[RFC] Plans for zh_TW: prioritize maintainability over coverage

[RFC] Plans for zh_TW: prioritize maintainability over coverage

[Weijie Yuan]

Hi Jon and Chinese documentation folks,

Recently I reviewed Jon's talk titled Kernel documentation: update
and future directions. In short, he said we Chinese people are
active, and he "even" thought why somebody is willing to read some
kinds of stuff in Chinese ;-) This led me to reflect on our current
situation.

                         Translations

      +-------------------+-------+----------------------+
      | Language          | Files | General state        |
      +-------------------+-------+----------------------+
      | Chinese           | 332   | Active               |
      | Chinese (trad)    | 97    | Almost abandoned     |
      | Korean            | 5     | Abandoned            |
      | Italian           | 67    | Occasional updates   |
      | Japanese          | 7     | Abandoned            |
      | Spanish           | 37    | Rare updates         |
      +-------------------+-------+----------------------+ Dec 2025

       https://www.youtube.com/watch?v=JXtx_zOHFgo (31:52)

       As of July 2026, checktransupdate.py reports the following:
       zh_TW: 51 translated files are out of date, involving 221
       distinct upstream commits in total.
       zh_CN: 171 translated files are out of date, involving 617
       distinct upstream commits in total.

First of all, I definitely would love to show my truly respect to
our friendly contributors, but considering the sustainability and
maintainability of the translation work, and more importantly, its
forward-looking nature (in line with the original English version),
I think we need to make some trade-offs.

It comes to me what the maintainer of Sqlite D. Richard Hipp said
about open-source or free software: there's free as in beer and free
as in speech. But there's another one: free as in puppies. When
people send you puppies as the gift, you're morally obligated to
take care of them. So, it's basically what's going on with our work:

  Nice and friendly contributors show up and give us their
  translations and walk away. If the original English page gets
  updated, you'll find that for most of the cases our friendly
  contributors won't show up again and maintain it, i.e. update it
  as expected.

The facts could certainly have been exaggerated. For both zh_CN&TW,
some articles have indeed remained unchanged for decades. So a
single translation is definitely a one-time solution, which means
they don't need to be touched often.

That's fine with zh_CN right now. But for zh_TW, we have a
significant gap and lag behind. You can easily see it from the above
table. So now we need to prioritize and get things done.

Therefore, for zh_TW, our expected plan is to monitor a list
containing important documents or folders (such as process/
submitting-patches.rst), and set a certain deadline for the latest
update (which can be discussed later, maybe 1/2 months): If no
contributors would like to update it, we as the maintainers will be
responsible for making the updates to ensure that readers who come
across this translation can be assured that the translation is
reliable and not worried about being out of sync with the English
documentation for too much/a long time. We may make the supported
set in the index page, and list out the current "well-cared"
directories or files.

That is to say, we will deliberately select relatively important
articles for translation and hope to maintain this consistently
first. By avoiding the large expansion of the translation coverage
we can't handle, it is possible to truly ensure the quality of each
article that has a translation.

Since the goal set for zh_TW was to catch up with zh_CN first, when
Chen-yu was working on the common articles this period, I was
extensively reviewing the existing content to determine the priority
of traditional Chinese docs, according to the importance and reading
volume, etc, and thereby deciding which specific parts should be
marked unmaintained temporarily.

Sorry for being so lengthy. In summary, our primary goal is to
ensure the timeliness and quality of the translations, and we will
also strive to expand our coverage.

Please feel free to leave your comments.

Thanks,
Weijie & Chen-Yu

Re: [RFC] Plans for zh_TW: prioritize maintainability over coverage

[Jonathan Corbet]

Weijie Yuan <wy@wyuan.org> writes:

> That is to say, we will deliberately select relatively important
> articles for translation and hope to maintain this consistently
> first. By avoiding the large expansion of the translation coverage
> we can't handle, it is possible to truly ensure the quality of each
> article that has a translation.

This seems like a good approach in general.

Of course, somebody is sure to point out that the *English*
documentation is often not maintained to such a standard.  We have a lot
of stuff that somebody dropped into the tree years ago and nobody has
updated since.  I shall continue to try to get the community to live up
to your level of diligence!

Thanks,

jon

Re: [RFC] Plans for zh_TW: prioritize maintainability over coverage

[Weijie Yuan]

On Mon, Jul 27, 2026 at 07:18:49AM -0600, Jonathan Corbet wrote:
> Weijie Yuan <wy@wyuan.org> writes:
> 
> > That is to say, we will deliberately select relatively important
> > articles for translation and hope to maintain this consistently
> > first. By avoiding the large expansion of the translation coverage
> > we can't handle, it is possible to truly ensure the quality of each
> > article that has a translation.
> 
> This seems like a good approach in general.

Thanks for your support.

> Of course, somebody is sure to point out that the *English*
> documentation is often not maintained to such a standard.  We have a lot
> of stuff that somebody dropped into the tree years ago and nobody has
> updated since.

That's very modest of you to say so. The English documentation is much
larger in scale, involves more complex workflows, and passes through
many more hands, so maintaining it is by no means an easy task.

The kernel community encompasses so many subsystems and people, and
reaching consensus is not always easy, so I completely understand the
constraints you are working under. (Yeah, The assisted-by trailer
discussion may be an example) By comparison, we have it much easier on
our side.

> I shall continue to try to get the community to live up to your level
> of diligence!

You're giving us too much credit. But it may be a little early to praise
us - we're only just getting started. ;-)

Since I´m new here, I´ve had to dig through the history of this part to
get up to speed quickly, and just watched Kernel Recipes 2016 - Kernel
documentation again, how time flies!

Thank you, Jon, for a decade of tireless dedication.

Re: [RFC] Plans for zh_TW: prioritize maintainability over coverage

[Weijie Yuan]

On Tue, Jul 28, 2026 at 11:59:39PM +0800, Weijie Yuan wrote:
> Thank you, Jon, for a decade of tireless dedication.

Sorry, decades, of course. ;-)