Read noise Astrophotography avatar

🛰️ Windows 11 Astro Imaging PC

Full Build Guide

From Bare Metal to Stable Observatory Appliance

This guide describes a complete rebuild from scratch of a Windows 11 PC used exclusively for astrophotography.

The goal is not convenience.

The goal is stability, predictability, and control.

If Windows Update or background services ruin even one clear night, this guide is for you.

This is NOT for daily-use PCs.

CORE PRINCIPLES

  • Windows is treated as an appliance

  • Updates are fully disabled

  • Drivers are installed once

  • Internet access is controlled

  • LAN/RDP monitoring is preserved

  • Imaging happens in a frozen state

PHASE 0 — REQUIREMENTS

  • Windows 11 Pro (Home is not sufficient)

  • USB installer created with Microsoft Media Creation Tool

  • Keyboard + screen attached

  • Internet available later, not during install

PHASE 1 — CLEAN WINDOWS INSTALL (CRITICAL)

1️⃣ Boot from USB installer

At the disk selection screen:

  • DELETE ALL PARTITIONS on the system disk

  • Select Unallocated Space

  • Click Next

  • Let Windows create partitions automatically

This removes:

  • Old drivers

  • Broken update state

  • Ghost USB enumerations

  • Corrupt recovery partitions

2️⃣ OOBE SETUP

DO NOT CONNECT TO WIFI

This step is critical and often missed.

When you reach:

“Let’s connect you to a network”

DO THIS:

Press:

Shift + F10

Command prompt opens.

Type:

OOBE\BYPASSNRO

Press Enter → system reboots.

After reboot:

  • Select I don’t have internet

  • Select Continue with limited setup

  • Create a local account

  • Do NOT sign in with Microsoft

  • Do NOT connect Wi-Fi

  • Do NOT plug Ethernet

This prevents:

  • Automatic driver downloads

  • Account sync

  • Early Windows Update corruption

PHASE 2 —BASE WINDOWS HARDENING (RUN ONCE)

Open PowerShell as Administrator.

Save and run this script once:

BASE_WINDOWS_HARDEN.ps1

# ================================

# BASE WINDOWS HARDENING

# ================================

# Kill Windows Update services

$services = @(

"wuauserv","bits","usosvc","dosvc",

"InstallService","WaaSMedicSvc"

)

foreach ($s in $services) {

Stop-Service $s -Force -ErrorAction SilentlyContinue

sc.exe config $s start= disabled

}

# Block updates via policy

reg add "HKLM\SOFTWARE\Policies\Microsoft\Windows\WindowsUpdate" /v DisableWindowsUpdateAccess /t REG_DWORD /d 1 /f

reg add "HKLM\SOFTWARE\Policies\Microsoft\Windows\WindowsUpdate\AU" /v NoAutoUpdate /t REG_DWORD /d 1 /f

reg add "HKLM\SOFTWARE\Policies\Microsoft\Windows\WindowsUpdate" /v ExcludeWUDriversInQualityUpdate /t REG_DWORD /d 1 /f

# Disable update scheduler resurrection

$tasks = @(

"\Microsoft\Windows\WindowsUpdate\Scheduled Start",

"\Microsoft\Windows\UpdateOrchestrator\Schedule Scan",

"\Microsoft\Windows\UpdateOrchestrator\USO_UxBroker",

"\Microsoft\Windows\UpdateOrchestrator\Reboot"

)

foreach ($t in $tasks) {

schtasks /Change /TN $t /Disable 2>$null

}

# Disable telemetry

reg add "HKLM\SOFTWARE\Policies\Microsoft\Windows\DataCollection" /v AllowTelemetry /t REG_DWORD /d 0 /f

sc.exe stop DiagTrack

sc.exe config DiagTrack start= disabled

# Power & sleep

powercfg /hibernate off

powercfg /setactive SCHEME_MIN

powercfg /change standby-timeout-ac 0

powercfg /change monitor-timeout-ac 0

# USB & PCIe stability

reg add "HKLM\SYSTEM\CurrentControlSet\Services\USB" /v DisableSelectiveSuspend /t REG_DWORD /d 1 /f

powercfg -setacvalueindex SCHEME_MIN SUB_PCIEXPRESS ASPM 0

# Defender noise reduction

Set-MpPreference -MAPSReporting Disabled

Set-MpPreference -SubmitSamplesConsent 2

Write-Host "Base hardening complete"

➡ Reboot immediately after running

PHASE 3 — REMOVE CONSUMER BLOAT

Remove:

  • Xbox services

  • Clipchamp

  • Teams (consumer)

  • Cortana

  • Widgets

  • Phone Link

  • Feedback Hub

Optional PowerShell sweep:

Get-AppxPackage | Where-Object {

$_.Name -match "Xbox|Gaming|Clipchamp|Teams|Cortana|Feedback|Widget|Phone"

} | Remove-AppxPackage

Reboot once.

PHASE 4

DRIVER INSTALL (CONTROLLED INTERNET WINDOW)

Now you may briefly connect to the internet.

Install order:

  1. Chipset driver

  2. GPU driver (Minimal install only)

  3. Reboot

Do not enable auto-updates in any vendor software.

PHASE 5

ASTRO SOFTWARE STACK

Install in this order:

  1. ASCOM Platform (will enable .NET 3.5 automatically)

  2. ASTAP + star database

  3. Mount control software

  4. Powerbox / hub software

  5. Camera / guider / focuser drivers

  6. SharpCap

  7. NINA

  8. PHD2

Reboot when prompted.

PHASE 6

CONFIGURE SOFTWARE (NO HARDWARE)

Before plugging in anything:

  • Configure NINA

  • Configure plate solving

  • Configure PHD2 profiles

  • Launch each app once

This isolates software problems early.

PHASE 7

LAN-ONLY IMAGING MODE (IMPORTANT)

We allow LAN + RDP, but block internet traffic.

Imaging Mode Script

# LAN-only imaging mode

Set-MpPreference -DisableRealtimeMonitoring $true

netsh advfirewall firewall add rule name="BLOCK_ALL_OUT" dir=out action=block profile=any

netsh advfirewall firewall add rule name="ALLOW_LAN" dir=out action=allow remoteip=LocalSubnet profile=any

netsh advfirewall firewall add rule name="ALLOW_RDP" dir=in action=allow protocol=TCP localport=3389 profile=any

Maintenance Mode Script

# Restore internet access

netsh advfirewall firewall delete rule name="BLOCK_ALL_OUT"

netsh advfirewall firewall delete rule name="ALLOW_LAN"

netsh advfirewall firewall delete rule name="ALLOW_RDP"

Set-MpPreference -DisableRealtimeMonitoring $false

PHASE 8

HARDWARE INTEGRATION

Rule:

One device → reboot → test

Order:

  1. Mount

  2. Main camera

  3. Guide camera

  4. Filter wheel

  5. Focuser

  6. Rotator

  7. Powerbox last

PHASE 9

SYSTEM IMAGE (DO NOT SKIP)

Before observatory deployment:

  • Create a full system image to external storage

  • Use Macrium Reflect or similar

This turns disaster into a 30-minute restore.

FINAL RESULT

You now have:

  • A stable Windows imaging appliance

  • LAN/RDP monitoring intact

  • No forced updates

  • No driver churn

  • Predictable nights

Clear skies should be your only variable not Windows.

Well Written Helpful Engaging
andrea tasselli avatar
Wonder how I managed all these years without any of the above.
Read noise Astrophotography avatar

andrea tasselli · Dec 18, 2025 at 06:15 PM

Wonder how I managed all these years without any of the above.

Andrea, hello again 🙂

No one’s saying these steps are mandatory plenty of people image successfully with stock setups.

The post is just about reducing background OS noise and edge-case instability, especially for unattended or long automated runs, like my fully automated observatory. If someone’s system is already stable, that’s great 👍

Clear skies!

Respectful Supportive
Noah Tingey avatar

Have you heard of https://schneegans.de/windows/unattend-generator/?

It lets you generate an autounattend.xml file that can automatically, heavily debloat your windows install. No need for opening command prompt in the OOBE yourself.

Quinn Groessl avatar

I too can ask AI how to set up a PC for Astrophotography.

Read noise Astrophotography avatar

Quinn Groessl · Dec 18, 2025 at 08:03 PM

I too can ask AI how to set up a PC for Astrophotography.

Nothing here requires AI it’s standard Windows housekeeping many of us have done manually for years.

If someone’s setup is stable already, great. This is just about improving reliability for automated systems.

Read noise Astrophotography avatar

Noah Tingey · Dec 18, 2025 at 06:45 PM

Have you heard of https://schneegans.de/windows/unattend-generator/?

It lets you generate an autounattend.xml file that can automatically, heavily debloat your windows install. No need for opening command prompt in the OOBE yourself.

Hi thanks for the link I’ll definitely check it out

Basically win update done something to my mini pc .. eventually it bricked . I spent 15 hours trying to rescue it. Had to fresh install. Lost all my data .. all my subs are gone. So anything that helps avoid that in the future. I tried win util but ended up in another reinstall.

SonnyE avatar

In a word, “Linux”.

A whole fresh set of problems to Concore!

I am considering when my current mount laptop dies to try and get a Linux based laptop. I dread the learning curve. (And the teddy bears, and cartoon crap)

Don’t you miss the old days when you could simply departion, repartion, and install your programs fresh to a virgin hard drive?

Windows has become a monster. A spoiled brat monster.

Read noise Astrophotography avatar

SonnyE · Dec 19, 2025 at 06:20 PM

In a word, “Linux”.

A whole fresh set of problems to Concore!

I am considering when my current mount laptop dies to try and get a Linux based laptop. I dread the learning curve. (And the teddy bears, and cartoon crap)

Don’t you miss the old days when you could simply departion, repartion, and install your programs fresh to a virgin hard drive?

Windows has become a monster. A spoiled brat monster.

SonnyE · Dec 19, 2025 at 06:20 PM

In a word, “Linux”.

A whole fresh set of problems to Concore!

I am considering when my current mount laptop dies to try and get a Linux based laptop. I dread the learning curve. (And the teddy bears, and cartoon crap)

Don’t you miss the old days when you could simply departion, repartion, and install your programs fresh to a virgin hard drive?

Windows has become a monster. A spoiled brat monster.

I’m in the same boat. I’ve seriously considered Linux as well, but I’m still hesitant to make the jump.

Windows today feels intrusive in a way it never used to constant updates, background services, and general bloat. It really does feel like you no longer fully own the machine. For me, Windows 7 was the sweet spot: simple, predictable, and it stayed out of the way. The era when you could clean-install, overclock, tune, and just get on with things.

What’s stopping me from switching right now is software maturity. There’s no real replacement for NINA on Linux. I’ve looked at the alternatives and they feel fragmented or underpowered by comparison. I’ve only just become comfortable with NINA (haven’t even gone deep into the Advanced Sequencer yet), plus the Touch ‘n’ Stars phone app integration works extremely well.

If NINA ever gets a solid Linux port, I’d seriously reconsider. Until then, Windows remains a necessary evil for imaging even if it’s become a spoiled brat of an OS.

Clear Skies !!

Tobiasz avatar

Read noise Astrophotography · Dec 20, 2025, 08:43 AM

What’s stopping me from switching right now is software maturity. There’s no real replacement for NINA on Linux. I’ve looked at the alternatives and they feel fragmented or underpowered by comparison. I’ve only just become comfortable with NINA (haven’t even gone deep into the Advanced Sequencer yet), plus the Touch ‘n’ Stars phone app integration works extremely well.

How can you state there are no linux counterparts to NINA if you only “looked” at the alternatives and checked how they “feel”? What did you use on the linux side and for how long?

Read noise Astrophotography · Dec 20, 2025, 08:43 AM

If NINA ever gets a solid Linux port, I’d seriously reconsider.

Linux does not need a NINA port, because there are strong alternatives already. It may be useful because competition drives innovation, but NINA is defintely not a necessity to do fully automated astrophotography in 2025.

Read noise Astrophotography avatar

Tobiasz · Dec 20, 2025, 08:58 AM

Read noise Astrophotography · Dec 20, 2025, 08:43 AM

What’s stopping me from switching right now is software maturity. There’s no real replacement for NINA on Linux. I’ve looked at the alternatives and they feel fragmented or underpowered by comparison. I’ve only just become comfortable with NINA (haven’t even gone deep into the Advanced Sequencer yet), plus the Touch ‘n’ Stars phone app integration works extremely well.

How can you state there are no linux counterparts to NINA if you only “looked” at the alternatives and checked how they “feel”? What did you use on the linux side and for how long?

Read noise Astrophotography · Dec 20, 2025, 08:43 AM

If NINA ever gets a solid Linux port, I’d seriously reconsider.

Linux does not need a NINA port, because there are strong alternatives already. It may be useful because competition drives innovation, but NINA is defintely not a necessity to do fully automated astrophotography in 2025.

I do realise there are other alternatives and I come from a Maxim DL background.

That’s fair, and I should clarify what I meant.

For context, I come from a MaxIm DL background (2012–2016) and in my view it’s still the benchmark for a mature, integrated imaging environment. My old observatory computer died back then, and although I’ve tried contacting Diffraction Limited to see if my licence could be reissued, I never heard back. I’m not particularly keen to pay for it again when my current setup is already stable and productive.

So this isn’t OS tribalism or lack of experience. I’m very familiar with what a settled, production-grade imaging workflow feels like.

I’m not claiming Linux can’t do automated astrophotography. I’m saying that for me, right now, Windows + NINA is where my workflow has reached that same “configure once, trust it, and collect data” stage. I’ve only recently got everything dialled in, and I don’t want to reset that learning curve while it’s delivering results.

I’ve looked at the Linux alternatives and I respect the ecosystem, but I personally prefer fewer moving parts and less supervision at this stage. That’s a workflow preference, not a judgement of capability.

One small but meaningful detail for me is the Touch ’n’ Stars phone app being able to sit next to my wife and casually browse targets and session status makes the whole experience feel less isolated and more enjoyable.

If a Linux solution reaches the same level of comfort and trust for my setup, I’d happily reassess. And if NINA ever gets a solid Linux port, I’d absolutely take a serious look. Until then, I’m sticking with what’s working.

Helpful Insightful Respectful Engaging
Bill McLaughlin avatar

Read noise Astrophotography · Dec 20, 2025, 08:43 AM

I’m in the same boat. I’ve seriously considered Linux as well, but I’m still hesitant to make the jump.

Agree! I have not made the jump yet but Windows seems intent to piss off it’s most knowledgeable customers. I suspect that my next non-astro home office PC will be Linux and once I am used to it in a less mission critical environment I might move my astro PCs to it as well.

SonnyE avatar

Read noise Astrophotography · Dec 20, 2025, 10:25 AM

Tobiasz · Dec 20, 2025, 08:58 AM

Read noise Astrophotography · Dec 20, 2025, 08:43 AM

What’s stopping me from switching right now is software maturity. There’s no real replacement for NINA on Linux. I’ve looked at the alternatives and they feel fragmented or underpowered by comparison. I’ve only just become comfortable with NINA (haven’t even gone deep into the Advanced Sequencer yet), plus the Touch ‘n’ Stars phone app integration works extremely well.

How can you state there are no linux counterparts to NINA if you only “looked” at the alternatives and checked how they “feel”? What did you use on the linux side and for how long?

Read noise Astrophotography · Dec 20, 2025, 08:43 AM

If NINA ever gets a solid Linux port, I’d seriously reconsider.

Linux does not need a NINA port, because there are strong alternatives already. It may be useful because competition drives innovation, but NINA is defintely not a necessity to do fully automated astrophotography in 2025.

I do realise there are other alternatives and I come from a Maxim DL background.

That’s fair, and I should clarify what I meant.

For context, I come from a MaxIm DL background (2012–2016) and in my view it’s still the benchmark for a mature, integrated imaging environment. My old observatory computer died back then, and although I’ve tried contacting Diffraction Limited to see if my licence could be reissued, I never heard back. I’m not particularly keen to pay for it again when my current setup is already stable and productive.

So this isn’t OS tribalism or lack of experience. I’m very familiar with what a settled, production-grade imaging workflow feels like.

I’m not claiming Linux can’t do automated astrophotography. I’m saying that for me, right now, Windows + NINA is where my workflow has reached that same “configure once, trust it, and collect data” stage. I’ve only recently got everything dialled in, and I don’t want to reset that learning curve while it’s delivering results.

I’ve looked at the Linux alternatives, and I respect the ecosystem, but I personally prefer fewer moving parts and less supervision at this stage. That’s a workflow preference, not a judgement of capability.

One small but meaningful detail for me is the Touch ’n’ Stars phone app being able to sit next to my wife and casually browse targets and session status makes the whole experience feel less isolated and more enjoyable.

If a Linux solution reaches the same level of comfort and trust for my setup, I’d happily reassess. And if NINA ever gets a solid Linux port, I’d absolutely take a serious look. Until then, I’m sticking with what’s working.

This was my next question, IF Linux could run NINA. I was casually looking at it and it appears to present like Winderz. So, my next question would have been can NINA be installed on the platform.

I guess not. “Linux does not need a NINA port, because there are strong alternatives already.” Well, thank you for such a broad-brush statement with NO explanation. Come on, inquiring minds want to know.

I know there are Linux users doing AP with it. But how? What are the Quirks and Qraps to Linux? Linux lost me a long time ago with the Teddy Bear and Nursery type Icons. I was a long time hold out on anything but windows classic screen because I don’t like caricature icons, for one example. But I’m so sick of Microsoft’s “Security Updates” that only makes things secure for them to shove their crap into our working computers and break things they do not offer.

Way back when I began researching how to do Astrophotography, I actually did look at what platform would be best. I would have switched to Apple IF it would have been better. But I discovered Windows was what most/all of the programs we typically use were written in. I suppose that may go back to when fit files developed which to this day dominates all AP files. But which are not compatible with our modern Internet because the data is too huge for everybody to post. Astrobin is the only place I’ve found where fit files can be posted.

I suppose Linux just does not, or has not, attracted the code writers to adapt or adopt NINA to its platform yet. Which gives room for hope. But other platforms, like Apple, try and adapt programs written for Microsoft, and they generally break the code. Such as what happened to ASTAP and wasn’t corrected until Han Klien himself fixed ASTAP. I still have my download from his repair, and it is the only ASTAP I use if repairing (reinstalling) ASTAP damaged by a Windows “security” update. (There have been several)

Right now, my best defense is to fight off Microsoft and to maintain my folder with programs that actually do work for reinstalling when a need arises. I was not aware Linux did not readily work with NINA (Or the other way around).

SonnyE avatar

Tobiasz · Dec 20, 2025, 08:58 AM

Read noise Astrophotography · Dec 20, 2025, 08:43 AM

If NINA ever gets a solid Linux port, I’d seriously reconsider.

Linux does not need a NINA port, because there are strong alternatives already. It may be useful because competition drives innovation, but NINA is defintely not a necessity to do fully automated astrophotography in 2025.

I guess not. “Linux does not need a NINA port, because there are strong alternatives already.” Well, thank you for such a broad-brush statement with NO explanation. Come on, inquiring minds want to know.

andrea tasselli avatar
I can't for the life of me see how any update to Windoze can "break" one piece of code to one user but not to many others. Besides that writing "breaking" is kid-wording as there is no such a thing in real world.
Read noise Astrophotography avatar

andrea tasselli · Dec 20, 2025, 04:43 PM

I can't for the life of me see how any update to Windoze can "break" one piece of code to one user but not to many others. Besides that writing "breaking" is kid-wording as there is no such a thing in real world.

The idea that an OS update can’t affect one user and not others assumes software exists in isolation from hardware, firmware, drivers, timing, and state which simply isn’t how real systems work.

Windows updates routinely modify:

  • kernel scheduling behavior

  • USB stack timing and power policy

  • driver load order and enumeration

  • device class filtering

  • background services and IPC timing

In hardware-attached workflows (cameras, mounts, hubs, rotators), those changes don’t “break code” they alter execution context. That’s enough to surface race conditions, power-state bugs, or enumeration conflicts that were previously latent.

This is why Microsoft documents update-induced regressions as environmental issues, not “broken code”. It’s also why identical setups on paper can diverge in practice.

This isn’t theoretical it’s standard behaviour in systems that depend on real-time I/O, USB topology, and vendor drivers.

Windows updates don’t need to “break code”.
They change drivers, power policy, and device enumeration.
That’s sufficient to affect hardware-dependent systems on some machines and not others a well-known class of environmental regression.

Helpful Insightful
Read noise Astrophotography avatar

SonnyE · Dec 20, 2025, 04:34 PM

This was my next question, IF Linux could run NINA. I was casually looking at it and it appears to present like Winderz. So, my next question would have been can NINA be installed on the platform.

I guess not. “Linux does not need a NINA port, because there are strong alternatives already.” Well, thank you for such a broad-brush statement with NO explanation. Come on, inquiring minds want to know.

I know there are Linux users doing AP with it. But how? What are the Quirks and Qraps to Linux? Linux lost me a long time ago with the Teddy Bear and Nursery type Icons. I was a long time hold out on anything but windows classic screen because I don’t like caricature icons, for one example. But I’m so sick of Microsoft’s “Security Updates” that only makes things secure for them to shove their crap into our working computers and break things they do not offer.

Way back when I began researching how to do Astrophotography, I actually did look at what platform would be best. I would have switched to Apple IF it would have been better. But I discovered Windows was what most/all of the programs we typically use were written in. I suppose that may go back to when fit files developed which to this day dominates all AP files. But which are not compatible with our modern Internet because the data is too huge for everybody to post. Astrobin is the only place I’ve found where fit files can be posted.

I suppose Linux just does not, or has not, attracted the code writers to adapt or adopt NINA to its platform yet. Which gives room for hope. But other platforms, like Apple, try and adapt programs written for Microsoft, and they generally break the code. Such as what happened to ASTAP and wasn’t corrected until Han Klien himself fixed ASTAP. I still have my download from his repair, and it is the only ASTAP I use if repairing (reinstalling) ASTAP damaged by a Windows “security” update. (There have been several)

Right now, my best defense is to fight off Microsoft and to maintain my folder with programs that actually do work for reinstalling when a need arises. I was not aware Linux did not readily work with NINA (Or the other way around).

I think we’re actually talking past each other a bit, so let me reset the frame.

My original point wasn’t “Linux vs Windows”, and it wasn’t “NINA is superior to everything else”.

It was about where my current workflow has reached operational maturity.

Yes, there are strong Linux-based alternatives (KStars/Ekos, INDI, etc.). I’m aware of them, I’ve looked at them, and I respect the people running observatories successfully on that stack. But those systems come with different trade-offs: distributed services, INDI servers, more moving parts, and a different supervision model. Some people enjoy that flexibility; others prefer tighter integration.

For me, at this point, Windows + NINA has reached the same “configure once, trust it, collect data” stage that MaxIm DL had years ago. That’s the entire decision. It’s not about ideology, aesthetics, or lack of awareness of alternatives.

Regarding NINA on Linux: there is no native port today, largely because of deep dependencies on Windows-specific frameworks and ASCOM. Running it under emulation or compatibility layers tends to reintroduce exactly the instability I’m trying to avoid, so it defeats the purpose.

If, in the future, a Linux-based solution offers the same level of comfort, predictability, and integration for my specific hardware and habits, I’d happily reassess. Likewise, if NINA ever gains a solid Linux port, I’d be very interested.

Until then, I’m not “defending Windows” I’m sticking with the platform that is currently delivering reliable data for me with minimal supervision. That’s a workflow choice, not a judgement of other ecosystems.

Well Written Insightful Respectful Engaging Supportive
Read noise Astrophotography avatar

The simple answer Simple answer I don’t like the UX/UI. I am not that keen on NINA either but I am not paying for Maxim DL again end of.

And for clarity!

There’s a bit of misquoting happening here, so let me reset this cleanly.

I did not say Linux has no counterparts to NINA, and I did not say Linux can’t do automated astrophotography. What I said was that for my current workflow, there’s no direct replacement for NINA which is a subjective, workflow-specific statement, not a claim about absolute capability.

I’ve evaluated KStars/Ekos/INDI at a high level and followed their development for years. They are powerful systems, but they are architecturally different: multiple services, distributed components, and a different supervision model. Many people run them successfully that does not automatically make them a drop-in replacement for NINA in every workflow, or something I personally prefer right now.

When I say “no real replacement for me”, I’m referring to:

  • level of integration

  • sequencing depth and flexibility

  • stability with my specific hardware

  • how much ongoing supervision is required once configured

That’s not a dismissal of Linux. It’s simply acknowledging that software maturity is contextual it depends on the user, the hardware, and the operational goals.

If someone’s Linux setup is settled and producing reliable data, there’s no reason to switch. Likewise, now that my Windows + NINA workflow has reached a “configure once, trust it, collect data” stage, I’m not interested in resetting that learning curve without a clear net gain.

Finally, this thread was never meant to be a platform comparison debate. I was responding to a specific stability issue and trying to spare someone else from the same problems I ran into. If people want to debate platforms in general, that’s fine but it belongs in a separate thread.

Different paths, same goal

Preference isn’t ignorance. It’s just preference.

bigCatAstro avatar

I do miss the virtual internal guider feature in EKOS, it was pretty effective.

Well Written
Read noise Astrophotography avatar

bigCatAstro · Dec 20, 2025 at 06:07 PM

I do miss the virtual internal guider feature in EKOS, it was pretty effective.

I miss MaxIm DL for my setup it guided really well no hassles at all .

SonnyE avatar

Read noise Astrophotography · Dec 20, 2025, 05:46 PM

SonnyE · Dec 20, 2025, 04:34 PM

This was my next question, IF Linux could run NINA. I was casually looking at it and it appears to present like Winderz. So, my next question would have been can NINA be installed on the platform.

I guess not. “Linux does not need a NINA port, because there are strong alternatives already.” Well, thank you for such a broad-brush statement with NO explanation. Come on, inquiring minds want to know.

I know there are Linux users doing AP with it. But how? What are the Quirks and Qraps to Linux? Linux lost me a long time ago with the Teddy Bear and Nursery type Icons. I was a long time hold out on anything but windows classic screen because I don’t like caricature icons, for one example. But I’m so sick of Microsoft’s “Security Updates” that only makes things secure for them to shove their crap into our working computers and break things they do not offer.

Way back when I began researching how to do Astrophotography, I actually did look at what platform would be best. I would have switched to Apple IF it would have been better. But I discovered Windows was what most/all of the programs we typically use were written in. I suppose that may go back to when fit files developed which to this day dominates all AP files. But which are not compatible with our modern Internet because the data is too huge for everybody to post. Astrobin is the only place I’ve found where fit files can be posted.

I suppose Linux just does not, or has not, attracted the code writers to adapt or adopt NINA to its platform yet. Which gives room for hope. But other platforms, like Apple, try and adapt programs written for Microsoft, and they generally break the code. Such as what happened to ASTAP and wasn’t corrected until Han Klien himself fixed ASTAP. I still have my download from his repair, and it is the only ASTAP I use if repairing (reinstalling) ASTAP damaged by a Windows “security” update. (There have been several)

Right now, my best defense is to fight off Microsoft and to maintain my folder with programs that actually do work for reinstalling when a need arises. I was not aware Linux did not readily work with NINA (Or the other way around).

I think we’re actually talking past each other a bit, so let me reset the frame.

My original point wasn’t “Linux vs Windows”, and it wasn’t “NINA is superior to everything else”.

It was about where my current workflow has reached operational maturity.

Yes, there are strong Linux-based alternatives (KStars/Ekos, INDI, etc.). I’m aware of them, I’ve looked at them, and I respect the people running observatories successfully on that stack. But those systems come with different trade-offs: distributed services, INDI servers, more moving parts, and a different supervision model. Some people enjoy that flexibility; others prefer tighter integration.

For me, at this point, Windows + NINA has reached the same “configure once, trust it, collect data” stage that MaxIm DL had years ago. That’s the entire decision. It’s not about ideology, aesthetics, or lack of awareness of alternatives.

Regarding NINA on Linux: there is no native port today, largely because of deep dependencies on Windows-specific frameworks and ASCOM. Running it under emulation or compatibility layers tends to reintroduce exactly the instability I’m trying to avoid, so it defeats the purpose.

If, in the future, a Linux-based solution offers the same level of comfort, predictability, and integration for my specific hardware and habits, I’d happily reassess. Likewise, if NINA ever gains a solid Linux port, I’d be very interested.

Until then, I’m not “defending Windows” I’m sticking with the platform that is currently delivering reliable data for me with minimal supervision. That’s a workflow choice, not a judgement of other ecosystems.

Read Noise,

I never took your post was a Windows Vs anything. However, for others to adapt what you found would be a very elongated approach to fix the inherent problems Windows does create in a working Windows/Nina environment. My battles with Windows have been years long, not something undercooked, or taken lightly. Years long of experiences’ with fixing my mount laptop and the ONLY thing that had changed was a Windows Security update.

Because since I began using computers in the early 1990’s I’ve been a devout Windows user. Since Microsoft 3.10 and 3.11. When Windows 95 came forth I was quick to join the march with my personal computer at home.

My Linux post is actually an inquiry to pick your brain. You appear to be way more deeply steeped in Windows than others I’ve encountered.

I hope one day that Microsoft can stop braking 3rd party drivers just because they don’t come from the Microsoft Store embedded in every copy of Windows XX versions. I’m certainly not alone in the summation. I’ve run across others who have found their typical 3rd party Astro Apps not working properly after a Windows Update.

So, I’m looking around for other alternative ways to get around their deliberate driver removals and damages just because they don’t have “ms” appearing in their code. Yours is likely the best I’ve seen up to now. But for me, trying to blockade myself and my mount computer from MS Updates, which are incessant, and keeping easily installable versions of solidly working programs when MS does foul up what was working yesterday. Like I have when Han Klien had to repair ASTAP after it was Apple modified, (an “Update”) by another. (Name withheld, but “Bob” knows who he is. How you like my hammer now?)

Maybe you do have a viable way to cleanse Windows so the code bangers at Microsoft can’t break out 3rd party, open sourced, Astro-Apps. I finally got NINA working by using Cuiv’s NINA tutorials. Any chance you could do a visual step by step of how you got Windows to finally behave for you?

I’d sure appreciate it and could stop having to repair my mount computer every blessed time a Windows Security Update rolls by and removes or damages (brakes) one of my useful Applications.

SonnyE avatar

Read noise Astrophotography · Dec 20, 2025, 05:45 PM

andrea tasselli · Dec 20, 2025, 04:43 PM

I can't for the life of me see how any update to Windoze can "break" one piece of code to one user but not to many others. Besides that writing "breaking" is kid-wording as there is no such a thing in real world.

The idea that an OS update can’t affect one user and not others assumes software exists in isolation from hardware, firmware, drivers, timing, and state which simply isn’t how real systems work.

Windows updates routinely modify:

  • kernel scheduling behavior

  • USB stack timing and power policy

  • driver load order and enumeration

  • device class filtering

  • background services and IPC timing

In hardware-attached workflows (cameras, mounts, hubs, rotators), those changes don’t “break code” they alter execution context. That’s enough to surface race conditions, power-state bugs, or enumeration conflicts that were previously latent.

This is why Microsoft documents update-induced regressions as environmental issues, not “broken code”. It’s also why identical setups on paper can diverge in practice.

This isn’t theoretical it’s standard behaviour in systems that depend on real-time I/O, USB topology, and vendor drivers.

Windows updates don’t need to “break code”.
They change drivers, power policy, and device enumeration.
That’s sufficient to affect hardware-dependent systems on some machines and not others a well-known class of environmental regression.

Thank You, Read Noise!

Windows updates don’t need to “break code”.
They change drivers, power policy, and device enumeration.
That’s sufficient to affect hardware-dependent systems on some machines and not others a well-known class of environmental regression.”

To me, if they change my drivers, screw with my power settings, or change things without me giving my permission, THAT IS breaking. (See #3) Because what worked before their incessant updates, DOES NOT after.

SonnyE avatar

andrea tasselli · Dec 20, 2025, 04:43 PM

I can't for the life of me see how any update to Windoze can "break" one piece of code to one user but not to many others. Besides that writing "breaking" is kid-wording as there is no such a thing in real world.

Break.

See #3.

andrea tasselli avatar
Read noise Astrophotography:
The idea that an OS update can’t affect one user and not others assumes software exists in isolation from hardware, firmware, drivers, timing, and state which simply isn’t how real systems work.

Windows updates routinely modify:

kernel scheduling behavior

USB stack timing and power policy

driver load order and enumeration

device class filtering

background services and IPC timing

In hardware-attached workflows (cameras, mounts, hubs, rotators), those changes don’t “break code” they alter execution context. That’s enough to surface race conditions, power-state bugs, or enumeration conflicts that were previously latent.

This is why Microsoft documents update-induced regressions as environmental issues, not “broken code”. It’s also why identical setups on paper can diverge in practice.

This isn’t theoretical it’s standard behaviour in systems that depend on real-time I/O, USB topology, and vendor drivers.

Windows updates don’t need to “break code”.
They change drivers, power policy, and device enumeration.
That’s sufficient to affect hardware-dependent systems on some machines and not others a well-known class of environmental regression.


How do you know if any if of this, assuming they are real and not just bluster, actually happens and why the vast majority of users do not feel any impact. Do you write your own drivers by any chance? Kernel scheduling behaviour? Really?
SonnyE avatar

andrea tasselli · Dec 20, 2025, 07:33 PM

Read noise Astrophotography:
The idea that an OS update can’t affect one user and not others assumes software exists in isolation from hardware, firmware, drivers, timing, and state which simply isn’t how real systems work.

Windows updates routinely modify:

kernel scheduling behavior

USB stack timing and power policy

driver load order and enumeration

device class filtering

background services and IPC timing

In hardware-attached workflows (cameras, mounts, hubs, rotators), those changes don’t “break code” they alter execution context. That’s enough to surface race conditions, power-state bugs, or enumeration conflicts that were previously latent.

This is why Microsoft documents update-induced regressions as environmental issues, not “broken code”. It’s also why identical setups on paper can diverge in practice.

This isn’t theoretical it’s standard behaviour in systems that depend on real-time I/O, USB topology, and vendor drivers.

Windows updates don’t need to “break code”.
They change drivers, power policy, and device enumeration.
That’s sufficient to affect hardware-dependent systems on some machines and not others a well-known class of environmental regression.



How do you know if any if of this, assuming they are real and not just bluster, actually happens and why the vast majority of users do not feel any impact. Do you write your own drivers by any chance? Kernel scheduling behaviour? Really?

Those who don’t feel any impact are probably not actually doing Astrophotography with things like PHD2, ASTAP, ASCOM, or programs that download their specific drivers to them. (ZWO)

You can live in denial, but others beside you DO experience issues.