Atlas Beta Program Launched | Hacker News Reader

Atlas Beta Program Launched

atlas-beta.heroku.com

46 pointswooby38 comments

I've been playing with this tonight. I had absolutely no problem paying for the beta - I've paid Microsoft and Apple for betas and they need the cash a lot less than these guys.

That said, I wouldn't quite recommend it if you're on the fence or just want a quick intro to Cappuccino. It is a beta -- there are a lot of polish issues with the cocoa app, there are some bugs, and these are understandable. But the biggest problem of all is the total dearth of documentation. There is a class guide, there are some non-Atlas tutorials, and that's about it. So, for instance, my app template comes up by default with a menubar, but there's no obvious way I can see to remove or edit it. MainMenu.cib has no menu in it, so I can't even see how to make connections to it and there's nothing to read to find out.

It's definitely impressive -- that they've done this in JS and running in a browser is a little marvel. However, I'm pretty used to Xcode and Interface Builder, and in comparison it's really not there. Trying to work with multiple embedded split-views really highlighted the differences: it's tricky enough in IB, and proved almost impossible here. That they've achieved so much does mean that you're into uncanny valley territory; if it looks like a cocoa app, I want it to work like one too.

However, the things I could figure out how to do, worked well. I defined outlets and actions, hooked them up, and was basically thinking and coding exactly as I would in Cocoa and Objective-C. Nifty.

It has got a fair way to go, but I think they well might get there. If you're already working with Cappuccino and don't need documentation, it could well be worth your time.

Thanks for the feedback. Accurate assessment. This is definitely still too early to use for most people. We're not really trying to push the beta, we're just looking for dedicated users to help give us great feedback.

Congrats you guys on launching the beta. I've been following Cappuccino for a while and, while I'm still on the fence about it as a developer, it is certainly an impressive technical feat.

Forgive my complete cluelessness, but what is "Atlas Beta"? I can't find any simple explanation for what this actually is.

http://280atlas.com/

I was wondering the same thing. From the actual website:

"Atlas is a development tool for building Cappuccino applications.

In addition to managing your project files and editing code, it includes a powerful visual layout tool for designing your interfaces without ever having to touch code.

Designers are empowered to interact with their designs instantly, which means programmers get to finished applications faster."

This confused me as well. http://280atlas.com/what.php says:

Atlas is a development tool for building Cappuccino applications.

Ah, and Cappuccino is apparently a clone of Apple's Cocoa in Javascript, using a preprocessor for Javascript called Objective J.

This sounds like the same approach Microsoft ASP.NET has: try to fit web apps in the desktop GUI event paradigm. Is this correct?

Yes. It doesn't work for every app.

Not exactly. Cappuchino tries to do it on the client side, while classical ASP.NET does it on the server side. The former has its uses, the latter is a terrible idea.

Sadly this seems to be more and more the norm. Someone launches their Foozwall product and a link gets on HN, but the website's page doesn't actually tell you what the Foozwall is.

I was really looking forward to signing up for the beta this evening, but unfortunately I'm on Windows and the beta is currently Mac only.

I'm surprised that more people haven't heard of 280 North or Objective-J and Cappuccino in general. I would consider myself moderately inclined when it comes to technology and I had heard of Cappuccino close to a year ago. I watched the demo of Atlas some weeks back and I was really impressed with what they've accomplished.

I think they're doing the right thing by charging for the beta. It's a small team of guys doing the development and I'm sure they could use the money. Not to mention, if you're paying for a beta then you've likely been following the development, have serious intent on using the technology long term and will likely contribute more than a casual user would - which from what I gather is exactly the target market they're after.

Looking forward to windows support in the coming month(s).

How is a signup form meant to be interesting? Perhaps I'm just out of the loop, but the only thing I know about Atlas now is that it's associated with Heroku and it will cost more than $20. That's not enough.

We're just hosted on heroku. If you'd like more info, there are a few videos available on http://280atlas.com

Don't worry though, this is a beta, and it's meant for people who already know what Atlas is and already want it.

Thanks. A blog post or something when you guys were ready would be great. But this link is a bit premature and not very interesting.

"Membership does not include a copy of the final version of Atlas, but we'll deduct the $20 from the final price."

Has there been any indication as to the neighborhood of the final price? I'd hate to invest a bunch of time learning a new tool to find out that it's priced outside of my budget when it releases.

We haven't set a final price yet. We're hoping to use our beta members as a sounding board for determining the right price point.

It's worth noting that all of the stuff Atlas builds is still usable without Atlas, though there will be no way to edit it without Atlas.

Of course, most of the editing technology is also open source, so building a cib2nib (the reverse of our nib2cib) should be very possible, and give you alternatives.

I think it looks like a really cool product; I definitely want to use it. You guys are making the right decision by charging for it for sure. My only concern is that I don't know how 280 North views pricing, so it gives me a little bit of pause.

Either way, I'm looking forward to seeing how it shapes up - good luck!

I can tell you that it will be a lot more than $20, and it will be less $1000. Big range, I know.

I take it will be a one time charge, not a subscription model?

Yeah, Atlas 1.0 is a downloadable app with a one time charge. Eventually we will have an online component.

Less than $1000? For any per-seat pricing that nears $1k, Atlas needs be so much better than the current tools that it'll shoot sunshine and rainbows out of our asses while we're using it.

Investing in Atlas is making a heavy investment in technology solely provided by yourselves: Obj-J and Cappuccino. The foundation being open source may alleviate some concern there, but to do so, the open source foundation will need to be of sufficient quality and usefulness that we feel we can work without your IDE in the future.

I'm also VERY surprised that you're charging for an expiring beta. Doing so will greatly limit user exposure, and I'm pretty uncomfortable with paying for the privilege of testing your software.

You're not Apple -- developers don't need to pay for the ability to test/update their software for an upcoming OS release.

Well, it's as unlikely to cost $1000 as it is to cost $20. That being said, Atlas is completely optional. Everything built with Cappuccino to date did NOT use Atlas. And this will always be a viable option.

This beta is not intended to get large exposure. It's intended to get a group of people who we are confident will be dedicated to helping us make Atlas a better product.

You wouldn't be the first company to start with these intentions, only to eventually widen the open source/commercial split, where features never reach the open source edition. It's a possibility, however remote you might consider it, that we have to anticipate and plan for.

If the tools save time and money, we'll buy it, but we also have to consider potential lock-in costs a very real part of the price.

This beta is not intended to get large exposure. It's intended to get a group of people who we are confident will be dedicated to helping us make Atlas a better product.

If we're supposed to be helping you, it seems inappropriate to charge.

If you feel it's inappropriate, please don't pay. We feel that we're providing something that a certain group of people want at a very reasonable price. If you don't think it's a fair trade, we'd rather have you wait until the final version to re-evaluate the product. Thanks.

You're right, I don't think it's a fair trade -- I'm not going to pay money to beta test your admittedly unstable, rough-hewn software.

I'm not telling you this to be offensive. I'm doing so because it's clear you're alienating people who would both test and evangelize, and I think that's an unfortunate waste for a product I think has great potential (but is much, much too veiled in secrecy and a general lack of public information).

If you think the opinion isn't valuable, you're also just as free to ignore it as you see fit (... but given that I've contributed patches to Cappuccino in the past, I'm pretty surprised you don't think I'm in your 'target audience' just because I'm not excited to pay $20 AND test your software for free.)

boucher seems to be running into some pretty stiff resistance here, and I totally get it. That said, I opted to pay for the beta. While it probably should have been marketed as an alpha instead, it was worth much more than $20. Really, many people probably blow more than that on coffee in day. The Atlas alpha/beta appears likely to provide me with more value than an afternoon's worth of coffee, so I subscribed.

If doing that doesn't float your boat, then working with rough, buggy, pre-release software that may never love you back probably isn't going to be your thing either. Then you're out $20 and wish you'd bought the coffee. For their intended market, I think this program was priced right.

I can see charging a modest amount for a beta program. If it were totally free, you'd get sign-ups from any random person who has the slightest curiosity in the product, and any feedback you get from such folks is likely to be next to useless. By charging for the beta, even just $20, you weed out this chaff, and also your beta testers now have some sense of investment in the product and will be more likely to give you feedback you can act on.

But aren't you punishing your target audience by making them pay money, no matter the amount, so you can "weed out the chaff?" If that is their motivation, I don't like it.

Like you, I'm concerned for the future of Cappuccino. I realize that 280 North has to make rent somehow, but dividing the platform into a free software component and a commercial proprietary component necessarily gives them incentive to focus on the proprietary component more. This is the sort of thing that can prevent the platform from ever gaining critical mass among developers.

Wow. Even NEXTSTEP, the whole OS, and dev tools, was less than $1000 I think. I may be remembering the .edu pricing.

And most of the $20 to $1000 range is less than $1000 also. Please realize they haven't set a price yet, and "between $20 and $1000" was apparently a big enough range to encompass what boucher thought might be possible.

This comment deserves between 0 and 1000 upvotes.

Hah. You got more than 0 and less than 1000 from me.

Well then IMHO he should have stopped with "we haven't set a price yet". That $1000 figure leaps off the page and causes a lot of potentially unnecessary consternation.

What? Apple's development environment is essentially free (unless you count the cost of the operating system, in which case it's still a lot less than $1,000). Even if you compare this with the iPhone Developer program, it's still way overpriced.

Apple profits from its development program in many other ways. This is 280's business model.

It's impossible for something that has no price to be overpriced. But thanks for the feedback.

Reply on news.ycombinator.com