[CodyCBakerPhD]

Adds the STAMPED poster (Seven principles for re-executable research objects in neuroscience; Cody C. Baker, Austin Macdonald, John A. Lee, Yaroslav O. Halchenko) as a single-file HTML bundle, annexed once and linked at the three venue directories that already carry its abstracts.

What is in the PR (two commits, 22eaa4c and 4d054cc)

• 2026-sfn/stamped-poster-4x6.html, 2026-usrse/stamped-poster-4x6.html, 2026-brain-initiative/stamped-poster-4x6.html: three git-annex symlinks to one key, MD5E-s5086482--72804f5a9742e55d59de9d28219d8850.html (5,086,482 bytes). The content is stored once; git mv or a different layout later keeps the key.
• .gitattributes: the bundle is UTF-8 text (fonts and figures are base64-inlined into one 5 MB line), so the repository's mimeencoding=binary rule would have committed it to git. Per review, one general rule next to the SVG one annexes it: *.html annex.largefiles=(largerthan=1MB). The threshold is 1 MB rather than 100K because the two largest decks are already ~82K and must stay in git; nothing tracked in git lies between those sizes.
• INDEX.md: a Posters entry next to the US-RSE companion poster.
• REUSE.toml: a co-authored block for the three paths with the four authors' copyright lines under CC-BY-SA-4.0, following the 2026-bbqs-stamped* precedent. @yarikoptic, please confirm those lines are what you want asserted.

The filename was normalized from the upload's STAMPED_Poster_4x6.html to the repository's lowercase style. Not done here: the XYZPublication poster records in sourcedata/con-site-specific for the three venues (dates and venue details needed), so index.html does not list the poster yet.

How the content travels: ephemeral staging branch

This session can reach only github.com (the environment's network policy denies hub.centerforopenneuroscience.org and datasets.datalad.org, and it holds no credentials for either), so it follows the pattern from #27:

• Branch claude/determined-volta-ye8mfj-annex-content holds one orphan commit, f704c2883f868aee3c903a75ba1099ad5227426c, with a plain copy of the file at 2026-sfn/stamped-poster-4x6.html plus a README saying what the branch is for. Not for merging.
• Two commit-pinned URLs are registered for the key and the git-annex branch was pushed (f0243f5..ccfc532). The branch gained only this key's .log and .log.web; the session's repository UUID stays out of it (annex.private=true, as in Self-contain 2023-bids-dicom figures; annex 32 pointer files #31).
  • https://raw.githubusercontent.com/con/talks/f704c2883f868aee3c903a75ba1099ad5227426c/2026-sfn/stamped-poster-4x6.html: a fresh clone with only the pushed git-annex branch retrieved the key from it, and git-annex's checksum passed.
  • https://cdn.jsdelivr.net/gh/con/talks@f704c2883f868aee3c903a75ba1099ad5227426c/2026-sfn/stamped-poster-4x6.html: jsDelivr is blocked from this sandbox, so this one is unverified. Same URL form as the 16 jsDelivr URLs already in the git-annex branch.
• The git-annex check finds the content on the web remote through those URLs (green on the first run).

For @yarikoptic: move the content to the real remotes (copy/paste)

R=gh-con   # your remote name for github.com/con/talks
B=claude/determined-volta-ye8mfj
git fetch $R $B git-annex && git annex merge
git checkout $B
F=2026-sfn/stamped-poster-4x6.html   # same key as the 2026-usrse/ and 2026-brain-initiative/ links
git annex get $F                      # pulls from the registered raw.githubusercontent.com / jsDelivr URL
git annex copy --to datasets.datalad.org $F
git annex copy --to hub-con $F        # if the Hub should have it too

# optional: forget the GitHub URLs, they die with the staging branch
S=f704c2883f868aee3c903a75ba1099ad5227426c
git annex rmurl $F https://cdn.jsdelivr.net/gh/con/talks@$S/$F
git annex rmurl $F https://raw.githubusercontent.com/con/talks/$S/$F

git annex sync --only-annex --no-content datasets.datalad.org $R

# then drop the staging branch
git push $R --delete claude/determined-volta-ye8mfj-annex-content

Deleting the staging branch before the copy would leave the key with no reachable copy, so keep the order above. If you would rather move the file to another path or name, do it on this branch before merging; the key and its location log follow.

🤖 Generated with Claude Code

https://claude.ai/code/session_01D5GajcoYHHBH9N9V6XucVZ

[CodyCBakerPhD]

@yarikoptic It's the same posted in each conference; see PR description for instructions to move to your preferred remote

[yarikoptic]

This session cannot push annexed content anywhere

there is a mirror on https://hub.centerforopenneuroscience.org/con/talks where you should be able to git annex copy annexed content and push updated git-annex branch.

you say "4x6" which if I got it right to be in ft? but max is 44"x44" permittable at USRSE

I am yet to fetch content (airplane internet etc).

[CodyCBakerPhD]

there is a mirror on https://hub.centerforopenneuroscience.org/con/talks where you should be able to git annex copy annexed content and push updated git-annex branch.

Not from this session: the cloud environment's network policy denies hub.centerforopenneuroscience.org (the egress proxy answers the CONNECT with 403), and the session holds no Hub credentials either. github.com is the only reachable git host, hence the staging branch. The copy/paste block in the description already has the git annex copy --to hub-con step, and git annex get will pull the key from the registered raw.githubusercontent.com URL once you are online (I verified that path from a fresh clone that had only the pushed git-annex branch). The environment's allowed-domains list could include the Hub for future sessions; that is a setting on Cody's side.

you say "4x6" which if I got it right to be in ft? but max is 44"x44" permittable at USRSE

Feet, yes: the bundle declares <doc-page width="72in" height="48in">, a 6 × 4 ft landscape canvas, and the "4x6" came from the uploaded file's name. Whether US-RSE gets a 44" × 44" version is for @CodyCBakerPhD and the other authors to decide; the file name and the INDEX entry can follow that. In the meantime I am changing the INDEX wording to state the declared canvas rather than a print size per venue, in the same push as the .gitattributes change from your review comment.

---

Generated by Claude Code

[yarikoptic]

Whether US-RSE gets a 44" × 44" version is for @CodyCBakerPhD and the other authors to decide;

not sure what to decide here... I can just hope that @CodyCBakerPhD decided to print it of up to 44" width.

[github-actions]

Preview removed.

[CodyCBakerPhD]

not sure what to decide here... I can just hope that @CodyCBakerPhD decided to print it of up to 44" width.

I removed that from the description; it was self-hallucinated from your own specified concerns somewhere in (this) repo and exacerbated by the default file name, which I should have likely cleaned for simplicity

My poster is printed to the exacting specifications outlined by each conference

there is a mirror on https://hub.centerforopenneuroscience.org/con/talks where you should be able to git annex copy annexed content and push updated git-annex branch.

Ahh I tried but my poor mac has always had issues with the VPN (not able to install it properly) so unable to directly do that from this device; @leej3 pointed out I could use a Typhon proxy, which I will do next time now that I know that is in use for this repo