Club Wiki

30 views
Skip to first unread message

Michael Wimble

unread,
May 7, 2026, 6:29:20 PM (8 days ago) May 7
to hbrob...@googlegroups.com
Well, so far it’s looking like there really is almost no interest in a club Wiki. Unless that changes, I’ll probably close the github repo I wrote about and just make it available to the few people who replied. For me, this would end the argument about whether the club should to build a wiki.

Stephen Williams

unread,
May 7, 2026, 8:44:57 PM (8 days ago) May 7
to hbrob...@googlegroups.com, Michael Wimble

I still think it's a good idea.

Stephen

Karim Virani

unread,
May 7, 2026, 8:55:28 PM (8 days ago) May 7
to hbrob...@googlegroups.com
lol, less than 48hrs ago you wrote:  YOU HAVE ONE WEEK TO PROVIDE FEEDBACK — GET TO IT.

even the 1 week turn-around is problematic for some of us - this landed as a pleasant surprise two days ago and i've got a loaded schedule going weeks out. so i may not be able to engage deeply with the idea even though it's warranted by your expertise.

as an itinerant/visiting member, i can't even speak to the needs of the group - but i don't think everyone is on the ROS2 bandwagon. i'm way too new to be aware of the need for a club wiki. whether the club needs a wiki and this is the correct focus - these can be thought of as fully independent from whether what you have produced is a valuable resource that others may want ongoing access to. 

the primary worry has to do with your "big caveat" - that this is largely AI-produced content and you wrote: "I have not yet read more than a fragment of all of the content." that last disclaimer is my main issue. i use AI as much as the next guy and it's extremely valuable - and your decision to accept a lot of it (likely as an interim bootstrap for this process) just points out how good it has gotten and how trustworthy it can be at its best. but, when i mix AI generated content into my own work product - i still need to read and fully agree with it before i publicly publish it - and that almost always includes post-generation corrections.

so i haven't engaged with it deeply, but i have skimmed substantial parts of the wiki, and it's clear to me that it is documenting the right questions being asked of frontier models by an expert in the field, with an eye toward getting beyond the beginner stages of ROS. to that extent i'm very interested in engaging it when i'm up against the boundaries of my slowly growing knowledge. but i'm more interested in the questions you asked, the guidance you gave, and the dragons you avoided, as you engaged with the LLM(s). as all documentation projects are - they are a frozen in time (until updated), short half-life view of the world. and freezing an AI generated slab of content seems less than useful as they continue to get better, and sharper. what i seek to develop more is the raw knowledge and the AI interaction skills that will get me the cleanest, most useful answer 6 months from now. this wiki might be a good resource toward that end. but if you're going to go all in with AI, think about how to put the maintenance burden on them as well.

that's my 2 pfennig.

for reference - the original thread: 

ps. if it's not clear, yes i'd like to engage, whether it's semi-private or public, but can't do much right now.


On Thu, May 7, 2026 at 5:29 PM Michael Wimble <mwi...@gmail.com> wrote:
Well, so far it’s looking like there really is almost no interest in a club Wiki. Unless that changes, I’ll probably close the github repo I wrote about  and just make it available to the few people who replied. For me, this would end the argument about whether the club should to build a wiki.

--
You received this message because you are subscribed to the Google Groups "HomeBrew Robotics Club" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hbrobotics+...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/hbrobotics/7649BB3F-9C25-4F0C-B274-87629D216F01%40gmail.com.

Thomas Messerschmidt

unread,
May 7, 2026, 9:02:49 PM (8 days ago) May 7
to hbrob...@googlegroups.com, Michael Wimble
Ditto. 

--
You received this message because you are subscribed to the Google Groups "HomeBrew Robotics Club" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hbrobotics+...@googlegroups.com.

Michael Wimble

unread,
May 7, 2026, 10:01:16 PM (8 days ago) May 7
to hbrob...@googlegroups.com
You’re right. Bad day in my life. I apologize. 

On May 7, 2026, at 5:55 PM, Karim Virani <ponder...@gmail.com> wrote:



James H Phelan

unread,
May 10, 2026, 1:37:05 PM (5 days ago) May 10
to hbrob...@googlegroups.com

Michael,

Something to recommend for all technical manuals - 

All acronyms need to be spelled out at least at first use.  Knowing what a bunch of initials stands for goes a good way to understanding what it is.

And give a link to a simple explanation of what it does and how it is different from its brethren.  The Distilled Takeaways and Selection Guides do a good job of this.

I was looking a "Choosing a Nav2 Controller" and saw "DWB" Controller Basics and Tuning.  Interesting.  What's a DWB controller?  Nowhere is it spelled out.  A prolonged Google search told me it was the 'B' version  of the DWA Controller.  Fine.  What's a DWA Controller?  Further search revealed it's a Dynamic Window Approach.  Which led me to a YouTube video that was a bunch of math.  But gathered that is was some sort of collision avoidance algorithm.  OK, fine.

James H Phelan
"Nihil est sine ratione cur potius sit quam non sit"
Leibniz

Michael Wimble

unread,
May 10, 2026, 1:48:26 PM (5 days ago) May 10
to hbrob...@googlegroups.com
Excellent suggestion. I also hate that everyone assumes you know what their acronym means.

--
You received this message because you are subscribed to the Google Groups "HomeBrew Robotics Club" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hbrobotics+...@googlegroups.com.

Karim Virani

unread,
May 10, 2026, 2:24:14 PM (5 days ago) May 10
to hbrob...@googlegroups.com
we can't predict the path ppl will take entering a wiki and skimming to the meaty portion they are looking for. so can't really rely on the first-use definition. but that's why wikipedia has hover popups for "glossary" type items. unfortunately github wikis don't support this, though simple tooltips are possible. you'd probably want a script to generate such tooltips from some common glossary page.

Reply all
Reply to author
Forward
0 new messages