My impression is that this app is looking for such a small overlap in demographics. It's looking for programmers who might not have the tools or the knowledge of deploying their own apps, which I think is relatively unlikely to find. I think programmers capable of writing their own apps would prefer to use their own tools. OH, BUT THAT'S JUST MY TWO CENTS.
I think there's a very wide spectrum of people who can appreciate AppJet. I would argue that most programmers don't have the knowledge to deploy a web app, especially if you define a "programmer" as someone who can hack together an AppJet app, which is very easy to do. Even for a very capable programmer, it's a lot of work to get an app hosted if you haven't done it before. Then there's programmers who CAN get a web app online, and have, but don't want to manage all the machines and software involved for every little project.
Before sites like Flickr, for example, it was possible to post photos online, just harder. Maybe the difference in perception is that hackers aren't expected to want such an "easy way out" when it comes to programming, their very area of expertise. But as a hacker myself I can tell you that using AppJet to whip up a quick app is a pleasure compared to going through all the typical production work.
I agree, and I find this idea fascinating. You can write code, and show testeable versions of it, together with their source to anyone who cares. And you can do some small code apps for your websites... I think there are plenty of applications for this tool.
You might very well be right that the overlap's small, but note that AppJet is potentially for programmers who do have knowledge about deployment/hosting issues. How so? Well, such a programmer might not want to deal with hosting issues for a particular project. If the project s/he is attempting really is 100 LOC, then getting the app all nice and hosted and scaleable might take an order of magnitude longer than the actual coding does. AppJet, if it delivers nicely, could be a real win here.
"Hosting issues", although I am not entirely sure what are you referring to, become issues only when a certain level of complexity is reached. By the time you get there, you already implemented several appjets yourself.
I mean the term "hosting issues" to be pretty general -- implementing a data storage system (perhaps a database), caching, setting up server space, getting Apache to proxy to your app server, etc.
One the one hand it is interesting, I guess mostly for mini-applications. On the other hand there is the risk that you can't take your app and run it elsewhere, you bind yourself to AppJet.
We don't want to get in the way of people hosting their own apps. We will at least release binaries for doing your own app serving and local development. You can already grab the source to the JS micro-framework. More details in our interview with Ajaxian: http://ajaxian.com/archives/appjet-develop-preview-deploy-al... - see the "Aren't you tied to the platform" question.
I love how close this is to what we (Ning) started with, down to the cloning terminology! Really interested to see where it goes. The other "JS Everywhere" folk-programming startup (name escapes me) went tits up recently :-( Not sure why.
not sure I get this appjet concept..If I have to do that much coding to build an app using appjet, what makes it better than the 200 million ways that already exist to do this?
I had the same question, but my interpretation of the claims on their site (http://appjet.com/about/) was that you'd save time if the app was simple enough, because you wouldn't have to configure a server and do all that nonsense. This is fair, I think, but the question becomes, how big is the set of applications that are both interesting and not so complicated that it'd be more appropriate to use a more traditional environment/language where you have more power/control.
I don't see any practical limit to the set of interesting apps that can be written in 100 lines of an expressive programming language like JavaScript. Of all the programs I have ever written in python, for example, I would wager that most were fewer than 100 lines. I wish these were all online instead of trapped on my hard disk.
Random thought: how big is the set of interesting essays you can write in 100 lines of English?
Our prediction is that there is a long-tail of web apps. We want to be kind of like the pen and paper of web apps. Maybe not great for writing a novel, but excellent for smaller works.
The key is the trivial deployment. If a user wants a little app, and he has no server and no hosting provider, and he doesn't know how to install software or configure a shared hosting account anyway, then he is likely to just give up before he even gets started. This at least gives him a chance.
There are a lot of business people that have become programmers by building increasingly sophisticated Excel spreadsheets or Access databases. This type of "casual" programming is becoming increasingly common.
That said, I think that DabbleDB might have a more successful approach--data first, instead of code first. That approach made MS Access very successful. Casual programmers care a lot about their data, but code is just a nuisance that is tolerated to get the data to "work." People want assurance that their data is always there, even when the code is broken. In MS Access it is trivial to back up your database and view/hand-edit your data.
how big is the set of applications that are both interesting and not so complicated that it'd be more appropriate to use a more traditional environment/language where you have more power/control.
Basically every Facebook app, for one.
There's also a big niche for hobby programmers who want to make one-off online tools. The code sharing makes that even easier.
Client-side Javascript is the QBasic of this decade--it is the first language of a lot of (most?) self-taught programmers. It makes a lot of sense to offer these people a place to expand into server-side software.
And, Server-side Javascript is spiking in popularity and utility. I am working on a project in this space as well.
I think this is a great idea. There is a lot of overhead in build web applications, and hosting is definatly one of them. Although I haven't checked out appjet yet, something that simplifies the process might would be valuable.
Niiice! It's a lot like heroku.com, but faster. With a rails like framework for serverside javascript with templates and database support, appjet could be the next big thing for small and medium sized sites.
You are telling me that I have to learn to code in Javascript? Are you serious? I've been avoiding Javascript for years. Abstractions such as libraries make this possible. This seems like a step backward.
What I saw "Make Simple Web Apps A Breeze," I expected Infogami done correctly. Drag and drop some buttons to make a website, publish it to an account, and voila, I have a website.
Don't get me wrong, the premise is very cool. It gets my computer science geekery excited. But this is a business here, and it seems like a tool built for geeks by geeks. I don't know what the business model, but based on every other venture YC has backed, it depends on the typical Web 2.0 paradigm: get users --> get funding --> ultimately get ad-based revenue or other branch out to other forms. If your only target is geeks, then good luck. Otherwise, I suggest a little more ease-of-use.
From the few examples I quickly browsed, nothing is out of the ordinary; a chat program, a wiki, a "notes" page, etc... All common Web paradigms that can be abstracted.
Again, very cool, and I look forward to playing around with it once my semester ends. But at the moment, I don't think anybody but the hardcore techies will enjoy it.
Interesting stuff there, I may have a poke around later.
Like the simple storage API too.
It seems against philosophy of this thing, but I'd like to see something similar that allowed for more separation of code and html/js/css etc. (I think I'm drifting into regular hosting territory though).
what about if you guys get together with the weebly guys and make an easy to use code-if-you-want drag-and-drop-if-you-don't web application maker (or back to viaweb but meta)?
This may not be the greatest website since sliced bread, but that "cool firefox pricing plugin" doesn't exactly have a lot of room for growth, whereas clearly there is plenty of room here.
Agreed. Although I wonder what the future plans are for this thing. It has some potential for virtual serving of web apps, but right now it seems more gimmicky than anything to me.
Congrats on launching.