Thanks for adding the doc strings, it is appreciated. The variable renaming makes more sense too.

In T21#2315, @wxl wrote:

@kc2bez what do you think about using one of the stock ones versus making something ourselves?

In T21#2311, @wxl wrote:

Maybe we should just pick one as the default rather than using random?

• cubicgrid
• drift, which was written by the author of fractal wizard Scott Draves, who also did Electric Sheep
• flurry
• goop
• hypnowheel

Thanks for picking up on this. LGTM

I put in a bug upstream.

I had a good test on my Acer Laptop. Laptop is somewhat low on resources. 1st gen i5 with 4G of RAM that is shared with the onboard Intel graphics. As far as resource usage it was almost identical with the original compton. 457 M for the original and 460 M for the new one. I had around 430 M with no compton installed. I used htop for all three measurements. I notice a bit of stuttering with both versions when moving windows around. I also tested the compton-conf that is shipped in Lubuntu and that seems to manage the config file fine. I see no downside currently.

• Remove shebang and executable flag from .desktop.
• Remove executable flag from non-executables.
• Adjust dpkg-architecture-variable.

• Fix lintian errors and warnings.
• Add lintian override.
• More lintian overrides.
• Add keywords to .desktop file.

It does it on a live iso, I wouldn't think there was any cache.
I also did try this on a different .desktop file and I tried a logout and login as well as a reboot. They all still show.

In T108#2111, @wxl wrote:

No, that's not it. $XDG_CURRENT_DESKTOP is LXQt as it should be.

@wxl I am confused here. I added NotShowIn=LXQt; to the desktop file yet it still shows in the menu. I re-read the spec and still don't see where I went wrong. Any ideas?

We should bounce this to 20.04 IMHO. There are some issues still upstream, this one impacts us. Then there is this, this and this.
To top it off, I can't see where you can remove packages, only add them, we'd have to tinker on that too. As much as I would like to see it, I don't think it is ready for us at this time.

In T99#2025, @wxl wrote:

Almost there!

In T58#2017, @wxl wrote:

Since Dan has what I think is the most sense of this, I'm giving this to him.

Bump. @The_LoudSpeaker how is this coming along?

In T56#1982, @guiverc wrote:

https://paste.ubuntu.com/p/Cvvx77Gv5r/
has the ~/.cache/Calamares/session.log file.
I hadn't exited the installer at that stage, it was grabbed when it was at the

• install alongside
• replace partition
• manual partitioning

question, ie. no erase disk & install option provided. Single 80gb disk; partitions for /, swap & /home only

• Fix changelog entry

In D29#605, @tsimonq2 wrote:

When you bump the debhelper compat, you also have to bump the dependency in debian/control. Please also say that no changes were needed in the changelog (obviously if there were, you would describe them).
Nice work!

In T58#1941, @wxl wrote:

So upstream has now implemented the packagechooser module which I think should fix this. @kc2bez would you be interested in both pulling down 3.2.12 and exploring this possibility? To be more clear there's two goals:

1. provide a way to do a minimal install, i.e. what lubuntu-core is rather than lubuntu-desktop
2. at least start providing some options to choose between different browsers, etc.

For this release I will change the patch so that NotShowIn=LXQt; is added to the stock .desktop that ships with Calamares leaving only our workaround visible in the menu and desktop. We can revisit it when there is an upstream change.

I forgot to mention, paste your buildlog so we can look at the errors.

In T99#1857, @The_LoudSpeaker wrote:

The patch doesn't apply. There are waaay too many changes since last release. around 60 something commits.
Should I ask for a new release upstream? Else will have to create a patch for each commit and apply and that will be too many patches.

I think some of the key to success here may be in having swap (see T22). I setup a VM with a 2Gib swap space and 478Mib. I left it running overnight and RAM utilization was 113 Mib. As a side note the web was browse-able, perhaps not optimal by any stretch but functional. Swap isn't a cure-all and may not be the ideal solution for everything but it can sometimes help. I know I just threw some more variables in the mix but I think we need to look at that too.

Should we wait to agaida to put it in sid?

Here is some interesting information.

In T26#1599, @wxl wrote:

I wonder if its even possible to be device-agnostic about this.

In T91#1605, @wxl wrote:

Oh, wait, that was a commit and not a differential. Ok, that's cool. 😄 Now assumedly you need me to sponsor it, eh? 😆

Transferring @wxl 's comment on the commit here into the task with my replies in bold:

In T21#1589, @wxl wrote:

So I did some tweaking. Installed xfonts-efont-unicode to get it to look this good. This is old style xlib font rendering so you have to use X11 fonts. xfontsel is your friend here. Not sure why the "new login" button is showing weird.

The mimeapps.list has been updated and is now in eoan. It would be great if we could get some tests. @guiverc @wxl

• Add changelog entry

@wxl I have tested by making a startup disk in every release from Xenial to Eoan. They all work very well for me, no issues.

I pulled down P10 and replaced my openbox lxqt-rc.xml
I then pulled down P9 and replaced my globalkeyshortcuts.conf

I have tested this in a VM and on bare metal. The bare metal machine had 2 displays and the full-screen window appeared only on the main display (expected behavior). The VM resolution was set to 800X600 and everything appeared on the screen and looked normal.

That looks very nice!

@wxl @tsimonq2 Is fullscreen the way we want to go here or do we want some other set size?
I like fullscreen for 2 reasons.

For the easy win we could set windowExpanding: fullscreen in the branding.desc file to force calamares into full screen mode.

I think we have to seed some of the extra packages.

Ammend the changelog entry to properly reference LP.

In T2#501, @tsimonq2 wrote:

As linked in T36: Fix full disk encryption on BIOS systems, we have functioning FDE on BIOS again.
Could someone test the package from -proposed on an EFI system and see if there's still the same results as before?

Ammend quilt header.

It looks like we already have ufw installed. I checked the live iso and did a fresh install, it was present on both.

In T36#354, @apt-ghetto wrote:

How is it handled by ubiquity? They should have the same problem.

My research has led me to this issue reported in Calamares. Basically there was a change to cryptsetup to default to LUKS2. LUKS2 and GRUB2 don't play well together. Which thing do we patch or do we abandon LUKS for now? @tsimonq2 @wxl what are your thoughts on this? Do you think I am on the right track here so far?

In T22#334, @tsimonq2 wrote:

In T22#335, @kc2bez wrote:

In T22#334, @tsimonq2 wrote:

I thought there was also the option of a swapfile, but it was unsupported. If we do swap, that's how we do it.

There is but I didn't test that, it said it was unsupported :P I will give it a try though.

This is going to take some more code spelunking. Adding - file # To swap file instead of partition (unsupported right now) didn't get presented in the dropdown.

In T22#336, @tsimonq2 wrote:

In T22#334, @tsimonq2 wrote:

I thought there was also the option of a swapfile, but it was unsupported. If we do swap, that's how we do it.

This does look super easy to implement now, there are even options for the size of swap that a user wants to implement. In the partition.conf module we would need to change neverCreateSwap: true to neverCreateSwap: false and then add:

userSwapChoices:
    - none      # Create no swap, use no swap
    - reuse     # Re-use existing swap, but don't create any
    - small     # Up to 4GB
    - suspend # At least main memory size

I tested this in Vbox from a daily iso and it seems to work as advertised with the exception of reuse, I wasn't able to see that as an option but perhaps I was doing something wrong. Any insight on this @tsimonq2?

Thanks to @wxl here is a link to the bug