Hacker News new | comments | show | ask | jobs | submit login
Banyan, the "GitHub of Science", is now in public beta (banyan.co)
143 points by twog 1490 days ago | hide | past | web | 102 comments | favorite

How does your product compare with http://www.openscienceframework.org/? As a scientist, I'm inclined to trust OSF more, since it's run by other scientists, fully open source, and doesn't need to make a profit. At the moment, OSF seems more focused on tools to facilitate collaboration and less focused on replicating functionality currently available on GitHub, although it sounds like they use git as a backend and eventually plan to implement functionality similar to what Banyan (and GitHub) currently provide.

Hi Simonster, Thats a great question & a valid concern considering the acquisitions of tools like Mendeley [1]. We are equally focused on tools for collaboration & reproducibility, but our efforts the past few months have been focused on building an MVP.

With Banyan, your data will always be open & you're free to take it with you anywhere. If you want to move your repos to github/bitbucket/another service, you're easily able to do so

[1] - http://techcrunch.com/2013/04/08/confirmed-elsevier-has-boug...

I'm a recovering PhD. OSF looks pretty cool.

By the way OSF costs money if it is done by scientists. Those scientists get paid by someone. They spent shit ton of money (in form of salaries + time) building it + publishing it. Science works on government grants, i.e. taxpayer money. It is actually a fairly profitable industry, and NIH grants are the biggest contributor to capital expenditure of US Research Universities. The 55% overhead coming out of each grant is basically the "profit margin". Calling something by a different name doesn't make it so. Your github profile says you're at MIT, around Cambridge. So am I. What do you think pays for all the shiny buildings around Kendall? It's not just philanthropy.

Sure, OSF costs money, but VC-funded startups are under pressure to monetize their customer bases, whereas grant-funded projects just need to make something useful enough to justify continued investment. There's the obvious risk is that a VC-funded company will make its money off of selling your private information. (I don't think this is very likely, but I trust Brian Nosek a lot more than I trust most startups.) The bigger issue is that VC-funded companies need to keep a tight leash on their product, whereas grant-funded projects are free to grow in a distributed fashion. While ordinary people are happy to live within the limits of their software, scientists like to explore new frontiers, and it sucks when our software can't take us there. This is why I firmly believe that software (and preferably hardware) that scientists use to do science needs to be open source.

I'm not sure I'd call grant income that gets reinvested in infrastructure and educational expenditures "profit." Not only do private corporations buy shiny buildings, but they also have a legal obligation to make as much money for their shareholders as they possibly can. MIT doesn't.

While I can agree with the spirit of your message, OTOH...

> Not only do private corporations buy shiny buildings, but they also have a legal obligation to make as much money for their shareholders as they possibly can.

This seems to be a popular belief, but it's simply not true. Private companies have no legal requirement to make profits or pay dividends. Even public companies, do not have to by law, maximize profits ("make as much money").[1] Corporate officers do of course, have an obligation to act in the best interests of the company, and have maintain certain fiduciary duties.

[1] http://skeptics.stackexchange.com/questions/8146/are-u-s-com...

Sorry, you're right, there is no direct legal obligation to make money. However, investors in startups frequently have a large degree of control over the direction of the company, and often care more about the bottom line than any broader mission to change the world. See e.g. Jason Hoyt's blog post on Mendeley: http://enjoythedisruption.com/post/47527556151/my-thoughts-o...

This is a great idea!

I've been trying to collaborate my professor on academic papers with git. We tried my own git server, github, but they still seem to be too much trouble for people who do not use git regularly. Online LaTex editor integrated with the git service would be great! It would be better if you could add code highlight in the editor though. Also realtime PDF preview (maybe not realtime.. it'll be overwhelming for the server. A compile button?) would be awesome!

And for sure we need a separate community from Github just for academic publications. Mixing softwares and papers altogether is not a good idea. Banyan is really cool!

Congratulations on launching!

Thanks songgao! Drop an email (mine is in my profile) and Ill get you signed up on our git-powered latex editor. It has most the features you just described, and is launching soon =)

Holy crap check this out: https://github.com/manuels/texlive.js It's pdflatex ported to javascript running in browser natively!

I would really like access to this too. I have been using sharelatex and the git support is still not there. My email is zmj@zmjones.com

Thanks! Will do. I will drop you a line this afternoon

Wow, thanks! I didn't find your email in your HN profile. Here's mine: song@gao.io

Whoops, sorry about that. I just re-added my email. Ill drop you a note this afternoon


Btw, LaTex compiler running in Native Client would be a potential solution to "realtime PDF preview". Not sure if it's a lot of work porting LaTex to NaCl though.

There were some articles a while back on why porting it to iOS was impossible/hard (loads of dependencies some of which are in a language knuth wrote) and I expect NaCl would be equally tough unless anyone has done a full C rewrite.

Great to see another product in this space (https://www.scigit.com/ launched not too long ago). As far as writing papers, my old-fashioned P.I. will probably never budge from Microsoft Word, so Google Docs is the best solution at the moment. The fact that they have "Google Scholar" integration for inserting citations into the editor makes it pretty much a no brainer for writing a draft.

Yikes! What field are you in? For most STEM folks, TeX seems to be pretty standard. I'd really love to see a collaborative TeX editor (web based?). Better yet... a collaborative emacs!

I'm in biology. The journals Nature and Science both specify that Microsoft Word is their preferred format for submissions and I'd be surprised if lots of others still do. Collaborative online TeX editors have been featured on HN before (https://www.writelatex.com/, https://www.sharelatex.com/).

TeX is what we are supporting currently, and should be launching shortly. If you drop me an email at toni (at) banyan.co, I will get you beta access in the next few days to our git powered, collaborative latex editor.

Rule 34 existed for emacs well before it did for pr0n.

Rudel (http://www.emacswiki.org/emacs/Rudel) is one example of collaborative editing in emacs.

collaborative web based TeX: sharelatex.com, writelatex.com, spandex.io.

Thanks for pointing out scigit, Im a big fan of what they are doing. We dont support Word yet, but I do recognize that older P.I. / professors wont leave Word easily/ever. We have some solutions on the way for that.

The easiest solution is to wait 10 years.

Your old-fashioned P.I. has written his fair share of papers in LaTeX thank you very much! :)

Hi HN!

We launched a private version of Banyan to a few of you here months ago. Since then, our team has been hard at work trying to make Banyan the future of open science. Todays launch is just the start.

We know we have alot of work to do. Over the next few months we plan to tackle science, reproducibility, and make existing tools more awesome through our API.

I want to be as transparent as possible. Im happy to answer any questions you might have.

Congrats! Unfortunately, your server isn't handling the load and the google cache is a SASS error. http://webcache.googleusercontent.com/search?q=cache:knsEMz3...

Got something more descriptive you can link to?

Sorry, we cross posted to reddit & HN and the response has been outstanding. Working on getting this fixed right now.

Blog is back up! Sorry about that everyone

This looks like a great and useful service. I'm going to give it a try. However, I tried signing up for "Banyan for Mac," but whatever email I provided was said to be invalid!

Back up now!

I am having the same problem.

Hi twog,

The blog appears to be down, but as an individual very interested in open science could you explain a bit more about the use-case Banyan targets? I imagine this is very focused on things like analysis or computational biology where the primary iterative factor is measured in lines of code changed.

Do you see this tool being used by adjacent sciences? Life sciences, human subject research, materials?


Our first target was initially bioinformatics, but we are moving into supporting all sciences. Our goal is to make sharing, collaborating & publishing research as simple as possible.

We are huge fans of git, and we want to leverage git to enhance every part of science. For us, that means integrations or improvements with existing tools like latex, ipython, etc.

I believe that open science hasnt taken off because there is no platform for it, no tool makes it easy/simple for the user, and the user isnt incentivized [1].

[1] Programmers apply for jobs now with their github/bitbucket profiles & there open source contributions. Why not provide scientists the same option?

"Our first target was initially bioinformatics"

Have you ever heard of the AAVSO? American Association of Variable Star Observers? It strikes me that you need a critical mass in the field you target and a small group to roll out. Bioinformatics sounds a little large. AAVSO sounds just about the right size. Of course your more detailed research will unsurprisingly have a different conclusion than mine, etc.

"I believe that open science hasnt taken off because there is no platform for it"

Well, the AAVSO has been around a lot longer than the web. Even if they're not eventually users, their experiences over the last century or so would make an interesting case study for you, one way or another.

Congrats, I like seeing startups from Tennessee!

Hi Mason!

Thanks, there are more of us every day! Check out http://noogastartups.com I think Tennessee is the next great startup state.

Really? That's great to here. I'm a student at UT and do some research at Oak Ridge. I've always been interested in startups and would like to start my own at some point or join a small team.

Banyan looks great. I've been mentioning Github to my advisor, but I think this is much more suitable.

I went to UT and started and did research at ORNL. I left in 2011 to start a startup in Chattanooga (Colocated with Banyan). Tenneessee (specifically Chattanooga) is a great place to start a company. There's good funding, low cost of living and a pretty nice talent pool.

Are you planning on releasing a Linux client?

When you say science, git, and LaTeX, Linux is often the next word in the sequence...

its kind of tragic that I read that whole post, spent another 5 minutes researching, trying to find some kind of more information, but I still have no idea what this tool actually does.

Didnt you get the memo? Its like twitter for facebookers using science with github

I don't get it. I can't see any information on the homepage that makes it different from just using github.

Our main two differences from github is that we built banyan to be version control for "humans". Our mac application for example is extremely simple to use, and was designed with an 18 year old undergraduate/55 year old lab veteran in mind. Banyan was designed to be the middle ground of collaboration for users who know git with people who have no idea what git is (and in most cases, cant be bothered to learn another tool).

Our second difference is our focus on scientific specific tools like latex, ipython, etc.

So, are you going to provide built-in things like: IPython viewer (like http://nbviewer.ipython.org/), LaTeX viewer for notes (like http://mathb.in/), support for slides (this thing would be awesome - to have possibility to view them and view/fork their source; a bit related to http://slidewiki.org/) or something in that line?

We have similar ideas for integrations with those tools, with the exception of latex. With latex, we have our own online editor with some neat git integration.

Do you have any examples of these online? Today, I don't see any reason I would use this over GitHub.

I gather that your Mac app may be simpler, but it's not available yet.

Since I don't already have dozens of projects on banyan, it's hard to see the value.

Perhaps it would be a good idea to link to some repos with examples of the fancy LaTeX or IPython formatting

Out of curiosity, why is underscore.js the sample repo on your homepage?

Have you thought about easing distributed collaboration for structured data? The use case I'm thinking of is multiple groups submitting data to one central informatics/analysis group (i.e. consortia, meta-analyses). Scientists still throw around data in excel sheets with ever-increasing versioning through file names (e.g., gene.data.v29.201230429.janes.final.Final.reallyFinal.xls). It's a huge time-suck that doesn't help anybody. Support for easy viewing of XML, JSON, or even some sort of hosted RDBMS would be incredible.

I think it's great to target the wider, non-developer audience. They could benefit immensely from version control, distributed systems, and smart editing tools. It could be a big benefit to the whole scientific community.

Im really interested in this but have some questions I would like to ask you. Do you mind dropping me an email at toni (at) banyan.co?

Congrats, this looks great! Some of the buttons are a little off (http://i.imgur.com/9CRvFQj.png) Mac OS X, Chrome 26.0.1410.65.

I'm very excited to tell people that github for science is now a thing. I'll be signing up for sure.

Thanks! Just submitted a fix for this. We will push this into production when things slow down on HN a little bit

No problem, the firehose that is HN is definitely exciting. Hope you guys are getting good traffic :)

Would love to try this but you'd have to let me have at least 1 private project. I don't have permission to share private research data with the world.

Also, I tried to submit my email to be notified for the mac app and was told it was invalid (it was just a standard @gmail address).

Hi Osmium, thats a good point. We will revisit the pricing model tonight.

The forms are broken unfortunately. I have submitted a fix, but we wont be deploying until later tonight when the HN craziness slows down. In the interim, you can drop me your email at toni (at) banyan.co and I will manually add it to the list

I don't know if you were inspired by a submission [1] I did almost six months ago, but if not, kudos anyway :) When the site is up again I will check it!

[1]: https://news.ycombinator.com/item?id=4712787

It wasnt! But we should talk. We are looking for engineers who want to change the way science works. We announced Banyan publicly last August [1] during the http://thegigtank.com

[1] http://us-ignite.org/2012/08/gig-tank-demo-day-delivers/

Oh, I see, and I am checking the website right now. It something different from what I thought, but it is cool anyway. My email is on my profile :) Congratulations again!

Is there a way to make repository which is partly private? For me it is a common use-case where some data aren't share for any reason (e.g. because they are not checked yet whether they are correct), while other files are public?

Not currently, but this is a great idea. We may just have to add this to the roadmap

The other thing is that, IMHO, private repo for this is not good for academic, instead, make it temporarily private. People write papers to get published. They surely want as many people as possible to see it. The reason they want it to be private is to hide it from others before getting published so the ideas won't get stolen. So I'm thinking maybe set up a time limit for repos being private would be better? After that time limit, the repo is automatically made public unless requested more time by user.

It is one reason. But e.g. in my field (theoretical physics) I am not afraid of theft (well, if someone is smart enough to understand and polish a half-written paper, then (s)he can write much better stuff) but of things that "as of now makes sense only to me" or "there are half-written" (e.g. in some places I changed indices i to i+1, but not in all, making it both frustrating and useless to read) or may simply be false and I don't want to publish it before checking (e.g. plots may be completely wrong because of a typo).

Anyway, a time limit is a good idea.

And anyway, when there are timestamps, it's harder to steal an idea. It actually may encourage people to write notes on things to do, share some partial proofs, etc

Why a time limit? Why not a "publish" button that flips it to public? (And maybe submits it to arXiv at the same time, optionally.)

Integration with arXiv (i.e. "sent to arXiv" or "get source from arXiv") would be great. And yes, just a button "publish" would be convenient.

That's true. A "publish" button makes more sense!

Clicking on either screenshot in the blog post takes me to a new page with a smaller version of the screenshot (http://blog.banyan.co/banyan-is-now-in-public-beta/dashboard...). Then I have to click the image again to see: http://blog.banyan.co/wp-content/uploads/2013/04/Dashboard-B...

Wouldnt a better interaction be to display a larger screenshot on the first click? Thats what I would expect?

Yes it should! Sorry about that. Just fixed it for you if check it now!

Cool. Thanks.

This isn't a publishing platform, right? I'm all about free data, but if some "scientists" decide to "publish" their paper on your site it seems like that would totally evade peer-review. I know that is not your intention, but do you worry that some people my try to put out results that have not been fully tested/peer-reviewed? There are non-peer-reviewed journals (my mom finds them all the time...) out there, but a free easy platform seems like some people might try to take advantage of this system.

"some people"

Oh, just call a marketing guy a marketing guy.

I'll have to look into how it manages trust, for things like this.

Great initiative! But I find the screencaps on the landing page's slideshow very misleading imho. It displays some code repo... ok, now I'm a coder and just like others I'd be then inclined to ask myself "why would I use this instead of Github or Bitbucket?". And my gf is a neuroscientist and her gut reaction was "well... I don't code, how is this of any interest for me? I don't get it...".

Just to be clear: we both are VERY interested by this site, and actually quite excited. It's better than any other related initiatives and we expect to try it out asap! It's just this detail made us tick...

I agree we need to clean that up. We use a rake command to seed our test server with test data. We will swap those out for some real screenshots today

We're registering. I've passed the word to other fellow scientist in my network. Looking forward to use it fully.

I am hopeful that this works, as my collaborators can definitely not be bothered to learn git. The UI elements are way too big IMO.

Hi zm. What do you mean by the UI elements? Do you mind snapping a screenshot for me?

The nav bar and the commits page are too big I think. There ends up being a lot of space wasted. The commits page especially so. 5 commits should easily fit on a page. There are also two date information fields. One listing how many days ago the commit occurred and the other indicating what date the commit occurred on.

Thank you for the design feedback. I take ux/ui feedback very seriously. I have passed this onto our designer

To Toni & the whole Banyan team, CONGRATS on launching the public beta! I'm excited to see Banyan on the gigabit web. :)

Thanks Will!

It's built on top of git?

Hi Maro, yes it is.

Within the first 60 seconds your site does not answer the question "why would I want to use this rather than Github?". I'm not sure most people will spend even 60 seconds trying to find an answer to that question, so you might want to figure out a way to show/tell.

Is there any way to create an anonymized link to a set of data and experimental results? Right now I'd love to use this, but all the conferences I submit to are double-blind, so there's no good way to let reviewers have access to my experimental data...

And, this is the other side of the trust coin. Who's data and is it the "real" data. (obviously I don't mean you personally but more like scientists generally)

Congratulations on launching! I applied to YC a few years ago with a similar concept. Your site looks gorgeous, and I'm excited to see it develop. I hope you succeed in disrupting a space that's been neglected for far too long.

Sorry the blog is down everyone! http://Banyan.co can provide a little more info while we scramble to get this back up!

How do i delete my account? I created one to check it out, but didnt realise i would not be able to hide my name or email address...

This is great. Too often we only get the original researchers' interpretation of the data and not much else. Good luck with it!

Thanks csmatt, I completely agree!

I first heard of you guys when I read an interesting post about the startup scene in Chattanooga. Congrats on launching.

If you haven't checked in lately the scene is Chatt is growing daily http://noogastartups.com/

Off topic: Dear banyan blog. I can't read your text!

No greater than #888 on a white background, otherwise you have to bold the text.

I tried to leave my email regarding the Windows port and it said my email address was invalid.

it looks like this just runs a local gitlab instance on your mac. how do you share with other users?

There is a central GitLab server.

So the blog post claims improvements to the webapp were made and has an image of the gitlab interface. Gitlab was recently updated and made big changes in the backend to improve performance.

This smells fishy.

Hi raffenet,

We have used gitlab components & also contributed back to gitlab. Its a great project, but I can assure you we have an entirely different product. Were just leveraging OSS projects like any other startup

from what I can see it is a Gitlab 5 install with a skin? What advantage is there for an organisation using your product vs a free gitlab install?

Banayan .. github ..

Hurray - VINES is finally open sourced!

"GitHub of Science"


I was wondering who else immediately leaped to Banyan Vines!

How could I not?

My first job in IT, I got tossed into the deep end of LAN administration - sink or swim - and Banyan was the pool.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | DMCA | Apply to YC | Contact