Codeless web development | Hacker News Reader

Codeless web development

codeless.org

18 pointscodeless30 comments

Sorry, I'm not buying into the concept. If selecting "Add -> HTML Element -> Body -> Div" on a menu is meaningful to you, you can write HTML and do it a lot faster than dealing with pulldown menus.

Likewise, if you understand the logic you need to implement well enough to correctly diagram it, you can write the code to do it. And (with experience) probably do it faster. We've had tools that generate code from UML diagrams for a long time, but I don't know anyone who uses them.

There is a lengthy history of Box-and-Line programming languages. Some people find them easier to understand than traditional languages.

[ http://en.wikipedia.org/wiki/Visual_programming_languages ]

But it's still code. "Codeless" is inaccurate.

Pre-order ?? How about a 7 day free trial before we get to the ordering stage?

Mind you, great respect for your focus on the monetization but I predict that with an unfinished product at a $65 per unspecified unit of time you may have a bit of an issue with chargebacks.

I don't think it's very good to have a blank terms page when you're charging money either.

> How about a 7 day free trial before we get to the ordering stage?

Or maybe a cut-down free version.

Thanks for the feedback jacquesm! More information will be provided on the specifics of the service offering shortly.

I don't mean to be overly critical, but I actually laughed out loud when I saw the image of the visual programming interface and read your description of it as "intuitive".

I first read on HN the quote, attributed to Bruce Ediger, "The only 'intuitive' interface is the nipple. After that it's all learned."

Say what you like about your "PHP logic editor": it's no nipple.

I'll try to get some videos up that show it in action since it can be challenging to gauge usability with only screen shots.

This look like an absolute nightmare to use, who wants to drag stuff around in a GUI interface instead of writing some easy to read HTML or script.

The same crowd that frontpage appeals to?

Yeah, but I can't image that crowd wanting to edit logical PHP diagrams. This looks like it is geared towards programmers, not your frontpage user.

Your frontpage user would be better off moving to a CMS like Harmony, SquareSpace, etc.

You would be surprised at what crap people can make using a drag and drop interface, just because people are better off doesn't mean they are more inclined that way.

Yeah, PHP logic editor = the mental models of programmers. It's still code, just trapped in boxes and arrows.

People that want to do it generally would use rapidweaver or similar tool like iweb

I heard this story with LabView: Aka Satan On Wheels for Instrument controls.

I'm going to have to see other people live it, love it, and convince me before I'll even clickthrough

People, please keep in mind that HN is obviously not the target audience for this product. You can put the scathing criticism on the shelf and try to be constructive as to how it might look to the people who this does appeal to (and they do exist). There's no need to get elitist about how great code is.

I think the target audience will be put off by:

   o no information whatsoever about who is behind the site. 
   o no way to evaluate what it would be like to use, just an option 
     to enter a credit card. 
   o "early beta"
   o no success stories or references from at least a few 
     people who have used it to build an application.

Great idea, certainly ambitious, but concerned at how intuitive it actually is to use.

For instance, I'm looking at your Logic Editor screenshots and I have no idea what any of them are actually doing. It seems to be on the minimalist side of graphical languages, like MAX/MSP, but those things are usually targeted toward specialists and not the casual user.

Also, I would need to use your layout editor to really know how useful it is, because doing HTML/CSS layout via a web app has been attempted in the past and just about every effort I know of has not caught on. It turns out laying out a page in a web UI is very hard to do right. Even if the editor is solid, the generated code is never pretty to look at.

I believe Visual Basic was spectacularly successful in part because it did visual layout; and there several tools that do the same thing for webpages.

I guess you're saying the difficulty is doing it in a webapp. Maybe with faster machines and better support etc and RIA this may now be workable. There's certainly no intrinsic reason for it not to work.

The intrinsic reason that web layout is particularly hard to do in an IDE is because the interactions between JS, HTML, and CSS are hard to fully represent in an editor. There is no simplistic, uniform layout system as you'd find in Cocoa or WinForms. I can't think of a way to represent CSS z-index, positioning, floating, rulesets etc. in a GUI without drastically oversimplifying all of them. If you try to fully express the power of CSS in an IDE you basically get a slightly enhanced code editor or a million property panels a la Dreamweaver.

You could also completely throw out the entire architecture of CSS and build your own layout engine. The best example of this is probably Atlas for Cappucino (http://280atlas.com/what.php). The problem is this winds up destroying many of the principles that make the web great like graceful degradation and progressive enhancement. You can't build a webapp in Atlas that will work without JavaScript or CSS, and the generated code is not SEO readable. That may be acceptable for a desktop-replacement web app but my impression is that Codeless is shooting for smaller things like blogs, CMS's, etc. which should be able to function as "good" web content.

I don't know, I will have to wait and see what others think as the pay-for-beta is not an option I will take...

You should know that there are tools such as NOLOH (http://www.noloh.com) that do take into account the evolution and architecture of the web and do gracefully degrade, and are SEO readable. You can also fully use your CSS with your objects if you like, though you don't need to. See: http://dev.noloh.com/#/articles/NOLOH-and-CSS/.

Furthermore, these concepts can in fact be represented in an IDE, once you have the sort of unification NOLOH provides, unfortunately I can't really say much more at the moment.

Disclaimer: I'm a co-founder of NOLOH.

My first impression was that this concept is the iPad of web development. The graphical development approach adds abstraction, which means it could potentially become popular with those with little technical knowledge but just want to (quickly) launch a website/web app (similar to the way that some people who use the iPad just want to type, use internet radio, TV shows etc. without having to mess about on a computer). HN is definitely NOT your target segment.

At least it's drawing a strong reaction, which is good. Far better than a collective 'so what'. Remember the iPod was criticised too when it was first released, and I seem to recall that did ok in the market.

All I see is a smiley. And that's with Chrome, Firefox, and IE8, because obviously other people are seeing something that works.

I feel like the biggest issue with visual programming is that they strive to eliminate code. The visual aspect of GUI design and even logical organization needs be augmented with code, perhaps a simplified pseudo-language, that would really help tie together what's going on beyond box and pointers.

Agreeing with some other comments, I think it's too complicated for the users you would otherwise attract. Being someone who could figure it out, I wouldn't use it.

What about doing pure drag n drop and getting rid of the code diagramming? You could abstract all CRUD operations away starting from the forms.

Or better, why not just build modules? You could do plug n play registration, authentication, comments, etc.

Imagine a drag n drop Django w/o any configuration of the backend that produces standards-compliant code. It would be challenging but not impossible.

And yeah, let me build with it before charging.

I wouldn't use this product and most serious web developers wouldn't. However, I think people here are being too critical given that it's not really designed for us. I can picture a lot of my naive cousins who are always telling me about their (terrible) internet ideas signing up for this and making a go at it.

We don't use FrontPage. We wouldn't use this. But there is a need for these products. Even if the end users don't end up being super successful with their endeavors.

"Codeless exists out of our desire to accelerate software innovation and thus its positive impact on society."

That's horse shit. And it's even worse by implying that you're a team instead of a one-man-show.

You can sell this thing as being cool and easy, but not as being one giant leap for mankind.

I'd be pretty curious to see the resulting code.

Most importantly, does it have Vi key-bindings?

I actually think this is a good idea.

1) I would want a trial before I use it.

2) HN is kind-of the wrong audience. You would probably get more bang for your post at digg or something.

3) While I personally won't use it, I know of a few people who would jump at the opportunity to use this (mostly designers).

Reply on news.ycombinator.com