This is an article that was written in November 2025 about a well-aged IED that was discovered in a backyard during landscaping at an address in Santa Rosa in January 2015. The gist of the write up is evidently to push speculation as to whether this could, in fact, have been The Zodiac's 'Buss Bomb' in situ. As to its size and design, it doesn't closely match anything from the drawing and description he gave in his April 1970 letter, but it was found in an area that was largely rural at the time in question, according to Tom.

The article also mentions the presumably well-known fact of bomb-making equipment having been found during a search of Allen's address in Fresno Street, Vallejo.

Although not claiming to be able to match the relevance of either of the above discoveries, I did come across a brief article from June 2011 that also cited "vintage explosives" having been discovered at a rented property in San Francisco, with a little more context added by NBC Bay Area here.

It may be that there are other such explosive finds around the Bay Area since 1970 and, if anyone has any stories or articles of this nature, perhaps you could post something on this thread?

Worth looking into. I hear there have been some tips called in, too.

Anyways, that's not even the fun part; I'd like to propose that the Doodler and the Zodiac Killer are one and the same. I have a question to start this off: why the ever-loving heck are the Zodiac's ears a different color than his face in the Presidio sketch? My theory is that the Zodiac is a black man, who wore white makeup as a disguise (note that the prime suspect in the case also did not need corrective lenses, as per his driver's license), since we know the Zodiac Killer loved to dress up, but left the ears alone. To mess with "the pigs," probably. He liked that.

Take it with a grain of salt. @zegalba on Tumblr and it looks like he has an Instagram, too. May as well do some digging.

"If Baber’s interpretation is correct, the Zodiac killer’s taunting letter and Z32 cipher pointed to his first potential victim - Short, Margolis’s former lover who was murdered two decades before the serial killing spree in the San Francisco Bay Area began.

However, cryptography experts who previously verified Baber’s solution to Z13 - and confirmed to the Daily Mail that they believe those findings to be correct - have been unable to confirm his solution to Z32.

Ed Giorgio, the former Chief US Codemaker and Chief US Codebreaker at the National Security Agency (NSA), described the finding as ‘plausible’ and ‘consistent with other people's theories’ about the map and coordinates, but said it lacked cryptographic evidence.

‘There’re two parts to it: the projection of the coordinates and the map. So all that is very intriguing,’ Giorgio told the Daily Mail. ‘But the issue for us is how do you decrypt those 32 characters?’

He has no code, no whitepaper.  The way to decrypt those 32 characters is to test all plausible combinations of radians/inches, then rank the crytographic survivors by proximity to known Zodiac activity.  That will yield IN THREE AND THREE EIGHTHS RADIANS TEN as the top ranked candidate solution, due to it's proxiimity to the Blue Rock Springs attack site (1.15 miles).  When you check that location on the map you will see a big triangle landmark on the ground (see my profile pic).

Ultimately it seems like this solution is designed to create a connection between the Zodiac and Dahlia case.  There isn't much substance to it, although it is being reported globally.  The narrative does well to achieve clicks, in that regard it is a good attempt.

We don't have even this much currently for the Z32. That said, however, it is still possible to propose the 'form' of the original artifact in this case also. Essentially, but by way of example only, uncomplicated logic must have us believe it would have been nothing any-the-less concrete and tangible than this:

But here's the thing - and this is where most attempts go astray right from the start. We have to first ask where it is that we are searching for our bomb location using what we have been given, and the ubiquitous answer to this appears to be "on the ground." Of course this is, however, not the case at all.

Surely, were we to one day reach the actual solution, we will have a point identified on the solid earth. That said, however, our process of searching involves only our proposed cipher solution and the Phillips 66 map.

In and of itself, the map we are given is very poorly detailed and sloppily scaled. Despite the almost universal assumption that a translation of any proposed solution into lat/long or GPS or any other coordinates on the ground will match the equivalent point as marked on the map with the same bearing and distance from Mount Diablo, it won't!

Even a cursory attempt to create a semi-transparent overlay of the Phillips map on top of the equivalent terrain on, say, Google Maps, will show just how badly the attempt fails. And, in 1970, we didn't have Google Maps either. Basically, to be authentic in our search we just need to step back from the modern day, close down our Maps application, and stick with the Phillips map that we were actually given.

There are, essentially, two possibilities for what we are looking for. Firstly, it may be the case that all we have been given in the cipher is the relevant coordinate pair for the feature we are guided to identify or, secondly, we are directed to a much more general area and given an additional clue as to the specific ground feature sought.

In the first instance, the map must give us everything. The feature looked for must be actually present on the map. Of course, there are very few features actually shown - the bomb site must be somewhere along a pretty major highway or main road - and thus this realisation alone must cause us to give especial consideration to the second possibility. In the latter instance, then, we require something extra, and very likely the last characters of the cypher would give us this further clue.

For example, suppose the intended location was back at Blue Rock Springs. With very little precision - since the map doesn't permit precision of any kind - we might find that the decoded coordinates point from Mount Diablo over Benicia toward eastern Vallejo in general. That, in itself, wouldn't give us much at all. But suppose something like 'LAZULUM' were added at the end, then now - and only now - we can pin down something not immediately present on the map.

The point is, however we are to proceed, we simply have to get away from thinking about coordinates, bearings, distances on the ground. If that had been the intention, we would have either been given a better map or, more likely, none at all. Essentially, in this one instance, if not in others, Google is actually your Nemesis, not your friend.

I wanted to share some of my findings regarding my work on Z13.

To begin, I must address the unicity distance problem, and the fact that multiple possible solutions exist for this cipher.

To account for this, I decided that I needed a list of all adult male names that lived in Vallejo between the years 1968-1969.  The list of possible names must be constrained to people that we know lived in the relevant area, during the relevant time period.  

I was able to obtain a scanned copy of the Vallejo 1968-1969 city directory, and OCR'd all of the names, and addresses.

I then augmented this list with common nicknames.  For example a name list as a "Robert Smith" in the city directory may also lead to entries for Bob Smith, Rob Smith, ect.  

I then took this list of names, and created another list that formats them differently.  For example, a name such as Robert A. Smith, might be tested in the forms Robert A Smith, Rob A Smith, Rob Smith, R Smith, ect.  We are testing all plausible ways of writing the name.

With this final list, names that do not meet the symbol repetition patterns of the cipher are structurally rejected.  This greatly narrows the list.  Then come the positional constraints of the cipher text symbol.  These are incredibly restrictive, rejecting every name in the city directory.

And so, I tested all of the names that meet the symbol frequency constraints using as many basic transposition steps as possible.  Now I am starting to get very close matches on a small subset of names.

However, I did not find a perfect match.

There was a single name in the city directory, whose name was a nearly perfect match on Z13, using both his nickname, and his given name.  I did research this person, and they worked for the city of Vallejo, on public infrastructure.  They played JV football for Vallejo High School.  Their obituary did not mention family members, and they passed away in Vallejo a few years ago.

I will not name them (unethical), and I do not consider them to be the Zodiac.  Everyone is innocent until proven guilty, and his name was not a perfect match on Z13, just one character off after a basic transposition. 

The only thing that would prove a Z13 solution is a DNA and/or handwriting match that is verified by an unbiased, qualified expert.  They will also need to have lived in the area during the right time (bonus points if they left around 1974).  They would need to resemble the sketch, and be within the age range of witness descriptions.

I wanted to share my work on the cipher, even if I am not presenting a solution.  I do have a copy of the city directory for those that want to do their own research.  You can send me a private message if you would like it.

I am sharing my whitepaper that details a computational methodology to solving Z32.  The code is also publicly available.

Paper:  https://zenodo.org/records/18645227

Code: https://github.com/dstampher/zodiac-z32-cipher

I reported the decoded coordinates to the authorities, and recommend a ground penetrating radar survey of the triangle anomaly.  

I am happy to discuss the computational approach and the findings described in the paper.

-David