Are We Dead Yet?

[marcov]

Simple: Pascal is not linked to some big multinational that spoon feeds it into  programmers throats to the point of forcing it down their throats.

[jamie]

nothing to worry about!
https://youtu.be/d-diB65scQU

[ASBzone]

It doesn’t do any good to have an excellent compiler when most people are already prejudiced against pascal language thanks to all the propaganda they have heard about pascal. Getting schools to stop teaching pascal as a beginner programming class was a major victory for pascal haters. Most people only want to program in a language that is guaranteed to have jobs which discourages new people from learning pascal.

This is very true.  Whatever the industry suggests is the sexy new thing, is what will get all the attention for at least a while.


Other languages will have to settle for their niche use-cases and communities.

Popularity is not something that can easily be generated (or sustained, for that matter), and is easily impacted by perception.

[ASBzone]

Simple: Pascal is not linked to some big multinational that spoon feeds it into  programmers throats to the point of forcing it down their throats.

And hasn't been for a while.   Borland had a good run for a bit.

The problem is that once you come off the radar, it is an order of magnitude harder to put you back on.

Even if both Microsoft and Google where to suddenly decide to throw money at Pascal implementations, I don't see that they would have nearly as much success overcoming the built up bias and disdain that already exists, and the loss of that hook into the educational system that Pascal used to enjoy.

The formula for rejuvenating a perceived dead or dying or obsolete language is very different, and requires more time and effort, than what it woudl take to make up a brand new language and get it entrenched.   (And both Microsoft and Google have experience with the latter endeavor.)

[ASBzone]

I am curious if any big corporations have ever tried to buy influence in fpc project though.,
Were they snubbed? I’m trying to figure out why we have such determined enemies .

I expect that the vast majority of any opposition to Pascal has absolutely nothing to do with FPC, or its developers.   There is overall disdain for the language, including the need to declare everything up front.   

And there may be hold over disdain for Borland / Embarcadero, which is largely the face of Pascal in the corporate realm.

Frankly, Borland resuscitated corporate interest in Pascal when it was already waning and sliding out of favor in educational circles.  If Borland has been as stable/profitable as Microsoft, the fate of Delphi and Object Pascal would be somewhat less tenuous today.   Microsoft's propping up of BASIC actually kept that in focus for a long time (along with BASIC remaining a focus in educational circles for a longer time than Pascal).

There are lots of factors, but my anecdotal experience (including two of my children who develop code for their hobbyist purposes [robotics]) suggests that the formal nature of Pascal is a deterrent to many, who want to just leap into things and crank out code.

And I remember the grief my friend and colleague experienced back in the early 90s when he first fell in love with Delphi 1.0, and was trying to champion it in our department to our boss who favored VB, and didn't care for any company that wasn't a behemoth so that it could survive financial pressures.

I remember those battles fondly.  My friend essentially made a deal that if he couldn't develop some internal department apps faster with Delphi than my boss could with VB, he would abandon the Delphi experiment.    He won -- not only because of the RAD, but because of the excellent database integration Delphi offered -- and our boss grudgingly allowed him to get training and move forward on other projects.

Fun times...

Pascal already had a negative perception by then, and I wasn't even a decade into having learned it and loved it at that point.

[marcov]

Simple: Pascal is not linked to some big multinational that spoon feeds it into  programmers throats to the point of forcing it down their throats.

Apart from the usage of fpc itself this is what I like most about fpc , it is not beholden to large corporations.

Yes. Even if they are benign they can force weird direction changes to align  with some other biz. (e.g. to push  ChatGPT or whatever)

Recently someone was trying to recommend rust as a better language than pascal and I looked.up the sponsors of it on the rust foundation website and it read like a who’s who in evil corporations.

I am curious if any big corporations have ever tried to buy influence in fpc project though.,
Were they snubbed? I’m trying to figure out why we have such determined enemies .

Not to my knowledge. Business involvement in FPC has happened but is relatively minor for the core project, and mostly by relatively small employers of developers.

Also I’d like to know has anyone involved with fpc project ever tried to lobby schools to try to get them to teach pascal again. Offering free compilers and technical help ?

All the time. I assisted with putting Lazarus on the image for the CS student laptops of my uni for a small decade. I'm not sure if they still use it thoguh. I think that subject is better something for a new thread.

[marcov]

This is very true.  Whatever the industry suggests is the sexy new thing, is what will get all the attention for at least a while.

Other languages will have to settle for their niche use-cases and communities.

Popularity is not something that can easily be generated (or sustained, for that matter), and is easily impacted by perception.

Not by projects the size of us (with a few small business partners like Blaise magazine and maybe some Delphi component builders).

I've always had more success by simply demonstrating what can be done (live in need be), rather than most "promote pascal" discussions that always seem to be about "modernizing" the dialect. In my experience, language details are only a small part of why people start to use Lazarus.

It is the set of features that makes it a fit for certain projects, e.g. IDE with designer, libraries, not too abstracted from lowlevel and the simplest of deployments of made apps.

Also don't rule out secondary users (users that are not fulltime using lazarus, but e.g. some web app). They often have a surprising need for a simple swiss army knife of a tool for all other stuff. (e.g. database pumps, file format conversions etc)

[ASBzone]

I've always had more success by simply demonstrating what can be done (live in need be), rather than most "promote pascal" discussions that always seem to be about "modernizing" the dialect. In my experience, language details are only a small part of why people start to use Lazarus.

It is the set of features that makes it a fit for certain projects, e.g. IDE with designer, libraries, not too abstracted from lowlevel and the simplest of deployments of made apps.

I agree.  At the end of the day, IMO, FPC/Lazarus is better served by demonstrating and promoting its suitability to specific projects and tasks, and keeping emotion and popularity out of the discussion.   As soon as you try to enlist people from a platform perspective, rather than just from a best-tool-for-the-job-at-hand perspective, then popularity and perception and biases all come into play.

The FPC/Lazarus community is probably best served by continuing to support the development needs and use cases that it is strong in, and keeping an eye on changes in the industry that it makes real sense to be able to support -- not just because they seem cool at the moment.

When the "customers" of FPC and Lazarus no longer have any interest in the tools, and when even core developers no longer have any interest -- even for their own projects -- then we can legitimately conclude that there is no viable interest, and then death/retirement of the ecosystem will make perfect sense. 


But this day won't be averted by blowing whistles and throwing parties and making noise and adding every popular feature found somewhere else.

[440bx]

anecdotal experience (including two of my children who develop code for their hobbyist purposes [robotics]) suggests that the formal nature of Pascal is a deterrent to many, who want to just leap into things and crank out code.

I think that has a lot to do with Pascal's falling out of favor and also explains why languages that do not promote good programming habits are popular.

In "big iron" circles, it used to be that software was _designed_ (admittedly, sometimes poorly) and lasted for a number of years (there is still plenty of COBOL code around, lots of it critical).  These days, "design" is often pulled out of polluted air to solve a poorly identified problem for a few months (if not just a few weeks.)

The consequence of this "just add water" life cycle is that it doesn't make much sense to invest development time into producing good software.  What makes sense is to use whatever 2-bit language (usually an interpreter with plenty of libraries) to produce "something" that the next rain will wash away.  No wonder Java and Python are popular.

[Curt Carpenter]

How long, do you all think, before something like chatGPT will be able to convert a large code base into any given target language, with annotations about where compromises have had to be made and trade-offs adjusted during the translation?  How would such a development change the language debates? (And what features of those debates would NOT be changed at all?)

Pascal won't die until there is no longer anyone making the intellectual investment required to learn it (not change or Improve it).  And even then, that AI translator may still be able to translate into Pascal much as Google Translate can do a pretty good job with Latin (not that Latin is "dead" ...)

[ASBzone]

The consequence of this "just add water" life cycle is that it doesn't make much sense to invest development time into producing good software.

At a commercial level, yes, this appears to be true.  Or, at least, to be perceived as being true.

But, what I inevitable find -- both in corporate America and in my home -- is that at some point, the lack of proper design catches up to the developers, and tech debt gets created very, very quickly.   I end up helping my youngest on a number of troubleshooting endeavors where at least 40-50% of the issue is that the lack of an upfront design, and the ability (and encouragement) to just changes gears along the way, results in logic bugs and even typos that are hard to troubleshoot or resolve.

But he's not going to give up that ease of development. 

No wonder Java and Python are popular.

Yes, I was around for the start of that too, in early 1995...   (It was cool, though, what it allowed you to do at the time)

Each time, these types of technological transformations doesn't just give more flexibility to existing developers -- they actually open the door to totally new developers who wouldn't necessarily have embraced software development before.  And they don't have the background and the structure, so they start off with bad habits from the very beginning.

[ASBzone]

How long, do you all think, before something like chatGPT will be able to convert a large code base into any given target language, with annotations about where compromises have had to be made and trade-offs adjusted during the translation?  How would such a development change the language debates? (And what features of those debates would NOT be changed at all?)

As someone who keeps an eye on the whole ChatGPT craze for both scripting/automation and systems administration (and cybersecurity, actually), ChatGPT is going to remain a tool that can help competent people achieve certain ends faster, by filling in some gaps, or putting some mostly viable code in place -- but it will burn those who can't actually determine the viability of the responses they receive.

We've had search engines for decades at this point, and some people can use them to quickly find out what they need to know or what they want/need to do, and others have no clue about sifting through what they have received for legitimacy.

ChatGPT will simply compound this divide for at least the next 3-5 years.

ChatGPT will not make dumb people smart.  It will make smart people more productive or efficient, and it will make dumb people more dangerous.

But that is a discussion for another thread...

[Curt Carpenter]

My question, in part I guess, is whether such a translation capability is possible at all.  If a large program P in language L can produce an executable C for a given ISA using some process F, is there a process F' which, given C, can produce P in some other language L' ?

[sketch]

In "big iron" circles, it used to be that software was _designed_ (admittedly, sometimes poorly) and lasted for a number of years (there is still plenty of COBOL code around, lots of it critical).  These days, "design" is often pulled out of polluted air to solve a poorly identified problem for a few months (if not just a few weeks.)

The consequence of this "just add water" life cycle is that it doesn't make much sense to invest development time into producing good software.  What makes sense is to use whatever 2-bit language (usually an interpreter with plenty of libraries) to produce "something" that the next rain will wash away.  No wonder Java and Python are popular.

This is part of why MIT switched from using Scheme to using Python. Today no one develops software from the ground up (rarely). What is fashionable is to use the thousands of libraries that have already been created and use them, as you mentioned. I think it is more akin to using Scratch and pushing blocks together to generate your desired outcome.

Of course, when that library turned out to be left-pad, which was used by React, was deleted it broke the internet. It is much easier to use an 11 line JavaScript library that pads characters instead of having to write it yourself. So a developer doesn't even know what is being used in their own software. And when that library is removed from npm by the author, the millions of applications written that required it, unbeknownst to the developer, stop working.

[440bx]

It seems like what used to be called spaghetti code is gaining popularity

Seems ???... polymorphic calls, a staple of OOP, is the ultimate in spaghetti code.

It is even worse than COBOL's ALTER statement (run-time dynamic goto, i.e, a polymorphic destination) which got programmers fired for using it, yet today, polymorphic calls are "kewl" and those  like me who point out the atrocity they are and, how anathema they are to good programming are looked at with one of those "what are you talking about looks ?"... gotta love it.