop-0.19.1
$ bin/hadoop dfs -copyFromLocal D:/eclipse/cygwin/home/313159/hadoop-0.19.1/hdfs              
copyFromLocal: Pathname /D:/eclipse/cygwin/home/313159/hadoop-0.19.1/dft/StartUnit.txt 
    from D:/eclipse/cygwin/home/313159/hadoop-0.19.1/dft/StartUnit.txt is not a  
valid DFS filename.
Usage: java FsShell [-copyFromLocal  ... ]

can someone suggest me…