Can FPC/Lazarus save the World?

[MarkMLl]

During the 2000s and 2010s we've had what might be called "The Second Software Crisis", with enterprise and government-scale projects almost always being delivered late and often never being delivered in the form envisaged.

Leaving aside the challenges of user demands that software must be accessible from a browser, and leaving aside the explosion of "Artificial Intelligence", and trying to learn lessons from mis-architected systems like Horizon in the UK. What fundamental technologies are possessed by the big players- IBM, Oracle, SAP- that conventional tools such as Delphi and Lazarus lack?

Would a consortium of developers using Lazarus and e.g. PostgreSQL be able to compete with e.g. Fujitsu, who are still landing significant government work despite being tainted?

And how could that consortium prove to the commercial and government communities that it was underpinned by robust technology, in a way that C++, Python and the rest weren't? Put another way, how can the community of Pascal developers overcome the undeniable prejudice that's developed against the language over the last twenty years or so?

MarkMLl

[Seenkao]

Eng:
Google translate:
Advertising, advertising and more advertising.

Advertising should not be everywhere and intrusive. Advertising should promote you, your products and the products (programs) you used. You must understand that in any program you create, you must indicate that Pascal was used (FPC/Lazarus/Delphi/...).

It is not necessary to create some kind of advertising website. It is not necessary to order advertising in the media.

Everyone can create their own advertisement.

The documentation in your library is an advertisement. It allows the user to quickly understand how to work with your library.
Demonstration of your examples using your library - advertising. This demonstration will also allow the user to understand how to work with your library and see the result of its work.
Educational lessons - advertising. If you are ready to create: text lessons, audio lessons, video lessons; who can quite simply convey how to use Pascal (FPC/Delphi/Lazarus) and show their capabilities, then this is a huge plus! In this case, it is important not to repeat what has already been covered, but to supplement what has already been shown before you, indicating the source.

The more each person contributes to the “advertising”, the more people around you will be informed about the capabilities of the Pascal language and the less they will talk about the fact that nothing can be done with it.

Attention! You must understand whether you are ready or not in order to develop a community! When in doubt, ask people. Write a preliminary paper and show it to the Pascal community. And, here you must be able to accept both praise and criticism! Criticism is especially important. Having corrected most of the shortcomings that were presented to you in criticism, your article can sparkle with new colors and you are more likely to present it to the entire community of programmers, and not just to those who program in Pascal.

Another important point.
Nowadays, a lot is lost on the Internet in the Internet landfill. If you post an article, the article is also lost behind other articles and becomes almost invisible or cannot be found at all.
There must be a place where this article can be reached quickly enough. Two or three clicks on a specific site.

Good luck! 

--------------------------------------------------------------
Rus:
Реклама, реклама и ещё раз реклама.

Реклама не должна быть везде и навязчивой. Реклама должна рекламировать вас, ваши продукты и продукты (программы) которые вы использовали. Вы должны понимать, что в любой созданной вами программе, вы должны указывать что использовался Pascal (FPC/Lazarus/Delphi/...).

Не обязательно создавать какой-то рекламный сайт. Не обязательно заказывать рекламу в средствах массовой информации.

Каждый может создать свою рекламу.

Документация в вашей библиотеке - реклама. Она позволяет пользователю более быстро понять как работать с вашей библиотекой.
Демонстрация ваших примеров с использованием вашей библиотеки - реклама. Эта демонстрация так же позволит пользователю и понять как работать с вашей библиотекой и увидеть результат её работы.
Обучающие уроки - реклама. Если вы готовы создавать: текстовые уроки, аудиоуроки, видеоуроки; которые достаточно просто смогут донести способы использования Pascal (FPC/Delphi/Lazarus) покажут их возможности, то это огромный плюс! В данном случае важно не повторять то, что уже пройдено, а дополнять то, что уже показали до вас, с указанием на источник.

Чем больше каждый человек внесёт своей лепты в "рекламу", тем более будут информированы люди вокруг вас о возможностях ЯП Pascal и меньше говорить о том, что на нём нельзя ни чего сделать.

Внимание! Вы должны понимать готовы вы или нет для того чтоб развить сообщество! Если вы сомневаетесь, то спросите людей. Напишите предварительную статью и покажите её обществу Pascal. И, тут вы должны уметь принять как похвалу, так и критику! Особенно важно - критику. Исправив большую часть недочётов что вам предоставили в критике, ваша статья может заиграть новыми красками и вы её с большей вероятностью можете предоставить всему сообществу программистов, а не только тем, кто программирует на Pascal.

Ещё один важный пункт.
Сейчас многое теряется на просторах интернета в интернет-свалке. Если вы выкладываете какую-то статью, то статья так же теряется за другими статьями и её становится почти не видно или вообще не найти.
Обязательно должно быть место, где до данной статьи можно дойти достаточно быстро. Двумя-тремя кликами по определённому сайту.

Успехов! 

[marcov]

What fundamental technologies are possessed by the big players- IBM, Oracle, SAP- that conventional tools such as Delphi and Lazarus lack?

Sales and middle management ? :-)

Basically just be big and established and having a sales department that is under the speed dial of civil servants.

Would a consortium of developers using Lazarus and e.g. PostgreSQL be able to compete with e.g. Fujitsu, who are still landing significant government work despite being tainted?

(or Atos, for that matter)

And how could that consortium prove to the commercial and government communities that it was underpinned by robust technology, in a way that C++, Python and the rest weren't?

I don't think this is really  ever about technology. It is about technology management. Popular languages allow easier hiring and replacing of developers with as little as possible internal education.

Education (schools) assist in this by delivering cheap graduates with basic skills in those languages so that they can be hired by big IT organisations at minimal investment. They underpay, so most move on after one or two years, but the student and/or the state paid for that education.

So it basically is about making hiring a as riskless and as cheap as possible investment, even though the model of high turnover but replaceable employees is not all that it is cracked up to be.

[VisualLab]

Does the World want to be saved by FPC/Lazarus? Does the World want to be saved (computer-wise) at all?

Maybe it will happen when governments start to push more for ecology in IT? Because there is no denying that certain quasi-technologies that have been fashionable in recent years waste energy (computers) and transfer (network) on nonsense (mainly people's whims). I mean:

- clouds - storing your own data on servers and sending it "back and forth" through the network,
- artificial intelligence - in fact more fanciful "finders" and "gluers" of text,
- pathological social networking sites of behemoth proportions (Facebook, Instagram, TikTok),
- cryptocurrencies (de facto "computer squeezing money out of naive people's wallets").

Everything has its end sometime. Maybe if the "ends" of these vastly overvalued quasi-technologies were to converge, the noise of their bursting bubbles might sober societies enough that people would go back to creating solid and truly useful software, instead of spending money on useless (clouds, AI) or completely garbage ventures (social networking, cryptocurrencies).

[MarkMLl]

What fundamental technologies are possessed by the big players- IBM, Oracle, SAP- that conventional tools such as Delphi and Lazarus lack?

Sales and middle management ? :-)

I'd suggest that there's more to it than that, and would be interesting in working out /what/.

To give two specific examples.

As the first, IBM used to have Lotus Notes, which- for all its flaws- was highly-regarded as a way to build groupware. They must have some successor even though they've divested themselves of the product: what do they have that we lacK?

As the second, a few months ago this community discussed what happened when a database link was broken (in the context of an application running on a domestic computer which got its IP address reallocated). I explored that using code at https://github.com/MarkMLl/testdb which demonstrates that the LCL's database controls became unusable without warning if the link was interrupted, and I suspect that this is a common failing of this type of development tool. Do the big players just add another level of middleware, or do they have some fundamental way of working around it?

Apropos Fujitsu, they got the Post Office job for political reasons despite having no significant experience in providing retail banking services; I suspect that their obvious competitors had got themselves onto an informal blacklist. I suspect that Pascal is by now in a similar situation.

MarkMLl

[gidesa]

I'd suggest that there's more to it than that, and would be interesting in working out /what/.

Another reason is the large number of "theorists", around '90, behind a language as Java: dozens of books on patterns, object oriented programming, UML, etc. Behind Pascal was a little group with Wirth and few others. Indeed, they supported other languages, as Modula 2, not Pascal. And they not supported much OOP. But OOP won... Although today many of that "astounding" ideas are archived, as mass practices. Think to UML.

Apropos Fujitsu, they got the Post Office job for political reasons despite having no significant experience in providing retail banking services; I suspect that their obvious competitors had got themselves onto an informal blacklist. I suspect that Pascal is by now in a similar situation.

But if an outsider  enters in restricted market, this has a name: competition! 

[MarkMLl]

But if an outsider  enters in restricted market, this has a name: competition! 

Provided they have a product, and that their failure isn't covered up by generations of politicians and venal civil servants.

They had no track record in retail banking, and in retrospect it's quite clear that they didn't have a clue how to write one. In fact while I'm very much looking forward to reading the public inquiry report, the impression I get is that Fujitsu (nee ICL) branded the whole idea of "database integrity" and "rollback anything that looks even slightly dodgy" as a vulgar IBMism which clashed with their traditional approach.

MarkMLl

[Thaddy]

e.g. Fujitsu, who are still landing significant government work despite being tainted?

That part of said company was British before taken over...
The criminals are Brits, actually English, not Japanese.
But as usual Englishmen Tories lay the blame elsewhere.
Not that I mind, Pound is dirt cheap and I booked a trip to London to show my wife where I worked. Vote with your brains on the 4th of July... Can't blame Brexit anymore. I could have written software to predict that, but it wasn't even necessary: Brits can save Brits, but usually a couple of years late... (like 14?)
This has of course nothing to do with me being a political scientist..

So yes, FPC/Lazarus can save the world, except for the English.. (I make an exception for Scots and Welshmen, the Irish already belong to Ireland so no need to interfear.)

[LV]

Sorry, I'm not a political scientist. According to the TIOBE index, Object Pascal saves the world in 11th place. It's not bad?
https://www.tiobe.com/tiobe-index/

[MarkMLl]

That part of said company was Brittish before taken over...

So what part of my "nee ICL" reference do you not understand?

Was it my omission of the accent, which offends your émigré sensibilities? >:-)

I also know that use of ICL equipment was mandated for all state affiliated enterprises. However exemptions were easily obtained when there was no real alternative to buying (usually) IBM equipment.

What I don't know in the current case is whether the Post Office applied for an exemption. If they didn't, I've belatedly come to realise that I might know something that contributed to that decision.

What I also don't know is whether the Post Office kit- or at least the "back-office" part of it- relied on "traditional" ICL mainframe technology, or if it was based on the "IBM-compatible" mainframes which Fujitsu had developed under contested circumstances. And I also don't know with absolute certainty what database technology the system was based on.

In short, I /know/ that the company, staff, and shyster lawyers were British.

So, Thaddy, which of the above gaps can you usefully fill in? >:-)

MarkMLl

[Thaddy]

The gap for people that still make spelling mistakes in their second language? After 60 odd years?

[marcov]

As the first, IBM used to have Lotus Notes, which- for all its flaws- was highly-regarded as a way to build groupware. They must have some successor even though they've divested themselves of the product: what do they have that we lacK?

Being a rather large company for nearly a century or longer ?

As the second, a few months ago this community discussed what happened when a database link was broken (in the context of an application running on a domestic computer which got its IP address reallocated). I explored that using code at https://github.com/MarkMLl/testdb which demonstrates that the LCL's database controls became unusable without warning if the link was interrupted, and I suspect that this is a common failing of this type of development tool. Do the big players just add another level of middleware, or do they have some fundamental way of working around it?

The ones with low connectivity problems occurrence probably implement some on error mitigation and tune the timeouts. Some with more widespread usage go 3 Tier (happening since Delphi 6, posting the datasets as TClientDataSets, and not hard wiring grids to remote sets).

Apropos Fujitsu, they got the Post Office job for political reasons despite having no significant experience in providing retail banking services;

For national systems, the competition often favours sizable domestic partners, based on criteria that are often already old when applied. Besides government this often also includes privatised former national companies. (like railway water and energy sector). Even after privatisation, many connections still keep existing. It is not just the UK who has pain there.

I suspect that their obvious competitors had got themselves onto an informal blacklist. I suspect that Pascal is by now in a similar situation.

Yes. So government and big corporations are mostly out. Go for SMB.

[Thaddy]

Lotus was bought by IBM. It is hardly IBM. It is HCL.
I still have Visicalc on my CBM64 (that still runs)

[VisualLab]

Sorry, I'm not a political scientist. According to the TIOBE index, Object Pascal saves the world in 11th place. It's not bad?
https://www.tiobe.com/tiobe-index/

The problem is that the TIOBE list "sucks". This has been discussed many times on this forum.

So, once again - the TIOBE list is garbage because:

• TIOBE is one of the IT companies, not an independent institution,
• their composition includes "peas and cabbage" mixed together,
• the results depend on everything the web spider digs up on the Internet about a given language.

Ad. 1. TIOBE is one of the IT companies, not an independent institution. So on this point they are not credible as one of the interested parties. They can falsify the results of the ranking so that they favor them in their business activities (and they probably do - they just need to "tweak them a bit").

Ad. 2. The list includes "computer" languages ​​of any category. And they are not equivalent in terms of purpose. There are categories of languages ​​used for specific application domains. For example:

• system-wide programming (Ada, C, C++, C#, Haskell, Java, Pascal, Rust, Scala, etc.),
• microcontroller programming (assembler, C),
• classic calculation automation (Matlab, R, SAS, Wolfram),
• generating web content (Perl, PHP, Ruby),
• support for interactions in web browsers (JavaScript, TypeScript->JS),
• cooperation with RDBMS (SQL and its mutations),
• scripts automating sequences of operations in programs or OS (AutoLisp, AWK, Bash, Lua, Perl, PowerShell, Python, VisualBasic),
• hardware description for CPLD and FPGA (Verilog, VHDL),
• educational and toys (Logo, Scratch).

Therefore, the popularity of a given language depends on which category is more or less mass. For example, many different websites have been created for a long time, which is why both JavaScript (in the browser) and PHP (on the server side) are very popular. AI has been fashionable for the last few years, so Python is used to "glue" data flowing through computational libraries for AI (written in C and C++) to make programming easier for AI "scientists" (because the vast majority of them have no idea about programming nor do I want to learn it). Year by year, more and more different small devices are produced that are controlled by microcontrollers, hence the still quite high popularity of C and the moderate popularity of assembler. And here's the problem, because there is no such thing as an assembly language - there are assembly languages, because each family of microcontrollers has its own. However, there are much fewer GUI applications, especially engineering or industrial ones. They require both knowledge of the field for which the program is being created (CAD, etc.) and experience in creating such applications (most of them are probably GUI programs). There are few new applications of this type, most of them are those maintained by manufacturers. And the lion's share of them were written in C++. It is also important that almost all AI libraries are written in C and C++. Hence the considerable popularity of both these languages. All this is compounded by suppliers' marketing campaigns, both positive (own solutions - advertising) and negative (competitor's solutions - FUDs).

Ad. 3. The point is clear here: "it doesn't matter what they say, what matters is that they say it." Whatever. And this will all add up and "it will be a great result". So numerical indicators like these are garbage. And it is probably not a lack of knowledge of the people from TIOBE - it is a deliberate action intended to mislead readers.


To sum up - it's as if someone made a list of the popularity of tools used, for example, by a car mechanic. And the first place would probably be a screwdriver. And then there would be: pliers (without separation into types - treated as an assembler in the TIOBE list), a jack, a compressor, a hammer, etc.

There are also other lists than TIOBE. And just like the one described, they don't make sense. Lists that would make sense would include:

• categories (applications),
• the number of applications actually created.

And this second point is practically impossible to determine (even with some reasonable approximation).

So I wouldn't count on Delphi/Object Pascal ranking much higher (but it would be nice if it happened).

[Thaddy]

• microcontroller programming (assembler, C),

You forgot Pascal there, it was once the most used language for micro-controllers.