Hi Eric,

> On Feb 1, 2026, at 4:11 PM, Eric Auer via Freedos-devel 
> <[email protected]> wrote:
> 
> 
> Hi Jerome, thanks for the update :-)
> 
> I found that VENTOY https://www.ventoy.net/en/ can help Linux and
> Windows users to make bootable USB sticks from ISO images, such as
> the live FreeDOS ones :-)

Good to hear about another viable alternated.

We could definitely use docs on how to create “real” media 
from the various images provided with a release. 

But, such docs are not something which I will likely be
creating. It is something which I simple do not have the
spare time to allocate to them. We really need volunteers
to submit them to the Documentation project in the 
FreeDOS GitLab Archive at https://gitlab.com/FreeDOS/docs 
<https://gitlab.com/FreeDOS/docs>

> Questions about the translation updates: It seems you updated
> several translations using AI, without notifications to the
> list where humans could proof-read the results. Which languages
> of messages for which FreeDOS packages are made by AI in T2026?

First let me say, I am not a translator. With the exception being
a couple of the programs which I have created, I do not create
or update translations for other programs either manually or 
using AI. 

What I try and accomplish is to provide a single location for 
translators to submit translations and include those updates
in the packages provided with FreeDOS. Also in the past,
I have also notified many developers who actively maintain
their projects that we have updated translations for them 
to include with their official download release. 

While I have mentioned this before, it has been a while. 
All translations that I include in the packages provided 
with FreeDOS are maintained in the FD-NLS project.

https://github.com/shidel/fd-nls <https://github.com/shidel/fd-nls>

You are correct in saying that the AI translations 
should be reviewed by humans. But honestly,
the same can be said for the Human translations.

Incorporating the AI translations presented a variety 
of issues. For example, we want to be able to 
distinguish AI from Human translations. We also
want to keep around any existing Human versions.
There are also issues involving the workflow and
tools used to manage all that stuff.

What I decided is the easiest and most reasonable 
solution to all of those issues is based on implementing
a simple directory structure to break up and 
preserve translations for the different programs that
received AI translations. It is a simple solution
which works well enough for now.

The different programs already contain a sub-
directory for the type of translation. For example,
NLS and HELP. The files in such directories are
the ones which get included in the package for the
program. This has not changed.

The programs that receive AI translations get
additional directories suffixed with ‘-ai’ or ‘-human’.
Those new paths is where any new contribution
work will go. Then, the latest version (wether AI 
or Human) is placed in then ls directory for
inclusion in the package.

This solution is fully compatible with the tools
I use to manage the translations. Plus, it keeps
any existing human translations around and 
permits integrating AI versions. Perfect? No.
But, works well enough for the time being.

A great example is to look at the newer method
used for translations to CHOICE at:
https://github.com/shidel/fd-nls/tree/master/choice 
<https://github.com/shidel/fd-nls/tree/master/choice>

:-)

Jerome

_______________________________________________
Freedos-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/freedos-devel

Reply via email to