433 karma · joined April 19, 2012
I’ve spent time in both academic research and industry.
Research is not supposed to be immediately applicable. The goal is to produce new knowledge - more importantly shared knowledge. Publishing is not a bad measure of that. Additionally, ability to secure grants provides incentive to focus on problems others want solved.
No incentive system is perfect, but I don’t really see how this is any different from any organization. And I don’t think it’s fair to judge an entire discipline by the negative examples.
Agreed it needs more attention, but - for academia - I think it's more of an incentive issue than a stigma issue. E.g. harder to benchmark the performance of two algorithms if they don't operate on the same dataset. Also to be fair, research into things like synthetic data mitigates the problem, just in a different way.
The paper you cited is interesting. Thanks for sharing. Hopefully that spawns more focus into understanding the subtleties of each dataset. IIRC Kaggle also had issues around generalizability, but for different reasons.
Anyways it's still early on... but we're currently building tools to help solve this problem. In particular simplifying the data collection / labeling process for vision systems. Would love to chat further w/ anyone interested in providing feedback. Email is sara@viewpointrobotics.com
Check out this tutorial for developing a node addon in VS, and let me know if you have any ideas on how to improve the experience.
http://computer-vision-talks.com/articles/how-to-debug-nodej...
Code and VS are different products, codebases, Etc. But if you see a specific feature you'd love to have in code or v.v., let us know, as we're always considering ways to collaborate wherever it makes sense.
dev/person-in-video here, happy to answer any questions we couldn't get to in the video :)
GitHub repo for those interested in playing around with the code. https://github.com/Microsoft/nodejstools
Let us know if you run into any issues!
Please comment so we can better prioritize the es6 feature work.
GitHub link: https://github.com/Microsoft/nodejstools
You can use the "From Existing Node.js Code" project template to create a project from a folder, but the project file itself is a requirement for visual studio and also allows us to include helpful metadata (whether or not to analyze a directory, etc - most IDEs/text editors include a project file for this very purpose - they just hide it a little better). That said, at some point we'd like to change things up so that the project does feel more similar to the folder/file experience you might expect.
This is to change the project file to be more "transparent". https://nodejstools.codeplex.com/workitem/1855
There's also another issue to directly open a folder without requiring the user to take steps to create the solution/project. https://nodejstools.codeplex.com/workitem/209
Additionally, like rlp mentioned - make sure that "show all files" is turned on so that you can see everything in the folder, not just files you've defined as being a part of your project.
Hope that helps - let me know if you have any other questions.
With regards to the technicalities of max_path itself... the issue is with the .NET file I/O API, which doesn't allow you to use \\?\ paths to work around it, and is too big for our team in particular to do much about. Here's a good article that might provide a little more context (it's a little old, and these "plans" they speak of never actually materialized.)
http://blogs.msdn.com/b/bclteam/archive/2007/02/13/long-path...
That said... we don't like it either, and we are still thinking about what we can do.
Thoughts? I'm especially curious to hear how you think we can improve the existing experience so that it caters better to your workflow. What are some of the abstractions that get in your way?
http://azure.microsoft.com/en-us/documentation/articles/web-...
- connect is for filing bugs
- uservoice is for "ideas and suggestions"
How people use it, of course, is a completely different story...
And then some teams (like us) have public issue trackers, which are the best approach if available because everything goes directly to the team: https://nodejstools.codeplex.com/workitem/list/basic
Otherwise, yeah - we don't like it either... here's the not-closed uservoice request ;-) https://windows.uservoice.com/forums/265757-windows-feature-...
EDIT: Re: fish: Indeed - we have some big fish to fry, but luckily msft is a big company that can fry multiple fish at a time. Personally, I'm excited to help improve the Node tooling story by frying this fish, and passing our learnings and feedback onto the the other teams @ msft so that they can improve their recipes. Ultimately, we strive to have a complementary set of fried fish. And maybe some vegetables too... :)
We have SO MANY PLANS, so please upvote or request the ones you think are important :) https://nodejstools.codeplex.com/workitem/list/basic
I don't think you need one of the paid versions, but I'll double check and edit this if I'm wrong.
1) warnings when you hit max_path that offer to run 'npm dedupe'
2) sometimes that doesn't work, so we also suggest people start their projects in a short path like c:\src (embarassing, I know)
3) patiently await a flat node_modules directory - http://blog.npmjs.org/post/111968476155/npm-weekly-6
4) groan very loudly when it happens - no good keeping those emotions bottled up ;-)
luckily #1/#2 cover most cases
This appears to be consistent with npm's data as well (see this infographic - % devs accessing npmjs.org http://blog.modulus.io/growth-of-npm-infographic).
We also try very hard to make sure that NTVS works well with people's existing tools and packages - so if you run into an issue, please let us know (or better yet, make a PR!)
Now watch this: My HN account? Nope... sorry to inform you, but I'm just weird. Anyone who knows me will attest to that. I'm generally far too full of energy, but depending on the occasion oscillate between a playful and pensive writing style. Here's my quora page, and a blog post I did earlier this year if you want to cross-check. You'll notice there's a healthy mix of both "caffeinated" and "decaffeinated" Sara. Let me know if you'd like to see any more writing samples :). - https://www.quora.com/Sara-Itani - http://blogs.msdn.com/b/lightswitch/archive/2014/04/17/scala...
Waitaminute... What just happened?!! CAN'T RESOLVE... COGNITIVE... DISSONANCE... MIND... IMPLODING!!!!
It's not so hard, actually. Both writing styles are pretty similar - one is simply more staccato and contains over-the-top faux-dialog with the reader. Usually if I'm trying to construct a convincing argument or reason about an idea while I'm writing it, my writing will turn out more pensive. If I'm super excited (as I was when people were posting awesome feedback,) or trying to get/keep someone's attention - it'll be more energetic/playful. If I want to do both, I'll combine the two.
Make sense? Awesome.
Apologies for the confusion :-). You may now return to your regularly scheduled HN thread.
EDIT: oops - got distracted trying to prove I'm me, and forgot to address the more important part of your message. Thanks so much for the feedback, even if you couldn't take the survey :)