I'm working on an OpenFlow controller in my spare time. There's nothing too cool about it other than, it's written in Go and GPLv3 licensed. It's useable but I have a big update in the pipes. Once that update is done I should be able to port some existing OpenFlow applications to Go.
https://github.com/jonstout/ogo
Using mininet right now to do verification of message parsing and link discovery. Other than the latency calculations, link discovery seems to be working just fine. After I get this next iteration done I'll take it to the office (InCNTRE SDN Lab). We have a lot of 1.0 hardware to test against. Next I'd like to add path computation based on user defined weight functions.
Working on a distributed document database. It's cool as it integrates a search and analytics engine, distributed file system and query language like SQL. Will be the first ever and will solve all database problems for the most part.
I keep thinking javascript in android apps is a great idea but I can never get it to work in a way that simplifies development. Good luck! I hope you can figure it out.
Just released a Q&A book about building community products. (I founded Forrst in 09.) Totally DIY effort, used ruby and the prawn gem to generate the PDF file. All questions were crowd sourced. Here's a discount if you're interested:
https://gumroad.com/l/obcp-book/saturday
I'm so sorry to hear it. This is why we're working in cancer.
We all work as fast as we can, but I've already seen too many cases with poor prognoses. The hope is that in growing this technology, we can catch cancers earlier when they are more treatable.
Currently we don't have that much to show publicly as we're working primarily with doctors; we need to validate our technology properly before it gets anywhere near patients or the public eye.
With that said, if you shoot me an email (see profile) I'm happy to answer questions and perhaps more.
Somewhere between my main interests in molecular+systems biology and my work in microscopy and software(building and using), I gained an appreciation for biomedical imaging techniques.
From there, much hard work learning from oncologists, radiologists, pathologists, and technicians to figure out what we could do to enhance their work with imaging technologies.
I am working on multiple things at the moment, though with my first year of college, progress has been slow.. After challenging C and a multitude of "introduction to programming" classes, I got into CISP 401 (java), so in order to get ahead of the class, I wrote a toy interpreter in java (deemed KjuScript). It is extremely slow, but based off the ruby language with full OOP implementation. Also recently I have been working on something in C++ with a friend that converts sound to color (to teach my dad music (he really wants to learn) who lost most of his hearing at 20, and so my mother who was born deaf, may enjoy my concerts in real time)
There is a prototype of it in python located on my github account:
It's cool because it's been an amazing ride keeping it nice and snappy in spite of an ever growing data set - over 50 million feed items retrieved so far and still growing :-)
http://pineapple.io - I think it's cool because I don't know of any centralized locations for development tutorials and tools. Reddit is filled with mostly jaded posts and HN is filled with lots of news. Mine is only tutorials, tools, and assets.
Self Experiments: Its mission is to make inhumane animal testing obsolete by using “open suffer” collaboration of volunteer testers from around the world.
Hey, this is an interesting idea. I have criticisms that are probably going to sound mean; but even though I don't like this idea, I like that you're thinking about these problems:
* Human subjects research (which the selfexperiments are) is governed by IRBs; if the research did not pass IRB approval, it can't be used in academic publications. It doesn't matter that the research was done outside of academia, it still has to pass IRB approval to be usable. I don't think any IRB would approve this at the moment (though maybe they should!), so it wouldn't be able to be used to replace research in publications.
* Of course, humans are not biologically valid testing replacements for animals if your goal was to test what happens in that other animal, so this can only replace research that was being done on animals because humans weren't available.
* It looks like up to one million animals a year are used for research in the US, whereas several billion a year (mainly chickens, pigs, cows, with increased capability of suffering over the average testing animal) are raised for food in factory farms. Why does it make sense to work on animal testing instead of factory farming?
Thanks for your comment Chris. Since you don't have the benefit of knowing what the experiments are, I don't blame you for being sceptical.
* Self Experiments isn't concerned with conforming to FDA specifications and obtaining IRB approval (yet). Also keep in mind this project is iterative. A lot of mistakes will be made along the way. Oh, and the experiments will be fun, or at least extremely interesting.
* Animal testing is often a proxy for testing on humans. Self Experiments will attempt to reverse this on a case by case, experiment by experiment basis. I hope Self Experiments will be preemptive and anticipate experiments that may be conducted on animals in the future - and conduct them on humans first. The trickiest part is designing the protocols for these experiments.
* Factory farming is deplorable, and I've mulled over how to solve the problem for several years now. I have ideas but that's all I'll say. In the meantime, I'm vegetarian and working towards vegan-ism (although it's harder than I thought).
You'll never make animal testing obsolete, nor would it be a good idea to do so as the benefits from using animals in research are so immense, and most such experiments are not transferable to humans (e.g. genetically modifying humans is not viable.)
But it sounds like an interesting idea nonetheless - what experiments are you considering running?
I understand that you must have invested a lot of your time and energy into this idea, so it may be difficult to accept criticism of its base concept. However by posting about it on a public discussion board you are implicitly inviting others to discuss and critique it, so I don't really see how that can be considered impolite.
Saying flatly "this could never work" is impolite, arrogant, and possibly wrong. "I don't see how this could ever work" suffers from none of these failings.
And unless it becomes ethically acceptable to genetically modify human embyros for testing rather than mouse embyros, then this is not going to change.
This isn't arrogance or impoliteness, it's just how it is.
Fun side project to scratch an admittedly very small itch I've had for a while. I've collaborated with a local company who is handling all order fulfillment.
Working on adding font chooser, image upload, and a more elegant customization form.
https://kwelia.com -- we are taking a quantitative approach to determining the market rental value of apartments. This has not been done accurately on a large scale in the past, and I think we are the first to do it well.
I just signed up for an account, only after the fact to find out that you don't have data for the Portland, OR market, which is unfortunate since this is definitely a tool I'd like to use in my upcoming apartment search. Two things I wanted to mention about the sign up process:
* Before giving you my email address, I'd like some reassurances that you won't spam me, etc.
* It took 10-20 seconds to create my account. Not sure what's going on there, but you might want to investigate that. This doesn't appear to be a fluke. When I logged out, then logged back in again, it also took 10-20 seconds to log in after entering my credentials.
Also, if I log in, then go to the Kwelia home page, then click on "Apartment Ratings", it brings up a login page stating that I'm already logged in, whereas I would expect it to take me to the Apartment Ratings page.
This tool looks really promising. Let me know if you ever add support for Portland, OR.
Thank you for the feedback! We're rolling out our models as we acquire customers for our B2B product, and we have none in Portland yet, but hopefully it's only a matter of time.
Sorry about the performance issues, I will be looking into them right away.
We're working on an app for developers that collates all of your saas products into one mobile feed. It helps you stay current on what's happening at work, both what your team and your machines are doing. It also helps you quickly dig deeper into issues & triage important events.
It's cool because you can keep track of a lot more than previously possible, communicate with your team, and act on it - from your phone. And it's damn good looking too (but we're biased :)
A legit way to get more followers on social media. Helps you build high quality one-to-one relationships with new people. Pick a keyword like "startup school". Our system will detect when "startup school" type conversations are going on in real-time and performs sentiment analysis. That's a highly targeted opportunity for you to start engaging with that user. And we deliver that user to you on a silver platter. It's up to you to keep the relationship going (put them in your sales funnel).
It was a bitch to manually get followers on Tumblr to sell t-shirts to. I had to seek users out, make sure they fit my target audience, and insert myself in the conversation. All without it looking like self‑promotion. That's a lot of mental energy just to reach out to one person. Once I built a fairly large sized audience boat loads of t-shirts started selling. Then I figured startups and businesses on Twitter would find it useful. And they did. Paying customers tell me it works 100 times better than Twitter ads.
I signed up to check it out and created a campaign. I can't tell if it's doing anything, though. Should I expect an email of some sort? Or do I need to check back to look at your page later?
Sorry it's so confusing. If you created a campaign then it is already working. There's nothing else for you to do inside the product.
Now all you have to do is interact with the people that are starting to engage with you. It's not an instant thing. Our system detects when relevant conversations are going on using fancy algorithms based around machine learning, sentiment analysis and POS tagging.
No, I'm sorry! The "social etiquette" on each social media channel is different to begin a conversation. For Twitter, it can be as simple as a "favorite". The system doesn't tweet or follow people.
Ah! I can see it at work now. I sure have favorited a lot of stuff recently :)
One thing I noticed is that it favorites some things that don't make a whole lot of sense. For example, I have a 'bitcoin' campaign. It favorites Russian spam links and random BTC arbitrage "results" along with some more reasonable type stuff. Is there any way for me to tailor these results a bit?
For example, does it recognize my behavior and favorite similar items? Or if I add keywords to the campaign, will it limit results a bit more?
Thanks for pointing this out. I do need more copy there stating exactly what the system will do.
In case you didn't catch my response to someone else, the "social etiquette" on each social media channel is different to begin a conversation. For Twitter, it can be as simple as a "favorite".
I'd be interested too! Why not just put up some screenshots on the website? For the time being (and for future readers of this thread), you could just use a temporary URL, such as
Toy project: an HTTP(S) brute forcing tool using Python as a templating language. Why is it cool: high performance using async IO, powerful templating for Python programmers, very easy to take HTTP requests and turn them into a fuzzing template that mutates request in a combinatoric fashion. Similar to features built into Burp proxy for those that are familiar with it.
Real project: A system that will help organizations understand their overall, and application, security risk and manage it across time. Why is it cool: because security is hard and this will make it easier in a non-snake oil fashion. Many organizations are flying blind about their actual risk. A good view of your risk can help you prioritize security budgets.
It, more or less, has a separate "Fuzzing Template" system which it uses to generate the brute force test cases. It was never meant to rise to the complexity of Sulley's fuzzing system. I wanted an in between complexity for the dumbest fuzzing and something completely flexible like Sulley and Peach. To solve that 80% problem of, "OK, I just grabbed an HTTP request, let's turn it into a quick and dirty fuzzing template." and from that, "And make sure it runs really fast on a single machine". In the time boxed assessment world you rarely have time to do all that you would like so this seemed to be a reasonable solution.
The beauty of the "fuzzing engine" I built is that there is nothing to it really. You put in "scriptlets", which are really just small bits of Python that generate lists or sequences, and it combines all of them. My goal was to just write up a lot of the common HTTP fuzzing scenarios (integer sequences, alphabetical sequences, demonstrate common encoding and other scenarios giving you a simple list of things you can copy/paste/modify into a template. And then it runs, logging it all into a SQLite database.
It will be. It has languished for a very long time, but I have resolved to get it out the door soon. I will put it up on a github repo (https://github.com/bitexploder) so watch there.
Working on a js library to create an ASCII canvas, and have drawing functions much like an HTML5 canvas. It will have the ability to create canvasObjects as well, which can be drawn and interacted with.
It's cool because it automates most of the sales process. It generates an NDA, and work-for-hire, invoices, and accepts electronic signatures for approval. Customers can review and approve milestones, make payments through Stripe. It even schedules calls!
I'm working on making virtual reality as compelling and comfortable as it can be with mostly off the shelf components. It is cool because seeing virtual reality finally work puts a smile on the face of just about everyone we try it on.
Just started working on something to attack SharePoint that is based on sensible standards (REST, JSON, etc.) with the front end completely in something like AngularJS or Ember.
This is cool as, in my opinion, SharePoint Must Die.
The idea that I am currently working on is very much to make the back end (i.e. the server part) do most of the heavy lifting (storage, security, searching etc.) and the UI to be completely decoupled in JavaScript.
Ideally what I would like is to have some kind of "standard" well documented interface between the UI and the server - so that different implementations of both could be possible.
http://www.collegeanswerz.com/ - better college reviews. Most websites just have a reviewer answer a few questions about their school. This doesn't work. I have a bunch of specific questions that reviewers answer.
http://xoxo.gl (hugs and kisses, good luck) - a web app to play family-friendly, traditional board games.
Still prototyping for now.
My goal is wide in scope: to build a site that allows you to take a few minutes to have a positive social interaction with friends and strangers. An anti-Zynga, if you like.
I still have a lot to do. With a goal of universal accessibility, I am working on progressive enhancement - making it work adequately for users with poor connections and noscript, and excellently for users with fancy AJAX.
It's cool and I think it fills a gap for people who want social games without spam viral marketing... as an anecdote: by playing Boggle online, I made friends and ended up visiting a couple on a different continent. I'd like to enable stories like that.
An easy way to enter the name of a specific drug or side effect to see its effects as reported to the FDA. One of the primary ways that the United States Food and Drug Administration monitors the safety of marketed drugs is the collection and analysis of reported adverse events (an event that was not the intended outcome of the prescribed drug and has a negative impact on health) through the FDA Adverse Events Reporting System (FAERS). These reports are submitted by physicians, healthcare consumers, lawyers amongst others, and then the FDA scientific staff will assess these events in the context of other databases to determine if a particular safety concern is associated, and possibly caused by, exposure to a particular drug. Since this is a public database and useful to prescribers and patients alike to know if "has what I'm experiencing been described in patients taking this drug before?" DrugCite has created a more friendly interface to answer that question. The data can also be sorted by Age and Gender in most cases giving a more detailed view.
785 comments
[ 3.3 ms ] story [ 284 ms ] threadWe apply a healthy serving of computer vision and a touch of machine learning to high resolution images (10Gpx) of cancer tissue.
Hopefully we'll be able to make diagnosis faster and more accurate, doing our small part to save lives.
We all work as fast as we can, but I've already seen too many cases with poor prognoses. The hope is that in growing this technology, we can catch cancers earlier when they are more treatable.
Is there anywhere I can follow the project?
With that said, if you shoot me an email (see profile) I'm happy to answer questions and perhaps more.
From there, much hard work learning from oncologists, radiologists, pathologists, and technicians to figure out what we could do to enhance their work with imaging technologies.
Http://www.github.com/pholey
It's cool because it's been an amazing ride keeping it nice and snappy in spite of an ever growing data set - over 50 million feed items retrieved so far and still growing :-)
We use it for working with local Python objects like lists, sets, queues and multisets, which are then backed by Redis.
https://github.com/stephenmcd/hot-redis
Here are some tags to start you off. I guarantee you will find hidden gems here :) http://pineapple.io/tags/all
http://apps.facebook.com/sortonsevents
Which is tied in with a page tab app I'm working on that watches a list of other pages for events they've created or posted:
http://www.facebook.com/UCDEvents/app_123069381111681
Most of the work for that being on the admin end:
http://imgur.com/f1BuKYA
And I've just started the android client.
I'm using GWT on App Engine and trying to keep generic Facebook stuff in a separate project, which will some day be amazing:
https://github.com/BrianHenryIE/GwtFBplus/
The first experiment is launching next month: http://selfexperiments.com
What could be more cool than that?
* Human subjects research (which the selfexperiments are) is governed by IRBs; if the research did not pass IRB approval, it can't be used in academic publications. It doesn't matter that the research was done outside of academia, it still has to pass IRB approval to be usable. I don't think any IRB would approve this at the moment (though maybe they should!), so it wouldn't be able to be used to replace research in publications.
* Of course, humans are not biologically valid testing replacements for animals if your goal was to test what happens in that other animal, so this can only replace research that was being done on animals because humans weren't available.
* It looks like up to one million animals a year are used for research in the US, whereas several billion a year (mainly chickens, pigs, cows, with increased capability of suffering over the average testing animal) are raised for food in factory farms. Why does it make sense to work on animal testing instead of factory farming?
* Self Experiments isn't concerned with conforming to FDA specifications and obtaining IRB approval (yet). Also keep in mind this project is iterative. A lot of mistakes will be made along the way. Oh, and the experiments will be fun, or at least extremely interesting.
* Animal testing is often a proxy for testing on humans. Self Experiments will attempt to reverse this on a case by case, experiment by experiment basis. I hope Self Experiments will be preemptive and anticipate experiments that may be conducted on animals in the future - and conduct them on humans first. The trickiest part is designing the protocols for these experiments.
* One million? Try 19.5 million animals are killed each year in research in the U.S. alone. This does NOT include mice, which is probably numbered in the billions: http://www.statisticbrain.com/animal-testing-statistics/
* Factory farming is deplorable, and I've mulled over how to solve the problem for several years now. I have ideas but that's all I'll say. In the meantime, I'm vegetarian and working towards vegan-ism (although it's harder than I thought).
But it sounds like an interesting idea nonetheless - what experiments are you considering running?
Feel free to sign up for when the first of many self experiments is launched: http://flowandzone.us2.list-manage1.com/subscribe?u=bc7fdf29...
And unless it becomes ethically acceptable to genetically modify human embyros for testing rather than mouse embyros, then this is not going to change.
This isn't arrogance or impoliteness, it's just how it is.
Fun side project to scratch an admittedly very small itch I've had for a while. I've collaborated with a local company who is handling all order fulfillment.
Working on adding font chooser, image upload, and a more elegant customization form.
* Before giving you my email address, I'd like some reassurances that you won't spam me, etc.
* It took 10-20 seconds to create my account. Not sure what's going on there, but you might want to investigate that. This doesn't appear to be a fluke. When I logged out, then logged back in again, it also took 10-20 seconds to log in after entering my credentials.
Also, if I log in, then go to the Kwelia home page, then click on "Apartment Ratings", it brings up a login page stating that I'm already logged in, whereas I would expect it to take me to the Apartment Ratings page.
This tool looks really promising. Let me know if you ever add support for Portland, OR.
Sorry about the performance issues, I will be looking into them right away.
It's cool because you can keep track of a lot more than previously possible, communicate with your team, and act on it - from your phone. And it's damn good looking too (but we're biased :)
It was a bitch to manually get followers on Tumblr to sell t-shirts to. I had to seek users out, make sure they fit my target audience, and insert myself in the conversation. All without it looking like self‑promotion. That's a lot of mental energy just to reach out to one person. Once I built a fairly large sized audience boat loads of t-shirts started selling. Then I figured startups and businesses on Twitter would find it useful. And they did. Paying customers tell me it works 100 times better than Twitter ads.
If you're interested: http://audience.goodsense.io
Now all you have to do is interact with the people that are starting to engage with you. It's not an instant thing. Our system detects when relevant conversations are going on using fancy algorithms based around machine learning, sentiment analysis and POS tagging.
One thing I noticed is that it favorites some things that don't make a whole lot of sense. For example, I have a 'bitcoin' campaign. It favorites Russian spam links and random BTC arbitrage "results" along with some more reasonable type stuff. Is there any way for me to tailor these results a bit?
For example, does it recognize my behavior and favorite similar items? Or if I add keywords to the campaign, will it limit results a bit more?
In case you didn't catch my response to someone else, the "social etiquette" on each social media channel is different to begin a conversation. For Twitter, it can be as simple as a "favorite".
Email me here! blog.goodsense.io/contact
goodsense.io/temporary-stuff/screenshots/...
Real project: A system that will help organizations understand their overall, and application, security risk and manage it across time. Why is it cool: because security is hard and this will make it easier in a non-snake oil fashion. Many organizations are flying blind about their actual risk. A good view of your risk can help you prioritize security budgets.
Might I also suggest an overall fuzzing engine? Sulley and Peach Fuzzer both have fairly ugly APIs and config formats, in my opinion.
A pretty DSL that lets you describe a template, also with the ability to add custom Python functions and integrate them on the fly, would be great.
I will put it up at http://github.com/bitexploder soon (a week? Maybe two?).
The beauty of the "fuzzing engine" I built is that there is nothing to it really. You put in "scriptlets", which are really just small bits of Python that generate lists or sequences, and it combines all of them. My goal was to just write up a lot of the common HTTP fuzzing scenarios (integer sequences, alphabetical sequences, demonstrate common encoding and other scenarios giving you a simple list of things you can copy/paste/modify into a template. And then it runs, logging it all into a SQLite database.
I'm coding something quite simular at the moment and would like to leech from / contribute to / test your project.
If you're interested - tkmop / lenta.ru
It's cool because it automates most of the sales process. It generates an NDA, and work-for-hire, invoices, and accepts electronic signatures for approval. Customers can review and approve milestones, make payments through Stripe. It even schedules calls!
1. Remembering people you meet is hard 2. The iOS (and Android) Contacts apps are horrible.
This is cool as, in my opinion, SharePoint Must Die.
Ideally what I would like is to have some kind of "standard" well documented interface between the UI and the server - so that different implementations of both could be possible.
http://www.collegeanswerz.com/university-of-pittsburgh/ is the only school with answers right now. I'm working on getting other schools.
My goal is wide in scope: to build a site that allows you to take a few minutes to have a positive social interaction with friends and strangers. An anti-Zynga, if you like.
I still have a lot to do. With a goal of universal accessibility, I am working on progressive enhancement - making it work adequately for users with poor connections and noscript, and excellently for users with fancy AJAX.
It's cool and I think it fills a gap for people who want social games without spam viral marketing... as an anecdote: by playing Boggle online, I made friends and ended up visiting a couple on a different continent. I'd like to enable stories like that.
If you sign up at http://xoxo.gl/ with an e-mail address then I'll let you know when there's more to see.
Otherwise you can contact me for a chat ronald@xoxo.gl
http://www.drugcite.com/
We use several data sets including FAERS, Meddra, Medical Device Data, UMLS/RXNORM and DAILYMED/Structured Product Labels to name a few.