Hacker News new | comments | ask | show | jobs | submit login
Light Table is in YCS12 (chris-granger.com)
401 points by ibdknox on May 18, 2012 | hide | past | web | favorite | 96 comments



I have a comment that I thought about since I read this in the morning but could be perhaps interpreted as negative, but I'm really curious as to what people are thinking about this:

I think this news is great, the quicker innovative editing tools like LT appear the better. But I have an uneasy feeling about an incubator (even though YC is the best one, IMO) or a VC scooping up projects from Kickstarter. I always thought KS was a way for the intelligent nobodies to bypass the whole incubator-VC-whatever system and to appeal directly to the masses. How to put it, there's a sense of Olympic amateurishness (in the original sense of the word) in KS projects. When they are picked up like this it sorts of negates that sense.

What do you think?


Hmm... that's an interesting perspective, I'm of the opposite opinion. I do get the sentiment though.

There's a big difference between imagining something might be a great idea because of intuition, A/B testing/market research, even launching and getting engagement but no monetization... vs. getting people to actually put their hard-earned dollars into it in advance of any available product.

To me, there's no stronger 'go' signal one can have. To an investor it takes away all the mystery... do you even need a pitch? The only thing left is whether or not the team can execute on the vision. And hence the value of a great incubator to reduce friction on that end.


Honestly, YCombinator provides near-zero money. It's a strategic investment because they gain networking, a position in the valley, and access to the best programmers. It will make deployment easier. Now, if they need a lawyer, an accountant, a new staff member, or anything else - the YCombinator backing will make everything smoother so that they can focus on the core product.


This is happening. I personally know of 3 Kickstarter -> VC-backed startups, and I'm sure it's more widespread than that.

Kickstarter serves projects that are on a continuum from "pure" art (uncommercializable) to pre-sales (already commercial), so a transition to a full-fledged commercial endeavor makes a lot of sense for many many projects.

There will still be a lot of amazing Olympic amateur projects, I hope/expect!


I've actually seen quite a few Kickstarter projects where VCs have used the Kickstarter as a gauge of consumer interest. In this, it still serves the same function I would like it to serve - it lets creators bypass the VC/publisher barrier of "we haven't seen something like this work, so we're not funding it", by showing in advance that there is market demand.


I think it depends on the project. With any sufficient amount of talent behind something this ambitious, the $200k they raised on Kickstarter will go quite fast...on one hand, they should take the VC money while they are "hot" - but on the other, if they could ride out their funds they would probably have more leverage with a stable-ish product in place.


  While being a part of YC doesn't affect our plans or change
  our use of Kickstarter as a means of accelerating the
  release of Light Table, many pledged with the explicit
  purpose of making sure that it had a chance. With YC that
  chance is now safe and we wanted to let our supporters know
  there will be no hard feelings if this changes your desire
  to pledge.
Light Table uses Kickstarter to demonstrate market demand, YC responds to demand with funding, Light Table effectively offers to buy back Kickstarter pledges with YC money just to be nice. Wacky.

There are a lot of ways to read the situation, but my favorite is that Light Table makes sense purely as a strategic acquisition for YC; there are enough YC companies using the languages Light Table targets that, should it yield magical productivity gains, the knock-on effects from being able to get their developers early access and input would be enormous. If they wind up changing the future of software development, all the better.


Please add double-lines to your comment so it's readable, thanks! :)


Hah, sorry. Forgot to read my own writing!


>There are a lot of ways to read the situation, but my favorite is that Light Table makes sense purely as a strategic acquisition for YC; there are enough YC companies using the languages Light Table targets that, should it yield magical productivity gains, the knock-on effects from being able to get their developers early access and input would be enormous. If they wind up changing the future of software development, all the better.*

This sounds like a really lame conspiracy theory. YC got LT to "yield magical productivity gains" to it's companies? Please. Even if the programmers dropped their editors (Emacs, etc) to adopt LT, the productivity gains would be marginal. And code productivity at that level is not the bottleneck for a startup success either.

A decent product people want to use, connections, and chance are far more important. For all the "we used Lisp so we did it faster/better" bragging, Yahoo could almost just as easily buy any other ViaWeb competitor for example, even one using the lamest of methodologies.

Downvotes? People believe that using a better editor/IDE will yield "magical productivity gains" for a company?

Or that ViaWeb succeeded because it was in Lisp? Well, how about the other 99% of companies, that succeeded based on lame technologies, from Facebook's PHP, to whatever. One company doing something successfully among thousands is as far from a proof as you can get. Post-fact rationalization sounds more like it ("Oh, we got bought. It was because we were ultra-productive programmers yielding Lisp").


While I sort-of agree with your point, I think you're being unnecessarily hostile here.


I might come of a little hostile yes. I dislike the cult aspect of the whole story, that is.

And I think the "ViaWeb/Lisp -> super programming capabilities over mere mortals -> Yahoo millions" thing that started the whole cult worship is both wrong and flawed (e.g statistically insignificant to prove anything).


But why does it anger you so much?

Besides, you seem to be ignoring that ViaWeb was built before there were ubiquitous open source libraries for practically everything. My guess is that they basically had to write the whole server/application/persistence/security/payment stack themselves from scratch. Considering that they were able to build it all with a handful (2-4?) of programmers and to beat out competition that had a head start and a much higher head count, I would wager that Lisp had at least a little something to do with it.


>But why does it anger you so much?

So much? I just made a comment, it's not like I'm running around foaming at the mouth, is it?

>Besides, you seem to be ignoring that ViaWeb was built before there were ubiquitous open source libraries for practically everything. My guess is that they basically had to write the whole server/application/persistence/security/payment stack themselves from scratch. Considering that they were able to build it all with a handful (2-4?) of programmers and to beat out competition that had a head start and a much higher head count, I would wager that Lisp had at least a little something to do with it.

I agree. But wouldn't something like Smalltalk be equally good for the same purpose? Or Perl?


I will agree with that the idea that the cult of Lisp started because Viaweb proved that super programming capabilities are better than mere mortals is both wrong and flawed.


Wow, man. Did a Lisp eat your cat or something?

Maybe you didn't, but a lot of people saw the Light Table demo and thought, "If that thing works right it looks like it might cut my workload in half." This would be the reason (you may have heard) they've raised a few hundred thousand in pre-orders in addition to funding from the most prominent startup accelerator in tech.

Okay, so you don't buy the hype. Bully for you. But "a really lame conspiracy theory"? Presumably, not only would Viaweb have been as dominant were it written in C, but 37signals would have been more successful sticking with PHP instead of wasting time writing Rails, and if all your Emacs users switched to using Notepad, it wouldn't cost the company any money.

That sounds pretty realistic.


>Wow, man. Did a Lisp eat your cat or something?

No, I even like it. It just didn't make my cat 2x more productive as a programmer, which is what I find a BS statement coming from many Lisp proponents...

>Maybe you didn't, but a lot of people saw the Light Table demo and thought, "If that thing works right it looks like it might cut my workload in half." This would be the reason (you may have heard) they've raised a few hundred thousand in pre-orders in addition to funding from the most prominent startup accelerator in tech.

I like the LT demo. I don't think in the least it will cut my workload in half --those that think so are IMNSHO delluded. At best it will make certain operations easier to carry out, but most of them will be marginal to my efficiency. Editing environment is not the bottleneck to my coding, thinking is. That's why thousands of programmers work just fine with crude old Vim and not something like a Smalltalk editor or a Lisp Machine (or even Emacs or VS or Eclipse). A competent programmer can get by even with vi or Notepad++, with only small loss of productivity.

>Okay, so you don't buy the hype. Bully for you. But "a really lame conspiracy theory"? Presumably, not only would Viaweb have been as dominant were it written in C, but 37signals would have been more successful sticking with PHP instead of wasting time writing Rails, and if all your Emacs users switched to using Notepad, it wouldn't cost the company any money. That sounds pretty realistic.

Actually there were programmers AS succesful at the time as ViaWeb that were INDEED writing in C. Case in point, Yahoo, the company that bought ViaWeb in the first place, used C. Later switched to PHP IIRC.

But, crucial to my actual point was the use of the world "marginal". C to Ruby might make a difference. Ruby to Lisp not so much (it could even be adverse, because of the lack of mature web frameworks and libs that Ruby has plently).

And certainly Emacs to LT, even less so.


So IDEs don't, can't, make programmers more productive enough to offset the cost of developing the IDE itself?

I mean, you can think that. But it ain't exactly Moon Science to think otherwise.


We normally would wait to announce this until it was time to drum up the press machine, but it was more important to us that you guys know what's going on.

If anyone has specific questions/concerns, I'll be lurking.


I initially donated to make sure light table becomes a reality, but I will not withdraw my donation because I think this is a very worthwhile project and I still feel proud to support it. In other words: SHUT UP AND TAKE MY MONEY!!!


I was planning to contribute towards the end of the campaign, but now that the kickstarter is successful AND YC is taking it under it's wings, I'm going to wait to be a retail customer.

Wish you the best of luck, this project is very interesting.


Seconded! I think you guys are building something great here, and I can't wait to take it for a spin. Keep it up!


Did you apply to YC, or were you invited to YC?

(I am curious how the process works for startups that already have significant traction before joining YC. Humble Bundle comes to mind.)


We had actually applied previously with a totally different idea, but then Light Table blew up and we resubmitted late.


I'd always been curious about what needs to happen for a late application to be accepted. Thanks for the transparency!


When we realized that multiple YC partners had already independently contributed to the Kickstarter project because they wanted to use Light Table themselves, it was not a hard decision.


I haven't seen Nathan talking about LT much. I'm assuming he's on the team from your previous YC attempt, am I correct? I was going to congratulate both of you, but I wasn't 100% sure.


Actually it's with another friend from high school :) He's dealing with trying to move out here right now, but you'll be seeing him a lot more - ibdthor is his username.


Question. Did this:

> I've been quiet over the past couple weeks as I've been putting together a demo to show the real power of what's to come and I can't wait to show it to you all. I'll give you a little hint: I've been making a lot of sssSSSSssssSSssss noises lately ;)

ever get released? Was the demo for the YC?


I assume it's going to have Python sssSSssSSssupport.


Alternatively hacking Minecraft on-the-fly via Clojure.


Look for it at the beginning of next week :)


We normally would wait to ... we believe we have the best possible chance ... We rallied together ... we wanted to let our supporters know ... We want to thank you all ... we were able to come together ... we can work to make it a reality.

Who is this "we"? Is it just you, Chris?


If it is just him, you may call this an old habit of an enterprise past.


I pledged because a) I like the fact that you're thinking outside the box, b) you're building something I wanted to build, as I think all IDEs are obsolete, c) you're building it for my favorite language first (Clojure) and d) I want it built as quickly as possible.

None of these reasons became invalid when you joined YC, so the pledge still stands.


Are the release dates for Kickstarter backers changing at all?


If they do, they'll be getting better, not worse.


Well do not even think of the term 'delay' or 'slip in schedule'. You have beaten up enough hype to keep people on toes expecting this thing to come out.

In fact it will be very interesting if we see commit logs on something like Github. A lot of people follow a certain project(s) which they like.

That may also help you test and flush out bugs as needed while soliciting feedback.

All the best and expecting to see a solid product out!


Great. Thanks for the transparency.


Epic! Alpha preview for Alums? ;)


The YC backing actually made me pledge. I held back initially figuring that I would easily be able to buy a license further down the road without taking the risk of seeing you guys give up (no offense intended!) However, getting into YC makes me appreciate how very serious you are and so I treated myself to an early beta.

Good luck, I'm REALLY looking forward to trying Light Table out :)


Ditto. If the timeline for releases accelerates, I want to have early access to this awesome tool!


Does this mean Python support in version 1 is more or less likely?


We're sticking with our goal on Kickstarter - if we hit 300k we'll definitely do it. There's a decent chance we will either way though.


Thanks for the reply :-) Another question if I may: YC funding amounts to $5k - what difference will that drop in the ocean make?


In this case YC is not about the funding but about the support and connections.


The end of Chris' post a week ago hints at Python support. What makes a hissing sound other than a python?

http://www.chris-granger.com/2012/05/12/we-made-it/


My question too. I backed Light Table with the hope they would hit the Python milestone.


I won't back them based on a maybe. We'll talk again when python support becomes more than a teaser :)


Wouldn't you think of this as slightly contrary to the purpose of backing a Kickstarter project?

What I mean is, say the project had been set at 300k and python was one of the base supported languages. That would actually be no different than it is now, except that now the software itself is already guaranteed(1).

Also, the python inclusion is certainly on the roadmap if it's mentioned before the project has been funded, right? So if you back it, then you're still probably backing a(n eventually) python-speaking platform.

1 - [Edit]: Well, "guaranteed", no. Funded, yes.


When backing a kickstarter, I pay for what is advertised and described. If they make the hard deadline, they will do it. If not, I'll get my money back. There is no such guarantee for Python support. I fund them, maybe I'll get Python support.


I dont see how light table is different than an good ide for java.

- docs everywhere

When i hover over a symbol, method or class i get the documentation right in my IDE

- instant feedback

Ok this is new but can only be done for simple calculations

- table as a unit of work

Saying that "we can de better than files" is true. In fact in my IDE i just dragndrop methods around, navigate the class hierarchy, etc. Its just coincidence that every class is represented as a file. But i do not care. The point is that i write code against the AST and not against "plain text". The IDE parses the code on the spot. This is nothing new.

So all in all i dont see a lot of new concepts.


Light Table isn't a Java IDE, so this isn't a fair comparison.

For a dynamic language, where a REPL will no doubt be part of your workflow, an IDE that aims to merge the process of editing and interacting is probably quite valuable. Like emacs with slime but less daunting to the unfamiliar.


This is some great news! I hope that will make Ruby and CoffeeScript support come sooner! :)


Congratulations!

Do you anticipate using the SLIME/SWANK style backends? Reason why I ask is because I do a lot of Common Lisp coding and this would be an interesting project to give a spin.


Does this change the story on it being open source?


Nope! The plan is still for the platform to be open source, likely under a GPLish license.


What will a "license" for LT then include?


The Kickstarter says that it might get you access to a packaged build, and that the price might be pay-what-you-want -- sounds a lot like the NeoOffice business model ( https://www.neooffice.org/neojava/en/mirrors.php?file=NeoOff... ).

But making money from open source projects is hard enough, and varied enough, that it makes sense to spend a lot of time thinking about and experimenting with this one before committing to anything. It looks like that's what they're doing. So I don't think there will be a firm answer to this until they're much closer to launch -- or that it would be a good idea to offer one.


If LightTable wants to be the 'Eclipse' of clojure and Python then it needs to be open source.

Mass adoption, patches, test cases, design, feedback and money comes only when it reaches crazy adoption and a lot of people hack on it.

That is why it is important for LightTable to release a minimal usable version ASAP and then improve it in further releases.

All this needs a open source set up. A Cathedral set up earlier but a bazaar set up later.


Indeed, I'm very curious about this one as well.

Going for YC is clearly a great starting point for building a scalable business, so I wonder what the business logic will look like. As in genuinely, positively interested to know. :-)


Please consider a BSD-style license. There are lots of people that can't or won't use GPL-licensed projects.

Alternatively, will you be releasing under a commercial license?


There are people who can't or won't use GPL code in their own projects our use GPL libraries, but I've never heard of anyone not using a program (rather than the code) because of a GPL license.

For a project like this, the GPL makes the most sense: an IDE needs people who use it and people who contribute to the core project much more than it needs people who use its code in proprietary projects of their own.


Eventually every successful IDE becomes extensible. Up to that point GPL is perfectly fine.

As soon as IDE provides extensibility one starts having problems with GPL and it's viral nature with linking.

GPL is forcing IDE into a "sandbox" where the whole distro must be GPL-compatible at the point of distribution. The workaround is often with the expense of user experience where one has to separately install extensions just because of licesnsing.


The Mozilla Public License v2 is a good compromise, as it is less aggressive than the GPL on links (it operates at the file level)


Ugh, please avoid the GPL. Go for Apache or BSD or something more free.


Like someone stated, it's a dev tool that will not likely need to be bundled with your end product. I use Emacs, which is also GPL, as do other people, with no hinderance from a license point of view.

If I were doing Light Table, I'd use GPL too - It also does not preclude someone from writing their own plugins and selling them as an add-on. I would assume there will be a way for easy/seamless plugin installation into the main program.


It's a standalone piece of software, not a library. How does it being GPL affect you?


Well, for one, you can't make small modifications to the code and then resell it as a commercial product with the original author's copyrights tucked away in an about box. You'd actually have to release the source code too. It will be harder to convince customers you've added value if they can see exactly what you've changed.


I have to know: Was this is an ironic statement?

No one is taking it as one, but poe's law aside I'm convinced they are wrong.


"It will be harder to convince customers you've added value if they can see exactly what you've changed."

I don't get it... is your added value simply an illusion?

If they can see what you've changed the code, they should be able to see why and how you added value.


We have some ideas on how we'll facilitate that. There will definitely be ways for people to sell plugins they create and or for large companies to keep them closed source.


Grandparent was being sarcastic:

It will be harder to convince customers you've added value if they can see exactly what you've changed.


Is this sarcastic, or did you just now realize the point of the GPL?


In the long term, the GPL kills the major incentive to improve the software's design and feature set, even though small improvements and ports may be made.

Sprinkling the GPL on top of a project no one wants anyway does nothing (I say this from experience), and applying it to a popular program just holds it back (by keeping it from getting the cash that people would otherwise pay).

As an end-user, I want to see Light Table take off, and to continue innovating, not to continually engage in campaigns for donations.


Chris Granger (@ibdknox) will be doing a talk in September at the Strange Loop conference - https://thestrangeloop.com/sessions/behind-the-mirror.



I appreciate your transparency in this.

I'm keeping my pledge in, though. :-)


OK now, this is a whole new business model and I like it. Early supporters pay for the demo and get a t-shirt. Later supporters pay for the continuing refinement of the demo and get a t-shirt. Later supporters pay for the demo and get a t-shirt a mention somewhere.

When the demo is awesome (but not open to the public) and also the market can be assessed based on early support, someone (who saw the latest demo and the early support) pays for the rest, but what they get is a cut from any potential future cash flows. Now, this, is, neat!!!

Incidentally, this is the problem with Kickstarter - if you want money, don't give me a t-shirt, give me a part of your future profit. Then people who back bad ideas suffer and people who back good ideas profit. As opposed to the current model, where everyone gets a cheap shitty t-shirt and the world gets another worthless r&b record that nobody wants.


I haven't followed the discussions much before, but will this support Android development sometime in the future?


Why exactly would you want support for Android development? As far as I understand, LightTable's target are languages which doesn't already have the tools they're offering (dynamic scripting languages like JS, python etc). The existing Android dev tools already have all the features that LightTable is bringing to them. We already have those in Eclipse, we don't need LightTable. Only the dynamic languages with poor tooling does.

(If you're an Android developer who misses some of the features of LightTable, you should look for some tutorials at www.eclipse.org, you'll learn eclipse can already do those and much more)


If Eclipse encompassed what LightTable wants to achieve, one would only have to write a plugin... I think its trying to be something completely different...


Cool. Now this definitely happens, likely sooner with more out of the box support, and I still get a T-shirt (assuming everyone on Kickstarter doesn't drop their pledges now). So, it's basically a win-win-win situation.


I just made a pledge anyway. Whatever we can do to help make this a reality seems like a worthwhile use of money to me. It looks like a great piece of software.


I am personally much more impressed with Code bubbles which as been around for 2 years and which I shamefully missed at a time of its appearance. Of course it's less talked about, as there is certainly less hype about Java than Clojure and Javascript, but a statically typed language benefits a lot more from this concept, rather than dynamically typed.


seconded. not to take anything away from LT, but i'd really like to see it deal with a large codebase.


I upped my pledge after reading this. Even more excited to be able to see the early betas and play with the future.


This is very exciting news! I see this as being a very good thing for both Light Table and Clojure in general.


Awesome. Way to go Chris!


Now, if only YC can invest in TextMate so that we can have 2.0!


TextMate 2 came out a while ago. It changed its name to "Sublime Text 2" and it is awesome.


You must be talking about the alpha release that was released end of last year? My joke about TM's new release turned out to be not so good.


I'm pretty sure that the GP just made a joke that at least didn't work for you.

The GP is likely referring to this sentiment: "I started to realize that Sublime Text seemed to be everything many of us had hoped for in Textmate 2, but in software that was available today in a polished, fully functional version, not a just a rather buggy alpha preview."

-- http://casperfabricius.com/site/2012/01/24/my-first-week-wit...


If only all these attempts at vim clones were instead spent on vim.


Sublime Text is most definitely not a vim clone. It may have a vi-ish mode, but it's fairly weak. Sublime Text is in the TextMate/Notepad++ space.


Wait did you just say that TextMate wasn't based on vim?




Applications are open for YC Summer 2019

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

Search: