So I set myself the task of picking five great works of software. The criteria were simple: How long had it been around? Did people directly interact with it every day? Did people use it to do something meaningful? I came up with the office suite Microsoft Office, the image editor Photoshop, the videogame Pac-Man, the operating system Unix, and the text editor Emacs.
Each person has his or her own criteria for these sorts of things, but in my view, this list is woefully inadequate. If it were up to me, I would pick these, in no particular order:
- A-0 System: the first ever compiler, written by Grace Hopper in 1951 and 1952, for the UNIVAC I.
- UNIX: This one’s a given.
- WorldWideWeb/CERN HTTPd: the first web browser and the first web server, both written by Tim Berners-Lee. Also a given.
- Xerox Star: this one is actually a tie between the Star, its research predecessor the Alto, and Douglas Engelbart’s NLS. These three combined still define the way we do computing today – whether you look at a desktop, a smartphone, or a tablet. I decided to go with the Star because it was the only one of the three that was commercially available, and because it’s so incredibly similar to what we still use today.
- Windows: you cannot have a list of the greatest software of all time without Windows. You may not like it, you may even hate it, but the impact Windows has had on the computing world – and far, far beyond that – is immense. Not including it is a huge disservice to the operating system that put a computer on every desk, in every home.
This leaves a whole bunch of others out, such as Lotus 1-2-3, DOS, the Mac OS, Linux, and god knows what else – but such is the nature of lists like this.
Without them, there would be no MS Office. It’s just too bad that even with them, there’s an MS Office to fight with everyday.
I really miss my WordPerfect. It’s too bad they dropped Unix support with 7? 8? And then dropped Linux support with 9.
If there was a modern Linux version, I’d still be using it (the last version no longer runs on recent versions of Linux). I still use WordPerfect 9 on Windows XP/7.
There really is no replacement for Reveal Codes, SGML-based file formats, or the PDF features (LibreOffice PDF export is good, but not as great as WPs).
Supposedly the MS Office formats have idiosyncrasies that can be traced back to compatibility with Lotus 1-2-3…
Yup. At least the traditional excel .xls format has a bug in it by design to remain compatible with lotus 1-2-3. The date format incorrectly assumes that 1900 was a leap year, or something like that. It stores the date/time as number of days since Jan 1, 1900, so every date would be off a day if it didn’t account for it. Open office/libre also have the same behavior.
No. Excel starts counting dates 1900/1/1. Openoffice starts 1899/12/31. After February 1900 the values are the same.
Sorry for the late update:
What I meant was that Open Office has the same bug for the XLS document type.
Not to mention every bug and design mistake made in the last 25 years.
Calling Windows, and especially Office, good software is like describing the U.S. tax code as a pinnacle of good legislation.
Uhm… If you’re going to follow that logic, shouldn’t we be saying Visicalc and Electric Pencil?
I mean, where do you draw the line?
Though honestly, I wouldn’t list WordPerfect over Wordstar — as at least today there are still people using ^K commands; as opposed to putting an overlay over the function keys…
I’m sorry, but on a list where you’re tracing back innovation to the originals like A-0, NLS, or CERN, Windows absolutely does not belong. Windows is in almost every respect a rehash built on top of broken standards, driven by marketing, not technical progress, and most of this readership knows it.
The graphical desktop? That was pioneered by PARC and brought to the public by the Mac. Desktop multimedia with pre-emptive multitasking? That’s the Amiga. Most of what’s good in more recent Windows iterations has been warmed over from VMS and Unix.
Why did the Windows world wait until 1995 catch up to the rest of the world in allowing anything beyond 8.3 filenames??
Putting Windows on a list like this is just painting a target on yourself on a site read by those knowledgeable on the subject.
I totally disagree, to not recognize the contributions Windows has made to the computing world is the worst kind of self imposed blindness. It, paired with the x86 processor, really did put a computer in every home, and the process improvements in the hardware, and the lessons that the insane hardware combinations Windows supports, set the stage for Android, which basically uses the same model as Wintel
Then it must be specified if the greatness refer to technology or marketing. Wintel is firmly placed in the latter.
That’s a monopoly’s marketing in action, not technological innovation. The market could have seen an ecosystem of competing platforms. The Amiga had long filenames, pre-emptive multitasking, hardware-accelerated video and sound, and multimedia in 1986. The Atari had amazing capabilities too. The Mac was the clear leader in GUI development, and BeOS had innovations still not fully realized in today’s OS. OS/2 had incredibly solid design, such that video driver crashes didn’t even bring down the operating system, and an awesome threading model.
All of these deserved a fair shot at the market denied to them by monopolistic practices and lock-down through OEM agreements with Microsoft. The fact that a PC wound up on every desktop was bound to happen. It happened belatedly and half-assedly and with security holes that have fundamentally affected the growth of the internet, thanks to Microsoft.
Everything else on this list was a first of its kind innovation; Microsoft invented nothing of significance, not even MS-DOS. In fact, everything they added was old news to the technophiles by the time it showed up in a Microsoft product, usually after MS’s marketing arm had argued against it and acted to kill it elsewhere.
Edited 2014-05-03 03:44 UTC
MIcrosoft didn’t have a monopoly when Win 3.1 came out, it sold millions of copies. People stood in line for Win 95, it created the monopoly.
Marketing maybe, monopoly, no. Either way, It helped bring computing to the masses.
You could be mistaking “popularity” with “monopoly.”
Microsoft already had a monopolistic position, at least OS wise, by the late 80s. MS had some ruthless, and indeed monopolistic, licensing terms with PC OEMs all through that decade. To the point that by the late 80s, PCs not running DOS were rarer than hen’s teeth.
A big deal of the success of Win 3.1 and 95 was due to the status of DOS as the (forced) defacto standard OS for the PC space.
http://news.bbc.co.uk/1/hi/business/600488.stm
This is why Windows should never be included in any list for greatest software – This anti-competitive behaviour of Microsoft’s is what lead them to dominate – Windows has always been a POS, make no mistake only reason why 7 was half way decent is because of the explosion of OSX in the west (developed countries) and Linux in the east / developing countries after the failures of rubbish like vista.
Its a common misconception that without Microsoft we wouldn’t have a computer in every home – this is complete baloney Microsoft would love you to believe this.
They had nothing to do with it Compaq reverse engineered the IBM Bios (I think? – someone did) Which meant generic oems started popping up all over the place – Microsoft was just the right bully at the right time to take advantage of it, otherwise it would have been the guy behind the Original DOS, before MS bought QDOS, or D.R. Dos or BeOS or OS/2 or FreeBSD/Unix and we would probably be in a place where computers would be seen as very reliable and the desktop market place would be full of healthy competition spurring on innovation.
Microsoft has done nothing but spend billions Marketing shite to the masses and holding back technology their contribution to Networking and the Internet was NetBUI FFS, an unroutable dog shit protocol. Yeah I can see how they made the list – Thom I’m sorry but your wrong here.
Edited 2014-05-04 09:09 UTC
Win7 is just Vista SE… if you have Vista with latest Service Pack and patches, there is practically no reason to switch to Win7.
(explosion of Linux? You’re imagining things, no web stats show it)
BeOS was too late (after they mocked about for half a decade with expensive proprietary BeBox and Macintoshes); OS/2 was about returning to IBM the control over the PC market …if you don’t like MS Windows, you would really hate IBM OS/2.
Quick clarification:
Your argument stands with OS/2, and BeOS. They wanted to be hardware agnostic and sell to OEMS, so consumers could buy computers from a variety of them. Microsoft’s anti competitive behavior did affect them to a huge degree. The biggest barrier to owning a computer in the win 3.1 days was the cost. Having multiple hardware vendors competing helped drive that cost down.
For Atari, Mac, Amiga… Well they only sold the OS on the hardware they made ( Ignoring the power mac clones for a second *). They were expensive, and had a lot of vendor lock in.
* I think the mac clones were a great idea, implemented too late.
But is Windows a great work of software though?
A lot of homes, at least in Europe, already had a home computer at, uhm, home. Granted it was probably mostly used for games but it was a computer in the home.
Having a great impact is, as others have noted, not the same as being a great work.
Does McDonalds serve great, or even good, food? Is it a great restaurant? Hell no. Has it had immense impact on the food and restaurant industry? Without a doubt.
I think the comparison with Android is the one thing you’ve got right here.
Saying that Windows put a computer in every home is about as accurate as saying that Android put a phone in every pocket.
Even worse; Windows was a success based on the illegal business practises for which Microsoft was repeatedly given huge fines for – only those fines were a fraction of the market the company ended up dominating.
Windows has always technically lagged behind one OS or another. It does not belong on this list.
I agree.
Windows was not a technical break through, and in all likely hood will disappear just like VMS.
It is rapidly disappearing now.
POSIX UNIX, and the ideas around UNIX will endure and command all Von Neumann machines for the foreseeable future.
The list should be about technical break throughs.
UNIX was built before Knuth finished all of his books, which makes it remarkable in the sense that a master work of computing science on Von Neumann architecture wasn’t finished yet.
This makes the accomplishment of the engineers all the more extraordinary.
By the time Windows, which is really rehashed MS-DOS 6.22 appeared, Gates and Co knew how to properly design a operating system, and they simply refused to do it.
Edited 2014-05-04 04:09 UTC
NT has nothing to do with DOS… (and is very technically competent)
To be honest, I would’ve left Windows out of there and brought the NLS ( https://en.wikipedia.org/wiki/NLS_%28computer_system%29 ) to the table instead. Windows didn’t really bring anything new to the scene, it merely collected many already-done ideas and designs to the same OS. The NLS, though, did much, much more, and the full list of computer-scene firsts is quite hefty:
* the computer mouse
* 2-dimensional display editing
* in-file object addressing, linking
* hypermedia
* outline processing
* flexible view control
* multiple windows
* cross-file editing
* integrated hypermedia email
* hypermedia publishing
* document version control
* shared-screen teleconferencing
* computer-aided meetings
* formatting directives
* context-sensitive help
* distributed client-server architecture
* uniform command syntax
* universal “user interface” front-end module
* multi-tool integration
* grammar-driven command language interpreter
* protocols for virtual terminals
* remote procedure call protocols
* compilable “Command Meta Language”
I mean, just look at the fucking list! That’s an insane amount of totally new inventions and designs and almost all of those are still in use to this day.
NLS is already on the list.
Well, you do say you picked Star over NLS in the list. I am saying you should’ve picked both and left Windows out of it. I don’t intend to start a fight about it, I am merely expressing my view that both NLS and Star were enormously important to software-technological progression and Windows was much less so.
One thing that Microsoft did and which was extremely important was help break IBM’s stranglehold on the PC-market, but I attribute that more to Microsoft’s marketing and business tactics than to their software itself.
Edited 2014-05-02 21:05 UTC
> ‘The great works of software’
> […] Windows: you cannot have a list of the greatest
> software of all time without Windows. You may not like it,
> you may even hate it, but the impact Windows has had on
> the computing world – and far, far beyond that – is immense.
The list was about the “great works”, not the ones that had more “impact”. One thing can be very used and, at the same time, not a “great work”.
It’s like:
‘The great restaurants’
[…] McDonalds: you cannot have a list of the greatest restaurants of all time without McDonalds. You may not like it, you may even hate it, but the impact McDonalds has had is immense.
That really demonstrates that one thing can be very used and not a “great work”. If it was otherwise, it would be a huge disservice to the real great works.
Edited 2014-05-02 21:16 UTC
To call that thing with the Golden Arches a Restaurant is an insult to almost every other establishment that uses that name on the planet.
The same goes for BK, KFC etc.
They are all in the same category as the Burger Van at a IndyCar meeting.
That does digress from the thread.
Windows copied lots from VMS. They got caught nicking stuff and had to pay (if my memory serves me right) about a Billion in damages. Then they hired Cutler and a few more from DEC. IMO there is no way Windows should be on the list. It is nothing more than a badly cobbled together POS. Sadly most of us have to use it on a daily basis but that latter reason is nowhere near good enough to get it included in this list.
But it’s pretty good.
Overall:
1) C
There is C and C-like then there is everything else.
2) UNIX
How many Unix clones are there? It’s simple, rugged concept that works really well.
3) Mac OS
Everyone’s first GUI. You’ll get indie points for naming obscure stuff, but to everyone else, this is the first GUI OS.
4) COBOL
It can’t be killed, even with fire. It’s 2014, and there is still a lot of COBOL code out there in production.
5) Internet Protocol suite
It’s how the world communicates.
Personal:
1) FreeBSD 4
My first real taste of a Unix-like operating system. I’d dabbled with Red Hat 5.2, but FreeBSD was the first one I really used.
2) BeOS
The little OS that is still makes modern operating systems look antiquated even though it’s over a decade old at this point.
3) AudioGalaxy
I downloaded a lot of live tracks with this, and it worked much better then other filesharing apps of it’s generation, or since really.
4) SSH
This one speaks for itself.
5) tmux
Screen was first, but tmux has vertical panes. This has really become an indispensable little tool. This plus ssh, and I am happy.
The golden goose of hackers, anti-virus and security tools vendors that will take generations to eradicate from our technology stack.
AmigaOS had multimedia and multitasking 10 years before MS Windows…
That list is not serious…
WWW is not a great work of software, it is a great work of networking.
Linux is the first massively collaborative software project in the world. That is why it makes the list.
You confuse WWW with the Internet. They’re two related but different works of engineering. Internet is the networking part. WWW is only concerned with how to easily and efficiently interlink documents over the Internet (or networks in general) based on client-server scheme using HTTP for client-server communication. You can have a complete WWW implementation (server and client) on a single physical computer, but you can not have ‘the Internet’ on it (‘Internet access’ is a different thing). Internet itself is short for internetwork; even that alone shows what the term refers to (a network of interconnected distributed autonomous networks).
The software behind the WWW is a giant pile of junk. What makes it popular is the tremendous viral networking phenomena behind it
TCP/IP is another great networking work. It is independent of the WWW and far predates it. TCP/IP is another viral network.
“As We May Think” (Vannevar Bush)
feasable.
As such I consider it great work.
pica
No, the WWW is software, at least implementations of the WWW are software. The WWW and the Internet are not the same thing, nowhere near the same thing.
The Internet is the networking, the WWW is the rendered pages you see before you, rendered by software. Even HTTP is just a protocol that runs on top of the network. HTTP and WWW is just stuff that runs on top the networking. The network was there before the WWW and will be there after it.
It depends how you define “great work”. The web may have had the biggest impact on the world than any other software, but the design and engineering of HTTP and HTML is an unmitigated disaster. Consider the humongous shitpile of code in the typical modern browser and how much computing power it requires to do basic tasks. The trainwreck continues to this day. The only reason it barely hobbles along is because TCP/IP was so brilliant and computers got exponentially faster. Now there’s a great work that’s missing from the list, TCP/IP.
Edited 2014-05-02 23:26 UTC
Come on guys….
VMS VMS VMS VMS VMS VMS VMS VMS VMS VMS
it broke huge amounts of ground in networking, clustering(including long distance clustering), distributed shared storage/filesystem, multiuser, distributed processing, distributed print queuing, distributed job queues, multitasking, shell scripting, integrated database, distributed lock manager, high levels of security, etc.. most large operating systems are STILL trying to get these things correctly and it was built right into VMS in the mid-80s, a full decade before Windows95 even got networking built in.
I am a unix/linux guy by my profession, but I cut my teeth on VMS in the early days and it was and still is an amazing system.
Edited 2014-05-02 23:35 UTC
VMS misses the list because even among nerds few people have heard of it. Like the Amiga, it was badly managed and badly sold, and there was no port to commodity hardware and/or open sourcing that could have saved it in time. Now VMS is scheduled for EOL next year and HP shows no signs of releasing source.
Not really.
The VAX/VMS combo was one of the reasons why DEC became at some point the second largest computer vendor after IBM.
Yeah, eventually minicomputers went the way of the dodo and so did DEC. But that is irrelevant of the fact that VMS had a very significant/seminal impact on the development of technologies such as networking, clustering, fault tolerance, etc.
Edited 2014-05-03 16:39 UTC
I would have picked zVM (CP/CMS), but I’m biased. You can’t ignore the first ever hypervisor long before the term was invented.
The software development world is still far from the live coding experience those systems offered.
I like that choice
pica
Windows is certainly one of the most successful pieces of software of all time. But it certainly does not belong in a list called “The great works of software”.
Windows didn’t bring anything new to the table, MacOS, AmigaOS and others already had a graphical user interface and Nextstep was much more advanced than Windows.
In my opinion, just because something is successful does not make it great. For me great software is well written software that brings something new to the table, like Visical, Lisp, Mathematica.
OK, let me review
Mathematica, IMHO an excellent choice
VisiCalc, the first spread sheet ever, IMHO also an excellent choice
NextStep, do not misunderstand me I consider it great, but I would prefer SGI’s IRIX to be listed instead. It had much more impact. Even Apple iOS uses SGI IRIX ideas like base the UI on a (Open)GL graphic engine.
pica
I have never used IRIX and I only briefly used NextStep after it was already irrelevant. I voted for NextStep because Objective-C and the NextStep Foundation and GUI classes were clearly ahead of their time and MacOS X and ios are based on NextStep.
But your points about IRIX and OpenGL are certainly valid.
I think there are a lof of operating systems that could be mentioned here: MacOS, NextStep, IRIX, AmigaOS and probably a lot of others I don’t know for bringing something new to the table.
Windows certainly was the most successful OS of its generation but to me it does not qualify as a great work of software.
* IMHO first scalable GUI
* IMHO first UNIX with GUI based administration tools
* (Open)GL
* XFS, the IMHO first journalling file system
* via “Jurassic Parc” great influence on film industies
pica
Nextstep features
* a MACH micro-kernel
* a Display PostScript based GUI
Mac OS up to 9.x features
* an application menu in a fixed place
Mac OS X features
* a MACH based hybrid kernel called XNU
* a Display PDF based GUI
* an application menu in a fixed place
As a result NextStep is the technical parent of OS X. OS up to 9.x is the GUI parent of OS X.
Yes, SUN now Oracle Solaris also features Display Postscript. And there are some NextStep GUI clones like GnuStep or AfterStep. But that is the only influence of NextStep I am actually aware of.
pica
EDIT typos
Edited 2014-05-03 12:06 UTC
AmigaOS doesn’t even have memory protection… classic MacOS was similarly technologically deficient.
In contrast, NT is very competent technologically.
In my contrarian view point, VisiCalc/spreadsheet should never have been invented. All they do is make white collar workers productive in creating information garbage that gets used with serious real life consequences.
Imagine if programmers worked like your typical Excel analyst MBA:
– All development would be done using live production databases with no backups.
– All code would be hidden from view as soon as it was written, making it impossible for anyone to catch mistakes later.
– Nothing would ever get tested, as long as the first result looked close to reasonable.
– No version control.
I could go on and on, but the point is that spreadsheets offer a great user experience at a horrible cost to usability (where producing correct results matters). Too bad there’s no hope for anything better replacing spreadsheets. A better system would make mistakes clearly visible and take away all fallacies of intelligence its users may have.
You always have the option of firing your spreadsheet monkeys and hiring real programmers, but nobody seems willing to do that yet.
Lisp is the proof code and data are isomorph.
Simula made object orientation feasible.
UML 2.x, please stay tuned …
pica
“There two kinds of people in this world. People who goes through the effort of making substantial things, and people who make lists of those things.”
– Alan Kay
So true…
Edited 2014-05-03 12:13 UTC
Yes, so true …
But in my view discussing on great systems communicates
* what others think
* what attributes & features are valued by others
* where others set the bar
* …
And as a result enables synergies that might result in new systems commonly considered great.
That aspect is my motivation to join that discussion. I can learn.
pica
PS Well yes, that diuscussion requires boring lists.
PPS Some make good statements, some cite good statements
Edited 2014-05-03 12:41 UTC
The actual quote goes like this:
“There two kinds of people in this world. People who goes through the effort of making substantial things, and people who make lists of those things. Oh shit, and of course people who then comment on how lame those lists are. They aren’t even worth mentioning so I stuck with two.”
Gratulation, I consider your statement on par with Arthur Schopenhauer’s essays.
pica
Edited 2014-05-03 13:21 UTC
I thought the quote was “There are 10 kinds of people in the world: those who understand binary, and those who don’t.”
visicalc as the first big spread sheet program deserves a spot. thewell or some other bbs from the dial up days should be included also. don’t forget cp/m.
Absolutely, Visicalc and Hypercard… I think those two are the most important and influential apps ever created.
A little disappointed by all the “TL;DR, let me vilify the original as stupid and give my own different list based on completely different criteria”.
From the article:
<blockquote>I propose a different kind of software canon: Not about specific moments in time, or about a specific product, but rather about works of technology that transcend the upgrade cycle, adapting to changing rhythms and new ideas, often over decades.</blockquote>
In case that is still too long for some, let me put it like this: this list is not about “without listed program A, we would not have programs B, C or D”.
Did A-0 System transcend the upgrade cycle? Did it adapt to changing rhythms and new ideas? Did it last decades? Did it last even one decade?
Did WorldWideWeb/CERN HTTPd transcend the upgrade cycle, adapting to changing rhythms and new ideas, often over decades? What version are we on today? Of course without WorldWideWeb/CERN HTTPd, we would not have Chromium, Firefox or Apache etc. But this list is not about inspiring other programs.
Did Xerox Star transcend the upgrade cycle, adapting to changing rhythms and new ideas, often over decades? Note: Not things inspired by Xerox Star: actually Xerox Star itself. How many people ever actually used Xerox Star itself? How many people still use Xerox Star itself?
The mouse: in a software canon??
Word Perfect and Lotus 1-2-3: Without them… the world has carried on regardless. Whereas MS Word is near-ubiquitous and has entered its fourth decade. (And if anyone points out that Corel is still selling Word Perfect, I ask “hands up all those who’ve used (let along bought) Word Perfect in the last 15 years?” Probably a similar number as those who’ve used a real teletype machine.)
Rather than deal with the actual list and the real criteria, people look at it quickly and go “but something I think should be on such a list isn’t, therefore the whole thing is wrong and stupid and poopy-faced”. But… such is the nature of lists like this.
Is this list only about pre-2007 great works of software? If not we must include iOS, and iOS like (Android)… After it everything changed in computing, isn’t it?
How long had it been around? Ok, may be not so long by now…
Did people directly interact with it every day? Yes, absolutely, much more than everything else
Did people use it to do something meaningful? Well, this is debatable
1) Windows – Bad: Reason, A monopoly. it kept at vibrant OS ecosystem from developing with OS/2, DR-Dos… among others. Caused the PC culture to become monolithic, stifle innovation and much more susceptible to mass virus outbreaks.
2) Office – Bad: Reason, Again a monopoly so we got almost no innovation for > 10 years.. still are not getting it.
3) TCP/IP – Good: Reason allowed for a robust open internet. Need IP6 soon!!!!! and government out!
4) UNIX – Nuff said.
5) iOS – Good: Reason, Freed people from the PC as the only real means of accessing a personal computer/network… (no it was not the first but the most user friendly and accepted) Really is the post PC era. I will never buy another PC (Mac) again. May not even buy another laptop…
putting Windows on that list equals you putting Apple and iOS on the list for greatest smartphone innovation of all time!