Hacker News new | past | comments | ask | show | jobs | submit login

Wikipedia provided this simple description, which would be nice to see on the RC post (but perhaps they assumed the only folks interested in an RC post would be folks familiar with the tool already):

Lazarus is a free cross-platform visual integrated development environment (IDE) for rapid application development (RAD) using the Free Pascal compiler.




I miss WYSIWYG IDE's like Delphi, VB-classic, and Lazarus. Dealing with HTML "autoflow" and Bootstrap is a royal pain. You sketch a draft GUI that seems a good fit for the application and users, and then realize you have to jump through hoops backward on a unicycle while chewing gum blindfolded to get it right on web browsers. Jetsons WYSIWYG technology yanked from my happy fingers by an LSD "standard". (I'm assuming "office" applications here, not those needing smart-phone UI's.)


I cut my teeth on VB classic, as I'm sure a lot of folks my age did. I was learning C/C++ concurrently, got through the basics, and of course hurried to GUI programming since that's what I made. It's been 20 something years, but I remember it being something like 31 lines of code to just get an empty window in Windows. With VB it was basically zero effort. I wish WYSIWYG IDEs for both desktop programming AND html (remember frontpage, or hotdog?) would make a comeback in a big way. To this day I shy away from GUI programming because of how tedious it is.


Re: I [now] shy away from GUI programming because of how tedious it is.

It's not really "GUI programming" anymore. It's making HTML/CSS/DOM act like a GUI.


I also did some VB classic and Delphi back in the day. I later moved on to C++ with Qt.

I've not been working on GUI programming much since, but I really enjoyed the Qt experience; good tools and superior documentation.

I never liked gtk - much for the same reasons as your Windows GUI C programming; Too verbose for trivial and mundane things.


It's only verbose if you use C. GTK is very nice with Rust, C++, D, Vala, Python, JS, Haskell, etc.


Yeah, me too. They were one of the steps on the pathway of enabling users and blurring the line between user and programmer. Sadly, the industry lost interest in personal computing in favor of turning users into a resource to farm and feed to advertisers.


Since it's more income to bill for rocket science than bicycle science, there's no incentive for many to simplify UI standards. Right now we are stuck with "rocket science" web-based UI's and JavaScript brain surgery.

Don't get me wrong, HTML browsers are fine for most documents and document navigation, but don't scale to real GUI's easily.


Since we're talking web and not native, your metaphor works only if your "rocket science" is performed with used swimming pool materials and your "brain surgery" with gardening tools.


The metaphor was more about complexity than success or quality.


HyperCard > VB-Clasic > Delphi... Then that's it. Somehow we stopped there. Our Computing Speed might have grown 1000x but We somehow manage to put in additional 1000x complexity to cancel out.

Not only have Software development not gotten any easier, most software are slo lots more resources intensive.


We just need a Lazarus target for webassembly and we'll be good.


Pretty sure this is being worked on! If you're really interested I can check with an ex-colleague who is involved in Free Pascal.


Hmmm, can Lazarus be used to build a "GUI browser"? For example, can one send a "create button" command or tag to it and have it draw one dynamically? What about new panels and forms? How "meta" can its GUI engine be at run-time.


Yes, it may be able to. At least, Delphi, the idea (and language) of which Lazarus is based on, was written in itself, at least for v1, I've read. Probably for later versions too, because Object Pascal (although IIRC the language is now called Delphi, like the product), is a powerful general-purpose programming language, with OOP support, module (units) support, and even some sort of generics (IIRC) in later versions.


Yes it can do that. That's how the IDE works and displays the GUI designer.


Can you recommend any documentation on the "meta" angle? Most documentation I find assumes the creation of "static" widgets that are activated or hidden at run-time (or at least act that way).


This is for Delphi however it should be very similar for Lazarus: https://stackoverflow.com/questions/1005373/creating-compone...

Nothing magic. Create it, set properties.


Thanks!


Heeeey, but actually, unicycle is real fun! Sorry for being OT, but I couldn't resist :D it's just that this exercise you mention seems as a seriously fun idea for one of the many unicycle trials competitions :) just I'm not sure if the chewing gum would up the challenge in any way? That said, I wouldn't really compare JS fighting as similar to really fun unicycle tricks ;D Again, sorry for OT :D

https://youtu.be/ePWN7Q4-7kI?t=156

https://youtu.be/VcwnpLv-iJo


Sorry, didn't mean to offend unicycle riders and jugglers. It was just an analogy about having to "juggle" (coordinate) many tricky activities at the same time.


Ahhh, ok, thanks a lot for the explanation; I didn't grasp the metaphor!



and for some Frameworks integrated in Delphi (first two sadly still not available for Lazarus)

http://www.unigui.com/explore/what-is-unigui

https://www.atozed.com/intraweb/

This one seems to be also available for Lazarus ! http://www.raudus.com/

I wrote an old desktop app for medical laboratories in Delphi (started with D2 ...) and am still quite unsure if this would allow me to port at least some part of it for the web in an (cost and time ) efficient and secure way ...

Does anybody have some experience really using them ?


Indeed. GUIs should be laid out visually, by hand. Because that's not tedious, and everybody has a display with the exact same size, resolution, and aspect ratio, right?

...Right?


WYSIWYG web tools still exist, but they went SaaS. They’re called “low code development platforms”. If you want to play around with one you can download mendix’s modeler for free.


It's not going to take off like HTML did if one has to depend on "cloud" vendors playing nice or lasting.


Can I interest you in a very Delphi-like IDE for building web apps with nothing but Python?

https://anvil.works has a WYSIWYG visual designer, with Python for in-browser code as well as server-side. (We compile it to JS.) It's even got a built-in DB if you want it (Postgres-backed). Everything in one language, full front-to-back autocompletion...all the things we missed about Delphi!


Don't know if it is you specifically, but anvil.works comes up in these contexts a lot so for once I'm going to give you some feedback: anvil.works is not the same thing.

You have some nice looking tooling, but you insist on hosting the result and locking the user into paying rent to use something they built. Your on-prem is locked behind "if you have to ask you can't afford it".

You'd have more interest from people like me if I could just buy or subscribe to your tool and know that even if we eventually drop the subscription the application I produced would still be usable.

But I suspect that your goal isn't to convince people like me, but catch people reading this who might be more amenable to your terms.


Nice product but it may have been better to release this as a opensource product because there will always be a fear that if your company goes bottoms-up, those companies who used your platform will be left high and dry.


The thing is, making Anvil open source now would virtually guarantee that outcome.

The market has spent the last few years demonstrating that the sustainable number of large companies whose main product is open source is...One. (Red Hat, if you were wondering.) Even trying to become that kind of company requires a highly speculative VC-fuelled trajectory, which is famous for turning sustainable 100m-scale companies into 1% unicorns and 99% rubble.

Even Docker, which utterly transformed its target market, is barely making any revenue ($25m/year on a $1.3bn valuation). All its value is busily being subsumed into Kubernetes, itself a loss-leader from a company (Google) that makes all its money from something else.

Anvil-the-company is sustainable, profitable, and growing, and Anvil-the-product is getting better all the time, because we charge for what we produce. We would love to be able to give it away and keep doing that - but basically, we can't.


What about releasing it like how Node.js was released? that is a full web platform but one which uses Python instead of Javascript? and with a repository for packages which people can download and use or even contribute to?

Ultimately unless there is an active and vibrant ecosystem, there is very little reason to consider your platform - especially when there are alternatives like Node.js, Django etc.

Also, for me I will be concerned about how I will integrate Anvil with third-party libraries needed for things like reporting, authentication/authorization, E-commerce etc.

Sure, the GUI wizard is not there for Django but what if someone were to develop such a wizard for Django? or what if someone were to say that it is best to stick with Django / Vaadin/Ketura/.NET/Go/Rocket.rs even if it takes another 25% more time to develop in because the ecosystem is much bigger and we can predict with more confidence that Django / Vaadin/Ketura/.NET/Go/Rocket.rs is going to be around?

HOWEVER, if your product was available open source, is a lot easier to setup and use than Django/Ketura/... (you could provide it as a docker image) and you had a business model where you provide hosting for that solution (similar to Heroku) and/or paid support, you may end up making more money because a lot more people would use your platform - lot more than your current model which very few companies will risk purchasing.


Did you consider a mixed approach like the one Telerik is using with Nativescript?


That's my fear also. Emulating GUI's in web browsers takes a ship-load of code. Therefore, it needs strong maintenance prospects to safely last into the future.


> all the things we missed about Delphi!

Except the insanely fast compile times, native WYSIWYG editor, etc, etc.


If you released this as a node.js type app which can be run directly on the user's desktop and accessible via the browser for some setup functions and via an editor like VSCode etc for the source code editing, this could be a great product.


Yeah, it is not a press release, for more info the home page is very informative IMO https://www.lazarus-ide.org




Consider applying for YC's Spring batch! Applications are open till Feb 11.

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

Search: