📷 `EXIF` Photo Blog

app-overview.mp4

Demo App

https://photos.sambecker.com

Features

Built-in auth
Photo upload with EXIF extraction
Organize photos by tag
Infinite scroll
Light/dark mode
Automatic OG image generation
CMD-K menu with photo search
Experimental support for AI-generated descriptions
Support for Fujifilm simulations

Installation

1. Deploy to Vercel

Click Deploy
Add required storage (Vercel Postgres + Vercel Blob) as part of template installation
Configure environment variable for production domain in project settings
- NEXT_PUBLIC_SITE_DOMAIN (e.g., photos.domain.com—used in permalinks and seen in top-right nav)

2. Setup Auth

Generate auth secret and add to environment variables:
- AUTH_SECRET
Add admin user to environment variables:
- ADMIN_EMAIL
- ADMIN_PASSWORD
Trigger redeploy
- Visit project on Vercel, navigate to "Deployments" tab, click ••• button next to most recent deployment, and select "Redeploy"

3. Upload your first photo 🎉

Visit /admin
Sign in with credentials supplied in Step 2
Click "Upload Photos"
Add optional title
Click "Create"

Develop locally

Clone code
Run pnpm i to install dependencies
If necessary, install Vercel CLI and authenticate by running vercel login
Run vercel link to connect the CLI to your project
Run vercel dev to start dev server with Vercel-managed environment variables

Further customization

Experimental AI text generation

⚠️ READ BEFORE PROCEEDING

Usage of this feature will result in fees from OpenAI. When enabling AI text generation, follow all recommended mitigations in order to avoid unexpected charges and attacks. Make sure your OpenAI secret key environment variable is not prefixed with NEXT_PUBLIC.

Setup OpenAI
- If you don't already have one, create an OpenAI account and fund it (see this thread if you're having issues)
- Generate an API key and store in environment variable OPENAI_SECRET_KEY
- Setup usage limits to avoid unexpected charges (recommended)
Add rate limiting (recommended)
- As an additional precaution, create a Vercel KV store and link it to your project in order to enable rate limiting—no further configuration necessary
Configure auto-generated fields (optional)
- Set which text fields auto-generate when uploading a photo by storing a comma-separated list, e.g., AI_TEXT_AUTO_GENERATED_FIELDS = title, semantic
- Accepted values:
  - all (default)
  - title
  - caption
  - tags
  - semantic
  - none

Web Analytics

Open project on Vercel
Click "Analytics" tab
Follow "Enable Web Analytics" instructions (@vercel/analytics already included)

Speed Insights

Open project on Vercel
Click "Speed Insights" tab
Follow "Enable Speed Insights" instructions (@vercel/speed-insights already included)

Optional configuration

Application behavior can be changed by configuring the following environment variables:

Site meta

NEXT_PUBLIC_SITE_TITLE (seen in browser tab)
NEXT_PUBLIC_SITE_DESCRIPTION (seen in nav, beneath title)
NEXT_PUBLIC_SITE_ABOUT (seen in grid sidebar—accepted rich formatting tags: , , , , ,  )

Site behavior

NEXT_PUBLIC_GRID_HOMEPAGE = 1 shows grid layout on homepage
NEXT_PUBLIC_DEFAULT_THEME = light | dark sets preferred initial theme (defaults to system when not configured)
NEXT_PUBLIC_PRO_MODE = 1 enables higher quality image storage (results in increased storage usage)
NEXT_PUBLIC_STATICALLY_OPTIMIZE_PAGES = 1 enables static optimization for pages, i.e., renders pages at build time (results in increased project usage)—⚠️ Experimental
NEXT_PUBLIC_STATICALLY_OPTIMIZE_OG_IMAGES = 1 enables static optimization for OG images, i.e., renders images at build time (results in increased project usage)—⚠️ Experimental
NEXT_PUBLIC_MATTE_PHOTOS = 1 constrains the size of each photo, and enables a surrounding border (potentially useful for photos with tall aspect ratios)
NEXT_PUBLIC_BLUR_DISABLED = 1 prevents image blur data being stored and displayed (potentially useful for limiting Postgres usage)
NEXT_PUBLIC_GEO_PRIVACY = 1 disables collection/display of location-based data (⚠️ re-compresses uploaded images in order to remove GPS information)
NEXT_PUBLIC_HIDE_REPO_LINK = 1 removes footer link to repo
NEXT_PUBLIC_ALLOW_PUBLIC_DOWNLOADS = 1 enables public photo downloads for all visitors (⚠️ may result in increased bandwidth usage)
NEXT_PUBLIC_PUBLIC_API = 1 enables public API available at /api
NEXT_PUBLIC_IGNORE_PRIORITY_ORDER = 1 prevents priority_order field affecting photo order
NEXT_PUBLIC_HIDE_SOCIAL = 1 removes X button from share modal
NEXT_PUBLIC_HIDE_FILM_SIMULATIONS = 1 prevents Fujifilm simulations showing up in /grid sidebar and CMD-K search results
NEXT_PUBLIC_HIDE_EXIF_DATA = 1 hides EXIF data in photo details and OG images (potentially useful for portfolios, which don't focus on photography)
NEXT_PUBLIC_GRID_ASPECT_RATIO = 1.5 sets aspect ratio for grid tiles (defaults to 1—setting to 0 removes the constraint)
NEXT_PUBLIC_SHOW_LARGE_THUMBNAILS = 1 ensures large thumbnails on photo grid views
NEXT_PUBLIC_OG_TEXT_ALIGNMENT = BOTTOM keeps OG image text bottom aligned (default is top)

Alternate storage providers

Only one storage adapter—Vercel Blob, Cloudflare R2, or AWS S3—can be used at a time. Ideally, this is configured before photos are uploaded (see Issue #34 for migration considerations). If you have multiple adapters, you can set one as preferred by storing "aws-s3," "cloudflare-r2," or "vercel-blob" in NEXT_PUBLIC_STORAGE_PREFERENCE.

Cloudflare R2

Setup bucket
- Create R2 bucket with default settings
- Setup CORS under bucket settings:
```
[{
    "AllowedHeaders": ["*"],
    "AllowedMethods": [
      "GET",
      "PUT"
    ],
    "AllowedOrigins": [
       "http://localhost:3000",
       "https://{VERCEL_PROJECT_NAME}*.vercel.app",
       "{PRODUCTION_DOMAIN}"
    ]
}]
```
- Enable public hosting by doing one of the following:
  - Select "Connect Custom Domain" and choose a Cloudflare domain
  - OR
  - Select "Allow Access" from R2.dev subdomain
- Store public configuration:
  - NEXT_PUBLIC_CLOUDFLARE_R2_BUCKET: bucket name
  - NEXT_PUBLIC_CLOUDFLARE_R2_ACCOUNT_ID: account id (found on R2 overview page)
  - NEXT_PUBLIC_CLOUDFLARE_R2_PUBLIC_DOMAIN: either "your-custom-domain.com" or "pub-jf90908...s0d9f8s0s9df.r2.dev" (do not include "https://" in your domain)
Setup private credentials
- Create API token by selecting "Manage R2 API Tokens," and clicking "Create API Token"
- Select "Object Read & Write," choose "Apply to specific buckets only," and select the bucket created in Step 1
- Store credentials (⚠️ Ensure access keys are not prefixed with NEXT_PUBLIC):
  - CLOUDFLARE_R2_ACCESS_KEY
  - CLOUDFLARE_R2_SECRET_ACCESS_KEY

AWS S3

Setup bucket

Create S3 bucket with "ACLs enabled," and "Block all public access" turned off

Setup CORS under bucket permissions:

[{
 "AllowedHeaders": ["*"],
 "AllowedMethods": [
   "GET",
   "PUT"
 ],
 "AllowedOrigins": [
   "http://localhost:*",
   "https://{VERCEL_PROJECT_NAME}*.vercel.app",
   "{PRODUCTION_DOMAIN}"
 ],
 "ExposeHeaders": []
}]

Store public configuration
- NEXT_PUBLIC_AWS_S3_BUCKET: bucket name
- NEXT_PUBLIC_AWS_S3_REGION: bucket region, e.g., "us-east-1"

Setup private credentials

Create IAM policy using JSON editor:

{
  "Version": "2012-10-17",
  "Statement": [
    {
      "Effect": "Allow",
      "Action": [
        "s3:PutObject",
        "s3:PutObjectACL",
        "s3:GetObject",
        "s3:ListBucket",
        "s3:DeleteObject"
      ],
      "Resource": [
        "arn:aws:s3:::{BUCKET_NAME}",
        "arn:aws:s3:::{BUCKET_NAME}/*"
      ]
    }
  ]
}

Create IAM user by choosing "Attach policies directly," and selecting the policy created above. Create "Access key" under "Security credentials," choose "Application running outside AWS," and store credentials (⚠️ Ensure access keys are not prefixed with NEXT_PUBLIC):
- AWS_S3_ACCESS_KEY
- AWS_S3_SECRET_ACCESS_KEY

Alternate database providers (experimental)

Vercel Postgres can be switched to another Postgres-compatible, pooling provider by updating POSTGRES_URL. Some providers only work when SSL is disabled, which can configured by setting DISABLE_POSTGRES_SSL = 1.

Supabase

Ensure connection string is set to "Transaction Mode" via port 6543
Disable SSL by setting DISABLE_POSTGRES_SSL = 1

FAQ

How do I receive template updates?

For forked repos, click "Code," then "Update branch" from the main repo page. If you originally cloned the code, you can create a fork from GitHub, then update your Git connection from your Vercel project settings. Once you've done this, you may need to go to your project deployments page, click •••, select "Create deployment," and choose main.

How do I edit multiple photos?

On desktop, select ••• menu in the top right next to site title and choose, "Select Multiple." On mobile, "Select Multiple Photos" can be accessed from the search menu. From there, you can perform bulk tag, favorite, and delete actions.

Why don't my photo changes show up immediately?

This template statically optimizes core views such as / and /grid to minimize visitor load times. Consequently, when photos are added, edited, or removed, it might take several minutes for those changes to propagate. If it seems like a change is not taking effect, try navigating to /admin/configuration and clicking "Clear Cache."

Why don't my older photos look right?

As the template has evolved, EXIF fields (such as lenses) have been added, blur data is generated through a different method, and AI/privacy features have been added. In order to bring older photos up to date, either click the 'sync' button next to a photo or use the outdated photo page (/admin/outdated) to make batch updates.

Why don’t my OG images load when I share a link?

Many services such as iMessage, Slack, and X, require near-instant responses when unfurling link-based content. In order to guarantee sufficient responsiveness, consider rendering pages and image assets ahead of time by enabling static optimization by setting NEXT_PUBLIC_STATICALLY_OPTIMIZE_PAGES = 1 and NEXT_PUBLIC_STATICALLY_OPTIMIZE_OG_IMAGES = 1. Keep in mind that this will increase platform usage.

Why do vertical images take up so much space?

By default, all photos are shown full-width, regardless of orientation. Enable matting to showcase horizontal and vertical photos at similar scales by setting NEXT_PUBLIC_MATTE_PHOTOS = 1.

Why are my grid thumbnails so small?

Thumbnail grid density (seen on /grid, tag overviews, and other photo sets) is dependent on aspect ratio configuration (ratios of 1 or less have more photos per row). This can be overridden by setting NEXT_PUBLIC_SHOW_LARGE_THUMBNAILS = 1.

How secure are photos marked “hidden?”

While all hidden paths (/tag/hidden/*) require authentication, raw links to individual photo assets remain publicly accessible. Randomly generated urls from storage providers are only secure via obscurity. Use with caution.

My images/content have fallen out of sync with my database and/or my production site no longer matches local development. What do I do?

Navigate to /admin/configuration and click "Clear Cache."

I'm seeing server-side runtime errors when loading a page after updating my fork. What do I do?

Navigate to /admin/configuration and click "Clear Cache." If this doesn't help, open an issue.

Why are my thumbnails square?

Absent configuration, the default grid aspect ratio is 1. NEXT_PUBLIC_GRID_ASPECT_RATIO can be set to any number (for instance, 1.5 for 3:2 images) or ignored by setting to 0.

Why aren't Fujifilm simulations importing alongside EXIF data?

Fujifilm simulation data is stored in vendor-specific Makernote binaries embedded in EXIF data. Under certain circumstances an intermediary may strip out this data. For instance, there is a known issue on iOS where editing an image, e.g., cropping it, causes Makernote data loss. If simulation data appears to be missing, try importing the original file as it was stored by the camera. Additionally, if you can confirm the simulation mode, you can edit the photo and manually select it.

Why do my images appear flipped/rotated incorrectly?

For a number of reasons, only EXIF orientations: 1, 3, 6, and 8 are supported. Orientations 2, 4, 5, and 7—which make use of mirroring—are not supported.

Why does my image placeholder blur look different from photo to photo?

Earlier versions of this template generated blur data on the client, which varied visually from browser to browser. Data is now generated consistently on the server. If you wish to update blur data for a particular photo, edit the photo in question, make no changes, and choose "Update."

Why are large, multi-photo uploads not finishing?

The default timeout for processing multiple uploads is 60 seconds (the limit for Hobby accounts). This can be extended to 5 minutes on Pro accounts by setting maxDuration = 300 in src/app/admin/uploads/page.tsx.

I've added my OpenAI key but can't seem to make it work. Why am I seeing connection errors?

You may need to pre-purchase credits before accessing the OpenAI API. See Issue #110 for discussion.

Can this template run in a docker image?

Possibly. See Issue #116 for discussion.

Name		Name	Last commit message	Last commit date
Latest commit History 1,521 Commits
.github/ISSUE_TEMPLATE		.github/ISSUE_TEMPLATE
.vscode		.vscode
__tests__		__tests__
public		public
readme		readme
src		src
.eslintrc.json		.eslintrc.json
.gitignore		.gitignore
README.md		README.md
jest.config.mjs		jest.config.mjs
next.config.js		next.config.js
package.json		package.json
pnpm-lock.yaml		pnpm-lock.yaml
postcss.config.js		postcss.config.js
tailwind.config.js		tailwind.config.js
tsconfig.json		tsconfig.json

sambecker/exif-photo-blog

Folders and files

Latest commit

History

Repository files navigation

📷 EXIF Photo Blog