linux: you cannot enter the virtual machine with putty

There are many threads about this … I've been reading them.

This is what I have, maybe someone can help. I am trying to build an OpenWRT image and I am struggling. I have a virtualbox machine running alpine. Root user (obviously) and a normal user, M00, created.

Virtual box network configuration:
Network: NAT
Port Forwarding: SSH Name, TCP Protocol, HostPort 2022, Guest Port 22

There is no problem running ifconfig on the virtual machine and get 127.0.0.1

run ipconfig on the laptop and get the address 192. etc. (without virtual box).

using putty, try entering as host: 127.0.0.1 Port 22, Port 2022, host M00@127.0.0.1, etc., but it was in vain.

If I do something with port 22, I receive a connection rejected message. If I do something with 2022, I get a black cmd screen with the ability to do nothing … without writing, nothing. Ideas?

Restore Time Machine Backup on a previous operating system

I did the same when using a USB installer and migrated data from the Time Machine backup during installation (it is one of the options during installation). In my experience, the data (minus the operating system) should return to Mojave, but they may not be visible in the backup list to choose from when choosing a backup source during installation. To explain further:

When I installed Catalina, I chose to migrate the data from the Mojave Time Machine backup, which was successful. In the brief period of time I had Catalina, she made maybe 1 or 2 backup copies of Time Machine. When I chose to erase the SSD and install Mojave again, Catalina seemed to have changed the file extension of the backup from .sparsebundle to .backupbundle. This prevented Time Machine in Mojave from showing me the backup to select (imagine my heart rate). Then, I deleted again, I installed Catalina, and after exploring the Time Machine backup in Finder, I could see the previous backup there, I changed the file extension from .backupbundle to .sparsebundle, I deleted the SSD again, I installed Mojave and soon, the previous Mojave. The backup was visible again. The data migrated successfully for the most part (except my contacts, which is another story).

Machine learning: what ML algorithm can "colloquialize" text bites?

I am looking to colloquialize short strings of natural language and titles for a more adequate grammar, and do the same in reverse.

For example, I would like to find an ML algorithm that, given the right data set and training, can turn "all of you" into "all of you" or "she was like" to "she said." Even professional titles: "Director of Human Resources" becomes "Director of Human Resources".

To be clear, I have many thousands of chains that must be colloquialized so I can't do this in a deterministic way.

Can anyone tell me what would be the right ML algorithm for this kind of thing so I can investigate?

At first I thought it would be a Naive Bayes classifier until I realized that we would end up with thousands of classification categories and that's not good.

Thank you!

How do I prevent the command prompt windows from starting on the console of an AutoSys virtual machine?

I have a Windows 2003 R2 virtual machine that runs AutoSys. The service runs under a domain user account. Most jobs involve the execution of batch files.

When someone logs into the VM through the Remote Desktop, the behavior is normal, but when someone logs into the VM console, he sees several command prompt windows that presumably belong to the AutoSys process execution.

The behavior is the same regardless of the credentials that were used to log in. Does anyone have an explanation of why this happens or how to prevent it?

I assumed that windows initiated by a process that runs as a particular user would only be visible when logged in as that user, but that does not seem to be the case.

ssh: port forwarding in the dockable container on the Linux AWS EC2 machine

I am a newbie and I need help. Currently playing with the following settings:
remote machine
an instance of AWS EC2 with Ubuntu 18.

I am trying to access the Jupyter laptop that runs in the Docker container on the AWS EC2 machine but I cannot access it.
Timeout error

I tried the following command for port tunneling that I read somewhere

ssh -L 8080: localhost: 8080 -i /home/user/admin_keypair.pem ec2user@ec2-xx-xxx-xx-xxx.us-east-2.compute.amazonaws.com

but getting this error

Warning: the identity file /home/user/admin_keypair.pem is not accessible: there is no such file or directory.
Permission denied (publickey, gssapi-keyex, gssapi-with-mic)

launchd: start the startup virtual machine before USB starts

I have a plist file located in / Library / LaunchDaemons that starts a headless virtual machine with Ubuntu on it.

However, the problem is that the USB probably does not start before the VM is started and, therefore, a USB memory located on the COM / dev / ttyACM0 port (on the VM) is not available.

Exit "VBoxManage showvminfo userver" when a normal startup is executed:

Currently Attached USB Devices:
UUID:                        73eb5973-ce2a-4410-8820-4ec7376e3f1d
VendorId:                    0x0658 (0658)
ProductId:                   0x0200 (0200)
Revision:                    0.0 (0000)

Address:                     p=0x0200;v=0x0658;s=0x0000000c8eba2f72;l=0x14200000

And with an automatic start:

Currently Attached USB Devices:    

Is there any way to delay the start of the script?

attacks: how to prevent access to local machine data

Sorry for the terminology since I am not an expert in information security.
My question is quite simple:

How can I guarantee * that I am the only one with access to my local machine data (screen, microphone, camera, files)?

* Assuming no one will physically enter my department and mess with the hardware.

I have a Macbook Pro with macOS. I've heard of cases in which a laptop
The data (camera / microphone, etc.) has been accessed remotely without consent. I wonder if it is possible to absolutely prevent such attacks.

Until recently, I thought the contents of the hard drive were safe, but I don't think so anymore.

The simplest procedure that occurred to me is to turn off the WiFi, read / write data to an external unit (such as a USB memory), disconnect the unit (and turn on the WiFi again).

Does the previous method guarantee that no one at my side can access the data on the external disk? Is there an easier method?

Thank you.

macOS Catalina: Time Machine backup is affected at the end

I am running macOS 10.15. For some reason, my backups of the time machine stagnate at the end and are not completed.
Any ideas?

The backup gets stuck at the end

Time Machine: the fastest and most economical 2TB external storage configuration (for example, internal SSD converted to external)?

Pre-introduction
This question may or may not be specific to Apple, so let me know if I should post to Superuser or post elsewhere.

Introduction
I need external storage for three purposes:
1) To make Time-Machine backups for a 1TB MacBook Pro Retina (2015 – Yosemite)
2) I also need external storage to store media
3) I will potentially install and boot / run MACOS Mojave dual and use it at the same time as the host operating system, Yosemite

It may make economic sense to use the same external storage device for all three purposes. I understand that I would need to make a separate partition for Time-Machine backups.

Potential problem
I assume that the external storage advertised as 1TB is not sufficient for a Time-Machine backup because its actual disk size will be <1TB and, therefore, smaller than the internal 1TB SSD of the machine ( MBPr). Based on this, you should use at least one external 2TB disk and partition 1000GB for the Time-Machine backup.

Potential options
I guess I have three options, but I'm not sure which one would be the fastest and cheapest:
A) 2TB external SSD (for example, Samsung T5)
B) An internal 2TB SSD (for example, M2, Sata III, NAND or NVME? – not sure what is best here) and then converted through (connected inside) a cart (for example, Orico / Ugreen) to an external SSD with a (USB 3.0 or and Thunderbolt 2.0 connection cable / adapter?)
C) 2TB USB 3.0 Flash

Query
I guess option B might be the cheapest one, but I'm not sure if there is any inconvenience with respect to the other easier options. I heard that there may be a problem related to TRIM, but it seems that this can be overcome through the bash script sudo trimforce enable according to Yosemite and SSD Trim incompatibility

Alternatively
There is another argument that says it makes no sense to use an SSD for a time machine backup, since it would not be accessed frequently. In this case, would it make more sense to buy a cheap external HDD with 2 or more TB (this would be much cheaper than an SSD) and buy a separate SSD to boot Mojave and store media there?

N.B
Some sources say that SSD would be faster than flash, while others say that SSD these days is flash or vice versa … Does anyone have a clear idea of ‚Äč‚Äčthis (with / without a credible source)?

Nb2
I guess for option B I would not use a USB 3.0 cart with a Thunderbolt adapter, since it would limit me to USB 3.0 speeds, which means that if I want to take advantage of the Thunderbolt 2.0 speeds (10 Gbps) I should use a native Thunderbolt 2.0 caddy .

Nb3
I guess I can start Mojave Guest twice from an external drive with Yosemite Host on an internal drive.

Recommended quality H.264 mp4 for computer vision, image processing or machine learning?

I need some quality advice to re-encode videos to H.264 mp4 by inference

With the handbrake, what constant quality number is a good compensation for the file size without compromising the image quality to run the analysis? (or average bit rate if that is better)

The focus is quality, but still reasonably compressed so as not to lose space unnecessarily

if you know a ffmpeg setting that is better than using the handbrake, it would also be great

I am re-encoding mp4 mainly for compatibility
I would like to know what everyone is doing or recommend, or if you think mp4 recoding is a bad idea in the first place