06 August 2021

Video Meeting Etiquette

I can't believe this hasn't blanketed the internet (or maybe I haven't seen the blanket coverage), but here are a few items of video meeting etiquette that might be tough to hear but is what everyone wishes you knew.

Always enter meetings on mute

You might be confident about starting every meeting with your video active, but never think charging into a meeting unmuted is what people really want.

If you are regularly late to meetings, expect a lot of loathing from your colleagues. This is equivalent to entering an in-person meeting already in progress while arguing with your spouse or sibling over some personal matter. Silence is golden.

Using only laptop (computer) microphone and speakers makes you "that person"

This creates the least awesome echo tunnel evar. Everyone will hate to ask you questions because you'll need to activate the voice tunnel to answer. No one wants the tunnel! No! One!

You can spend about $12 on Amazon to buys a four-plus star rated headset with earpiece and boom mic. This will win points with your colleagues by leaps and bounds. They might even throw you a party.

Stay on mute when not actively engaged in discussion

If you never involve yourself vocally in a meeting, there is no reason to not be on mute. Your keyboarding, food chewing, slurpee slurping, dog barking activities are so much less subdued when a microphone is added to the scenario.

Please mute yourself if you know you need to cough or sneeze. At least move away from mic if you can't click a button in time. If we learned anything from Zonday, it is to move away from the mic when merely breathing.

And if you are guilty of the situation above, just stay on mute...forever. Your mic will likely try to pick up all sounds to make sure you are heard. You will be, as well as every other sound around you.

Add a profile picture to your account

This is less of an annoyance than the first three, but it can really help people quickly identify you. Most meeting platforms will switch to your account's representation when you speak. Video is cool but not always possible. A picture to represent is a nice welcome. It helps to have an actual photo of you, especially for meetings where people don't know the whole audience. Even if you choose to upload an image of your cat, it's still better than just a name.

Note: This is very beneficial for other communication mediums such as interactive chat/messaging. Just having the system-generate initial(s) for your name will help others assume you are someone else at a quick glance. I tend to quick glance at a chat when multitasking.

One of my smaller chat groups involved five other people all with the first initial "S." The chat tool defaulted to a color circle with the initial of the first name...and the colors were duplicated. To make matters worse, the colors didn't help. It would have been nice if the two green circles were the men and the two pink circles were the women or vice versa. But no. Green was one man, one woman. Pink the same. Color didn't really help. It took extra time to figure out which "S" I was answering. "Is the green 'S' Sharon or is the pink 'S' Steve? No, that green 'S' was Sharanya!"


If you know of other things which I didn't list, feel free mention in the comments. I know these are the recurring things which should have been said now that we've all experienced a year communicating in this fashion.


11 February 2020

Robot Authenticity

I've gone back and started watching "Mr. Robot" from the beginning again. I always believed that a lot of the code was real and the technology was legit and something with which I could identify.

And then I reached an episode where Elliot typed in nano to edit a file and realized it's just Hollywood garbage again.

No real hacker uses nano.

I would have respected him for even starting up Emacs. I wouldn't have agreed, but he would have had my respect.

29 October 2019

No More Captions

I am a bit late in posting the follow up to the previous rant about the persistent captions on Google Play media. They did in fact release an update fixing the captioning problem, but it took about a month to get it.

I ended up using the Google Play Movies & TV app on Roku to avoid the captioning problem. The app for Roku is not the standard app found on Android devices. It was uneffected by this phenomenon. Something I will keep in mind if things go wonky again.

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.