It's that time again when Ubuntu posts a new long term support release. That happens every two years, in April. Sometimes the alternate flavors of Ubuntu lag that release a little, such as Xubuntu and Lubuntu and the server version.

I upgraded the OS on my personal computer, which was running Xubuntu 18. I installed Ubuntu 20, Manjaro 20, Debian 10, and finally settled on Lubuntu 20, for now. Here are some of the highlights from the various upgrades, starting with our attempt to upgrade the Mac Pro.

MacOS
Our Mac Pro failed in its attempt to upgrade to the latest version of MacOS. Physically it is a plenty capable machine by today's standards and it was very expensive when new. We even upgraded the video interface to a newer more modern one for this upgrade, because the upgrade program said it was needed. However, due to the age of our expensive Mac Pro, Apple decided that it cannot be updated. This is a lesson for anyone who plans to spend the huge amount of money that a decked-out Mac Pro costs. At some point it will become obsolete by decree from Apple and there's nothing you can do. This is the second time in my life that I have found myself cursing Apple and swearing that I will never buy another of their computers; I should have listened to myself the first time.

I read it is possible to load Linux onto the old Mac Pro. Paula currently uses it as her main work machine. If she upgrades to a new box, I might try loading Linux onto the old one, just for fun and because its a beautiful machine. My big Alienware box was bought at the same time as the Mac Pro. Comparing the two boxes, inside and outside, is like comparing a Ferrari to a mid range Ford. Don't get me wrong, I love Fords after/while owning several of them but they aren't Ferraris.

Speaking of which, and really off topic, we had the opportunity to see the movie Ford vs Ferrari. It was a great movie, kind of made the Ford guys look like a bunch of corporate asses. I highly recommend seeing the movie.

For now, I have dropped support for the MacOS build of Shores of Hazeron. That could change in the future but its not likely unless someone requests it. Then it is only a matter of building and posting the software so don't be afraid to make that request.

Notably, all the latest versions of Linux were quite happy to upgrade my vintage Alienware computer.

Ubuntu 20
Competent and stable. This would be a fine OS for most purposes. It woke up to my three monitor configuration without any trouble.

Shores of Hazeron built and ran just fine. As usual with a new version of the development system, there was a wave of new compiler warnings that needed fixing.

I was pleased to notice when running SoH that most emojis in the text were now rendered in full color, not the text color, including the ones that appear in the on-screen chat text.

I didn't care for the new Gnome desktop. I lamented to Paula that I would like to see them roll back all the stuff they've done in recent years, back to when Linux felt cooler.

I had grown to like Manjaro 18 after using it in a virtual build machine for the last couple of years. I decided now was a good time to give it a try.

Manjaro 20 Xfce
Competent and stable. This would be a fine OS for most purposes. It woke up to my three monitor configuration without any trouble.

The Xfce desktop was more traditional and likeable.

Running Shores of Hazeron produced an error message about not being able to set a custom style. Then the style of many dialog boxes was borked, particularly the system supplied dialogs such as the file browser.

Worse, the OpenGL scene was transparent through the window to the desktop below, and whatever other windows were behind the SoH window, like having a peep hole through the SoH window. I concluded this could be a bug in my own window configuration so I made some attempts to eliminate the problem. Nothing worked. This error happened when Manjaro 20 was running as the main OS of the computer and it also happened when Manjaro was running in a virtual machine under Ubuntu 20. Both were running on the same physical computer/video hardware.

For this reason, support for Manjaro was dropped for now. I can/will build a version of SoH for Manjaro if anyone requests it. Perhaps your computer won't exhibit the problem.

Debian 10
Competent and stable. This would be a fine OS for most purposes. It woke up to my three monitor configuration without any trouble.

The installer allowed me to install Gnome, KDE Plasma and Xfce desktops. There were others but I only picked those three because they were familiar. Then you could pick your desktop when logging in, a nice feature.

Shores of Hazeron built and ran just fine but the frame rate was extremely low. This is because Debian does not install any non-free software, such as custom video drivers. I probably could have located and installed the necessary video driver but I didn't want to mess with it.

Xubuntu 20
I started downloading Xubuntu 20, which is Ubuntu with the Xfce desktop. While I was waiting, I started reading about Lubuntu, which is Ubuntu with the LXQt desktop. Wait what? I love Qt. LXQt is a desktop built using Qt?

I canceled the Xubuntu download. I'm sure it would have been a fine OS for most purposes.

Lubuntu 20
Competent and stable. This would be a fine OS for most purposes. It woke up to my three monitor configuration without any trouble.

Shores of Hazeron built and ran just fine. My only disappointment was that the emojis are back to being text-colored, instead of full color. At least we know somebody is working on it, maybe next version.

Somebody was listening to my plea. The LXQt desktop is clean and simple and everything seems to work. Not only that, a few of the cool fun things are back in the distribution, silly stuff like having zillions of pointless screen savers, and a terminal window background texture.

I tried using the QtTerminal window. It was very nice but I had to give it up because I've grown too used to using Alt+number to switch screens quickly. It was easy to install the good old Gnome terminal but I miss one of the fun features in the QtTerminal. Alas, you can't have it all.

Work on Shores of Hazeron feels like I'm back to my old pace. OS version upgrades are disruptive and a bit stressful until everything is back in its place and working as intended.

I suppose I could have simply upgraded the Xubuntu I was using but where's the fun in that? Sometimes you have to live it and do it to find a better way.

The final line up of officially supported platforms is...

• Debian 10
  
• Mint 19
  
• Ubuntu 20
  
• Windows 64 bits
  
• Windows 32 bits
  

Dear Haxus,

Would you be so kind as to upload the icons for the planet types to the wiki?  
We have WorldRingworld.png, WorldRing.png, WorldPlanet.png, WorldMoon.png, and WorldGasGiant.png.  
We're missing icons for Titan and Planetoid.

Thanks!

Navigator Fail
Ship still sits and wiggles when following a travel direct to solar system order.

Fixed.

Hold and Fuel Cell Repairs
Someone asked how to manually fix damage to the hold or the fuel cell.

Hull, hold and fuel cell are all considered to be parts of the same "Hull" system. Manual repairs to the hull system are applied to the hull portion until it is fixed, then repairs are applied to the fuel cell, and finally to the hold.

I made a slight change in this update. When you are standing in a room with access to the fuel coupling, manual repairs are applied to the fuel cell first. When you are standing in a room with cargo hold access, manual repairs are applied ot the hold first.

Launch and Recovery Systems
Launch and recovery system of a ship should be split up like other systems with subsystems, so damage to the system doesn't disable all launchers and landers at once.

Launch and recovery systems are now split into separate individual systems.

Launch and recovery systems now appear on the damage report window. No module is associated with them. Each is represented by a different spare part, for part removal and replacement.

The launcher uses a Mechanical Part. A torque wrench is used to remove/replace parts and to repair hit points on a vehicle launcher. Each launcher has only one removable part.

The recovery system uses an Electronic Part. A soldering iron is used to remove/replace parts and to repair hit points on a vehicle recovery system. Each recovery system has only one removable part.

Electronic parts and mechanical parts are now included in the category of Spare Parts.

Manual repairs to a launcher are performed at the launch pad, which is the large icon placed in the blueprint at the beginning of the launch tube. It is not visible in the final model. A visual cue must be designed into the blueprint if one is desired.

Manual repairs to landing systems are performed at space vehicle parking spots. It is impractical to use the landing system icon itself because it is usually invisible outside the space ship. However, the landing system requires space vehicle parking spots for recovered vehicles. Those parking spots are the points of access for repairs to the landing system(s).

When a ship has multiple landing systems, a specific landing system cannot be targeted for manual part removal or repair. Repairs are applied to each of them in turn and parts are removed from each of them in turn.

Launch and recovery systems now fail to operate when they are broken. Previously they never stopped working regardless of damage.

Launch and recovery systems now appear on the power relay station. They can be disabled there. They can also be fixed there, by routing power from a different system.

Delete Manufacturing Process
When a manufacturing process is deleted, it should return the parts to inventory.

When a manufacturing process is deleted, it now returns its parts and money to the inventory of the building. Spacecraft processes already did this but others did not.

Magmium Heat Sink Patent
Magmium heat sink patent should require the cryo heat sink patent as a prerequisite.

Added cryo heat sink patent as a prerequisite to developing the magmium heat sink patent. The magmium heat sink patent was already a prerequisite to developing the vulcium heat sink patent.

Jurisdiction Overlap
New capital and police jurisdiction allowed to overlap existing police jurisdiction.

Fixed jurisdiction overlap test.

Add Patent Process
When adding a patent process, init the values on the dialog to one workshop and limit the output to one.

This is the normal configuration when I set up patent developments and I got tired of always setting them. If this is a bother to anyone, I'll put it back the way it was.

Tool tips were updated to explain how patent processes are affected by the number of workshops and the output limit.

Add Process Confusing
Add production line dialog is confusing when no more can be added. Asks for number of shops and stuff but says I can't do it.

Unneeded controls are now hidden when a process cannot be added.

NPCs Staring at Ceiling
NPCs spend too much time looking at the ceiling or the floor.

NPCs now straighten their head tilt sometimes.

I feel like doors are a laborious task that could be programmatically simplified for the users without limiting creativity.

My (limited) use of the designer finds me always creating the same type of door - a simple room to room via a small space in the hull or an external to internal door.  I always create the jig, cut the door, create the voids of about the same size, create the navs, associating everything and then doing any customizing like creating custom switches, states, tweaking voids, etc.


One way that would simplify door creation is to create all the necessary pieces and associations by default and let designers modify the auto-generated pieces.  For example:

• Given a valid selected door jig
  
• And two rooms are selected selected
  
• And the option for default part creation is ticked
  
• When the jig is cut
  
• Then the door direction is inferred by locations of the two rooms selected with the jig
  
• And a default path object is placed from one side of the jig through to the other extending slightly beyond the jig's bounds in the inferred direction of the door
  
• And a new room void is created for the door that encompasses the jig shape
  
• And the room void and path parts are associated with the new door's proper state
  
• And optionally the user chooses the thickness of the door much like the end of an extrusion, but using stretching mechanics
  

• Given a valid selected door jig
  
• And a selected room and a hull
  
• And the option for default part creation is ticked
  
• When the jig is cut
  
• Then the door direction is inferred by locations of the room and the opposite end of the jig volume from the room
  
• And a default path object is placed from one side of the jig through to the other extending slightly beyond the jig's bounds in the inferred direction of the door
  
• And a new room void is created for the door that encompasses 80% of the jig shape from the room side of the jig
  
• And a new hull void is created for the door that encompasses 25% of the jig shape from the opposite side of the jig from the room
  
• And the room void, hull void, and path parts are associated with the new door's proper state
  
• And optionally the user chooses the thickness of the door much like the end of an extrusion, but using stretching mechanics
  

We really like the idea of having EV suit storage at Airports, but there is no setting limiting the storage of various airport commodities.   
For example, most would agree that Fuel should simply fill to max capacity, but that same number of units for space suits is too much.   Perhaps  the armory production limit could equal the suit storage limit at airports?     We just don't have a need for over 100,000 EV Suits and Helmets at each airport.

At a large building with ~30k shops, all are allocated, but only a few thousand citizens ever show up to work. The city has more than enough citizens to fill the positions, and the allocated shops are even counted towards the city employment. 

While it might be possible that the positions are being allocated very slowly, each one reliably reaches a few thousand, but that rate would be reduced to a crawl. I've never had a building see 10k workers in my months of play.

This is reproducible at all large buildings. See the 'Inferno' smelter design by Krusseldorf and use it as an example.

In my case, a smelter had ~7400 workers for metal. The building has an enormous ~49,000,000 storage capacity, but due to the exponential bonuses bug reported here, this process was producing an absurd ~55,000,000 metal every 2 seconds. This exceeded the storage and didn't output.

Workaround: reducing the workers to 4000 caused output to be 16,040,000 per run, and the inventory began to fill.




With Workaround:

Herald - TTFXCB

Game will work for about 5-10 mins and then will just lockup and the client becomes 'Non-Responsive' and closes. This happens with all avatars i use across multiple empires and sectors.

Hazeron is the only window open and processes are kept to a minimum. So not sure why client is failing to run

Do in game assets have anything to do as im currently running game with full particles and such ?

Any suggestions.

I've noticed that some ships get partially refueled, and others never even begin. Initially it seemed to have no pattern but I think I've figured it out: It tends to happen when you either a) have a ship whose fuel cells are larger than its hold and you use the 'percent of hold space' option to harvest hydrogen, or b) use the 'actual quantity' option and specify a quantity that is lower than the current amount of fuel in the fuel cells.

Let's say hypothetical case: you have a ship where the hold is 5,885,777 units and the fuel cells are 52,971,999 units. So, fuel cells significantly larger than hold's capacity.

In this hypothetical ship's case, when you queue a fuel harvest job a few things can happen based on the state of your ship's hold.

Case 1: You tell the crew to harvest '100% of hold space', but you already have more fuel than the hold space is capable of holding. In this case, you immediately receive a message from your officer: "All harvester operations are completed as ordered. Super job crew." Nothing happens, the job does not even begin. 

Case 2: You tell the crew to harvest 'actual quantity'. If you enter a quantity lower than the current amount of fuel in the cells--say, 40,000,000 as the quantity when there are 41,736504 units already in the cells--you get an instant:"All harvester operations are completed as ordered. Super job crew." Same as before.

If, however, set the actual quantity to a higher value that the current amount of fuel in the cells, like say 53,000,00 units, then the crew will successfully start/finish the mission.
This is the ONLY way I can get my current ship model to refuel successfully every time.

For newer players the seemingly random/unpredictable behavior may lead to some frustration.

The tl;dr problems: The harvest hydrogen by '% of hold space' appears to be using the hold space in its calculation when it should be using fuel cell capacity, since hydrogen doesn't get stored in the hold. Additionally, the 'actual quantity' option appears to have a hidden conditional check which looks at the current amount in the fuel cells before deciding whether to harvest which can lead to even more confusion.

Possible fix:

a) Add special case for harvesting hydrogen by % of hold space to check the fuel cells instead of the actual hold when performing the calculation.

b) Change 'actual quantity' to not look at the current hold space before starting. Instead, have it blindly attempt to harvest <actual quantity> of resource, stop when hold/fuel cell (in case of hydrogen) is full or when quantity has been reached.

I think seeing the amount of people who want a simpler designer again might help move things along. Post your name here as a pseudo-petition if you stopped playing entirely or almost entirely because of how janky or complicated or pain-in-the-ass-y the new features are, like the new designer, the new city system, ect.

I'll start!

Minty.