Remove Migrated Lecture ar1_processes

[HumphreyYang]

Hi @mmcky,

This PR removes lectures that have been migrated to the intro series (QuantEcon/lecture-python-intro#334).

The AR(1) lecture is now in lecture-python-intro/pull/335

Do we still have a template for the migrated lecture? Apologize that I cannot find the template anymore in the commit thread.

[mmcky]

Do we still have a template for the migrated lecture? Apologize that I cannot find the template anymore in the commit thread.

@HumphreyYang I am not sure what you mean by template? (a process?)

We will want to also setup a redirect using sphinx-reredirect

• setup redirect

[mmcky]

https://documatt.com/sphinx-reredirects/

[HumphreyYang]

Do we still have a template for the migrated lecture? Apologize that I cannot find the template anymore in the commit thread.

@HumphreyYang I am not sure what you mean by template? (a process?)

I remembered that we have a small admonition at the top that says the lecture has been migrated and will be taken down after a month. Please correct me if I were wrong.

[mmcky]

Next Steps

• merge Migrate AR(1) Lecture lecture-python-intro#335 and make live
  scalar_dynam has been migrated (@HumphreyYang were you able to diff them to make sure all changes were transferred?)

Once the new locations are live and published

• merge this PR with redirects in place for _config.yml

[HumphreyYang]

Do we still have a template for the migrated lecture? Apologize that I cannot find the template anymore in the commit thread.

@HumphreyYang I am not sure what you mean by template? (a process?)

I remembered that we have a small admonition at the top that says the lecture has been migrated and will be taken down after a month. Please correct me if I were wrong.

I found it for the heavy_tails lecture. I will put this in place now.

[mmcky]

thanks @HumphreyYang.

I guess we need to choose if we redirect automatically to the new location OR use that admonition to indicate the lecture has moved. Let's go with the admonition for now (as that is what we have done in the past) and think about the redirect process. We could add the admonition to the top of the new location (for a month instead) to say this lecture location has changed and this is the new location.

[mmcky]

@HumphreyYang

We could add the admonition to the top of the new location (for a month instead) to say this lecture location has changed and this is the new location.

Actually - let's give this a go.

We will need to:

• setup sphinx-reredirects (@mmcky)
• setup the redirect for appropriate pages (@mmcky)
• setup admonition on the new lecture (@HumphreyYang)
• open an issue on the new lecture series to remove the admonition (@HumphreyYang)

[HumphreyYang]

Many thanks @mmcky,

I think it is a better design choice. Readers cannot see the migration notice for the heavy_tail lecture as they are redirected to the new page : )

[mmcky]

Readers cannot see the migration notice for the heavy_tail lecture as they are redirected to the new page : )

We couldn't have both as the md file can't exist when setting up the redirect. So I think the admonition in the new location is the better option. You end up in the right spot and then we let them know the location is new. I think that works well. 👍

[HumphreyYang]

scalar_dynam has been migrated (@HumphreyYang were you able to diff them to make sure all changes were transferred?)

Many thanks @mmcky,

I just ran a diff and found one section is deleted in the intro version. Perhaps we should leave it for now, and I will consult John in our next meeting.

[HumphreyYang]

The warnings are complaining about the missing ar1_processes in the intro series.

We should get green ticks once ar1_processes is merged and live in the intro series.

[mmcky]

@HumphreyYang this lecture is now live

https://intro.quantecon.org/ar1_processes.html

[mmcky]

@HumphreyYang I have added the redirect for the ar1_processes lecture to https://intro.quantecon.org/ar1_processes.html

[github-actions]

🚀 Deployed on https://65b331b9e3fde16f8cba5402--nostalgic-wright-5fa355.netlify.app

[mmcky]

@HumphreyYang well this is interesting. Are these the warnings to showed up last time?

/__w/lecture-python.myst/lecture-python.myst/lectures/finite_markov.md:677: WARNING: unknown document: intro:ar1_processes
/__w/lecture-python.myst/lecture-python.myst/lectures/inventory_dynamics.md:286: WARNING: unknown document: intro:ar1_processes
/__w/lecture-python.myst/lecture-python.myst/lectures/linear_models.md:47: WARNING: unknown document: intro:ar1_processes
/__w/lecture-python.myst/lecture-python.myst/lectures/kesten_processes.md:41: WARNING: unknown document: intro:ar1_processes
/__w/lecture-python.myst/lecture-python.myst/lectures/kesten_processes.md:173: WARNING: unknown document: intro:ar1_processes

[HumphreyYang]

Many thanks @mmcky,

I fixed them in 646cf0b

and we have green ticks after that.

Perhaps redirects overrules intersphinx? We can try deleting intersphinx tags at the front.

[mmcky]

Many thanks @mmcky,

I fixed them in 646cf0b

and we have green ticks after that.

Perhaps redirects overrules intersphinx? We can try deleting intersphinx tags at the front.

thanks @HumphreyYang yeah there definitely seems to be an issue using both intersphinx and reredirects.

I know inter sphinx doesn't do connections to external website (at the page level) so I will take a closer look at https://documatt.com/sphinx-reredirects/usage.html to see if perhaps we should use these for internal links as well?

I am surprised though as inter sphinx is using a domain so should be independent. Will have to figure this one out.

[HumphreyYang]

Many thanks @mmcky,

I experimented removing the tag or adding an empty ar1_processes but no luck.

I was also thinking about its implications on our new series. I think we can prioritize intersphinx in our new lecture series as we might be able to create an empty repository (with only the config file) hosting the redirect mappings between old intermediate series and new series. In this case, we do not have internal intersphinx references in the redirect repository, and intersphinx will also work in our new series.

[mmcky]

I think we can prioritize intersphinx in our new lecture series as we might be able to create an empty repository (with only the config file) hosting the redirect mappings between old intermediate series and new series. In this case, we do not have internal intersphinx references in the redirect repository, and intersphinx will also work in our new series.

Thanks @HumphreyYang I don't fully understand what you are proposing. Can you explain?

We will need page redirection support to make sure anyone with python.quantecon.org/ar1_processes ends up at intro.quantecon.org/ar1_processes. This is important as 25K a month will be making use of the old addresses until they update.

[HumphreyYang]

Many thanks @mmcky

I am not sure whether it will work but I am proposing to have a fresh repository with old intermediate series URL. It will host the redirect config and intro page only (without lectures). In this case, it wouldn’t conflict with intersphinx since lectures are hosted in our target repositories.

Please let me know your thoughts on this.

[mmcky]

@HumphreyYang are you replicating this locally -- I can't so the mystery depends. I have triggered a new build run to confirm.

[HumphreyYang]

@HumphreyYang are you replicating this locally -- I can't so the mystery depends. I have triggered a new build run to confirm.

Yes, I can replicate this locally. I will have another go today.

[mmcky]

@HumphreyYang I am definitely getting no local build errors for either html or pdflatex using a fresh environment of the base environment.yml. It would be helpful if you can confirm. I am going to do a test on this PR by removing the reredirect setting to see if this changes -- then will restore.

[mmcky]

@HumphreyYang the error still persists without ca5a20a so maybe intersphinx isn't working as it should?

[HumphreyYang]

Hi @mmcky,

This is getting more interesting as 646cf0b passed, and the only change afterwards was adding redirects. I am running a fresh build locally and will report my results once they are available.

[mmcky]

@HumphreyYang I think I have figure this out 🤞 -- I think the settings are misspecified in the context of the yml file. I'll push a test commit in a minute or two.

[mmcky]

@HumphreyYang I am rebuilding the cache to check there isn't some form of corruption.

[mmcky]

@HumphreyYang no idea why cdf81e8 built OK when the cache hasn't finihsed rebuilding yet. Clearly some instability. I will re-run this once cache is finalised to check all is green after that.

[mmcky]

@HumphreyYang I think this is working nicely now.