This blog is also a living document for myself so I can improve and reference this working pattern in the future.
Model selection
If using ChatGPT for coding tasks, especially maintenance tasks, opt for o1, other models are crap and will hallucinate or forget more of the original code.
Avoiding regressions
If the goal is to alter existing code, ChatGPT risks breaking existing code by removing as part of the end result, a good prompt is something like below, to reduce the amount of feedback loop and future debugging.
[insert-here your own description of your problem to be solved by ChatGPT/LLM, with instructions]
Here is the code to be modified. I want the complete code as the final result, with the same number of functions as in the original.
First, tell me the list of ALL the functions you see, and which ones you will modify. Show me all the resulting code from each modification. Show me the final file with the modifications. Do not omit code for brevity, keep untouched code same as before.
[Insert here the original code to be maintained/modified/debugged by the LLM]
TLDR: In December 2023 I picked this low-cost VPS provider. I do not recommend them, and I hope this writing will save people from wasting any time or money on Contabo.
They claimed "German Quality, Always". It's neither good quality, neither always.
For the most part, from December 2023 to September 2024, I could almost live without caring too much about managing my server, whose setup is mostly delegated to Cloudron. And I had no problem using Cloudron for my web apps on this cheap VPS; my Cloudron backups were also configured in Cloudron to use Contabo Object Storage. No problem on the surface. Until Murphy's law enter the game:
An outage in September 2024 broke availability in Nuremberg Data center and so my backups and my Cloudron instance were unavailable. I entered panic mode because there was nothing I could see or do. Contabo offered no support because their customer portal was impacted as well and no fail-over mechanism had been put in place. That tells a lot. Fortunately this was solved after a few days.
Quickly after this outage resolution, I noticed too that rclone and Cloudron backups were acting weird, mostly due to the slowness or rate limiting of Contabo Object storage API. I couldn't find an explanation so I just contacted Contabo support, which fixed my problem without saying much, sadly.
Few weeks later still in September 2024, my DNS config on my VPS starts to cause issues, but only to the light of problems encountered in WordPress, Miniflux and Changedetection. All those issues exploded to my face in the form of timeouts or unreachable URLs. Fortunately I found this write up https://www.thomasmartens.eu/contabo-dns-lookup-timeouts/ and could just fix my default NetPlan configuration and switched from Contabo DNS to 1.1.1.1 (CloudFlare) and 9.9.9.9 (Quad9).
I noticed too that my object storage space was filling up constantly, causing me some anxiety. And I couldn't clean anything. The backup cleanups were also silently frequently in Cloudron as a result. In the end I couldn't find the root cause but switching to Hetzner storage instead of Contabo Object Storage fixed and made the whole process faster.
Unfortunately I went through those issues. But I could have avoided them. When one have to pick a product or service, it's better to do more research and ask around for advice.
Use ChatGPT to be brutally honest when giving you advice and saving you money and pain.
Or just visit forums and communities like Reddit;
Lessons learned
Some products and services are cheap for reasons: quality for free is often a scam.
I should have checked internet before. Be more picky, check the community forums and reviews. Avoid unneeded pain.
Do not put all your eggs in the same basket, i.e pick different locations and providers for your compute and data.
Monitor the health of your backups and storage location.
Both are in the same region. The amount of CPU and disk space vary, but overall there is a clear winner about performance.
-------------------- A Bench.sh Script By Teddysun -------------------
Version : v2024-11-11
Usage : wget -qO- bench.sh | bash
----------------------------------------------------------------------
CPU Model : AMD EPYC Processor (with IBPB)
CPU Cores : 6 @ 2799.996 MHz
CPU Cache : 512 KB
AES-NI : ✓ Enabled
VM-x/AMD-V : ✗ Disabled
Total Disk : 593.5 GB (53.7 GB Used)
Total Mem : 15.6 GB (4.9 GB Used)
Total Swap : 4.0 GB (268.2 MB Used)
System uptime : 6 days, 17 hour 31 min
Load average : 0.67, 0.85, 1.09
OS : Ubuntu 22.04.3 LTS
Arch : x86_64 (64 Bit)
Kernel : 5.15.0-124-generic
TCP CC : cubic
Virtualization : Dedicated
IPv4/IPv6 : ✓ Online / ✗ Offline
Organization : AS51167 Contabo GmbH
Location : Nürnberg / DE
Region : Bavaria
----------------------------------------------------------------------
I/O Speed(1st run) : 131 MB/s
I/O Speed(2nd run) : 98.0 MB/s
I/O Speed(3rd run) : 94.4 MB/s
I/O Speed(average) : 107.8 MB/s
----------------------------------------------------------------------
Node Name Upload Speed Download Speed Latency
Speedtest.net 399.41 Mbps 399.48 Mbps 11.37 ms
Los Angeles, US 378.05 Mbps 401.30 Mbps 155.48 ms
Dallas, US 390.21 Mbps 402.25 Mbps 127.98 ms
Montreal, CA 44.98 Mbps 400.75 Mbps 105.08 ms
Paris, FR 404.24 Mbps 397.38 Mbps 18.16 ms
Amsterdam, NL 392.33 Mbps 396.11 Mbps 23.08 ms
Beijing, CN 275.61 Mbps 410.68 Mbps 241.41 ms
Shanghai, CN 129.99 Mbps 362.16 Mbps 428.46 ms
Hong Kong, CN 224.22 Mbps 386.99 Mbps 284.37 ms
Singapore, SG 93.82 Mbps 362.53 Mbps 329.19 ms
Tokyo, JP 302.55 Mbps 393.36 Mbps 256.67 ms
----------------------------------------------------------------------
Finished in : 5 min 43 sec
Timestamp : 2024-11-13 09:18:58 UTC
----------------------------------------------------------------------
-------------------- A Bench.sh Script By Teddysun -------------------
Version : v2024-11-11
Usage : wget -qO- bench.sh | bash
----------------------------------------------------------------------
CPU Model : Intel Xeon Processor (Skylake, IBRS, no TSX)
CPU Cores : 4 @ 2099.998 MHz
CPU Cache : 16384 KB
AES-NI : ✓ Enabled
VM-x/AMD-V : ✗ Disabled
Total Disk : 75.0 GB (3.5 GB Used)
Total Mem : 7.6 GB (632.2 MB Used)
System uptime : 39 days, 14 hour 22 min
Load average : 0.03, 0.02, 0.00
OS : Ubuntu 24.04 LTS
Arch : x86_64 (64 Bit)
Kernel : 6.8.0-41-generic
TCP CC : cubic
Virtualization : Dedicated
IPv4/IPv6 : ✓ Online / ✓ Online
Organization : AS24940 Hetzner Online GmbH
Location : Nürnberg / DE
Region : Bavaria
----------------------------------------------------------------------
I/O Speed(1st run) : 846 MB/s
I/O Speed(2nd run) : 795 MB/s
I/O Speed(3rd run) : 929 MB/s
I/O Speed(average) : 856.7 MB/s
----------------------------------------------------------------------
Node Name Upload Speed Download Speed Latency
Speedtest.net 2292.48 Mbps 2473.56 Mbps 0.66 ms
Los Angeles, US 605.99 Mbps 2986.01 Mbps 157.65 ms
Dallas, US 702.97 Mbps 3420.20 Mbps 128.18 ms
Montreal, CA 175.26 Mbps 937.60 Mbps 90.89 ms
Paris, FR 7119.88 Mbps 9681.24 Mbps 13.65 ms
Amsterdam, NL 4993.84 Mbps 7353.62 Mbps 10.41 ms
Beijing, CN 627.18 Mbps 2611.88 Mbps 187.03 ms
Shanghai, CN 1.34 Mbps 1791.62 Mbps 325.14 ms
Hong Kong, CN 352.36 Mbps 2526.89 Mbps 234.43 ms
Singapore, SG 181.44 Mbps 670.17 Mbps 194.45 ms
Tokyo, JP 313.71 Mbps 2686.38 Mbps 236.80 ms
----------------------------------------------------------------------
Finished in : 5 min 48 sec
Timestamp : 2024-11-13 09:25:24 UTC
----------------------------------------------------------------------
Conclusion and what's next
I still have my main VPS at Contabo, however my storage and backups are with Hetzner, migrating to Hetzner should be a no brainer, thanks to the built-in quality of Cloudron.
I've also configured another VPS with Hetzner, with satisfaction, for running Gitea Actions, because there is no way I rely on GitHub.
So you know. Move away from them, unless for cheap, non business critical workflows where performance and reliability are not a concern.
I'm addicted to Changedetection for spying on website changes and internet search results for specific keywords, Occasionally also for monitoring price changes. It's quite handy to discover new links added to web directories, or stay updated with some websites that do not provide any RSS feed.
Context
I'm watching hundred of URLs.
I often spy on webrings and blogrolls to discover new interesting links, and also on search engines results for specific keywords.
I'm self-hosting Changedetection through Cloudron.
I'm mostly following through those watches via my RSS Reader, Miniflux.
For some specific changes, like weather bad conditions, I subscribe via ntfy.
Anyway, I've developed a few habits that fit my workflow so well for every new watch, which are:
Settings > General
This is where we set defaults for all future watches, it's pretty obvious you must start here. Here is my current setting:
Time between check: By forcing a convenient interval between checks, you try to find a balance between information overload and staying current. Pick your poison, but don't hesitate to override this setting at per-watch level.
Extract from document and use as watch title: it's convenient to let Changedetection take care of naming your watches based on the webpage titles rather than leaving the sometimes very long and non human-friendly URL as a default description.
Random jitter: this is handy to avoid stressing your I/O too much.
General > Group tag
This one is mostly for better organizing stuff, as I mentioned I follow those changes through RSS, I noticed it was harder to distinct between important and less important stuff because I was following the default RSS feed, but Changedetection provides distinct RSS feeds per groups/tabs of watches, and that's my preferred workflow now.
I'm trying to always set a label, I have around 15 in total, some for specific interests (privacy, discovery aka list of links, devops, music, ...) or specific people, locations and business updates. The rest is generally less important and is labelled with things like FOMO, misc, ...
Those group tags appear as labels next to the URLs you are watching.
If you want to watch a whole group through RSS, link is at the bottom right of the page on the group tab.
Filters & Triggers > Remove elements
It's common on bloated rich web pages to want to focus on specific parts, like everything between <header> and <footer> sections, so I sometimes have to add footer and header. It's mostly needed for sites like eBay, 2ememain, where we can buy and sell things.
Filters & Triggers > default filter and triggers
This is purely for spam reduction as I mostly want to know when something new is made.
Sometimes I also enable Sort text alphabetically depending how the page is updated by its author.
🆕 Those new settings have been added recently and I'm also enabling them on new watches:
Extension
Try the web browser extension for Chromium based browsers, it makes watches one-click away.
Next
I've opened a discussion in Changedetection's repository to talk about how repetitive it feels to me, in the hope we can see something like template settings be proposed in the future, at least for the filters & triggers which I consider is not too hard to start with.
I'm following Joy of Tech comic via RSS in Miniflux but the image was never loading.
I found half a solution on this blog post of Jan-Lukas Else, unfortunately the proposed solution fails probably as a consequence of some changes in the format of Joy of Tech pages.
The fix is quite simple actually. Edit the feed settings, set the scraper rules to the following:
p.Maintext > img[src$=".png"]
And of course enable "Fetch original content" in the feed options.
Kissed Google Keep goodbye and embraced Obsidian even more, thanks to the Importer plugin.
Trying to escape the WhatsApp surveillance state. I'm axing useless groups left and right.
Scrubbing my old web presence. It's like digital housekeeping.
Using Syncthing now. Real-time sync across devices without cloud middlemen. Dropbox, you're on notice.
Deploying FDUPES for disk decluttering – it's a duplicate file slaughterhouse. Throwing inotifywait into the mix for smart folder monitoring, because who likes manual mess management?
Cut down on RSS. I've dropped Wallabag (done by November 2024). I'm still using Miniflux. and Shaarli with their weaknesses. Thinking about it. Bookmarking tools still suck somehow and I can't see a better alternative (yet) for my needs... yuck, so I'm working on my own thing, Tapas. Looking at the market for knowledge and bookmarks management tools, there is room for improvement in how we manage and consume information. Most of the hard work is on you for years with tools that connect to information.
Harnessing RSS-Bridge and Miniflux for streamlined info feeds. Using Changedetection for the unRSS-able stuff, i.e to monitor some indexes, lists, legal terms, release pages.
My tab hoarding was legendary, now capped at 18 with Tab Limiter. Browser zen achieved.
I have been known by my colleagues and partner to keep too many tabs open. My nerves cracked at reading other folks suffering same issue. So I decided to close a number of them, and limit each Window to 18 tabs with Tab Limiter.
Exploring and creating.
Blogging's up, but it's a discipline game. Need to turn Obsidian hoarding into public wisdom. Notebooks over phones, knowledge over scrolling.
Taming my Brave extension zoo with Context. It's like a digital bouncer for my browser. Funny, now I've more UX/Privacy oriented extensions than tabs.
Eyeing Geeqie to outsmart duplicate photos. Even my pixels need to be minimal.
cron jobs executing php scripts on a Cloudron LAMP instance is new digital butler, still a bit rough around the edges. Coding my own automation magic because their recipes are just appetizers for my needs.
Diving back into Rust. Cooking up something for productivity and knowledge management. Stay tuned.
And of course, some snowballs and video gaming to keep things balanced and fun.
