Unironically, I had the most success with old windows programs not when using wine directly but using proton with steam. I personally use umu[1] to use proton without directly needing to run steam. I wrote a small KDE script for .exe files so I can just double click them and they run lol. Or for setups I can right click them and just install them as a setup and it automatically creates a app shortcut I can open.

[1]https://github.com/Open-Wine-Components/umu-launcher

Got a link to the script?

Yes! I do here [1]. But keep in mind I am using NixOS and so it looks a bit weird. You should be able to just copy paste the script part and it should work when you have umu at path.

[1] https://pastebin.com/JRaK4uRV

Might want to try again, Wine progressed a lot in the past couple years.

> Might want to try again, Wine progressed a lot in the past couple years.

You could even go as far as suggesting SteamOS once they release the OS to more devices. Gaming themed sure but it's a flavor of Arch and you have full control over what gets installed.

It works well, though its messages could use some TLC:

    libEGL warning: pci id for fd 31: 10de:1ff0, driver (null)

    pci id for fd 33: 10de:1ff0, driver (null)
    pci id for fd 34: 10de:1ff0, driver (null)
    libEGL warning: egl: failed to create dri2 screen
    libEGL warning: pci id for fd 31: 10de:1ff0, driver (null)

    pci id for fd 33: 10de:1ff0, driver (null)
    pci id for fd 34: 10de:1ff0, driver (null)
    libEGL warning: egl: failed to create dri2 screen
    libEGL warning: pci id for fd 31: 10de:1ff0, driver (null)

    0124:fixme:nls:RtlGetThreadPreferredUILanguages 00000034, 0313F66C, 0313F6DC 0313F674
    0124:fixme:nls:get_dummy_preferred_ui_language (0x34 0x1009 0313F66C 0313F6DC 0313F674) returning a dummy value (current locale)
    0124:fixme:heap:RtlSetHeapInformation HEAP_INFORMATION_CLASS 1 not implemented!
    0124:fixme:nls:RtlGetThreadPreferredUILanguages 00000034, 0313F9D4, 0313FA44 0313F9DC
    0124:fixme:nls:get_dummy_preferred_ui_language (0x34 0x1009 0313F9D4 0313FA44 0313F9DC) returning a dummy value (current locale)
    0124:fixme:shell:InitNetworkAddressControl stub
    0124:fixme:richedit:editor_handle_message EM_GETLANGOPTIONS: stub
    0124:fixme:richedit:editor_handle_message EM_SETLANGOPTIONS: stub
    0124:fixme:ntdll:NtQuerySystemInformation info_class SYSTEM_PERFORMANCE_INFORMATION
    0124:fixme:win:RegisterTouchWindow hwnd 000100E0, flags 0 stub!
    0124:fixme:msvcrt:__clean_type_info_names_internal (7853A300) stub
    0124:fixme:msvcrt:__clean_type_info_names_internal (7B4F6BE4) stub
    0124:fixme:msvcrt:__clean_type_info_names_internal (79410E54) stub

wine whatever.exe 2>/dev/null

At this point, you might want to consider throwing an LLM at it and just letting it reimplement the thing so that it runs on linux.

They can actually do that. They may not like it, but they can.

Reverse engineer a binary proprietary database format? I have some doubts.

Why?

Certainly not unattended, but no AI should ever be unattended.

But if you closely guide it, support it with tools like Ghidra and force force force force force a process with many sanity checks and quintuple-checks, it's possible.

What previously needed a whole team and months might be one guy, a lot of tokens and 1-3 weeks. Doable, fun, and interesting.

__

Judging by the downvotes, I guess people mistook the initial comment for HN VC fueled AI delusions and I can't blame them for that. That's not what it was tho.

How do you guide it when you have no idea yourself? You mean you must first reverse engineer it manually and then hint it to the solution? Why would you do that at that point?

I feel like you might still be perceiving the AI suggestion as some magic box that just does things. It's not. It's a very fast very eager junior.

Why and how do people hand work to juniors?

You have a rough idea where you want to go, you have a rough idea what needs to be done, and then you iterate.

For example, you think "okay, I need some way of validating this" and then you tell it to run the software to generate test data.

While errors may creep in, you should be able to validate that step.

And then you use that test data to validate whatever next steps it should be doing.

It's essentially the same workflow as any mapping out of something new, but with the individual mapping steps being done by a drone. You do still draw the actual map, do logistics and strategize.

It's not easy work, but it's also not difficult work. What it is is laborious, and that's where LLMs help.

__

To stick with your database format question:

If the software uses that data, it also needs code to parse and interact (with) it.

This code resides within it, and you can forcefully pull it out.

That step of "write me a spec for this data format based on this code. Build a parser in python and build test cases using this DB file" can be done agentically (when sliced into small logical blocks of work orders).

You see how you went from "it's easier to use ai instead of using wine on linux" to "to use ai you must do all this work"

Well then using wine was the correct solution to begin with?

> This code resides within it, and you can forcefully pull it out.

What code? If you had the code you could port it.

The slop engine can drive IDA just fine. Does hitch a bit on obfuscated code, but with some poking it does fine there too.

VMs were not an option?

Can’t wait till Fable 6 can just decompile and reimplement old software like that.

Great for the americans. What are the rest of us going to do?

Use the distilled chinese models.

Maybe get your governments and citizens to innovate and create their own instead of relying so heavily on other countries. I thought that's the direction other countries were trying to go.

Good thing that the US frontier labs only download data from the US internet created by US citizens /s

Try Opus 4.8? It's just a language translation task. LLMs should be good at it.