Recent

Author Topic: The future of Free Pascal  (Read 228514 times)

Thaddy

  • Hero Member
  • *****
  • Posts: 14205
  • Probably until I exterminate Putin.
Re: The future of Free Pascal
« Reply #285 on: March 22, 2017, 06:44:04 pm »
Yes, laptop.
Specialize a type, not a var.

jacmoe

  • Full Member
  • ***
  • Posts: 249
    • Jacmoe's Cyber SoapBox
Re: The future of Free Pascal
« Reply #286 on: March 22, 2017, 06:52:43 pm »
So, what you guys are saying: just throw more RAM/CPU at it, and all is well.
I suppose that's one way of dealing with it.

Just keep in mind that for the majority of the world - read: non-white, non-rich regions - there isn't an abundance of RAM/CPU/GB.
more signal - less noise

vfclists

  • Hero Member
  • *****
  • Posts: 1013
    • HowTos Considered Harmful?
Re: The future of Free Pascal
« Reply #287 on: March 22, 2017, 07:55:39 pm »
I can totally understand were Graham is coming from. I must absolutely have 16Gb  on  my next laptop and 32Gb on  my desktop and they both must be using SSDs, i7 and preferably dual CPU E3/E5s for the deskop. If I set it to  anything less I will be wasting my time watching the hourglass.

Having to open browsers to look up information kills performance more than anything. If you have all the information in your at your fingertips and don't have to look up anything online that is fine. Lazarus will work fine on 2Gb. Open your browser have a few VMs running and everything will crawl to a halt. Try working across different technologies and OSs simultaneously and you will  need those. Visual Studio here, Eclipse there, Android Studio there and you are in for a rough time.

When you recollect having to pay 1000s for 386  and 486 systems with 16Mb of RAM, 32Gb systems in 2017 for 1000s are an absolute steal. The problem is development tools have become bloated when in the past they were tightly coded. C++ and Java and the culprits here.

https://www.xkcd.com/303/

Graeme you mentioned recently that you were compiling Firefox. How long did it take?

FWIW computer developers have become cheapskates who undervalue their work and/or allow their work to be undervalued. Other professionals pay far more for their tools and they don't even whinge. Even Uber drivers invest more in the cars and they don't gripe about costs so much.
Lazarus 3.0/FPC 3.2.2

vfclists

  • Hero Member
  • *****
  • Posts: 1013
    • HowTos Considered Harmful?
Re: The future of Free Pascal
« Reply #288 on: March 22, 2017, 08:03:39 pm »
When it comes to performance I think Lazarus and FPC will not be fully optimized until all the modern/Delphi functionality bells and whistles that Maciej and co have been asking for have been added. It is hard for small teams to optimize software when they are adding new functionality to it.

It is only when things get stabilized that they can concentrate and squeezing every last drop out.

I am looking for automatic garbage collection but I don't see it  happening soon. Perhaps Maciej's ARC will help when it comes out, which will be nearer 2020 if not after.
Lazarus 3.0/FPC 3.2.2

marcov

  • Administrator
  • Hero Member
  • *
  • Posts: 11383
  • FPC developer.
Re: The future of Free Pascal
« Reply #289 on: March 22, 2017, 08:09:22 pm »
When it comes to performance I think Lazarus and FPC will not be fully optimized until all the modern/Delphi functionality bells and whistles that Maciej and co have been asking for have been added. It is hard for small teams to optimize software when they are adding new functionality to it.

Those feature decrease performance not increase them.

Quote
It is only when things get stabilized that they can concentrate and squeezing every last drop out.
I am looking for automatic garbage collection but I don't see it  happening soon. Perhaps Maciej's ARC will help when it comes out, which will be nearer 2020 if not after.

GCt is the biggest performance sink of all.

john horst

  • Jr. Member
  • **
  • Posts: 68
    • JHorst
Re: The future of Free Pascal
« Reply #290 on: March 22, 2017, 08:35:05 pm »
FWIW computer developers have become cheapskates who undervalue their work and/or allow their work to be undervalued. Other professionals pay far more for their tools and they don't even whinge. Even Uber drivers invest more in the cars and they don't gripe about costs so much.

In the US programmers are generally in the top 1% when it comes to income. I don't see how you claim undervalued. Some of us have software that goes out around the world, to the average joe. If my constraints are 16/32 GB to write and ship this software, well I might as well close shop. I do understand the need to have the resources to run a VM. I personally have never seen a company have a group of developers all using multiple VMs. That would generally be a special use case, as was stated before.

Artlav

  • New Member
  • *
  • Posts: 36
    • Orbital Designs
Re: The future of Free Pascal
« Reply #291 on: March 22, 2017, 08:56:23 pm »
FWIW computer developers have become cheapskates who undervalue their work and/or allow their work to be undervalued.
Most programmers are in entry level and/or minimum income jobs due to the market oversaturation these days, and can't just spend a $1000 on a top-of-the-line computer simply for the sake of reducing compile time from 60 to 30 sec or something.

The ones who decide on developing big name tools, on the other hand, work for obscene money in places like the USA and have top-of-the-line hardware as a given, and so can't really understand what's the big deal is with making the requirements a little higher.

If I set it to  anything less I will be wasting my time watching the hourglass.
Case in point.
Time-money tradeoffs are a luxury.

RegNatarajan

  • New Member
  • *
  • Posts: 28
Re: The future of Free Pascal
« Reply #292 on: March 23, 2017, 04:43:41 am »
Well, Delphi is as good as dead. And, to be honest: good riddance.
I'm very late to this thread but I am interested why people feel Delphi is dead.  I was toying with buying it, even at its exorbitant price.

TIA
Reg


marcov

  • Administrator
  • Hero Member
  • *
  • Posts: 11383
  • FPC developer.
Re: The future of Free Pascal
« Reply #293 on: March 23, 2017, 06:59:52 am »
Well, Delphi is as good as dead. And, to be honest: good riddance.
I'm very late to this thread but I am interested why people feel Delphi is dead.  I was toying with buying it, even at its exorbitant price.

Delphi isn't dead, but price hikes and periods of neglect have reduced its userbase. Delphi has been declared dead continously since before 2000, and survived a lot of the suggested alternative. If only because desktop apps are also continously being declared dead. If it aint a cloud running on IOT with VR glasses attached it doesn't exist nowadays.

It is a silly kind of remark that shows ignorance on behalf of the author.

mtanner

  • Sr. Member
  • ****
  • Posts: 287
Re: The future of Free Pascal
« Reply #294 on: March 23, 2017, 07:49:42 am »
programming languages never seem to die. Fortran is still around after all, if only as libraries of useful scientific routines. I remember that I got into Turbo-Pascal on very early PCs because it was the only reasonably-priced option for serious programming on PC's then. And Delphi gave a productive IDE and Windows access. But is has now got very expensive, especially when you need to pay for add-ons like TeeChart (if supplied version not enough). I still think it's a great language. C/C++ have rather tortured syntax, though obviously the dominant language. Python is fine, but again the non-free-form syntax can be limiting. And I want to stay away from C# because of it's dependency on MicroSoft environments ( and data marshaling can be a nightmare for interworking with other languages). Pascal is a great language for actually focussing of the task to be accomplished. And Lazarus now provides a path away from Windows, plus the very excellent TAGraph.

Groffy

  • Full Member
  • ***
  • Posts: 204
Re: The future of Free Pascal
« Reply #295 on: March 23, 2017, 11:14:49 am »
This thread is now almost going for a year, and without reading/remembering every single message I can't remember that I read something what suggests/leads into the future of this language. What do we mean when we are talking about the "future" of Free Pascal? Is it the further development of the language/compiler itsself? Is it the question whether we will have enough maintainer on volunteer base in the future? Is it the question whether the language itself will stay alive in form of a new young user generation which might be still in school or university right now? My generation who started on Z80 CP/M computers with Turbo Pascal might be out of profession within the next 10-15 years, in which timeframe we should think when we are talking about the "future"? The language "Object Pascal" will not die, but the user base is continious shrinking and there will be the day when it makes no sense anymore developing products like delphi (whoever will own it in the future) on a professional base. I'm pretty sure, that companies which are developing VCL products commercially already noticed the shrinking market years ago and its still going on year after year. 

In my opinion the future of Free Pascal is strongly connected to Lazarus and the ability to organize a structure around that package which signalizes that there will be a continious maintainance and development. Companies like TMS components gave a huge credit to the Lazarus project when they decided to enter the market for LCL products on a professional base. I just hope, that more companies from the classical VCL market will follow and trust Lazarus/Free Pascal.

I'm working with the Firebird database (moved with it to the .net environment 10years ago) for many years now, the users/developers founded the Firebird Foundation to bring the future development of this RDBMS into a kind of professional structures. Maybe thats a possible way also for Free Pascal/Lazarus(?). Lazarus will continue the legacy of TP/delphi one day. Nobody knows when that will happen, the project should be well prepared.

With best regards
« Last Edit: March 23, 2017, 11:49:19 am by Groffy »
Linux Mint / Windows 10 / Lazarus 3.0.0 / trunk -qt

Thaddy

  • Hero Member
  • *****
  • Posts: 14205
  • Probably until I exterminate Putin.
Re: The future of Free Pascal
« Reply #296 on: March 23, 2017, 12:29:53 pm »
The future always means future. So it doesn't matter how long the thread runs. If a goal is achieved, there's a new future feature..., etc.
Specialize a type, not a var.

Groffy

  • Full Member
  • ***
  • Posts: 204
Re: The future of Free Pascal
« Reply #297 on: March 23, 2017, 12:50:47 pm »
The future always means future. So it doesn't matter how long the thread runs. If a goal is achieved, there's a new future feature..., etc.

Thank you for your philosophical point of view. Can you propose goals?
Linux Mint / Windows 10 / Lazarus 3.0.0 / trunk -qt

Graeme

  • Hero Member
  • *****
  • Posts: 1428
    • Graeme on the web
Re: The future of Free Pascal
« Reply #298 on: March 23, 2017, 01:02:34 pm »
Just my 2 cents:
10 years ago 8 or even 16 GB was common for a developer in the Netherlands..
Maybe the USA is more backwards than the rest of the world thought. 10 Years ago I was based in a 3rd world country - South Africa. I worked in Science & Academics for a 30+ employees company. Developer systems were all desktops (laptops were simply not powerful enough) and RAM was a mix between 4GB and 8GB - *never* lower that 4GB. And yes, 2 monitors per PC was common too.

To put this in context, my personal laptop back in 2007 was a 15.6" DELL Inspiron (weighs a ton) with a screen resolution of 1920x1200 and 4GB RAM and some 3Ghz CPU and a 128MB Radeon video card. I still have that laptop today (fully working, but not used any more). Pity laptop designs went backwards since then, now sporting shitty 1366x768 resolutions as common place. WTF! And the consumer seems happy with that?? WTFx2!

It seems some people here forget that developers are not the standard end-user, so standard end-user equipment (like laptops with limited upgradability, or stock off-the-shelf PC's) simply doesn't cut it.

Maybe some here should read some of the "Joel on Software" blogs. If you don't have the time to read all the excellent articles, read this one and take a particular note of Item 9.

    https://www.joelonsoftware.com/2000/08/09/the-joel-test-12-steps-to-better-code/

I can't agree more with that point.
--
fpGUI Toolkit - a cross-platform GUI toolkit using Free Pascal
http://fpgui.sourceforge.net/

Graeme

  • Hero Member
  • *****
  • Posts: 1428
    • Graeme on the web
Re: The future of Free Pascal
« Reply #299 on: March 23, 2017, 01:14:09 pm »
Graeme you mentioned recently that you were compiling Firefox. How long did it take?
7 minutes if I remember correctly - which I believe was faster than usual. I went and made a cup of coffee while it was churning away. :) Compiling LibreOffice takes magnitudes longer (35+ minutes)! Imagine the pain in debugging that.

Quote
FWIW computer developers have become cheapskates
Definitely. Every second I wait for a compilation is time wasted and costs somebody money.
--
fpGUI Toolkit - a cross-platform GUI toolkit using Free Pascal
http://fpgui.sourceforge.net/

 

TinyPortal © 2005-2018