What are we missing?

[marcov]

2. Not easily possible (also not in C) The compiler governs the registers, not the asm blocks, during inlining. C often ignores inline in such cases, just like FPC: inline is a hint, not a must to the compiler.

There are possibilities but that will require specially crafted code that uses virtual registers and  metadata that says it expects parameters in certain registers, and results in others. gcc has something like that, and LLVM too.

[Leledumbo]

I don't understand one thing, when you could do similar or better Application in Lazarus why is Python popular? When you can do more and better in Lazarus?

Python is backed by corporations and organizations, this is where the big advantages come from. Python is used by Microsoft and Google, for instance, and they provide in-house private or public solutions written in it. Pioneers are not many but followers are abundant. Followers don't want to take risks, they just use whatever pioneers provide. The side effect, this makes the language extremely popular and abused in fields it wasn't designed for (have you ever tried data mining tool called Orange? It has a GUI that loads forever, if you wanna try). Python was never meant for heavy fields like data mining, because it's too slow. But its popularity drives people to force it anyway. Python was supposed to be an easier LISP (it borrows the functional programming facilities) with Modula-3 strength (modules, exception handling, named arguments) to interface with Amoeba OS (as scripting language), not any wider.

[Greeniemax]

You guys gave me confidence, when so many of you answered overwhelmed me.

Lazarus to future and beyond.

[Trenatos]

What we need

* Commercial backing by big companies
* Reliable, up-to-date documentation
* Quick tutorials for how to do things
* Commercial support
* "Made with" graphics, spread awareness
* High profile projects using it to replace other tech
* "Success stories", not just a list of software, but presentations on why companies go with Pascal and how it helped them win compared to other languages and other competitors.


Commercial backing
----------------------
Java has Oracle
Python has Python Software Foundation (Non profit but still)
Go has Google

We need companies willing to put time and money into using and improving things.


Reliable documentation
--------------------------
The Wiki has tons of outdated information, some packages are just not documented at all, and tutorials/how-to's are either outdated or not there either.

When I evaluate a language, framework, or library, one of the first things I look at is documentation, because with poor docs, it'll be so much harder to get up and running, and the whole point in evaluating is to see what kind of an effort I think it'll be to solve a specific problem.


Quick tutorials for how to do things
---------------------------------------
Videos, pages, whatever, just lots of tutorials showing how to do common and some uncommon things.

A lot of people are working with the web today, we should have solid documentation and tutorials for how to write servers, both just simple non-connect servers to fully Enterprise ready setups.

A simple example is http://wiki.freepascal.org/Executing_External_Programs compared to https://helloacm.com/execute-external-programs-the-python-ways/

The FPC one has all the info needed, if you already know some FPC, doesn't put one and one together or help you along the way.

The Python one starts at the beginning and works with you to learn more, stepping through the various ways, increasing complexity, and explaining things.


Commercial support
----------------------
Another big one for corporations to consider using something, is there commercial support? Can they call a company and rent 40 hours of expertise immediately?


"Made with"
-------------
Simple, people don't know FPC is used if they don't know it's used.


High profile projects using it to replace other tech
-------------------------------------------------------
A lot of languages and frameworks became a thing because Amazon, Google, Netflix, etc. decided to use them and then presented something along the lines of "How Netflix boosted their throughput by 40% by switching from <some tech> to <other tech>"


Success stories
-----------------
Similar to the above, but doesn't have to be really big companies and projects, smaller ones are great too.

Present at local hackathons, conferences, spread the word about how you went from PHP to FPC and saved money, made development faster, and grew your company as a result, record, and spread online.


We need to consider what other languages and frameworks are doing, because those are our competitors.

We need people do adopt it, and we need corporations to adopt it.

For that, they need to know about it, they need to want it, they need to succeed in using it.

[Thaddy]

What we need

* Commercial backing by big companies

Commercial linux distros distribute FPC/Lazarus and support it.

* Reliable, up-to-date documentation

The real documentation is exemplary, best in class.. Ignore wiki's...

* Quick tutorials for how to do things

Almost everything comes with examples, otherwise refer to the - otherwise to be ignored -  wiki or this forum.

* Commercial support

Is available....

* "Made with" graphics, spread awareness

Is done

[Trenatos]

Every time I read about Linux and FPC, the first thing that happens is that someone says "Don't use what comes in the repo, it's an ancient version", so no.

Commercial Linux distros may include some version of FPC/Laz, but I'm talking about big companies using FPC/Laz as one of their main languages that they put time, money, and resources into expanding, fixing, improving.

The real documentation is exemplary? What real documentation?

Commercial support is available from where?

[Ondrej Pokorny]

Commercial support is available from where?

http://forum.lazarus.freepascal.org/index.php/board,41.0.html

[Trenatos]

I'll add some thoughts regarding documentation.

Looking at python.org and http://www.lazarus-ide.org/

Python has a link that says Docs right at the top
Laz has one called Wiki

There is a Start Learning link on the page, but that's not the same as "Docs"

Under the Python Docs is a Beginners Guide, right from the first page.

The Laz Start Learning takes you to a page which.. doesn't actually get you going, it just goes over the Gui tool and menus

The Python has several levels of getting started guides, from "I've never coded before" to "I'm a software engineer".

Python Tutorial actually guides you through some code, gets you making (simple) things.

Python.org has "Success Stories" right at forefront, the stories are well written and laid out.

Laz has Application Gallery, little info given per application.

Python has better sales copy. Right on the front page: Python is a programming language that lets you work quickly
and integrate systems more effectively. Learn More

Laz has a block saying "What can it do?" with the answer being "All the things"

Python has a Search right on the frontpage.

Laz does not.

The Laz Highlights mention things about the Pascal and Laz, not what you can DO with it or WHY it's awesome.

Design: Python website looks nice, Laz could use an overhaul.

Why does Screenshot have a top link, but not Getting Started?

[Phil]

Problems / obstacles

* Identity crisis / “brand” issues. What does Lazarus stand for, what is its reason for existence? Maybe 10 years ago you could say it was a refuge for disaffected Delphi users, but that doesn’t make much sense today. In any case, you normally don’t want to define your brand in contrast to another brand. In the Ricky Gervais movie, The Invention of Lying (a world where no one lies), you see a billboard for Pepsi that says “For when they don’t have Coke.” You don’t want the Lazarus slogan to be “For when you don’t have Delphi.”

* Lack of farm league / feeder system. Historically in U.S. professional baseball, the farm teams were towns and small cities with baseball teams that were each linked to a big league team. This provided not only a reliable base of fans outside of the big league city, but the best players from the farm teams eventually got called up to play in “the show.” What’s the Lazarus feeder system? And don’t say retired and former Delphi users. That’s not sustainable.

* Communication problems. As an outsider, the workings of the Laz project have always seemed pretty secretive. No blogs from the project owners, no white papers, few announcements.

* Have stable releases outlived their usefulness? This has been a chronic problem with Mac, where both the Carbon and Cocoa widgetsets in the stable releases often have serious problems. Sometimes these problems have been fixed in the trunk, but users who download a stable release don’t know anything about that. Even now, before 1.8 is even released, the Carbon and Cocoa code is already 6 months out of date.

* Euro-centric nature. Let’s face it, for better or worse, the eyes of the world’s developers are mostly focused on a single area that’s about 8-9 time zones away from where most of the Laz contributors live. This is even a problem for U.S. companies. Mapbox, a company whose products I use, started out working with NGOs in Washington, D.C, but now has shifted much of its offices to San Francisco. Geography still matters. It’s not just about access to capital, it’s also about access to talent, visibility, being part of the “hustle.” How does Lazarus join the hustle?


Possible solutions.

* Put a countdown timer to the next release on the home page. See how QGIS does it: www.qgis.org. Not only does this provide valuable information at a glance, it speaks volumes about the project itself: we’re confident, we have a plan, we’re coming, we’re inevitable.

* Actively solicit funding from users. Money is both lubricant and fuel. Don’t be bashful about asking for it when it’s truly needed. The current bounty system is too timid, too passive, and doesn’t provide general operating funds or targeted funding for essential things that the owners and users have identified. QGIS solicits money all the time and then publishes a big report on what it was spent on. This is not only a way to be accountable, but also a great way to promote and publicize what’s happening with the project.

* Get a free architectural analysis. Look at this fascinating chapter on Mapbox that a bunch of mere students did:

https://delftswa.gitbooks.io/desosa-2017/content/mapbox-gl-js/chapter.html

All you have to do is recruit some students like these who are interested in doing one for Lazarus.

* Co-opt other languages and projects. Don’t see them as competitors that you have to shun or criticize. The students who profiled Mapbox came up with some revelations. One of them is this sentence: “When there is a tool that is an alternative to one of the Mapbox tools, this is not seen as a threat.” Mapbox doesn’t recognize competition, only opportunities to extend their own products by embracing other products.

* Emphasize the experimental stuff in FPC and Lazarus. If it’s good and interesting, tell the world about it.

[marcov]

Problems / obstacles

* Identity crisis / “brand” issues. What does Lazarus stand for, what is its reason for existence?

It is an development environment, so what do you think ?

FPC/Lazarus does suffer and benefit from the Delphi connection at the same time. 

* Lack of farm league / feeder system. Historically in U.S. professional baseball,

Oops, you lost me there, I don't understand baseball analogees.

* Communication problems. As an outsider, the workings of the Laz project have always seemed pretty secretive. No blogs from the project owners, no white papers, few announcements.

Last weekend there was Lazarus conference with the lazarus main devel doing a talk. Didn't you go ?

* Have stable releases outlived their usefulness? This has been a chronic problem with Mac, where both the Carbon and Cocoa widgetsets in the stable releases often have serious problems. Sometimes these problems have been fixed in the trunk, but users who download a stable release don’t know anything about that. Even now, before 1.8 is even released, the Carbon and Cocoa code is already 6 months out of date.

I don't know much about the current state of Mac, but these things was why I abandoned Mac in 2005. It seems that Mac doesn't really have grassroot developer potential, and always looking for others to fix them.

It doesn't help when major users in Mac-pascal list recommend people to use swift. This creates a vicious circle.

* Euro-centric nature. Let’s face it, for better or worse, the eyes of the world’s developers are mostly focused on a single area that’s about 8-9 time zones away from where most of the Laz contributors live.

No, only the venture capitalists eyes are, and if you didn't went to (Ivy league) school with them, your chances are slim anyway. So while some form of big bucks top is there, it is not really an achievable goal, so better ignored.

Specially if you are part of an 25 year old project outside the current favourite fields.

I'm more in favour of actually continuing developing own strengths rather than going on a wild goose chase to try a realistically unachievable Silicon Valley jackpot.

This is even a problem for U.S. companies. Mapbox, a company whose products I use, started out working with NGOs in Washington, D.C, but now has shifted much of its offices to San Francisco. Geography still matters. It’s not just about access to capital, it’s also about access to talent, visibility, being part of the “hustle.” How does Lazarus join the hustle?

Simple. Get Vinod to call us, that worked out for Google. The details can be worked out later 

* Put a countdown timer to the next release on the home page. See how QGIS does it: www.qgis.org. Not only does this provide valuable information at a glance, it speaks volumes about the project itself: we’re confident, we have a plan, we’re coming, we’re inevitable.

And it is pointless if you can't marshal the resources to actually make things happen. Currently the FPC 3.0.4 release hangs on a disappeared OS X maintainer, who is apparently to busy to check in.

* Actively solicit funding from users. Money is both lubricant and fuel. Don’t be bashful about asking for it when it’s truly needed. The current bounty system is too timid, too passive, and doesn’t provide general operating funds or targeted funding for essential things that the owners and users have identified. QGIS solicits money all the time and then publishes a big report on what it was spent on. This is not only a way to be accountable, but also a great way to promote and publicize what’s happening with the project.

We have had donation systems and applied for grants for decades. GSOC, private investments etc etc. We even set up a foundation. It does bring in something, but that is usually indirect (a core developer starts working somewhere and can do some FPC/Lazarus work in employers time).

* Get a free architectural analysis. Look at this fascinating chapter on Mapbox that a bunch of mere students did:
https://delftswa.gitbooks.io/desosa-2017/content/mapbox-gl-js/chapter.html
All you have to do is recruit some students like these who are interested in doing one for Lazarus.

Personally I think there are better ways to spend the developer hours. Student or not.

Still it would be fun what they would categorize under technical debt. (probably 99.999% of the project).

* Co-opt other languages and projects. Don’t see them as competitors that you have to shun or criticize. The students who profiled Mapbox came up with some revelations. One of them is this sentence: “When there is a tool that is an alternative to one of the Mapbox tools, this is not seen as a threat.” Mapbox doesn’t recognize competition, only opportunities to extend their own products by embracing other products.

Generic and sad. Again comparisons with projects in "hot" areas are pointless. But the problem is as usual not getting the fact that FPC/Lazarus is not a company. There is no labour to (re-)assign to other projects or promising fields. People work on something because they want to.

* Emphasize the experimental stuff in FPC and Lazarus. If it’s good and interesting, tell the world about it.

Go ahead. Nearly everything in Lazarus/FPC is done on personal initiative.

In summary, you are comparing a company in a fairly hot area of IT to an old volunteer effort with minimal organization with a foundation for sponsoring.   IMHO that is pointless. Compare like with like.

[cpicanco]

I'll add some thoughts regarding documentation.

Looking at python.org and http://www.lazarus-ide.org/

Python has a link that says Docs right at the top
Laz has one called Wiki

There is a Start Learning link on the page, but that's not the same as "Docs"

Under the Python Docs is a Beginners Guide, right from the first page.

The Laz Start Learning takes you to a page which.. doesn't actually get you going, it just goes over the Gui tool and menus

The Python has several levels of getting started guides, from "I've never coded before" to "I'm a software engineer".

Python Tutorial actually guides you through some code, gets you making (simple) things.

Python.org has "Success Stories" right at forefront, the stories are well written and laid out.

Laz has Application Gallery, little info given per application.

Python has better sales copy. Right on the front page: Python is a programming language that lets you work quickly
and integrate systems more effectively. Learn More

Laz has a block saying "What can it do?" with the answer being "All the things"

Python has a Search right on the frontpage.

Laz does not.

The Laz Highlights mention things about the Pascal and Laz, not what you can DO with it or WHY it's awesome.

Design: Python website looks nice, Laz could use an overhaul.

Why does Screenshot have a top link, but not Getting Started?

I think you should blame FPC, not lazarus. Lazarus is the IDE and docs about how using the graphical part of the ide are in the wiki.

FPC official documentation is excelent. So your blaming against them would not be ok.

[Trenatos]

But there's no real separation.

If the Laz wiki didn't have FPC information, I wouldn't be blaming it for its bad/lacking FPC information.

FPC and Laz website have similar issues, they're geared towards people who already know FPC/Laz, not beginners, not people who are curious, not people who are evaluating it.

[Trenatos]

People care less about languages than they do about solutions to problems, it's simple.

Don't sell the language, or the IDE, sell the REASON why people should use it.

Powerful. Fast to develop with. Resource efficient. Easy to deploy binaries. Easy to integrate into build systems. GUI designer and IDE in one. Scalable solutions.

Those are much more interesting than "Over 200 components", or "Many frameworks".

[cpicanco]

Free pascal docs here: https://www.freepascal.org/download.var

I would be very happy if I have found this folder earlier:
https://www.freepascal.org/~michael/articles/

This last folder does not seem to be easily reachable from the main site, though. Am I blind? Please make something like that:

STUFF WRITTEN BY US

[cpicanco]

People care less about languages than they do about solutions to problems, it's simple.

Don't sell the language, or the IDE, sell the REASON why people should use it.

Powerful. Fast to develop with. Resource efficient. Easy to deploy binaries. Easy to integrate into build systems. GUI designer and IDE in one. Scalable solutions.

Those are much more interesting than "Over 200 components", or "Many frameworks".

I think you should rephrase like that:

Don't focus on the language, or the IDE, focus on the REASON why people should use it.