Fossil Data Part 1: Paleontologic Data Fossilized on IBM 8” Floppies

We meet scientists with a 300 Million year old data problem in need of retro-computing help. What it can tell us about our distant past is amazing.
Our sponsor for PCBs: www.pcbway.com
Support the team on Patreon: / curiousmarc
Buy shirts on Teespring: teespring.com/stores/curiousm...
Learn more on companion site: www.curiousmarc.com
Contact info: kzbin.infoa...

[@justinthehedgehog3388]

For anybody interested, Tom Phillips is interviewed and gives a demonstration of his coal ball procedure in Ep01 of the 1989 BBC documentary: "Lost Worlds, Vanished Lives", hosted by David Attenborough. It still can be found on DVD.

[@stevesloan7132]

The original professor was a pure or basic researcher producing a massive data base which later researches could add to, analyze, make sense of, and use. This was in fact his life's work - producing a database which he could hand-on to future researchers. This was indeed a noble effort and one of the reasons that I believe that pure researchers are the unsung heroes of science. Your teams effort to recover this man's trove of discovery brought tears to my eyes. Thank you all so much for your efforts and for filming the effort and sharing it here.

[@CuriousMarc]

Well sort of - this is not quite an accurate representation. I was told he used portions of the data, the summaries we see here, to publish many articles and make important discoveries during his lifetime. But he was never able to work with the dataset in its entirety as we could far more easily do today. It was big data way before its time.

[@1funnygame]

Crazy that somebody can spend decades painstakingly recording invaluable data, only for the next guys to have no ability to use it

[@Damien.D]

The painful truth of obsolescence.

[@macguyver209]

Why I still like paper/books... No batteries/tech required...

[@andersforsgren3806]

​@@macguyver209 To some degree I do so also, but images are digital and now the computer caught fire in December I had another blow to my research. Most is saved, but I did loose some - as I do not have records on what I save on laptop or on the main computer my backups turned out to be less than complete.

[@gaspardeelias2485]

This reminds me of Tycho Brahe, that recorded every position of every celestial object for years. Data that was used later by Kepler to find the Kepler laws that govern the solar system

[@johnps1670]

Computer gear is not for conservation. You better use acid free paper or clay tablets.

[@merseyviking]

Ben's excitement when he starts talking about the discs is amazing. As an archaeologist, this is all too early for me, but as someone who loves reverse engineering file formats, I really can't wait for the next episodes!

[@PistonAvatarGuy]

6:43 - "Except to us." ... I was thinking very much the same thing. I *LOVE* this, it reminds me of the kind of things that could be seen on PBS back in the '90s.

[@15743_Hertz]

Would it be neat to find out that they might have used COBOL to analyze the coal balls? Nah! Probably Fortran.

[@edgeeffect]

I just came down here to comment that... and found that you already had!

[@AndyHullMcPenguin]

It might even be Snobol... en.wikipedia.org/wiki/SNOBOL

[@JohnDlugosz]

Cute and a funny pun, but nothing to do with irony.

[@15743_Hertz]

@@JohnDlugosz Corrected. Thanks!

[@edgeeffect]

@@AndyHullMcPenguin No, because this is the Carboniferous era, that's aeons after the Snowball-Earth en.wikipedia.org/wiki/Snowball_Earth ;) SNOBOL was always one of those languages that I heard about at school and college in the early 80s (where our comp. sci. syllabuses were rather "behind the times") but never actually saw... thanks for the Wikipedia link that's an important bit of history I'm caught up on now.

[@Tedd755]

3:12 "Well, their wingspan was... up to two feet" Gee wee!

[@leqin]

So nothing like as annoying as Scottish Midges :.)

[@MikeBramm]

Cool, a cliff hanger. I can't wait to see what you find.

[@PapasDino]

Break out the popcorn, this is gonna be a great ride!

[@Strothy2]

I really hope you will make a follow up once they established the database, and what they learned from it, that would be so awesome!

[@rearusato7545]

Another fantastic series unfolds! Thank you Marc & Co for your efforts! Just goes to show the importance of not only documenting and restoring old equipment but also keeping some in working conditions. How else would you retrieve the data off those floppies? Keep them coming!

[@ekaa.3189]

Oh man, I wish I'd known about the professor's data analysis needs. My first data analysis programs, when I was a preteen, were for statistical analysis of my mom's thesis data set. A few mega bytes in size. We first used university computers, then later ones we bought. That was the mid 70s. Oh, I should mention I now have two 8" floppy drives that I've successfully read data from using a reconfigured PC floppy controller under Linux. I now use them as bookends.

[@sydneybiscuit]

I need more! Shoot I'd love to help covert those data to a modern sql database with a front end for public use. That would be such a fulfilling endeavor. I could imagine at some point creating and storing individual digital images of the samples and linking them to the dataset with a visual overlay. That would be neat

[@Echin0idea]

I wonder if the slices were close enough together to do tomography on (i.e. reconstruct the 3D structures from the 2D slices)? Having a fully annotated 3D model of these would be even more amazing than having the 2D images.

[@digitalshackonthelane]

@@Echin0idea Do you mean like a digital "staining" type process? Like you would do say for lab slides, only digitally?

[@Echin0idea]

​@@digitalshackonthelane Yeah, that's a pretty good analogy, though something like this could go further than that. They have the slices, and will hopefully have the data for what's at each grid square. They also mentioned that there is data on which structures on other slices are connected to each grid square. Depending on the quality of that data, that would potentially allow reconstructing complete structures in 3D, using images from multiple slices, and then annotate these structures with what they are, since the manual labour to classify and connect them has already been done. You could then ask questions of the dataset like: "show me the volume distribution of stems from x species" or "cluster rootlets from y species by degree of branching and show me some typical examples of different branching types". How much you could get from it depends very much on the quality of the slices and the accuracy of the annotations, but it could potentially be a really powerful dataset.

[@digitalshackonthelane]

@@Echin0idea Yes exactly! That would be so exciting to be a part of!

[@digitalshackonthelane]

@@Echin0idea I could not agree more! I would totally be stoked to work on something like that!

[@chuckinwyoming8526]

I guess I am just as much an odd ball, I still have a working 80286 machine with 3.5" 5.25" and two 8" floppy drives....but it looks like a BUNCH of floppy disks to go through. That one cardboard box of 256K byte floppy disks would fill part of a DVD disk or take several seconds of high speed internet to transmit. Hope the oxide has not broken down. Some of the older media has proven not too stable.

[@simontay4851]

They'll find a way. Even if they have to manually correct errors in the binary. Bit by bit.

[@Diamonddavej]

In the Arigna coal mines in Ireland, the miners hated coal balls because they could fall from the roof of the mine injuring or even killing miners.

[@amylaughnan2410]

I was present when all of this original work was completed in the Paleobotanical Laboratory at the University of Illinois-Champaign between 1966-1975. Dr. Phillips was like a second Father to me; took care of me when I was in the lab along with my parents. He later taught me when a was a grown up Lady. I was born at the University of Illinois on April 3, 1967. Needless to say he was in my life from my birth until his death in 2018. I will forever love you Dr. Phillips, and I miss you, and Mom, and Matt. You all were the best of times in my life. All other experiences shrink by comparison to those days in the lab, those long days in the lab, and Trenos, and Jesus Christ Superstar.. You, and my parents are etched in the deepest parts of my heart for eternity; as old as our beloved rocks; even older. God blessed you all. Love, your precious Amy.

[@hedgehog3180]

Stuff like this could end up being genuinely revolutionary. This kinda data about ancient environments is rare to find and we might be able to learn some incredible things about ancient environments because of this. Once modern techniques like machine learning are applied to this who knows what we'll learn.

[@beefchicken]

I grew up in coal mining country, my parents have coal balls decorating their garden. I had no idea they contained this stuff.

[@hanniffydinn6019]

I always save my data in the most obscure old formats so future generations can have great adventures figuring out how to recover the data ! 🤯

[@TheCondoInRedondo]

Those look like old Displaywriter diskettes, circa 1981. The Displaywriter was actually an IBM PC before there was an IBM PC. It was a dedicated word-processor device (8086 CPU instead of 8088) from which Don Estridge's crew "borrowed" portions of the first IBM PC's design. Some folks at IBM Austin (where the Displaywriter was made) even hooked up the 8-inch floppy drive to an IBM PC just for grins+giggles.

[@TonyLambregts]

Fascinating. Looking forward to the next episode.

[@burkantorun]

I am very happy to bump into your channel. Fascinating job.

[@BrendaEM]

Interesting project and process. If they have one extra coal-ball slice that aren't doing anything important with, perhaps they could fashion a CNC-rig to photograph, and then grind a layer off of it, clean it, and scan it again, and so on. That way they could see a section of the plant. The tomography scans may be stacked for viewing in Paraview. It's free as in Free Kittens! I believe that this could be done with a diamond coated tool, perhaps a angled, cup-wheel, but what comes to mind is constructing a finely lathed drum, and having it coated with fine diamond dust. The grinding drum could be angled from the work direction to reduce streaks, like an old VRC flying head. It might be possible to cool the drum using compressed gas, or a vapor-phase system, perhaps CO2, so that it would not need a wet coolant. The slice might even be ground upside down to reduce the material that might embed itself. The scanning system would be retracted into a dusk-free enclosure while grinding. A timed blast of air could clean the sensor. An enclosure could cut down on noise, a bit. A HEPA filter could keep the dust down. (I suspect that there are other methods for polishing silicon wafer, which may work, too.)

[@MrPocketrocketgaming]

This looks super exciting! Can’t wait for the next episodes!

[@xephon3000]

This is so interesting! What a great collaboration opportunity!

[@hoofie2002]

Fascinating video Marc. Thanks for not dumbing down the content. Really looking forward to this series.

[@digitalrailroader]

Marc, you and your friends are doing the Lord's Work when it comes to preserving computer technology and the data that is a part of it; but i find it sad that Dr Phillips will not be able to witness the fruits of his labor of 50+ years ago. i bet everyone in the Paleobotanical field are hoping you succeed in extracting that old data so it can be finally analyzed with modern computing tecniques.

Fascinating! It shows again how small on the Earth's timeline we are....

[@CrudeOYL007]

Amazing! Great job guys!

[@skfalpink123]

That was absolutely fascinating! Thank you Marc, for these wonderful programs..

[@nagarev]

Woow! This is AMAZING! Looking forward to see the rest of it! You are amazing guys!

[@johnwilson2250]

So glad you are working on such problems!

[@scsirob]

Cool! I started with 8" floppy's on a Grundy CP/M system. When I got it, it came with one boot disk and a Z80 assembler on it. Had to write my own floppy formatter routine on the single disk I had for the system. Scary stuff! If you need information on those, I still have some of the source code for the formatter and the BIOS routines to read them. Floppy controller chip was a WD 1793, with some luck you can still find those. 128 bytes per sector, and you'll have to find out if these are soft- or hard sector disks. Good luck on digging up the data! EDIT: I just went through my old-chip collection and found 2x WD 2793 Floppy Controller chips! If you need one to build a controller for these 8" disks, let me know.

[@Rob2]

Coincidentally I am currently in the process of getting my old TRS-80 going again after it has been sitting in storage for 25 years. Back then I built my own floppy controller (around 1982) which could control both 5.25" and 8" drives and it also used that kind of controller (1771, 1791 and 1691). Those were the days... I found the schematics for the controller as well (my own hand-drawings on large fanfeed paper sheets). First steps are to carefully powerup things and make everything running, the 5.25" cabinet blew the fuse due to a shorted mains filter. That happens... Always nice to follow Marc!

[@RichterPavel]

I can't wait to see next episode! Thanks guys!

[@digitalshackonthelane]

Go Marc and team! This appears to be shaping up to be a super exciting set of episodes!

[@pjs199]

Can feel this is going to be a great adventure, thx for sharing-and such a teaser for the next episode!

[@estherlied]

My good sir, you just created a new meaning of 'Data archaeology!'

[@portlyoldman]

This is really exciting. Can’t wait to see how this all works out!!

[@robcfg]

This is great! Looking forward to see part 2. I'm really interested to know how you'll be making that 8" drive work, as I have one and would like to get it working.

[@ashpowell9451]

So great, can't wait for the next episode!

[@arrbam02]

Awesome, just awesome. Can't wait for the next epoisodes.

[@dereketnyre7156]

Lots of computer history was made at that school. I know it as where Ray Ozzie studied, worked with the PLATO system and laid the ground work for Lotus Notes....

[@Tedd755]

Normally when a video takes so long to finally get to the crux of the title, I'd say Wadsworth's constant applies, but this is so fascinating! I knew you were going to turn into the obsolete technology A-Team! DataSavers? TechRescue?

[@Joel-st5uw]

Haven't anticipated a sequel this much in years!! Cannot wait for the next episode! Creepily ironic that I watched the new PBS Eons video "When the Rainforests Collapsed" just prior to this.

[@TheOnlyDamien]

I am so excited about this series, I also really want to learn more about Antoine, that shot of all of those chips looked like pure heaven.

[@kaitlyn__L]

Wow, generating all that data and having no way to analyse it en masse. Essentially he dedicated his studious life to generating all this data just so later generations could do something with it. So it's highly excellent that you're getting the data back out of these disk archives, to be backed up across many different formats instead.

[@160rpm]

I've never seen such a huge stash of 8" disks before, he must have spent ages and ages entering all that information

[@franglish9265]

This looks like a histology slide. Amazing that they were able to come up with a procedure to generate specimens with this much detail.

[@DanielPalmans]

I love all you videos Marc. you are all amazing!

[@hamishgrove7722]

Wow what great interests you have. From r2d2 to moon lander computers to ancient plant life. Keep up the great work. Thank you.

[@TheJimbodean67]

I love a good mystery. Good luck and fortune on your quest.

[@GoldSrc_]

Wow, amazing. I can't wait for the next one :D.

[@MarcelHuguenin]

Wow! What an amazing story! I am emotional an excited at the same time. So wonderful to see how era's and scientific fields come together. And the idea of collecting that vast amount of data without knowing how to analyze it later. Great project, you are all heroes! Can't wait for the next episode. (great to see some fellow dutch people involved as well ;-)

[@chutipascal]

Thank you, another epic episode.

[@dufflepod]

Simply wonderful!

[@carlossantiago4845]

This is great work.

[@jamesdecross1035]

What an astounding story!

[@c0olcast]

You guys are just a bunch of retro-computing heroes. And, I just find it amazing that people actually take the time to "thumbs down" something like this, what a bunch of jerks.

[@materialsguy2002]

Cool, cool, cool. I’m on board for this one. 👍🏻

[@joopworst]

Yeah the Dolch PAC back in action. I love the series were you repair that quirky machine. Hopefully you have an OS to suit;)

[@rkan2]

I need an episode on the OCR. You'd imagine there is still much more of data that could be scanned from mainframe prints. (any prints, not just this video's stuff) Thus exploring OCR'ing everything would be cool.

[@SidneyCritic]

Maybe that CZUR book scanning ep has something to do with this.

[@Meric_N]

What an exciting story !

[@johanrg70]

Very interesting. That area of work would not be for me but looking forward to the next part.

[@sheep1ewe]

This was awsome! If he's interested of course we would love to see more of this guy!

[@jimengr]

This is gonna be so cool! UIUC BSME 1988

[@numlockkilla]

Cant wait. Wahoo thanks guys

[@kingcrunch85]

This is going to be a veeery good series. Watched it twice.

[@joed2392]

Marc, I just realized something....... When the Phd's were describing the coal deposits, they mentioned the coal-balls composition. And hen it stuck me......! When mankind arrived and progressed into the iron age, could the process of making Steel be a Stroke of Luck ??? For instance..... someone was trying to make some refined iron from ore, and just happened to have some crushed coal-ball in with the regular coal and fired it up and Bingo !! Instead of regular iron, you now have a low grade Steel ! If you lookup how steel is made now days, you'll see limestone in the mix !! This might explain some anomalies in the history of metallurgy ! Could be the foundation for someones dissertation...... !! And Thanks Again, for all of your hard work !!

[@swebigmac100]

Curious Marc: please never stop being curious. 😘

[@Quasam]

Can't wait for the next one!

[@jeffreyplum5259]

Marc, Could it be this is a COBOL coal ball dataset. Fortran is more likely, but the possibility is interesting.> My grandfather was a coal miner. I have seen the plant and fossil traces left in stone bordering a coal deposit. This is exactly the sort of data the WEB was invented to spread. I can imagine maps of the sites where the data came from on the WEB. Now we can support detail and imagery those who collected the data could not dream of. Good luck!!

[@blenderbuch]

Wow such an interesting topic and that clifhanger! 🔭📺

[@littlejason99]

From computers in space, to computers for archaeology.... The only thing that's going to top this is Marc & crew going to Norad to stop a rogue 70's mainframe from starting WWIII.... lol

[@meteor8076]

wow this is so exciting !

[@givemeakawasaki]

Ahhhhhhh... i cannot wait till the next episode!!!!!

[@Lucius_Chiaraviglio]

When you have the problem of data storage format obsolescence, you really need to have nerds who specialize in restoring old computer equipment -- historical preservation really pays off, even if only at infrequent intervals.

[@jmpattillo]

What a great story!

[@charlestonyank2067]

I have a small box with a bunch of 8" floppy. I put them aside years ago, maybe in 1971? from a 360 computer room that I programmed for. I have no idea what is on them, but probably some boring business data. Still don't understand how I managed to keep them all these years.

[@chiIinviIin]

I cant wait for the next episode!

[@andersforsgren3806]

Good work mate, and thank you - though I am not a palaeontologist - only an ordinary dull biologist who work on current day species. I do view your contribution as valuable. :)

[@MichaelLloyd]

Very interesting! I'm looking forward to seeing the results. If anyone can do it, you can!

[@knietiefimdispo2458]

40 years ago i repaired and maintained those floppy disk drives in nixdorf 8860 computers. hard sectored and 78 kbytes of endless space :- )

[@emmanuelf1]

Waiting the next episode with great interest (argghh I can't wait!!!) I hope you have something like a Kryoflux to low level sample the raw discs in case of damaged ones to recover what is recoverable.

[@simontay4851]

Whats a kryoflux.

[@emmanuelf1]

@@simontay4851 See www.kryoflux.com/?page=kf_features

[@gfr2023]

absolutely love this video !!! i'm a geologist but never found a job as geologist so i recycle in GIS... i land a job in GIS thanks to FORTRAN that i learn for geophisics ... remember i write a program that do the work of six months in 2 hours and thay hire me !!! :) the program was in FORTRAN.

[@roberthayes6329]

Great episode Mark, I hope those 8" floppys don't have disk rot. I seen a little moisture exposure on the one in the video. Floppys that old might be made out of something resistant to rot.

[@ml.2770]

This is so satisfying.

[@gurueddy]

What an absolutely rare and unique data set, from a very short era when coal mining and computing technology co-existed! Just think in the 19th century we had coal mining but no electronic technology to record any observations... in the 21st century we have a digital society but we’ll probably see the end of coal mining. Ps: did you just create a new science? Paleo Data Biology?! Brilliant. 👍

[@pdppanelman5889]

Now that is interesting!!

[@osgeld]

fascinating

[@ppser818]

It seems more reasonable to me that coal seams are the result of floating log mats which slowly drop their bark to the bottom. this accounts for the perfectly flat top and bottom of each seam, at least in the Kentucky region of the U.S. - a gentleman wrote his dissertation on this view in the late 70s (and successfully defended it).

[@jagardina]

Man what a tease. This is awesome. Can someone explain how that data helps anything? I didn't quite follow what each entry actually represented, it's not scans or anything. Did it say what they saw on each sector of the microscopic survey?

[@Dreamlgider]

Cool stuff indeed 😎

[@TheCj71984]

Amazing

[@ElmerFuddGun]

10:23 - Perfect example for when someone else pays for your paper supplies!

[@physnoct]

I thought they found fossilized floppy disks.

[@Dust599]

as an old fossil myself I can't take these cliff hangers!!! your going to cause me a heat attack!

[@mikus4242]

This is going to be good!

[@mitfreundlichengrussen1234]

WOW - !?! speechless.

[@abelouellette8051]

Give a all new meaning to the term: data rot. Hope you can recover the data form the floppy disk keep the good work

[@RaymondHng]

kzbin.info/www/bejne/a3e5qJubibuKodE

[@A3Kr0n]

I think you've reached a whole new level of cool.