Ubuntu 17.04 with Gnome

I spent the day Saturday upgrading from Ubuntu 14.04.02 to 17.04. I always know that there will be hours spent tweaking things, reinstalling programs, cussing a lot, and wondering WHY DID THEY DO THAT?!  Once I had it installed, I installed the gnome-session-fallback to get out of the restrictive Unity shell. I just don’t like it.

Gnome has been a nice environment for me for years. Except this time they have moved some of the button locations. I don’t mean the Minimize, Maximize, Close buttons for which everyone and their uncle have fixes. I mean buttons like the final SAVE button when saving a picture from a browser. It used to be on the bottom right, now it is on the top right and the bottom right has a dropdown list of image types that used to be in the lower left. Grumble, murmur, @#$*%&!

When users become used to using windows a particular way, it becomes a royal pain in the ass when developers move things around because THEY think it is better and fuck you users for thinking differently. At least give us a straightforward way to move things around where we want them so our productivity doesn’t suffer for your whims of design fancy.

Another thing is that any time I copy a file from one disk to another, BOTH disks have a popup telling me the progress of the transfer instead of the old “progress” window that was an unconnected popup. And get this – the popups both STAY ON THE SCREEN UNTIL I MANUALLY CLOSE THEM!!!

I tried the gnome tweak tool, gsettings many keys and values, and haven’t yet found a way. I figured if there is a setting for those three buttons in the upper right, then perhaps there are settings for the rest of the window layout. Not so much, it seems. I even installed Gnome 3.24 and no change.

Oh, and you can’t just resize windows by dragging the edge like we’ve been doing for the past 27 years. Now you have to press a key combination to enable that first. Alt+F8 for me. Who thought that was a good idea?! Stupid shit like that… I did find that I can do Alt Middle-click and drag.

So, if any of you knows how to tweak this, please let me know.

Other gripes:
Themes that look decent are apparently hard to find. The default icon theme Adwaita makes folders and documents look the same in this shell. The default orange icons used in Humanity are gross. Orange? I am settling for “Tango” icons now that are blue, which is at least tolerable.

Atomic Tanks 6.5 sucks shit compared to the previous versions. The fun has been sucked out of it and there is little chance of winning when they kill you with 1 or 2 perfectly accurate shots every time. I uninstalled it. I tried installing older versions from the Ubuntu software repository, but they choked on 17.04 and the lack of dependencies. I tried installing the dependencies, but no go. Boo hoo!

UPDATE: I installed XFCE4 as an alternate shell at login, and I love it so far. The save button is back on the bottom right, the appearance is nice, and things are mostly running smoothly. The thing that prompted me to install it was when a 1994 graphics program I was running under Wine locked up the Xwindow so that the mouse clicks and keyboard inputs were ignored. But this also happened in XFCE, so it looks like Wine 1.8.7 may have an issue with Ubuntu 17.04. (I know, I should get with the times, but PSP 4.12 works so damn well and has no bells or whistles.) I’m installing Wine 2.0.1 tonight, so I’ll see if that fixes things.


Audio hiss/noise on PC

Once again I’m having horrible sound quality on my PC and spent hours on forums searching for possible resolutions. I checked several software solutions and am convinced now that it is a motherboard or electrical issue. I tried a different outlet across the room, but heard the same noise. I tested the outlets wiring and they seem to be wired correctly.

The noise started very suddenly today after being quiet for a month or so. Even during the boot process, before any drivers can load, there is an odd hissing back and forth in the headphones with odd bloops and beeps now and then. This gets worse after the OS loads (Linux or Windows). Even when sound is muted on the OS, I still hear the hissing and blooping. I did test the headphones on a different device (battery powered) and they work fine. I also tried a much older PC on the same outlet and it only had a very faint rustling sound in the background, leading me to think it is a motherboard issue.

I may try the EB Tech HumX device, but it is around $70, so is an expensive experiment. Even a ground-loop isolator doesn’t change anything (I didn’t really expect it to on headphones).

I also want to get a UPS for the PC to further isolate the power source, but when I went to Amazon to get the one I wanted, the reviews for the last month say that the latest batch of CyberPower units have a terrible strong odor that doesn’t go away. So looks like I get to wait until CyberPower takes charge, forces some quality control on their China factories, and purges these bad units from the market.

In the meantime, I have to record onto a handheld digital recorder and listen to files the same way.

UPDATE: It did turn out to be the motherboard having a problem. I bought an inexpensive PCI-e audio card (ASUS Xonar DSX) and the sound is perfectly clear now. It took a bit of looking at settings to make it work in Linux, but it was recognized right away and was mostly a matter of adjusting settings in alsamixer and pulse-audio. Mobo is a 10yr-old “ASUS P6X58D Premium”. Still works fine for what I need, but probably time to look at building a new box. I disabled audio in BIOS so that nothing would interfere with the add-in card.

However, in Windows 7 64-bit, the card installed fine but has no sound at all. The only other sound device is the video card which has HDMI audio available, but which I’m not usning. Tried reinstalling the driver and enabling the HSMGR.EXE file and another similar 64-bit one that kept popping up asking for permission. Still nothing. I’ll have to work with it another day.

UPDATE: For whatever reason, after removing the drivers again and reinstalling, rebooting, the ASUS Xonar DSX now works in Windows 7 64-bit. This time it didn’t prompt me to allow any exe files. (I did notice how damn slow Windows boots. Maybe that’s a sign of the age of the motherboard. Linux boots fine.)

Uber – We don’t want to hear from you

A couple of days ago, I began receiving Uber (the private taxi service) emails for a guy named “Troy” to my gmail account that clearly has my own name, not his. I tried to find some kind of link in the email for “Hey, if this isn’t the right email, please let us know”, but there was no such link. If I reply to the email, Uber replies “Hi there. We’re sorry. You’ve contacted an address that does not accept incoming email.” Then they direct me to their FAQ. Oh, and I can sign up if I want to contact their tech support.

Hey Uber, since you didn’t bother to use email verification when people sign up, I could put anyone’s name in and have them receive annoying useless emails from you. Whee.

Same goes for Lyft. I started getting emails from them also, with no verification that they had the right address.

Then I got an email for a Jon White who purchased an Uber gift card, and he has an email address that bears no resemblance to mine at all. Yet I received it.

Uber, your technical abilities suck publicly. Start using email verification when people sign up like everyone else has been doing for 20 years. And make it easier for people to contact your tech support with issues without signing up for your service. I’m not only not a customer, now I never want to be one.

So for now, I will just block them and their gift card service.

Update: I stopped receiving Uber and Lyft messages, but someone named Ron White (probably not the comedian) is still trying to use my email to sign up for things. Or he’s stupid and isn’t checking his auto-correct.

Honeywell NetAXS security card panels

I work at a facility that uses a LOT of Honeywell card-access panels, mostly the N1000-4x type, and for several years we’ve been happy with those (though seriously, only 4 doors per $2000 panel?!). When they came out with a new panel with built-in Ethernet, we started purchasing those instead. The NetAXS 4 panels saved us from having to buy a separate Ethernet to RS-232 and RS-485 devices, and we loved how quickly they would initialize compared to the old panels.

However, we ran into a pretty major brick wall recently. Our campus has several buildings, with several floors, and several departments. One building houses a number of different tenant organizations. This means that we have a lot of different access levels, 108 to be exact. The software we use to manage the panels is called Winpak SE 4.4, and it began complaining immediately after we updated from version 3.3 that some of our NetAXS panels may not function properly, and that they are limited to 128 access levels. WTF?! None of us recalled seeing anything in the adverts about a limitation on how many access levels could be used. And why in the world would a newly designed panel have such a tiny limitation when its older version had virtually no limit?

Now 128 may sound like a lot until you find out that every card given a custom access to a particular door counts as an access level, we tend to run out very quickly on a popular door. This seems to be the primary way that levels are used up. A single card can only have a single assigned access level, so we have to do custom levels quite often.

We could replace the new panels with the old panels. But at $2000 a piece, that is a chunk of change.

ANOTHER HUGE ISSUE: We recently had a lot of difficulty getting the panels to download access levels. Through several calls to tech support and trying new firmware and OS files, and several emails sent, we found that our database (which started back around 2003, I think, and has gone through several upgrades) probably has a bit of corruption. How it manifests is that we see cards repeatedly being given “Host Grant” on the Event View. Host Grant is designed to check with the server and download a card to the panel if it is authorized, but not yet downloaded. The cards were being downloaded, but because no access levels are ever downloaded to the panel, the panel is continually using Host Grant. That is, the cards are downloaded, but because they have no permissions, they can’t do anything. So the panel checks with the server and it says, “Sure, let that card through that door” or “No, it isn’t authorized”. This creates about a 2 second pause between the card being read and the lock opening.

Tech Support had us create a new access level with appropriate permissions for each of our existing levels, isolate the cards in each level and assign them to each new matching level. This creates new fresh access levels that the panels can read just fine. They immediately were downloaded to the panels and everything began working as designed. That solved some big issues when people couldn’t get through doors when the network was down. Now the panels store the information and can work independently of the server most of the time. I also had to go through each custom access level that touched a new panel, remove the custom setting, then add it again. 5 hours later everything is working normally.

ANOTHER ISSUE: If we have an “active shooter” situation, we’d like to be able to shut down access to a site, or lock all the doors. That isn’t an option in the Winpak software without doing some major cartwheels. The software has been out for at least 15 years, so they’ve had ample time to work on it. All Honeywell ever seems to do is release a new version without new benefits other than support.

ANOTHER ISSUE: The Winpak database isn’t a pivot table. If you want to search for specific things that Honeywell doesn’t offer as a prefab search, you are out of luck. It runs off of a SQL Server Express database, but the forms they offer for searching often leave out things we’d like. If it is in the database, we should be able to search for it, and create custom searches.

So, we did get some good help eventually from Tech Support that resolved a big issue. But we still would like to see some major improvements in the software soon.

Choosing a microphone for singing (and mic-shaming)

I sing jazz standards as a hobby. Doing so puts me in contact with a lot of singers, some professional, most not. I’ve found that a lot of people don’t really know much about choosing a microphone, and tend to rely on others recommending a brand. Conversely, I’ve found that some will “mic-shame” others for their choice, ignoring the actual sound produced by the singer, or overlaying their actual sound with mental prejudice.

Choosing a mic is a personal decision, and finding one that fits your voice is essential to producing the sound that you desire. There really is no substitute for going to a music store and trying out at least a few mics (and actually singing in the store). If they are interested in sales, they should have a way for you to test a few mics at a time with a mixer board and quality speaker (or speakers).

There are generally two kinds of handheld microphones, dynamic and condenser. Dynamic mics are the more common kind, can take some wear and tear, and do not require any special power source. Condenser mics are said to be more sensitive to sound, but subject to damage and require that the mixer board provide 48 volts of “phantom power” over the mic cord to the microphone or have a battery installed.

You may also want to consider a wireless microphone. Such a mic also requires a receiver station, and there are considerations for the frequency range that the mic uses. If you intend to use your mic at vocal jams, keep in mind that setting up the receiver station in a safe location with a power cord is a lot more work than just hooking up a mic to the venue’s mixer board. Wireless mics require a battery, and may encounter interference with other electronics.

Comparing Microphones Equally
There is a common misconception that one equalization fits all mics, or should fit all mics. This is balderdash. Each mic has different components and electronics and will require that you pay attention to the sound quality you are producing with that particular microphone. I’ve seen Youtube mic comparisons where they use the same settings for each mic, allegedly to show an unbiased approach to the sound. Again, this is nonsense. Each mic requires that you set it up for the best sound for your voice and that particular mic. Only then can each mic be compared fairly for your choice. Unfortunately, not all shops carry a wide selection, so you may want to start your search online for popular microphones (I can save you a bit of time by saying that the most popular mics I’ve encountered are the Shure Beta 58, the Sennheiser e935, and the Heil PR35).

1. Each mic you are testing should be set up with its own cord to the mixer board. While speaking and singing various styles with each mic, tweak the equalization knobs (often just  High – Mid – Low knobs) on the mixer board starting with each at the middle position. Try each mic with and without some reverb. Your goal is to find the best sound for your voice with THAT mic. Once you have found the best setting, turn the volume down for that mic and try the next mic, again tweaking the settings until you find the best sound for your voice and that mic.

2. Once you have set up each mic with the best sounding settings, begin comparing the sounds that you produce over each mic, both speaking and singing various styles. Notice if there is “handling noise”, that is the sound of simply holding the mic in your hand and moving your fingers. Some mics pick up a LOT of handling noise, and you may need to use a low-cut or high-pass filter on the mixer board for that mic (oversimplified, this cuts out some of the low range of the mic while allowing the range for human voices through). Try the mic with windscreen (sometimes called a “clown nose”) and without. Windscreens can help control plosives (spoken “P” and “B” words that have a sudden burst of air directed at the mic) and sibilance (hissing “S” words which the mic can pick up strongly). Learning how to handle the mic and to direct plosives and sibilance away from the mic can be a better solution than the windscreen.

3. Narrow your choices to two mics and then decide based on sound, cost, and durability of the microphone.

I’ve seen singers go through this process and pick out a perfectly fine microphone, only to then encounter another singer (sometimes a pro) who decided years ago that their microphone was the best and only mic that should ever be used by anyone and that all other mics are inferior regardless of sound and components. Some singers arrive at this simply based on “I’ve always used this and it works, and I don’t like change”.  Sometimes the shame factor is enough to make a singer change mics, simply on the basis of alleged authority. Do the same comparison of mics as before with the suggested mic and the mic you already chose. Ignore the “authority” and focus on the sound quality, then make your choice.

My microphone is the Heil PR35. I chose it after looking online at hundreds of reviews and forums, after using a few different mics at a mic-comparison session with fellow singers, and based on the specs of the microphone. The PR35 has a larger diaphragm than most other handheld mics, and this gives it both sensitivity to low volume singing (for a sultry whispery sound) and to low frequency sound which is helpful for conveying my baritone voice. I cannot get right up on the mic as I would with the Shure Beta 58, but then again I don’t have to because it picks up subtle sounds easily. It does have rather a lot of handling noise, so I find myself compensating with a high-pass filter and slightly boosted low EQ. This mic gave the most natural sounding version of my voice, to my ears. It is more expensive than some mics, and most shops don’t carry it as a mic you can try. I tend to use a windscreen to help avoid plosives, and haven’t noticed that it “muddies” the sound, though it does dampen the overall volume a bit, which can be compensated at the mixer board.



Update on Electronic noise during recording

After we moved to a new home, I had not been able to record vocals on my computer due to a constant electronic noise in the background, something like a high-pitched whine and chatter.

After a lot of troubleshooting, I purchased a Ziocom Ground Loop Noise Isolator from Amazon ($9). It is a small non-powered device that plugs into the microphone input (3.5mm) on the computer, and the output from the mixer board plugs into the ground loop isolator. Just that quickly, all the chatter stopped and the recordings are clear again, except for a background hiss. I’m going to try a DAC to convert to USB digital prior to the PC and see if that fixes it.

The concept is that the ground point of my computer and the ground point of the mixer board are different, even though they are in the same room. This creates a slight “potential” (DC current) which is then amplified by the mixer as electronic noise. This gizmo filters out that current and lets the audio signal through.

I used an XLR female to 3.5mm audio cable from the mixer to the isolator.



(NOTE: I am not an electrician. I am familiar with basic wiring, having done simple projects for about 30 years. If you are not entirely comfortable doing such a project, please hire a licensed electrician who can easily do it for you, or just change out the entire fixture.)

The light fixture in my closet was under-powered so much that I couldn’t tell the colors of the shirts. I removed the glass cover and discovered that it was using orange-colored compact fluorescent bulbs with an odd base I hadn’t used before (GX24Q-2).

I searched online to see if there were LED lights that used this base, and found that while there were some, there were also adapter bases made to allow a standard bulb to be used. I purchased 6 off of Amazon and looked at the wiring description which wasn’t a lot of help.

It said to remove the fluorescent ballast entirely, but leave the wires attached to the GX24Q-2 sockets. On each socket there are 4 wires, 2 wires will be unused, 1 will be attached to the 120V white neutral wire and 1 will be attached to the 120V black hot wire.


I shut off power to the fixture at the breaker to prevent accidental shock. I removed the fixture from the ceiling and unscrewed the 3 wire-nuts holding the fixture to the house wiring (Ground, Hot, Neutral).

I used my multimeter to test which of the pins on the new adapters went to the screw socket and which went to the base contact. The side of the screw socket will be connected to the white neutral, and the base contact will connect to the black hot.



I found that 2 of the pins on the adapter were not connected (these correspond to the 2 unused wires), and then labeled the 2 remaining pins either B or W for Black or White. I also labeled the sockets attached to the light fixture so I could easily match them. (I’m not sure that it really matters for a bulb, but since I’m not sure, better to just err on the side of caution).

I used a marker to put a dot on each wire that went to either the B corner or the W corner. The other two wires I bent down and attached a wire-nut to keep them from touching anything electrical. I arbitrarily decided to make the yellow wires attach to Black 120V and the others to attach to the White 120V.


I used wire-nuts to attach the colored wires to the house wires. I turned on the power again and tested the lights which worked perfectly. I reattached the fixture to the ceiling.

All in all, the directions that came with the adapters left a lot out. It would probably have been easier to just get a new fixture, but it was a lot cheaper to do it this way. As I said at the beginning, if this is beyond your skill and understanding, please do not attempt to do it yourself. Hire a licensed electrician to do the work.