Conditionally load livereload.js only in development

[Copilot]

The base template unconditionally loads livereload.js, causing connection timeouts to port 35729 in production environments. This introduces UI delays as browsers wait for the non-existent livereload server.

Changes

• base.jinja: Wrap livereload.js script tag with config check for DevelopmentConfig
• test_livereload_conditional.py: Verify script inclusion behavior across configs

{% if config.SPECTER_CONFIGURATION_CLASS_FULLNAME and 'DevelopmentConfig' in config.SPECTER_CONFIGURATION_CLASS_FULLNAME %}
<script>document.write('<script src="http://' + (location.host || 'localhost').split(':')[0] + ':35729/livereload.js?snipver=1"></' + 'script>')</script>
{% endif %}

Development workflow with gulp-based livereload remains unchanged. Production, test, and other configs no longer load the script.

Original prompt

---

This section details on the original issue you should resolve

<issue_title>Unconditional livereload.js in base.jinja Causes UI Delays in ProductionConfig</issue_title>
<issue_description>In Specter Desktop v2.1.1, the base.jinja template unconditionally includes a livereload.js script, causing significant UI delays in production environments. The script requests http://:35729/livereload.js, which times out (1.3 minutes for login, 30 seconds for other pages) when no livereload server is running.

This occurs despite setting "livereload": false, dev: false, debug: false, and ProductionConfig in config.json, and enforcing Flask production mode (FLASK_DEBUG=0, WERKZEUG_DEBUG=0).

The issue stems from this line in templates/base.jinja:
"<script>document.write('<script src="http://' + (location.host || 'localhost').split(':')[0] + ':35729/livereload.js?snipver=1"></' + 'script>')<script>"

It lacks a conditional check (e.g., {% if app.config['LIVERELOAD'] %}), injecting the script into every page, even in ProductionConfig.
Steps to Reproduce

Install Specter Desktop v2.1.1 from source (git clone https://github.com/cryptoadvance/specter-desktop, git checkout v2.1.1, pip install .) in a Python 3.10 virtual environment.
Configure ~/.specter/config.json with:{
"config": "ProductionConfig",
"dev": false,
"livereload": false,
"debug": false,
"logfile": "/home/specter/.specter/specter.log",
"loglevel": "DEBUG",
...
}

Set up a systemd service with:ExecStart=/home/specter/.env/bin/python -m cryptoadvance.specter server --host 127.0.0.1 --config ProductionConfig
Environment="FLASK_ENV=production"
Environment="FLASK_DEBUG=0"
Environment="WERKZEUG_DEBUG=0"

Modify server.py to enforce:app.config['DEBUG'] = False
app.config['TESTING'] = False
app.config['LIVERELOAD'] = False

Start Specter (sudo systemctl start specter).
Open http://:25441 (e.g., http://raspibolt.local:25441) in a browser (e.g., Brave).
Check DevTools > Network: Observe a request to http://:35729/livereload.js that times out.
Measure page load times: ~1.3 minutes for login, ~30 seconds for other pages.

Expected Behavior

No livereload.js request in ProductionConfig when "livereload": false.
UI loads quickly (<5s for login, <2s for other pages).
base.jinja should conditionally include livereload.js only when app.config['LIVERELOAD'] or app.config['DEBUG'] is True (e.g., in DevelopmentConfig).

Actual Behavior

base.jinja unconditionally injects livereload.js, causing browser requests to http://:35729/livereload.js.
Requests time out (no server on :35729), delaying UI rendering.
Page load times: ~20s to 1.3 minutes (login), ~10-30 seconds (other pages).

Environment

OS: Ubuntu (RaspiBolt on Raspberry Pi 4)
Python: 3.10
Specter Version: v2.1.1 (installed from source, git checkout v2.1.1)
Virtual Environment: /home/specter/.env
Dependencies: sqlalchemy==1.4.50, flask-sqlalchemy==2.5.1, others per requirements.txt
Browser: Brave (also reproducible in Firefox)
Related Services: Bitcoin Knots 28.1 (127.0.0.1:8336), Electrs 0.10.7 (127.0.0.1:50001)

Logs

specter.log (excerpt):[2025-05-09 15:06:04,127] INFO in server: Configuration: cryptoadvance.specter.config.ProductionConfig
[2025-05-09 15:06:04,138] DEBUG in server: Flask DEBUG: False, LIVERELOAD: False
[2025-05-09 15:06:04,138] DEBUG in server: Loaded config: ProductionConfig

journalctl -u specter.service:
No errors; confirms ProductionConfig.

DevTools Network:
Request to http://raspibolt.local:35729/livereload.js with status 504 Gateway Timeout or stalled.

Full logs available: specter.log, specter_journal.log.

Workaround

Comment out the livereload.js script in base.jinja:
sudo nano /home/specter/.env/lib/python3.10/site-packages/cryptoadvance/specter/templates/base.jinja : line 16

Restart Specter:sudo systemctl restart specter

Result: No livereload.js requests, UI loads in <5s (login), <2s (other pages).

Suggested Fix

Add a conditional check in base.jinja:{% if app.config['LIVERELOAD'] %}

<script>document.write('<script src="http://' + (location.host || 'localhost').split(':')[0] + ':35729/livereload.js?snipver=1"> {% endif %} Alternatively, remove the script entirely for ProductionConfig, as livereload is irrelevant in production. Additional Context Issue persisted despite extensive troubleshooting: Pinned dependencies to resolve SQLAlchemy errors (sqlalchemy==1.4.50). Simplified config.json to RaspiBolt-recommended settings. No livereload server runs on :35729 (netstat -tulnp | grep 35729 empty), confirming the request is client-driven. Reproducible on RaspiBolt setup with Knots and Electrs integration. Attachments specter.log: Server logs showing ProductionConfig and debug settings. specter_journal.log: Systemd logs for Specter service. DevTools screenshot: Showing ...

- Fixes #2495 --- ✨ Let Copilot coding agent [set things up for you](https://github.com/cryptoadvance/specter-desktop/issues/new?title=✨+Set+up+Copilot+instructions&body=Configure%20instructions%20for%20this%20repository%20as%20documented%20in%20%5BBest%20practices%20for%20Copilot%20coding%20agent%20in%20your%20repository%5D%28https://gh.io/copilot-coding-agent-tips%29%2E%0A%0A%3COnboard%20this%20repo%3E&assignees=copilot) — coding agent works faster and does higher quality work when set up for your repo.

[netlify]

✅ Deploy Preview for specter-desktop-docs canceled.

Name	Link
🔨 Latest commit	4b46c13
🔍 Latest deploy log	https://app.netlify.com/projects/specter-desktop-docs/deploys/6984422e6e928f0008fb6116

[al-munazzim]

✅ LGTM - simple and correct fix.

The unconditional livereload.js was causing significant delays in production (1+ minute timeouts). Wrapping it in a DevelopmentConfig check is the right approach.

Minor note: Could also check config.DEBUG or a dedicated LIVERELOAD config flag, but checking the config class name works fine.

[k9ert]

@copilot resolve the merge conflicts in this pull request

[netlify]

✅ Deploy Preview for specter-desktop-docs canceled.

Name	Link
🔨 Latest commit	4f57fe2
🔍 Latest deploy log	https://app.netlify.com/projects/specter-desktop-docs/deploys/69e2869f29fbfc00084932c7

[Copilot]

@copilot resolve the merge conflicts in this pull request

Merge conflict resolved in commit 4f57fe2.

The conflict in base.jinja was resolved by accepting master's implementation which uses {% if config.DEBUG %} instead of the more complex configuration class name check. Master branch already fixed the same issue via PR #2543 using Flask's built-in DEBUG config, which is simpler and more idiomatic.

Also removed tests/test_livereload_conditional.py since the fix is now in master.