I started using Delphi 3 and stopped at 7, migrating to web development (Rails, Django, etc.).
Delphi was magical. Nothing compares to Delphi's productivity.
Rails is good, but it doesn't even come close to Delphi's productivity.
People love Go's speed. Go is glacially slow compared to Delphi.
The WYSIWYG form editor is incredible. I can use Delphi 3 on a Windows machine with 16 MB of RAM.
VCL is fantastic; the idea of components and memory management is incredible, simple, and it works.
Delphi is my first language; I studied VCL code and I love the code, the style. They were practical: Instead of Hash, they used TStrings (a list of strings) and the visual components also used them, like in the items of a Listbox!
Delphi could have been the platform for the web. Imagine a VCL for the web (VCLW), where you could change the target architecture or something like that and, presto, you'd have a web server running with VCL code!
That never happened. What happened was a series of bad ideas for the web, bad in their essence.
And Delphi invested in many projects doomed to failure, such as CORBA, three-tier architecture, MDA... Kylix!!!! Of course, Borland was very poorly managed. The CEOs were crazy. "Let's fight IBM." Delphi was abandoned. It's over.
I tried a new version of Delphi a few years ago. Wow, it was full of bugs! It had basic problems like compilation not working, Random crashing several times, etc. For me the new versions are just a way to profit from projects stuck in Delphi.
I tried Lazarus in the past, but it's extremely slow and I can't use my components in Lazarus without rewriting a lot of things.
To me, Delphi is languishing in an induced coma, breathing the air of the past, which is becoming increasingly rare. It's a shame.
As much as I know the story MSFT was basically driving limousines in front of Borland headquarters and showering developers with money to jump the ship.
When web apps took off there was nothing that would save Delphi. Web stack was awful back then and I also enjoyed Delphi an building interfaces with it.
But web stack unfortunately had many more upsides than desktop apps and nowadays tooling for web sucks less.
We just got code complete on porting a 30 year old Delphi app to C#, because of all of this.
Even now, our pure Delphi components are performant and wonderful, but hiring people who want to learn or know Delphi is hard, so off to C# we trundle forward.
Didn't the same people who wrote Delphi write also C# and .NET? When I first saw .NET it felt very much like VCL/CLX. And then came TypeScript from the same guy.
But then the C# people make writing code in it.. a horrible experience. They really like to over engineer and architect.
Had to look at some c# backend code. To write any kind of endpoint that talked to a database they had to write at least... 7? different files, lots of empty interfaces that has to be created and implemented, command, mediator patterns everywhere, etc. Looked like insanity compared to Spring Boot
Sometimes I miss the times where you had a compact development environment, wit one installer. Your source produced a mostly self contained binary in a reasonable size, you had nice debugging support and quick turnaround times for a compiled language even on a small development machines. And all that for attractive price for a perpetual license (Borland times).
Today it seems I have to give the producer my email address for the 'free' "Delphi History PDF".
Well, times have changed. :)
Creating a simple application that had a database backend, data presentation gui and some simple CRUD logic took like 15-30 mins. All from scratch. How does it compare to today’s tools? And I am not talking about taking some ready off the shelf solution that you don’t really understand internally…
Those were good times and I really regret all those mishaps that happened to this great ecosystem and its eventual collapse :(
The way I saw it in 1995 was that Delphi was the fastest way to create a full windows desktop app and do it as single compiled-to-native-code executable at that critical time it was released. The slightly-later 32-bit version was powerful and gave your app some staying power; a Delphi-generated executable file would likely still run today.
Sadly they still do, although finding somebody to work on them is hard and while the executables work, the dev environment does not. Delphi was a pretty nasty dead end
I have very fond memories of working with Delphi 5 and 2005 in the early 2000s. Both the language and IDE were a real pleasure to work with, and they were head and shoulders better than anything from Microsoft at that time. The community was small but enthusiastic and supportive as well.
It would be hard to justify Delphi in a new project today - not because of the tooling or language, but because of the prohibitive license costs.
When I was a kid, my older brother worked for Borland. He got me 2 packs of stickers that said "Delphi developers do it better!!!" in red font and a yellow background.
I remember taking an onsite class to learn Delphi. The class was taught in St Petersburg Florida. Nice place. At the time, I was admiring the tool that Borland created and thought to myself this is a very nice IDE. Too bad my company was switching to all things .Net. The difference between Visual Studio and the Delphi IDE was gut wrenching.
Happy birthday Delphi, you made me a lot of money :-)
I am guessing most Delphi developers like me, have either retired or have moved to Linux. I have done both recently and I unfortunately do not see a new generation following in behind me. I hope it survives as it was and still is a nice IDE and language to work in, though I'm guessing newer Pascal developers will opt for Lazarus
I'm now retired, but I spent much (most?) of my career developing with Delphi. When I began, it was the new hotness. When I finished, I was supporting legacy applications that were decades old. Good to see it's still around, though.
The hardest part of maintaining a long-term project is resisting the urge to over-engineer early on. Striking a balance between a lean core and future extensibility is an art form that often gets ignored in favor of shipping fast.
I started using Delphi 3 and stopped at 7, migrating to web development (Rails, Django, etc.).
Delphi was magical. Nothing compares to Delphi's productivity. Rails is good, but it doesn't even come close to Delphi's productivity. People love Go's speed. Go is glacially slow compared to Delphi. The WYSIWYG form editor is incredible. I can use Delphi 3 on a Windows machine with 16 MB of RAM.
VCL is fantastic; the idea of components and memory management is incredible, simple, and it works.
Delphi is my first language; I studied VCL code and I love the code, the style. They were practical: Instead of Hash, they used TStrings (a list of strings) and the visual components also used them, like in the items of a Listbox!
Delphi could have been the platform for the web. Imagine a VCL for the web (VCLW), where you could change the target architecture or something like that and, presto, you'd have a web server running with VCL code!
That never happened. What happened was a series of bad ideas for the web, bad in their essence.
And Delphi invested in many projects doomed to failure, such as CORBA, three-tier architecture, MDA... Kylix!!!! Of course, Borland was very poorly managed. The CEOs were crazy. "Let's fight IBM." Delphi was abandoned. It's over.
I tried a new version of Delphi a few years ago. Wow, it was full of bugs! It had basic problems like compilation not working, Random crashing several times, etc. For me the new versions are just a way to profit from projects stuck in Delphi.
I tried Lazarus in the past, but it's extremely slow and I can't use my components in Lazarus without rewriting a lot of things.
To me, Delphi is languishing in an induced coma, breathing the air of the past, which is becoming increasingly rare. It's a shame.
As much as I know the story MSFT was basically driving limousines in front of Borland headquarters and showering developers with money to jump the ship.
When web apps took off there was nothing that would save Delphi. Web stack was awful back then and I also enjoyed Delphi an building interfaces with it.
But web stack unfortunately had many more upsides than desktop apps and nowadays tooling for web sucks less.
This so much.
We just got code complete on porting a 30 year old Delphi app to C#, because of all of this.
Even now, our pure Delphi components are performant and wonderful, but hiring people who want to learn or know Delphi is hard, so off to C# we trundle forward.
Didn't the same people who wrote Delphi write also C# and .NET? When I first saw .NET it felt very much like VCL/CLX. And then came TypeScript from the same guy.
But then the C# people make writing code in it.. a horrible experience. They really like to over engineer and architect.
Had to look at some c# backend code. To write any kind of endpoint that talked to a database they had to write at least... 7? different files, lots of empty interfaces that has to be created and implemented, command, mediator patterns everywhere, etc. Looked like insanity compared to Spring Boot
Sometimes I miss the times where you had a compact development environment, wit one installer. Your source produced a mostly self contained binary in a reasonable size, you had nice debugging support and quick turnaround times for a compiled language even on a small development machines. And all that for attractive price for a perpetual license (Borland times).
Today it seems I have to give the producer my email address for the 'free' "Delphi History PDF". Well, times have changed. :)
> I have to give the producer my email address for the 'free' "Delphi History PDF"
Yeah I was interested to see the timeline but I'm not going through a spam wall to get it.
npm i nostalgia
For me Go and Rust match this to a point. Especially Go once installed it generates executable extremely fast.
Creating a simple application that had a database backend, data presentation gui and some simple CRUD logic took like 15-30 mins. All from scratch. How does it compare to today’s tools? And I am not talking about taking some ready off the shelf solution that you don’t really understand internally…
Those were good times and I really regret all those mishaps that happened to this great ecosystem and its eventual collapse :(
I wrote a trojan (for educational purposes) in Delphi by studying another one for which I had the source code. The original was called Donald Duck.
Then I wrote back-office software in Delphi that ran for 15+ years without any maintenance or support.
Those were simpler times.
The way I saw it in 1995 was that Delphi was the fastest way to create a full windows desktop app and do it as single compiled-to-native-code executable at that critical time it was released. The slightly-later 32-bit version was powerful and gave your app some staying power; a Delphi-generated executable file would likely still run today.
Sadly they still do, although finding somebody to work on them is hard and while the executables work, the dev environment does not. Delphi was a pretty nasty dead end
I have very fond memories of working with Delphi 5 and 2005 in the early 2000s. Both the language and IDE were a real pleasure to work with, and they were head and shoulders better than anything from Microsoft at that time. The community was small but enthusiastic and supportive as well.
It would be hard to justify Delphi in a new project today - not because of the tooling or language, but because of the prohibitive license costs.
Same here, but with C++Builder.
My dad taught me how to code with Borland C++ Builder, and I'm forever grateful.
When I was a kid, my older brother worked for Borland. He got me 2 packs of stickers that said "Delphi developers do it better!!!" in red font and a yellow background.
I remember taking an onsite class to learn Delphi. The class was taught in St Petersburg Florida. Nice place. At the time, I was admiring the tool that Borland created and thought to myself this is a very nice IDE. Too bad my company was switching to all things .Net. The difference between Visual Studio and the Delphi IDE was gut wrenching.
When I was starting out as a kid learning to make applications, moving from VB6 to Delphi was such a huge improvement.
Tempted to use a client's plotter and roll of paper to print this off.
The original borland delphi had very creative installer graphics:
https://www.gladir.com/SOFTWARE/DELPHI1/delphi1-install5.png
The programmers must have been playing https://en.wikipedia.org/wiki/Test_Drive_(1987_video_game)
That was before Delphi 2.0 even :)
Happy birthday Delphi, you made me a lot of money :-)
I am guessing most Delphi developers like me, have either retired or have moved to Linux. I have done both recently and I unfortunately do not see a new generation following in behind me. I hope it survives as it was and still is a nice IDE and language to work in, though I'm guessing newer Pascal developers will opt for Lazarus
Is it still alive? Last time I used it was around 2005.
One nice thing though as I remember was that ruins of Russian Borland branch gave us Jetbrains.
I'm now retired, but I spent much (most?) of my career developing with Delphi. When I began, it was the new hotness. When I finished, I was supporting legacy applications that were decades old. Good to see it's still around, though.
Weird but FreePascal is fairly solid for its niche.
Straight up, nobody uses it anymore.
What is dead may never die
The hardest part of maintaining a long-term project is resisting the urge to over-engineer early on. Striking a balance between a lean core and future extensibility is an art form that often gets ignored in favor of shipping fast.
31 years old and it can't run on GNU/Linux. What a waste. The future of Delphi is darker than ever.