27 August 2019

Google Play Movies & TV App Version 4.14.22-tv Enables Closed Captions by Default

I started watching a moive the other night on my Sony Bravia Android TV via Google Play Movies & TV and found the closed captions active. I turned them off and watched the movie thinking it was fluke.

But then, I switched over to watch a few episodes of some televsion shows I purchased through the Google app to find the closed captioning enabled for each of these. No matter how many times I turned it off, the next episode/movie would have it enabled again.

After calling Sony (and having them point to Google) and calling Google (who thought I had turned them on myself), I was at the end of things promised a call back the next day. I think it might have come two days later, but I had already determined the problem.

I had jumped through the hoops of clearing data and cache and force stopping the app with no effect. It later dawned on me that a tech had me remove the updates for an app and reapply them for a different matter in the past. I gave that a shot.

What I discovered is that version 4.14.22-tv of the Google Play Movies & TV app automatically turns closed captioning on for everything. Switching it off only applies ot the current show (episode or movie). The next selection will take the new global, unctrollable setting.

There is a small reddit thread started by someone six days before I found it. I documented whole experienece of removing the updates, having normal closed caption behavior, then reapplying the updates and going back to frustration land.

If you downgrade (remove the updates) in the meantime, you can enjoy everything as normal. My Sony Bravia will auto update the app, so I will be continuing the practice of downgrading until Google fixes this (I assume this is a bug and not some new default setting).

Reddit: https://www.reddit.com/r/GooglePlayMovies/comments/crf3uc/subtitles/
YouTube: https://www.youtube.com/watch?v=wAOOrGev3cQ&t=7s

15 August 2019

Amazon: New or Used

My niece is a gamer. When she pinged me a couple of months back during E3 to tell me how excited she was about the "Zelda: Breath of the Wild" sequel, I went ahead and bit the bullet to get her a Nintendo Switch for her birthday.

I jumped onto the Amazon app, searched up the device, made sure all was legit, and purchased the unit. It came promptly, and I put it aside until her birthday a couple of months later.

The day finally arrived. I knew she was going to be so excited to get this. And she was! She was very happy. Since we had spent the day out celebrating her birthday, she was tired and planned on opening the Switch the next day.

Midmorning the following day, I pinged her to ask her how the Switch was going. She told me she had just gotten up and was starting it up at that moment. A few minutes later, she messaged me asking if I ordered the unit was new or used. She said there were accounts already on the device.

I checked the order from Aamzon, and it was listed as New, sold by Amazon Services, Inc. I Googled the account names and found nothing to indicate any dummy/default accounts preinstalled. I opened a chat session with Amazon. A tech joined shortly, and I began to ask about the situation.

In the meantime, my niece was still trying to connect the unit to the WIFI with no success.

I had asked her to send me the serial number. I thought it might come up in the conversation with the tech. She gave me the number. I did another quick search only to see if there were any reports with that serial number. But the number didn't seem to match a Nintendo Switch. A little more digging informed me that the serial number my niece provided was for the charging unit. When I told her this and where the serial number should be, she told me it had been torn off leaving only little bit of white sticker where it once was.

So, now I have discovered that although I ordered New and my invoice reports New, I was sent a Used device which isn't even functioning as refurbished.

To make matters more frustrating, the return window for an expensive device like this is only one month. When I mention this to the tech, she tells me that she can make an exception to the return window. This is a concern to me since I tend to buy gifts in advance, especially around Christmas. I have purchased many items a couple of months before the event. With a one month window for returning items, I am now realizing that I am going to need to either wait until it is less than a month and take the chance it might be sold out or order early and open it make sure it works.

What really makes me the most angry is that although the device was listed as New in all the places for the order, I received a less than Refurbished item. I could even understand if this was sent from an Amazon seller, but it was Amazon Services itself.

Amazon has made things right, something I expect for an item this caliber. A one month return window on this type of device seems a little small. (A friend suggested a one-month-from-open and not one-month-from-receipt. That seems smart.) And how does a unit with non-functioning WIFI end up in a box in their New-status stock?

This experience will change how I handle gifts from Amazon.

31 July 2019

Ryzen 5 2600 Browser Latency

I have recently built my first Ryzen desktop. I am running the Ryzen 5 2600 processor with 32GB RAM, the Sapphire Radeon Pulse RX 580 8GB graphics card, and a couple of SSDs.

My goal was to get system which could handle the output of one 4K monitor and an HD second monitor without any sluggishness. The graphics card seems to be meeting this goal. The 32GB RAM also helps keep the machine spiffy.

What I am finding, though, is that at times when choosing functionality within a website, operations for that activity pauses for a moment or two. Simple actions such as opening a page from a bookmark or pulling a list for a dropbox are met with the same pause. It's more than data retrieval on the backend. The all browser interaction is suspended.

I do see a spike for the browser process in my system monitor, but I've seen that on Intel processors running under the same or lesser load with no latency. The spike happens, but there is no lag with Intel. And the Intel boxes do not have the graphics horsepower or increased memory this Ryzen box has.

And it is not just a single browser. I have seen this behavior with four different browsers on the same system. Is this just a Ryzen thing?

I want to blame Javascript because we all know how much of a glutton for memory it is. But it is more than just a single page or tab within the browser. I cannot immediately switch among my tabs during that pause. Maybe it really is Javascript impacting the whole browser. I have not done any investigation to determine if that is the case.

If it is the Ryzen chip, maybe the Ryzen 5 2600X might perform better. Overall, I am pleased with Ryzen and would like to do more with it. This is the only real issue I have encountered.

24 July 2019

NMAP Adventures

Today, I spent some time checking some firewall rules to see if any were still valid and which ones could be cleaned up. (These are internal firewall rules between our cloud account and our data center.) I was using nmap.

At a point, I wanted to be able to check both TCP and UDP protocols for a specific port for a specific host. I read through the help and man pages to see that I can use -sO. This is where things got a little wonky. (Number have been changed to protect the innocent.)

$ nmap -sO 1-17 -p 52001 10.10.10.5
Starting Nmap 7.70 ( https://nmap.org ) at 2019-07-24 13:16 EDT
Protocols specified must be between 0 and 255 inclusive
QUITTING!

Huh?

$ nmap -sO 1,6,17 -p 52001 10.10.10.5
Starting Nmap 7.70 ( https://nmap.org ) at 2019-07-24 13:16 EDT
Protocols specified must be between 0 and 255 inclusive
QUITTING!

This goes on for a little longer using different syntax. I dig further into the man page. After hashing it out, this sentence reveals the confusion:

Yet it still uses the -p option to select scanned protocol numbers, reports its results within the normal port table format, and even uses the same underlying scan engine as the true port scanning methods. So it is close enough to a port scan that it belongs here.

Wait, so, the -p option is used differently than everywhere else? Seems like it. I feel like points should be deducted here.

Using the following gets me the list of my protocols, however they are not completely what I need to know:

$ nmap -sO 10.10.10.5 -p 1,6,17
Starting Nmap 7.70 ( https://nmap.org ) at 2019-07-24 14:06 EDT
Nmap scan report for 10.69.55.214
Host is up (0.00086s latency).

PROTOCOL STATE         SERVICE
1        open          icmp
6        open|filtered tcp
17       open          udp

Nmap done: 1 IP address (1 host up) scanned in 1.48 seconds

Here's the real truth, though:

$ nmap -sU -sT -p 52001 10.10.10.5
Starting Nmap 7.70 ( https://nmap.org ) at 2019-07-24 13:32 EDT
Nmap scan report for 10.69.55.214
Host is up (0.00081s latency).

PORT      STATE  SERVICE
52001/tcp open   unknown
52001/udp closed unknown

Nmap done: 1 IP address (1 host up) scanned in 0.49 seconds

And it leads me to this final bit of confusion: Why does the scan using -sT -sU -p 52001 not give me the same as using the -p U:52001,T:52001 syntax? This is what the latter gives me:

$ nmap -p U:52001,T:52001 10.10.10.5
Starting Nmap 7.70 ( https://nmap.org ) at 2019-07-24 14:06 EDT
Nmap scan report for 10.69.55.214
Host is up (0.00084s latency).

PORT      STATE SERVICE
52001/tcp open  unknown

Nmap done: 1 IP address (1 host up) scanned in 0.49 seconds

This is very confusing. Comment if you have some insight into this.

27 February 2019

recreating a user

a short while back, i made an attempt to switch my home shell from bash to xonsh. in the process of doing this, i took more of a raw approach. despite the learning process that was, it left me with scars.

these scars are specific to my system's user management and my account (see the last paragraph of the shell switch post). i no longer had a recognized user requiring me to type in my username rather than just typing in my password.

today, i made the attempt to fix that. now, i will admit there might be a better way out there to do this, but i took this (raw) route again.

with superuser privileges, i edited the /etc/passwd file and appened a digit to my username and changed my user id to 1001. i did this in the /etc/group and /etc/shadow files.

i then created a new user with my original name using this command (as root):

# useradd mock -d /home/mock -M -u 1000 -G wheel -U

i set the password for the replacement user, and i rebooted the system for all the things to be recognized. (i came back later to add the other groups to which this user belonged.)

the command created the new user with the established name assigning the original user id to the new user and the original user's home directory also to the new user. because the user id is the same, all filesystem permissions still belong to the appropriate user.

i then removed the old, modified username from the /etc/passwd and /etc/group files.

now, things are back to normal.

08 February 2019

when to use less instead of vim

i am baffled that so many people feel like vim is the correct tool for viewing files on linux systems. using vim to view files you'd edit is fine. when i am viewing code or a config file which i may change, i might use vim but only if i plan on making changes to the file.

but when i am going to view a file which i don't ever plan to edit or ever need to edit, i use less. log files are ones where i see this happening very often. unless you expect to change the log file, i see no reason for using vim.

even when viewing files which could be edited, it's best to use less. scenario: someone wants to view a file. they use vim. i jump on the server and need to actually change the file. i get the message that the file is already being edited. maybe i can say that i want to edit it anyway, but there are no guarantees someone won't :wq or ZZ out of the file and undo my changes. if they used less, there would be no threat of this.

less lets me move through a file much quicker for pure examination purposes. spacebar works like CTRL+F to scroll a page at time, yet all the vim moving commands are available too. searching text is the same as vim. and i can exit less with a simple tapping of the q key.

"but, but i need line numbers and vim lets me do that!" so does less. use the -N option on less to get line numbers.

"what if learn that i need to edit a file after i have entered less?" just press v in less to enter vim mode.

and a side annoyance are any leftover swap files vim might leave behind on a dirty exit of vim. swap files for log files? how first-year cadet is that?

if you feel the outstanding need to use vim to view files which should never be edited, then use the -R argument to ensure the file is never changed. if you practice using the -R, i am willing to bet you'd find typing less much easier.

and before you ask me, "why not use more?" i will let you figure out why less is more. (and if you are viewing files with nano, there are no words to describe that insanity.)

bonus assignment: find out what the view command does.

25 October 2018

changing shells

i don't know if this a fedora-specific thing or a newer linux thing, but in the past changing your default shell was as simple as editing the /etc/passwd file. today, there is the chsh which can perform this for you. there is a catch, however. the shell must exist in the /etc/shells file to be allowed to be changed.

to make this work, i needed to add my shell to the /etc/shells file. then i issued the chsh to make my shell change.

$ chsh -l
/bin/sh
/bin/bash
/usr/bin/sh
/usr/bin/bash
/usr/bin/tmux
/bin/tmux
/usr/bin/xonsh

$ sudo chsh -s /usr/bin/xonsh mock
Changing shell for mock.
Shell changed.

my previous post about xonsh is what led me here. i had manually changed the shell directly in the /etc/passwd file, but then my gnome session did not recognize my user. this was the strange side effect of changing the default shell not on the /etc/shells list. gnome (gdm) started the gnome-initial-setup process which runs when gnome is freshly installed.

and thing are still messed up. the gnome settings > users section shows no one, even though i am logged in under my account. (figuring out how to fix this will be another lesson for a future time.)