Bit disallusioned about julia awareness and adoption

Hi all
I am retired from silicon valley and have some idea about how to ask pertinent questions about a technology. My first startup’s chairman and ceo was Tom Perkins and so I learnt some things right from day one. An example of this is that Mr Perkins made sure we had Dr Jim Gray to guide us in our architecture. I say these things not to brag but to point out that I learnt from some pretty astute people.

Using my silicon valley senses I went to a manufacturing show to find out what the market penetration of julia was in an area I thought it would be examined by the data crunchers. Not one of the software people had heard of it and I had MANY conversations about C# and Python, actually in ALL of them. Not one had heard of julia. I pointed them at some youtube videos I like ( as a noob) and they seemed interested given the scalability and ease of use. I even managed to show coding in the REPL which the little group who gathered liked.

I also wandered around the machine shop guys ALL of which are producing data via IOT and NOPE, never heard of julia. C# and Python all the rage.

I then went to chat with AWS and hooked up with their core developers. after swapping stories of Turbo Pascal ( they didn’t know C# is the baby of the same wonderful chap) we moved on to C# then python and fracking COBOL which the poor people are having to examine to translate to another tech. So I spotted a moment and asked about julia. CRICKETS! not one of the AWS developers had heard of it and no one at AWS had approached them on the language. I asked them to watch the same youtube videos as before AND to consider writing a COBOL to julia conversion package and open sourcing it. I pointed out that it would be fun to do, VERY therapeutic and career enhancing. I asked them to look to see if there was an existing package, I can’t do it because I had the cobol part of my brain removed as well as RPGII and ( zarquon help me) APL). Fingers crossed.

BUT the final straw for me was the MIT booth, neither of the MIT people on the booth had heard of julia. I understand why they use C# and Unity for their wonderful VR app but NOT TO HAVE HEARD OF JULIA!!! I know MIT is a big place, I spent time there BUT something is amiss here. I don’t understand how two people active in software devolopment and photonics had not even HEARD of julia.
I asked one of the booth staffers to watch Phillip the corgi’s TA on parallel julia and to contact Dr Edelman to progress this. I’m not holding my breath.

In short I am wondering what I am doing wrong. In Chicago I have offered ( on discourse) food and coffee for a adhoc meet ( not using meetup as there already is a julia group but they don’t seem to be active) NO Takers. I have pointed out I have MANY chums who privately fund FUN projects and I have one. NO TAKERS. I have asked people who are coding in other languages ( C++, Python, C# et al) when I am asked to audit Chicago startups for people and NONE of them have shown any knowledge of juila. NOT one. SOME of the prop traders are all over it but… Last time I was at the Trading show ( 2020 I think) NOT ONE of the companies I spoke to had exposure to julia. I “think” the NAG guys had looked at it but nothing significant. I am attending it again this month so I will give it another shot.

I wanted to tell you about my practical experiences in Chicago and mention that I am a little disillusioned right now. I can see so much scope for julia, right now I am looking into hooking up a live financial feed to Pluto using pluto hooks, zmq, dataframesmeta.jl, pluto web api, tufte level data representations and I “was” having fun until I realized it was just me a nd discourse. I thrive in what I experienced in silicon valley. A bunch of people sitting around a long table coding the next big thing in a front room. Bouncing ideas of each other, joking around enjoying the thrill of the chase You can’t do that unless you have people and I can’t find them. For the record I am NOT a fan of remote and never have been since we tried it out in the mid’ 90’s. I believe in face to face augmented by github. I see the value in remote for social interaction but for startup coding and motivation it’s face to face the whole way. ONE bright spot on my julia calendar is Bogumil is coming to Chicago to do a workshop. I hope to meet some people there.
end rant.
theakson

11 Likes

I think these things just take time… and more time the more different a language is from what’s previously popular. And Julia is (IMO at least) very different from pretty much all previous languages other than CL and Dylan. To me it’s a whole different programming paradigm (dispatch-oriented) and it took me years to wrap my brain around that. But I’m starting to teach classes in it and have some grad students using it for their theses; slowly the number of people who “get” it grows

9 Likes

ok but julia has been around since 2012.

It’s not that different to code in julia than python. At least in my experience. Once the FABULOUS discourse community have pointed out my mistakes ( MANY) it’s a dream to code in.

grad students are GREAT well done!!! but we need people NOW! we need the maker community.
one of the MIT people at the show said they were active in the maker community especially with kids but had never heard of julia!!! that was upsetting to me hence this thread. I don’t want to be specific but the people in question don’t seem to be too far away from the julia core developers on campus.

In the valley I would host pizza and drinkies in a component store for ALL people to share ideas, goals initiatives. It was a source of MANY startups that got funded. You cannot believe the Chicago version of this! not good not good at all. The “incubators” are a joke, the exodus of the best and brightest finance techs is depleting the fun.

We need to stimulate the people who ARE NOT GRADS, who didn’t get it at school and want a second shot. The finance people in Chicago are GREAT at this and don’t care what your qualifications are. You make a reasonable attempt at the CODING test they ask you pre interview and it’s $50,000 to start for python. In Chicago if you are from Pilsen (60608) that’s more than your dad or mum makes! You spread THAT message in Pilsen and you’ve got a LOT of happy people. NOTE THE PYTHON NOT JULIA!!! so it’s Python, C#, C++

If I were in the valley right now and mentioned julia the FIRST thing my chums would do is tiobe it, figure out it was ABOUT to break 20 in ranking. set up a breakfast and get a bunch of people together who can use matt’s WONDERFUL compiler explorer to give their analysis of julia and it’s attributes. The goal would be to find out what THEY knew about julia not use the tool to analyze it. Then, if they got a tingle there was something here. they would fund a small project ( $100k) and set the ground rules ( plutohooks, pluto web api, Cuda, ZMQ, Dataframesmeta) and the failure conditions. They’d be authorizing the buying of huge amounts of cheap graphic cards from Nvidia been dumped after the merge. Use the app to blow the minds of the private money and thus help the spread of the good word. We can’t let Google and the rest of the walled gardens vacuum up our young ones. Watch tv show silicon valley HULI CANNOT WIN

NONE of this will happen in Chicago as far as I can see and I am hoping that other cities are doing better. I’m retired and so happy to just carry on developing my “killer app” ( at least in my mind) alone but I would like others to share my journey.

my chums in the valley just told me that recent “guesses” are that, by 2025, 70% of new apps will be NO CODE or LOW code. The rise of the citizen developer. Wouldn’t you like to have those new citizens to have written the foundations in julia? To share the goals and aspirations of the julia community? Do you really want Huli, Google and the rest of the gatekeepers to win?

My time is over and I want the new bugs to succeed. But right now, sitting looking at Lake Michigan I am a little disillusioned.

Well done getting your students involved, people like you give me hope. Thank you for your observations.

1 Like

It’s a marketing issue.

Julia Computing does have a Senior Director of Marketing, Mischa St. Amand. Julia Computing is trying to sell into Pharma, Banking & Finance, Energy, Medicine, and Government markets. I would assume marketing efforts are focused to get attention of developers for those markets.

Other markets are served by publicity, I think. Articles in trade magazines, hacker news posts, and enthusiastic users (including one enthusiast at Data Science Central).

If they were to market to others, I would hope they would focus on use R, MATLAB, SAS, and Python users.

you sure about that? I think it’s a KILLER app ( python Pandas) married with FUN advocates, see below.

I was working in finance when python was just a gleam in someone’s eye. We got one of the first presentations in a brown cafe in Amsterdam and within two days I was handing it over to the fine folks on sand hill road for analysis and projection. I don’t remember ever hearing from or talking to a director of marketing for python. What I DO remember is traveling around and enjoying spending time attending random techs meeting up to have a blast with their new toy. I was trying to make up for the vc’s I worked with creating java ( my boss funded Sun and java… dear god what have they done… :-)) I actually wrote an Oppenhiemer level letter to the great and good saying we could be destroying a whole generation of computer scientists. I was almost right but then along came Dr Upton and the raspberry pi to save the day.

I think this is ON US, yes US! to get this moving. We need more of the likes of julia for talented amateurs or dj office hours.

we need the julia equvilent of python’s dave beazley I am a huge fan of phillip the corgi’s TA ( Prof Edelman) or Fons and nick and my personal heroes bogumil and adrian and Helmut.

If you set up a coffee evening and show things like the above an how could you NOT be stimulated. Add a little funding, set up a MVP and…

this is on us! not the director of marketing. we’re not a corporation are we :slight_smile:

3 Likes

I have given this some more thought and think it’s not enough for me, a noob, to just say this is on US. I don’t speak for anyone but myself so I’m amending it.

This is on ME. If I want to get something done I need people, to get people I need to find them, when I find them I need to show them the possibilities. This is true of an investor and a founder.

Last week I was at the biggest international manufacturing show auditing it for my old chums in the vc community. Looking for whatever had been missed by everyone else. Sometimes even a blind squirrel finds an nut ( especially in a masting year :-))

they were handing out 64gb 3.0 usb’s. Most of the machines at the show have a test cycle that simulates a successful run, a run with flaws, a run that fails and so on. They all dump the info to either IOT devices or USB ports. I could have paid them to spend a little time in the off hours to run the data and give it to me under the condition I could prove I could anonymize it and would not use the data other than as a sample set.
SO I have my data as a CSV.

using the MARVELOUS , what I consider KILLER APPS, genie builder married with dataframesmeta I could have put together an adhoc presentation to take to the AWS developer group who were in the show. I could have offered the devs a slap up pizza meal in any one of the meeting rooms I can use in Chicago to make my point. There’s a career/money in this so consider it. They MIGHT have gone back and started to augment the julia awareness that HAS to be present in AWS. Amazon is one of those key players that has the ecosystem ( vc arm) the sales and delivery system to create a viable app environment. For the record I don’t like the walled gardens of google and apple. I respect them but NOT what they are doing to the maker environment.

I could have approached the show organizers to take up any no show/dead space in the presentation schedule to give my pitch BUT with NO questions! I am not competent to answer them.

I could have added the cherry on the cake with pycall NOTICE that temple is a trading shop in Chicago and they get it. As I suspect others do given the edge julia gets them so julia isn’t unknown in Chicago and having worked in massively distributed high speed trading systems, I can see why they aren’t open about it’s use in house. Sometimes prop shops are a tad cagey :slight_smile:

LOTS of could there! I know I know. But cut me some slack, I’m retired. but julia has stimulated the founder in me once again.

The MIT encounter was the trigger for me. I just didn’t get the fact that they didn’t know about julia. Yes the campus is “big” but so is Stanford and it seems they are 2 buildings away from the math lab.

If I were still in the valley and this happened to a Stanford initiative then some very serious questions would be asked and solutions found. My point in this case is that if they didn’t know about it then it couldn’t have factored into their design decisions for the app they were presenting. That’s how C# wins. I have an idea, from practical experience, what makes a language prevalent. I was there when the vc company, run by my first ceo and chairman, lit the fuse for java after funding Sun microsystems. I watched, and helped a VERY small amount, python develop into what it is today. Let’s face facts, python 1980-2008 gestation period julia 2012 -2018 gestation period ( rough guesses) so which is the more successful? why? look at the developer adoption rate, how many books are written about python vs julia? These are all questions I “would” have been asking myself sitting in a park in Atherton, then I would binge watch the tv show “silicon valley” which I love and use to stimulate my creative juices.

I won’t be caught like this again, I’ll have my MVP on my trusty X1 so someone gives me a csv and I’ll put on a pitch that will blow them away. Screw barcharts hello bump charts, here’s the abstraction layer that gets you the gateway to AR…

I wanted to not just be the ranty guy but make suggestions that might help. I’m NOT going to put a package together as I don’t have the time to support it BUT I can put the above on my x1 so the next time it’s not just a repl demo using debugger.jl which is what I did at the show to the people from the sov wealth fund who were nice enough to give me din dins at my fav Chicago restaurant.

soon I have to audit the trading show in Chicago, I will be looking to see what happens there ( last time NO one knew about julia) BUT will not approach anyone from MIT ( if they are there), I AM trying to find out if the Stanford guys are interested and if we can get them a flight over to have some fun. I think I’ve left if too late but it pays to try. There’s a LOT, and I mean a LOT of dev talent in Chicago, more private investment money that you can imagine on the sidelines, and the message that computers = money has been operational since Yra and Uncle Leo digitized trading in the CME.

theakson

4 Likes

Really it’s not just Julia. IMHO the entire world is suffering a serious mis-allocation of resources, and my own explanation is decades of pumping free money into the finance industry and that trickles into various tech and pharma and whatever, but basically no one is going to solve problems that everyday people have unless everyday people have money, and they don’t.


(in 1995 we turned on the tap and money supply grew by 6%/yr ever since, while CPI inflation grew like 1-2%, meaning all that money was going somewhere else… namely asset prices: stocks, real estate, hedge funds etc)

Anyway, I think decades of futzing with the economy has come home to roost in now no resources available to address many important issues.

2 Likes

A quick comment, there’s a lot of different thoughts in the post, I will make one point about the folks at AWS and MIT who had not heard of Julia: both are extremely large orgs where I personally know many folks using Julia.

Python is essentially the defacto programming language for everything so we should assume 99% of people know it and don’t know Julia.

On the Chicago scene, I am now semi-long term in the area and will be trying to connect with more folks. Sorry we haven’t been able to connect yet! Very busy working on many threads. We should perhaps (as I have mentioned on slack I believe) get a regular Julia meetup going in the city.

2 Likes

hi there

not suprised to hear that AWS is using julia, they are very astute. My observation was that the devs at the show hadn’t even heard of it. My experience with the Amazon people is that they encourage anything that can give them the edge, especially in the dev world. I asked the devs to watch some youtube videos and find out what resources are available inside Amazon to give it a shot. I also made some suggestions regarding forming valley type startup groups. They “seemed” to be very excited.

I have spent time at MIT and understand it’s size but the distance between the MIT show team labs and and the math labs ( Dr E) is 2 BUILDINGS. It’s stunning to me that they had no knowledge of a language that might be so important to their work. One of them works in Photonics. Something is very wrong here.

I know that Python is the defacto language and so was Cobol. If you go back to my OP ( I have enboldened the relevant section just in case you want to follow up with your AWS contacts) you will see that the Aws Devs have Cobol to (something) task in store for them. My suggestion ( again in the Op) is to ask the Aws people you know to consider julia as a target language to convert to. There is no harm in asking and the results could be dramatic.

As to the Chicago scene I have done my best and am giving it one last try at the trading show. Last time NO ONE knew anything about julia. I was wandering around with two board members of the CME and three partners from the vc community. There was no julia presence there at all. There were devs from most of the trading group ( free swag and food), recruiters from the prop shops and a mate of mine who runs the shop that ranks in the top 10 forex trading globally. My point is that the show is small but relationships are made that carry on in Gibsons. Last time the head of it for Northern Trust was there and I was chatting to him about Python scaling ( his choice) I had to do this because I had no grounds to suggest julia. All he would have done is ask all the other people had they used julia…
there were the people who plumb out the high speed networks who need distributed scalability. I’m going again this year to audit for my chums in Atherton and I’ll ask the same question
“you heard anything about julia?” Just for laughs I’ll put up a flag on a table and put on it a bunch of links to videos just in case someone bites.

As a fall back condition I’m aiming at Indiana, my chums have asked me to chat with the state representatives about incubators. Same thing for Wisconsin. I don’t use slack just discourse and julip for the GREAT pluto community.

Chicago is the place that DEFINES RISK, they have some VERY sharp devs who know how to eke the most out of a compiler. I am certain that julia is used in the shops but they don’t seem to be too active in stimulating a julia subculture which, in my experience, is what they do if they are interested. They did it for python, C# and C++ so where’s julia?

the “smart” money thinks by 2025 70% of development with be NO CODE or LOW code. Shouldn’t julia be the building blocks for that? shouldn’t that dev community of citizen developers share the same ideals and aspriations so evident in this discourse community? I vote YES. Fons is headed that way with Pluto so lets help out as best we can.

don’t want to get into a discussion about economic theory in this thread. There is PLENTY of money on the sidelines ready to invest in projects ($1 trillion last time I got briefed). The people who will be making the decisions are not idiots. They can read between the lines and spot social inequity a mile away, especially if they fund research in Berkley or Stanford to get an idea of the harsh reality.
Or fund the people on the sharp end of the problems.

There is very little people can tell we Chicagoans about getting it wrong and, given we define risk (cme), we have some understanding of what makes the financial world work. To do this we need a language that can be socially adopted by all (like Python ) but can scale. If we, in Chicago, can get young ones programming in julia in the libraries then getting them jobs in finance would be easier. I am an advocate of tar getting an area called Pilsen in Chicago given my experiences there. So I agree, everyone needs a shot and julia can help with that.

agreed, shouldn’t derail the topic.

sorry to be rude about the not wanting to play the econ stuff.
But we are BOTH on the same page. I can’t tell you how far away from the America I came to we are right now. It makes me VERY angry but I’m going to try to help fix it. It’s the least I can do for a place that has given me so much. I’m not someone with clout but I think I know people who do who are just as angry as I am. Julia is the great leveler and I want to do everything I can to help. There is about to be a LOT of bitcoin mining rigs on Ebay ( all hail the merge) and I wonder what the young ones in Pilsen could do with julia, a bunch of rigs and a cunning plan. It would make a huge difference in a job interview to walking in with a rig and saying “watch this”. A lot of the old floor traders live down in 60608 I think they could help out with the plan :slight_smile:

“Never underestimate the power of a small group of committed people to change the world. In fact, it is the only thing that ever has.” ~ Margaret Mead

2 Likes

Thank you for the Margaret Mead quote.

It seemed right for my response. I was fortunate to be around such people for all my working life. Thank you for your reply.

this quote kept resonating with me. You are completely correct and this might be part of the solution.

I don’t want to just address the cs equivalent of how many angels can dance on the head of a pin. I think it would be interesting to have a form of apollo project. I was fortunate to spend time with people who were in mission control during the landings, I learnt a lot from them but the most important thing, to me, is that people matter, opinions matter and unless you have a common goal that is understood and agreed on AND a mechanism to address that goal then you don’t stand a chance. I think the people who are not interested in compiler design or the internal workings of the language should have a shot. This discourse forum is by far the best I have seen in doing just that and perhaps it’s time to bring those values into the physical world.

Python has solutions that cross the spectrum. It’s not just math or science. I use it to kill my linux machine :slight_smile:

I happen to have a lot of friends who know things about trading. Not directional trading but statistically driven trading. I’m going to spend some time with more competent people that I to see if it’s possible to come up with a trading augmentation app that will educate and facilitate people in the approach. So you grab peoples interest by making them money, they MIGHT drill down and figure out how it works ( how the math works, how the process works) and, if I design the architecture properly ( Dr Jim Gray guiding hand) then they can contribute. Clearly defined goal, people can contribute and building a future whilst making a ( insert currency here).

If they are successful I would hope that they consider contributing some of their gains to local initiatives ( I could build that approach into the system) that actually help people.

Maybe fund a public library based julia group mentored by a retired teacher. Using julia to solve local problems that, I suspect, might be globally relevant. The julia founders have given us a wonderful ecosystem to develop. The garden is open to all but right now seems to be, to me, scientifically orientated. One of the killer apps for python was Pandas which came out of a hedge fund. We have the analog in dataframes.jl and dataframesmeta.jl , we have no/low code possibilities in pluto and geniebuilder. So the tools are there for young and old. We just have to seed the garden a tad to get the folks to come in and lend a hand.

Who knows maybe a local venture capitalist might swing by the library and help out. Perhaps a hedge fund owner, an exchange owner. Chicago is an interesting place and we have cornered the market in social inequity.

Thank you for making me think @dlakelan

It may be a bit of a stretch to think you could just go around asking people in finance/tech space if they know Julia and expecting to find many that do. Think about the difference in market shares between Julia and the combinations of {Python, Fortran, C/C++, OCaml, R, (COBOL?)}. You are essentially asking about Julia versus “the field” and the field is quite large, established, and already embedded into those firms. That doesn’t mean that Julia isn’t a worthy competitor or isn’t gaining market share.

1 Like

I’m not finding ANYONE that has heard of julia. I have tried to get people to meet up about it in Chicago by posting here ( discord), by going to venues ( IMTS/trading shows).

I am versed in tracking interest levels, it used to be my job. I can use tiobe just as effectively as others and, yes, julia is about to break the top 20. Yes it’s perfectly positioned for low/no code solutions. I covered those topics in other parts of this thread. I wouldn’t be wasting my time considering advocating for it if I believed that it wasn’t a viable contender.

Before Python or java became mainstream it was possible to detect a growing swell in interest in the developer community. I am not finding that in Chicago. I don’t see universities, in Chicago, posting on this forum open lectures or meetups. I have people asking in apartment buildings full of developers.

Of course the trading shops are looking at it and using it. I linked to a presentation a Chicago firm made at juliaconn. I’m not saying that it isn’t used in Chicago, I AM saying that I am concerned that I can’t find people who know about it.

As I said. I’m giving this one last shot at the trading show but I will make sure I consider your valid points when I do so. Thank you for taking the time

Showing what it does that is of relevance to the people of interest, and how easy it is to have well-designed packages interoperate in service of one of their focused goals, may be more effective than asking people if they are using Julia. The specifics are important, of course.

the showing is already been done and has BEEN done for sometime now.

It’s on youtube, juliaconn, sometimes on hacker news. There are packages that can be used off the shelf ( geniebuilder, dataframesmeta, pluto). It has the best package management system I have ever seen, the installation process is wonderful thanks to juliaup.

I ask in places like the Ryan ability lab research group, nope never heard of it. I ask all the people who work with my doctors at Northwestern ( they mostly head up the departments) again never heard of julia. I ask in coffee shops near the trading floors still nothing. I know it’s been used here, I know the prop shops and labs who get it. This thread is about MAINSTREAM awareness. That is my concern.

In a few days time I am having a Carbon briefing and I’ll know a little more about the universe after that.

Julia is well-regarded by and (relatively) widely used in the Economics community.

Likewise with Pharma.

MIT, Stanford, others use Julia in their courses.

1 Like