My friend has been making a working replica of the Patinho Feio. As I happen to know Portuguese, I translated some of his materials about this project, including making English subtitles for his (very wide-ranging) university lecture about the project.
(Maybe I'll feel silly about having done this at some point, because AI translation is getting so good...)
There's an amazing amount of material about this computer available online now. Lots of manuals, interviews, images, and so on. It must have been really exciting to be involved with that project in the early 1970s. (As I understanding it, an American electrical engineer who had worked with computer architecture was a visiting professor at the Universidade de São Paulo at that time, and led the students through many of the necessary steps -- but all of the implementation and design, down to the instruction set with its Portuguese-language mnemonics, was the students' own work.)
In his lecture, he notes that ITA (Instituto Tecnológico de Aeronaútica) actually turns out to have made its own computer even earlier than USP's Patinho Feio, but there's almost no documentation of any kind about ITA's machine available, whereas USP students published various theses related to the Patinho Feio and preserved extensive documentation and even physical artifacts and software related to it. He thus calls the ITA computer the "0th Brazilian computer".
There are two stages really, right? Pre-microprocessor, it takes some effort to put together a programmable computer. Although now even amateurs implement computers based on 74-series and memory chips without a microprocessor.
After the microprocessor, it was truly very manageable for an amateur anywhere to build a computer. A handful of chips get you there, with no requirement for elaborate PCB or routing. Wikipedia lists the Mark-8 as first published and available as kit, in 1974 https://en.wikipedia.org/wiki/Mark-8 - from a chemistry grad student. So, any country is now in. No problem. And "more interesting" university or research projects shift to LISP machines, multiprocessors, coprocessors, etc, or the VLSI chips themselves. Which does not prevent countries from promoting "national microcomputers" (including OS and programming language translated in the local language). Around microprocessors what is also interesting in this second phase are the commercial and "industry" development attempts in various countries as described by Asianometry.
Thanks for sharing. What surprised me was that there was one group that tried to pull of parallel LISP on Z80s whereas the rest seem to be keen to use existing technology (no own operating system, no own microprocessor chip designs attempted).
"Unfortunately, [...] the lack of collaboration and coordination among the various groups designing computers inMexico, and the lack of communication between industry and academia hindered the country's further development in this field."
I found this sentence and subsequent passages to be the most interesting; how come Brazil could pull things off when a similar environment existed in Mexico and Brazil? Are Brazilians "more collaborative" as a tendency than Mexicans?Would be interesting to get more detail on good v. bad policy decisions in both countries that led to it.
Minor corrections: Table 1. "intepreter" -> "interpreter"; "Hasta 64" -> "up to 64"
Brazil had government incentives for fomenting the national industry as well as high taxation to deter imports.
Nowadays only the high taxes remain (effectively a 100% tax), the current government is trying to implement new incentives to rebuild the national industry though.
Also, I feel Mexico by being under strong influence of the US has been kind of sabotaged economically, while Brazil together with neighboring countries tried to develop a local market (MercoSul).
I wonder if there are just geopolitical or demographic issues. Brazil is larger than Mexico by a bit, there are a lot more Spanish speaking countries. Spain and Mexico are similar sizes, while Brazil is much larger than Portugal, plus Mexico has to deal with the US, and our endless appetite for talented people, being right next door…
This paper is wrong if the article "History of computing in South America" at Wikipedia is right [1]: the first computer arrived in Chile in 1957, not in Mexico.
There is another issue. Here [2] where the first computer in Brazil also arrives in 1957 and it is not included in [1]:
> 1957 - A Univac-120 arrived, the first computer in Brazil, acquired by the Government of the State of São Paulo, it was used to calculate all water consumption in the capital. It occupied the entire floor of the building where it was installed. Equipped with 4,500 valves, it performed 12,000 additions or subtractions per minute and 2,400 multiplications or divisions at the same time.
So, we need to check which of those two computers arrived earlier.
BTW, there is an article about Argentina and computers in a complex political moment [3] (as usual). It also touches the computer science field itself beyond the hardware involved.
The first computer built in Latam was in Argentina, it started in 1958 and was finished around 1962 [4].
Chile (South america) is not in the same continent as Mexico (North america).
The way it is phrased is prone to confusion but article says:
"This date marks a milestone in the history of computing in Latin America, as the IBM-650 was the first electronic computer to operate on this continent, south of the Rio Grande."
"This continent" meaning North America, not Latin America (which is not a continent).
The paper notes: "Thus, an annual conference called 'Computers and Their Applications' was organized. It is interesting to note that the third edition of that conference, held in 1961, featured lectures by professors John McCarthy, Marvin L. Minsky6, and Harold V. McIntosh [9]." That conference was also known as the First International LISP Conference -- see https://mcjones.org/dustydecks/the-first-international-lisp-...
I used to assemble Dell's Optiplex 486s at factory in North Austin back in the day (early 90s) before DELL move to Round Rock. Some of the components like the daughter board (a riser board to connect your LAN card or Souncard) was made in Mexico, among other parts.
Great article.
> include a small but rich (and sometimes astonishing) variety of systems ranging from research and teaching-oriented computers to high-performance personal computers.
You'd have to ask the author, perhaps in 2024 he found the notion that Mexican engineers built a parallel computer based on data flow that executed LISP programs in 1979 (and onwards for a few years) to be astounding.
[+] [-] thiagoharry|1 year ago|reply
I was not aware about these Mexican computers. The fact that they had computers that could be programmed in LISP, not only in Assembly, is very cool.
[+] [-] schoen|1 year ago|reply
https://forum.fiozera.com.br/t/historical-preservation-of-th...
https://www.youtube.com/watch?v=UaDzfq42B-g
(Maybe I'll feel silly about having done this at some point, because AI translation is getting so good...)
There's an amazing amount of material about this computer available online now. Lots of manuals, interviews, images, and so on. It must have been really exciting to be involved with that project in the early 1970s. (As I understanding it, an American electrical engineer who had worked with computer architecture was a visiting professor at the Universidade de São Paulo at that time, and led the students through many of the necessary steps -- but all of the implementation and design, down to the instruction set with its Portuguese-language mnemonics, was the students' own work.)
In his lecture, he notes that ITA (Instituto Tecnológico de Aeronaútica) actually turns out to have made its own computer even earlier than USP's Patinho Feio, but there's almost no documentation of any kind about ITA's machine available, whereas USP students published various theses related to the Patinho Feio and preserved extensive documentation and even physical artifacts and software related to it. He thus calls the ITA computer the "0th Brazilian computer".
[+] [-] glimshe|1 year ago|reply
[+] [-] hulitu|1 year ago|reply
[+] [-] adl|1 year ago|reply
[+] [-] creer|1 year ago|reply
After the microprocessor, it was truly very manageable for an amateur anywhere to build a computer. A handful of chips get you there, with no requirement for elaborate PCB or routing. Wikipedia lists the Mark-8 as first published and available as kit, in 1974 https://en.wikipedia.org/wiki/Mark-8 - from a chemistry grad student. So, any country is now in. No problem. And "more interesting" university or research projects shift to LISP machines, multiprocessors, coprocessors, etc, or the VLSI chips themselves. Which does not prevent countries from promoting "national microcomputers" (including OS and programming language translated in the local language). Around microprocessors what is also interesting in this second phase are the commercial and "industry" development attempts in various countries as described by Asianometry.
[+] [-] jll29|1 year ago|reply
Minor corrections: Table 1. "intepreter" -> "interpreter"; "Hasta 64" -> "up to 64"
[+] [-] hcarvalhoalves|1 year ago|reply
Nowadays only the high taxes remain (effectively a 100% tax), the current government is trying to implement new incentives to rebuild the national industry though.
Also, I feel Mexico by being under strong influence of the US has been kind of sabotaged economically, while Brazil together with neighboring countries tried to develop a local market (MercoSul).
[+] [-] huevosabio|1 year ago|reply
Brazil strikes me as having a decent history of industrial policy with the government partnering with private industry.
That's why you have what I think it's a freer market in Mexico but stronger tech-heavy corporations in Brazil (eg embraer).
[+] [-] bee_rider|1 year ago|reply
[+] [-] wslh|1 year ago|reply
There is another issue. Here [2] where the first computer in Brazil also arrives in 1957 and it is not included in [1]:
> 1957 - A Univac-120 arrived, the first computer in Brazil, acquired by the Government of the State of São Paulo, it was used to calculate all water consumption in the capital. It occupied the entire floor of the building where it was installed. Equipped with 4,500 valves, it performed 12,000 additions or subtractions per minute and 2,400 multiplications or divisions at the same time.
So, we need to check which of those two computers arrived earlier.
BTW, there is an article about Argentina and computers in a complex political moment [3] (as usual). It also touches the computer science field itself beyond the hardware involved.
The first computer built in Latam was in Argentina, it started in 1958 and was finished around 1962 [4].
[1] https://en.wikipedia.org/wiki/History_of_computing_in_South_...
[2] https://tpilmg.wixsite.com/tpilmg/en/historia-dos-computador...
[3] The First Decade of Computer Science in Argentina: https://sedici.unlp.edu.ar/bitstream/handle/10915/24033/Docu...
[4] https://mapmeld.medium.com/first-computer-by-country-81fa479...
[+] [-] jecel|1 year ago|reply
Most other sources just mention the year.
[+] [-] prmoustache|1 year ago|reply
The way it is phrased is prone to confusion but article says: "This date marks a milestone in the history of computing in Latin America, as the IBM-650 was the first electronic computer to operate on this continent, south of the Rio Grande."
"This continent" meaning North America, not Latin America (which is not a continent).
[+] [-] pmcjones|1 year ago|reply
[+] [-] CoBE10|1 year ago|reply
[+] [-] agumonkey|1 year ago|reply
[+] [-] cgannett|1 year ago|reply
[+] [-] moralestapia|1 year ago|reply
https://news.ycombinator.com/item?id=28337064
[+] [-] flobosg|1 year ago|reply
[+] [-] copperx|1 year ago|reply
[+] [-] makz|1 year ago|reply
[+] [-] javier_e06|1 year ago|reply
[+] [-] akira2501|1 year ago|reply
Why exactly would it be "astonishing?"
[+] [-] defrost|1 year ago|reply