Prueba con ayoze #515
Prueba con ayoze #515
Conversation
❌ Deploy Preview for pauseai failed.
|
|
Apologies, this change was not meant to be suggested for global website |
|
Can you note what you were trying to do, how you made the changes (code / by hand) and your understing of what happened @danilupion ? My reference in #447 above might help in sharing context. |
|
I was planning to explain during Wednesday meeting, but I might as well do it here @anthonybailey we forked the site because there is interest to have control on the content. My priority right now is to help local Spanish group to have a simplified version of the page in Spanish, right after that I would like to continue helping global with i10 and Spanish. In case you are curious you can see what has been done here My previous comment regarding this PR being accidental, really meant that the target should have never have been this repo but the one I shared before -and even there this was not meant to be merged, as this PR was created as part of a hands-on git training for Ayoze to be able to contribute and was meant to just be an example-), again... Apologies for the slide. |
I read this as: in aggregate, the team want to manually edit completed translations. Is it just in the short term, or going forwards? If the latter: member requirements matter most, and I want to support you. But I have to understand properly to do that best. So I would like to ask: why? Assuming the former: that kind of matches what I understood you were going to do, although cloning a new repos from the global website repos was a bit of a surprise. I don't see any notes on how you are going to do this end-to-end - is this pull request on the global site repos the best we have? How are you intending to:
There are multiple options, and I've love to help advise and to learn from how it goes. But of course, I'm a bit unreliable with time at the moment, and exploring unconstrained may well suit your own tastes and learning experience. Do what you judge best! Just let me know. |
I looks like going fordward (and it is not only about the translations, i will explain in the following section)
I will try to share what the group's view is in case you see some middle ground (and/or there something we are not taking into account in terms of what is expected/not expected). The group understood:
The group would like to:
The end goal is to have something parallel to https://pauseai.rs or https://pauseai.ro but having been exposed to the spanish version of the global site, to me it looked like the obvious version to start with for some content fast (just in case this is sensitive, we were planning on respect/enforce attribution)
Nope, is the later, hope that explains the forking (and apologies if you would have expected a headsup). In all honesty it was my first time working with forks in github and i was not aware that from a branch in the forked repo the default PR goes against the original repo.
For now what I had in mind is to dockerize the forked repo and serve in a server I currently have at hetzner
Any feedback (in any area) that you might have is extremely valuable to us |
|
Right...
Cassandra here predicts this will be regretted, but will try not to be stubborn, and does value learning from experience.
Really appreciated!
Great that they will be happy to do that. But also suggests they don't think it will be ineffective? OK. I read "the l10ns are basically fine, it's all about the remaining differences."
I get that. Fwiw this is a change we envisage on the global site too. It's always time that pushes it down the backlog. I think there is a nice opportunity here. If your folk are already wanting to put effort into that, then could they suggest the change in terms of en content? If it is generally better, the world net improves not just Spain. (We could also imagine localising simpler es back to simpler enough, although more fiddly.) Interlude. I will share a dream: I would actually like simplification or "summarize, with optional details" to itself be a l10n-like task. We can go all the way to "explain like I'm five". One source, a few tags and case-specific prompts and we end up with an easily editable website that can present itself in hundreds of reader-friendly ways. (I think my dream does not extend to "persuade particular political factions" but it's clearly doable.)
Localising the logo is quite cute: definitely doable in general for l10n of global site. Different choices re theme mechanics is the previous bullet, I think: "we don't like some choices global made and want to do things ourself". This Is Fine. You Do You. We just have to be clear it has costs on the es side too. See later.
These are nice, although I'm kind of distressed there are no links in either navigation or text to see more in English on the global website. It seems obviously value dropped on the floor. (Interested to know how the join and petitions options work. I'm hoping those do go back to global DBs rather than also being forks! But don't want to test in prod and lazy about reading code.)
Absolutely agree with that: will be interested to see how much you want to cut away. As mentioned: if you want a pretty minimal version of the site where you are only changing a few pages, adding those and building/serving via global as Australia did is another viable alternative. Russia and Estonia have the same option. Hmmm, I wonder whether we should get in touch and ask them how it's going?
Don't even remember where that applies. Do you mean some affiliate links? If you just mean credit the global en source, not necessary, only encouraged as a way to link back!
Only re the change in plans. No worries.
Sounds good for now! You didn't answer the questions re (if staying forked) how you expect to keep up with improvements and additions on the global site, or how we can get your good ideas back! You know the coding analogy of managing ongoing forks: you need decent tools and processes and it is still painful anyway. Forking often means "goodbye forever". And it's worse with content, because there you don't have unit tests to share!
Ok I won't be shy until I fall over or you tell me to back off. 🙂 |
|
Sorry for the delay, here I go :)
My current goal is to have something working that makes local spanish group happy, as soon as possible. If we are able to align in "single source of truth", me personally see that as the best possible outcome. So far I sense "friction" in the sense that from one side i see appetite for custom content (even for "general content" suchs us learn) and on the other I sense appetite to join initiatives so that everyone benefits from whoever's work. I think this could be a little challenging but my (naive) take is... let's see what the spanish group would produce, if we think that the content is valid for global... maybe this is a no brainer and we just want to adapt that content.
It is not about remaining differences but rather content (more leaning towards a way lighter approach, but again... I am not 100% sure, thus I still think that having the outcome (spanish only content in a forked version) to analyse is overall netly benefitial. (and from there we decide that the goal is)
I share your vision but also do not want to be in the way, right now it feels (at the local spanish group) that we have an tangled set of interdependencies and my goal is to unblock what I can (that does not mean we stop there).
I think this is pretty much what the spanish group would love to see
Amazing :)
Apparently after a quick poll... this is just about potentially making default theme light instead of dark... so in my mind became a no brainer (as in it should not be a determining factor at all)
We are commited to backlinking also 100% clear that the join form should be feeding your data (one of the items i wanted to discuss is i18n of that Tally form, which based on a 5 mins search is more complex than I thought)
Will keep you posted the moment i have something and if needed will share english translations
I think this would make sense so that just because someone became verbose in the spanish group, we do not steer direction just "for" them (as in take into account their requirements)
I mean https://creativecommons.org/licenses/by/4.0/ and for sure I would influence if needed (i really doubt there is a need) so that there are link backs
It really depends on what the outcome is, i hope to have a better understanding when i get the sense of what the desired content looks like.
I agree and personally would prefer if we eventually merge back (willing to work in that direction if an agreement is met)
|
|
Time-limited now - book reading! - but quick response is that this all
sounds like we are thinking very similarly. Go for it! More notes later
maybe.
…On Mon, 20 Oct 2025, 10:41 Dani Lupión, ***@***.***> wrote:
*danilupion* left a comment (PauseAI/pauseai-website#515)
<#515 (comment)>
Sorry for the delay, here I go :)
I looks like going forward (and it is not only about the translations, i
will explain in the following section)
Cassandra here predicts this will be regretted, but will try not to be
stubborn, and does value learning from experience.
My current goal is to have something working that makes local spanish
group happy, as soon as possible. If we are able to align in "single source
of truth", me personally see that as the best possible outcome. So far I
sense "friction" in the sense that from one side i see appetite for custom
content (even for "general content" suchs us learn) and on the other I
sense appetite to join initiatives so that everyone benefits from whoever's
work.
I think this could be a little challenging but my (naive) take is... let's
see what the spanish group would produce, if we think that the content is
valid for global... maybe this is a no brainer and we just want to adapt
that content.
I will try to share what the group's view is in case you see some middle
ground (and/or there something we are not taking into account in terms of
what is expected/not expected).
Really appreciated!
The group understood:
The only way spanish group was meant to influence content was by giving
feedback so that the prompt got updated. (We are still happy to do so to
improve spanish content for global page)
Great that they will be happy to do that. But also suggests they don't
think it will be ineffective? OK. I read "the l10ns are basically fine,
it's all about the remaining differences."
It is not about remaining differences but rather content (more leaning
towards a way lighter approach, but again... I am not 100% sure, thus I
still think that having the outcome (spanish only content in a forked
version) to analyse is overall netly benefitial. (and from there we decide
that the goal is)
The group would like to:
Be autonomous in the content (as in make the content they see fit -for
instance the group thinks that global website might be "a little too much"
and would like have simplified texts (at least in the main sections -that
might reference longer more in depth content-)
I get that. Fwiw this is a change we envisage on the global site too. It's
always time that pushes it down the backlog.
I think there is a nice opportunity here. If your folk are already wanting
to put effort into that, then could they suggest the change in terms of en
content? If it is generally better, the world net improves not just Spain.
(We could also imagine localising simpler es back to simpler enough,
although more fiddly.)
I share your vision but also do not want to be in the way, right now it
feels (at the local spanish group) that we have an tangled set of
interdependencies and my goal is to unblock what I can (that does not mean
we stop there).
Interlude.
I will share a dream: I would actually like simplification or "summarize,
with optional details" to itself be a l10n-like task. We can go all the way
to "explain like I'm five".
I think this is pretty much what the spanish group would love to see
To have the possibility to have slightly different designs (I do not
foresee anything big, but for example exiting the auto theme, using custom
logo and such)
Localising the logo is quite cute: definitely doable in general for l10n
of global site.
Amazing :)
Different choices re theme mechanics is the previous bullet, I think: "we
don't like some choices global made and want to do things ourself".
This Is Fine. You Do You. We just have to be clear it has costs on the es
side too. See later.
Apparently after a quick poll... this is just about potentially making
default theme light instead of dark... so in my mind became a no brainer
(as in it should not be a determining factor at all)
The end goal is to have something parallel to https://pauseai.rs or
https://pauseai.ro
These are nice, although I'm kind of distressed there are no links in
either navigation or text to see more in English on the global website. It
seems obviously value dropped on the floor.
(Interested to know how the join and petitions options work. I'm hoping
those do go back to global DBs rather than also being forks! But don't want
to test in prod and lazy about reading code.)
We are commited to backlinking also 100% clear that the join form should
be feeding your data (one of the items i wanted to discuss is i18n of that
Tally form (which based on a 5 mins search is more complex than I thought)
having been exposed to the spanish version of the global site, to me it
looked like the obvious version to start with for some content fast
Absolutely agree with that: will be interested to see how much you want to
cut away.
Will keep you posted the moment i have something and if needed will share
english translations
As mentioned: if you want a pretty minimal version of the site where you
are only changing a few pages, *adding* those and building/serving via
global as Australia did is another viable alternative.
Russia and Estonia have the same option. Hmmm, I wonder whether we should
get in touch and ask them how it's going?
I think this would make sense so that just because someone became verbose
in the spanish group, we do not steer direction just "for" them (as in take
into account their requirements)
(just in case this is sensitive, we were planning on respect/enforce
attribution)
Don't even remember where that applies. Do you mean some affiliate links?
If you just mean credit the global en source, not necessary, only
encouraged as a way to link back!
I mean https://creativecommons.org/licenses/by/4.0/ and for sure I would
influence if needed (i really doubt there is a need) so that there are link
backs
For now what I had in mind is to dockerize the forked repo and serve in a
server I currently have at hetzner
Sounds good for now!
You didn't answer the questions re (if staying forked) how you expect to
keep up with improvements and additions on the global site, or how we can
get your good ideas back!
It really depends on what the outcome is, i hope to have a better
understanding when i get the sense of what the desired content looks like.
You know the coding analogy of managing ongoing forks: you need decent
tools and processes and it is still painful anyway. Forking often means
"goodbye forever". And it's worse with content, because there you don't
have unit tests to share!
I agree and personally would prefer if we eventually merge back (willing
to work in that direction if an agreement is met)
Any feedback (in any area) that you might have is extremely valuable to us
Ok I won't be shy until I fall over or you tell me to back off. 🙂
[image: Keep it real @Gifcept - Where gifs live!]
<https://camo.githubusercontent.com/7f2880810139d03c4cfe6322c0be41386b7f3ecc49348b75d8597eac1d1c5be9/68747470733a2f2f676966636570742e636f6d2f3337374b497a526a4f2e676966>
—
Reply to this email directly, view it on GitHub
<#515 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AAAXXDOPM2L5Q3EF22MPQIL3YSU2ZAVCNFSM6AAAAACJRODJEWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTIMRRGI3DMMRTHE>
.
You are receiving this because you were mentioned.Message ID:
***@***.***>
|
2 participants
No description provided.