Re: i18n'd boot-floppies ?
Michael Sobolev <email@example.com> writes:
> BTW, one of the way for saving the disk space is static linking...
No comment! Seems like it's too late to rely on this.
The problem is to reduce the font sizes, isn't it, not to squirrel out
a little space from root.bin.
> > - rescue disk -- syslinux messages (i386 at least)
> syslinux does not allow to have same messages in different languages! So
> should we want to allow a non-English speaker see syslinux messages in her
> own language, we have (!) to create a localized version of rescue disk.
That's what I thought.
> > - root disk -- dbootstrap msgs, addressed by your work
> It's not ready for use. Again, I rely on a very specific things: a framebuffer
> capable kernel and dbootstrap running under bterm.
Interesting. Are there test versions built this way?
Do you think we could provide a root disk compiled with i18n as a new
'i18n' flavor on i386?
> > - base -- already i18n'd I think
> I know nothing about it.
> > - documentation -- already i18n'd
> Well, Russian variant is not up-to-date. I believe that other documentation
> may be not up-to-date as well.
Yes, but that's a translator problem. I am only concerned with
removing and *structural* barriers to i18n.
> > What's the story with the rescue disk?
> To my mind, a reasonable approach is to have English language speaking rescue
> disk (I mean syslinux messages). If user does not need to to do anything
> special those messages are of no use for her. (If I understand correctly,
> this is the only thing on rescue disk, no?)
> > Are we planning on building one rescue disk for each language supported?
> No! I want to avoid this!
Well, we have two choices:
- no i18n of rescue disk
- build one rescue disk per language
Neither are very nice.
Oh well, we can just let the local teams provide l10n'd versions of
the rescue disk unofficially if they want...
.....Adam Di Carlo....adam@onShore.com.....<URL:http://www.onShore.com/>