R.I.P. Niklaus Wirth

[Jurgen Keller]

On January 1, Niklaus Wirth died at the age of 89. For a long time, Mr. Wirth was a professor at the Federal Institute of Technology (ETH) in Zürich. He designed the programming languages Pascal (1970), Modula (1980) and Oberon (1988), and the workstations Lilith (1980) and Ceres (1986), as well as their operating system.

Wirth received a PhD from the University of California at Berkeley in 1963. He was awarded the ACM Turing Award in 1984. Maybe you know his article "A Plea for Lean Software" or the book "Algorithms and Data Structures" or "Compiler Construction", or you remember Wirth's Law, which states that software is getting slower more rapidly than hardware is becoming faster.

His influence is still around, for example, in the Go programming language. Robert Griesemer, one of the inventors of Go, did his doctorate under the supervision of Hanspeter Mössenböck and Niklaus Wirth on the subject of a programming language for vector computers. If you know Go and the other Wirth languages, you will notice the relationship. BTW, Mr. Griesemer grew up in a village near to mine and for two years or so, we were at the same secondary school. He and his colleague once grabbed my HP-41C and gave it back with some strange characters on the display, created by synthetic programming. After that I ordered the Wickes book which was not that easy back in those days. Good memories.

R.I.P., Niklaus Wirth, and thank you for the fish (programming languages, design principles, amazing computers, etc.). I really like your work!

[Maximilian Hohmann]

Hello!

He must have been one of the very few computer scienentists widely known even to not-at-all-computer-science persons like myself. Maybe the only one. His trademark " := " will survive him, at least as long as people write programs in Go.

Once I had to briefly do some work using Pascal, so his ways and mine crossed for a short period.

Regards
Max

[jthole]

That is sad news; Wirth, Knuth, and Dijkstra bring me right back to my first year in university.

[Namir]

Sad to hear the news about the father of Pascal and Modula-2 which I had programmed in both. He has given us a lot of tools to sharpen our skills in structured programming!

R.I.P. Professor Wirth!

Namir

[WillMarchant]

Sad news. His algorithms book is brilliant.

[Namir]

In the mid 80s I used to write for Computer Language magazine. The only in-person interview I did for that magazine was with Ada's creator Dr Jean Ichbia. I never had a chance to meet Professor Wirth. I did ask him a question through the intermediary of one of the main editors of the magazine. The question was related about my interest in creating a new programming language. Professor Wirth replied to my editor that creating a new programming languages is not as easy as it seems.

I remember asking Dr Ichbia, in my interview, about the difference between Ada (which was influenced by Pascal and was adopted by the US DOD as the standard programming langauge) and Modula-2. Ichbia's reply that solving real-world applications like those for naional defence requires a complex progamming language like Ada. He added that one cannot build complex real-world applications with simple solutions like Pascal or Modula-2.

I guess both Dr Ichbia and Professor Wirth can discuss and compare their programming languages in the great beyond.

Namir

[Jurgen Keller]

(01-05-2024 03:16 AM)Namir Wrote:  ...
Professor Wirth replied to my editor that creating a new programming languages is not as easy as it seems.
...
He [Dr Ichbia] added that one cannot build complex real-world applications with simple solutions like Pascal or Modula-2.

I guess both Dr Ichbia and Professor Wirth can discuss and compare their programming languages in the great beyond.

Namir,

thank you for sharing this anecdote. I fully agree that designing a new programming language is a very complex task. A big challenge is to make things easy on the surface. However, under the hood that could need complex algorithms. Example: garbage collector. A very convenient feature because you don't need to think about deleting objects anymore. However, implementing an efficient and fast garbage collector needs very sophisticated algorithms.

At work, I have to use Java. This programming language added so many features over time that you often have to think more about how to solve a task best in Java than to think about the actual problem you want to solve. Google and Stack Overflow are a software engineer's best friends ;-)

If you are interested in how to design a new programming language, then Go is a good candidate to read about. There are also many videos on YouTube about the motivation to invent a new programming language and what the design decisions were. In the end, as it's the case with many things, it's also a matter of personal preference. As you might guess, my preference is more on the Wirth side.

[Eddie W. Shore]

Rest in peace and gratitude, Mr. Niklaus Wirth

[BruceH]

(01-05-2024 01:59 PM)Jurgen Keller Wrote:  I fully agree that designing a new programming language is a very complex task.

Someone once said that every new (procedural) language created since about 1980 is simply Algol re-implemented badly.

Personally, I wish that integrated systems like Oberon or SmallTalk had succeeded and the world hadn't gone down the DOS & Windows path. If Microsoft had put half as much effort into adopting one of those instead of trying to impose de facto standards and 'defeat' perceived enemies then the IT landscape now would be radically different and all the better for it.

[ijabbott]

(01-04-2024 05:28 PM)Maximilian Hohmann Wrote:  His trademark " := " will survive him, at least as long as people write programs in Go.

The := operator was used in ALGOL before Wirth got hold of it though!

Quote:Once I had to briefly do some work using Pascal, so his ways and mine crossed for a short period.

It was used as a teaching language on Manchester's freshman CS course when I was there in the early 80s, with an emphasis on structured programming. I didn't use it again (in its Borland Object Pascal variant) until the late 90s.

[harryp]

I got into programming when I returned from the service, and thought about going in that direction. I was the only non-programming major in the Pascal class and got the highest score. I appreciated the 'tank proof' nature of Pascal. He made a major contribution to cleaning up the mess of unstructured programming. Along the way I dabbled in Macro Assembler, Basic, C, and some C++.

[Peter Klein]

I was never a professional programmer, but I learned several programming languages in the course of my self-education. Those were the days of the marketing battle between C and Pascal. Computer people were often asked which side they were on. I would always say that I preferred Pascal, because “I like my variables to be what they are, rather than where they are.” ;-)

RIP, Dr. Wirth, and thank you for your contributions.

[BruceH]

(01-06-2024 11:12 AM)ijabbott Wrote:  The := operator was used in ALGOL before Wirth got hold of it though!

Wirth was on the Algol-60 committee. So he may have proposed it, or simply recognised its advantages as an assignment operator and continued to use it.

[Wes Loewer]

(01-06-2024 08:07 PM)Peter Klein Wrote:  I would always say that I preferred Pascal, because “I like my variables to be what they are, rather than where they are.” ;-)

I cut my teeth on Pascal but soon after became a fan of C. In spite of that, I'm afraid I still don't get your "what" vs "where" joke. Please explain.

[jthole]

(01-13-2024 02:22 PM)Wes Loewer Wrote:  

(01-06-2024 08:07 PM)Peter Klein Wrote:  I would always say that I preferred Pascal, because “I like my variables to be what they are, rather than where they are.” ;-)

I cut my teeth on Pascal but soon after became a fan of C. In spite of that, I'm afraid I still don't get your "what" vs "where" joke. Please explain.

Passing a value or a pointer, I guess.

[Albert Chan]

Both C and Pascal support pointers, but C is "weakly typed", in the sense we can type-cast to whatever.
C is great if you know what you are doing ...

gcc:1> double x = M_PI;
gcc:2> printf("%.17g, bits = %#016I64x \n", x, * (uint64_t *) &x);

3.1415926535897931, bits = 0x400921fb54442d18

I guess tihs is "what" vs "where" (just some memory) joke comes from.

[ijabbott]

(01-06-2024 11:39 PM)BruceH Wrote:  

(01-06-2024 11:12 AM)ijabbott Wrote:  The := operator was used in ALGOL before Wirth got hold of it though!

Wirth was on the Algol-60 committee. So he may have proposed it, or simply recognised its advantages as an assignment operator and continued to use it.

The latter. As far as I can tell, he was not involved in the preliminary IAL (International Algebraic Language) report that became ALGOL 58. (The name "IAL" was rejected as an "'unspeakable' and pompous acroynym". Oh, how times have changed!)

I'm not sure about your statement of Wirth being on the ALGOL 60 committee. He is not mentioned in the revised report by Naur, et al.

[BruceH]

(01-13-2024 05:26 PM)ijabbott Wrote:  I'm not sure about your statement of Wirth being on the ALGOL 60 committee. He is not mentioned in the revised report by Naur, et al.

Checking again, I see I mis-read one of the obituaries. It said he was on a working group to replace Algol 60 which I misread as being on the Algol 60 group. Sorry for any confusion.

[HrastProgrammer]

Thank you, Mr. Wirth ... I am still making most of my living by programming in Pascal, Object Pascal / Delphi, my favorite programming languages, next to machine language.

[0db]

Nelson H. F. Beebe of the University of Utah has compiled an extensive bibliography of works by and about Wirth: [url=https://math.utah.edu/pub/bibnet/authors/w/] takes you to an FTP server; scroll down to 'Wirth'. I came across this in volume 45 of the TeX User's Group TUGboat journal; Wirth and Pascal had connections to Donald Knuth and LaTeX.