4 ms·
Coherent made a fatal choice at the wrong time. Mark Williams Company started to ramp up their marketing right before Linux really started to his. When they had
by cowmix 6y ago
Coherent made a fatal choice at the wrong time. Mark Williams Company started to ramp up their marketing right before Linux really started to his. When they had to make a choice to add either XWindows or a TCP/IP stack to Coherent, they chose XWindows. Coherent could have been a great Internet OS at the dawn of the commercialization of the Internet but they blew it.
- derekp7 6y agoA buddy at work used to work for MWC, and always talks about what might have been. This, among other stories...
- zerr 6y agoIsn't Linux being free (including as in beer) one of the main reasons of Coherent fail?
- arexxbifs 6y agoIn the mid-90s, downloading Linux wasn't feasible for most people and getting a manual with the install media was a plus because you couldn't just google for help the same way you can today. But I guess Coherent could have been competitive for no more than a few years, at best.
- anthk 6y agoIn the mid-90's you have Slackware 96.
- xbar 6y agoAnd books like Linux Unleashed with Slackware CDs included.
- icedchai 6y agoI remember going to a "Linux install fest" at a local ISP. We downloaded all of the SLS Linux floppies over a 56K leased line! This was late 1993 or early 1994, I think.
- guenthert 6y agoYou didn't have to download it yourself. There were vendors which offered media with it for a small fee (remember RMS was Ok with people charging a copying/service fee for GPL'ed software, as long as access to the sources was provided). Yggdrasil was one I remember.
- pjmlp 6y agoFor me Linux was only a thing, because it was the only way I could do the university homework at home, while the computer center was running DG/UX and Solaris. Red-Hat Linux was eventually adopted when both of them died, until then it was available in a couple of desktops that would dual boot alongside Windows 95. Had Windows NT offered a solid POSIX story, that would have been what the labs would have switched to instead.
- rjsw 6y agoSeems like the wrong choice to me too. I was running System V/386 at around that time, it wasn't very difficult to port X11 to a new platform but hard to add a network stack to a binary only operating system.
- guenthert 6y agoVery early Linux didn't have TCP/IP support either. People used (for a short time) KA9Q (http://www.ka9q.net/code/ka9qnos/ http://www.ka9q.net/code/ka9qnos/), which was also available for other OS. Anyone remember (the now glorified) Bill Gates claiming the Internet would just be a fad and announcing that Windows '95 would come w/o support for TCP/IP?
- justin66 6y agoIf they'd chosen to skip X and create a TCP/IP stack, in the early-to-mid nineties Unix battles they would have had to compete with Xenix, BSDI and the other BSDs, and Solaris on x86. Not to mention Linux, the non-x86 options, and whatever else I forgot. In hindsight, it's not like there was some kind of very powerful move they could have made.
- cstross 6y agoPer memory, Coherent 4.0 showed up around 1992-3, with TCP/IP and pseudoTTYs. X11 came along a little later. But by then it was too little too late: the Linux distros were already eating their lunch. Same with SCO. When I worked there circa 1991-95, the company lost $250 in license fee payments to assorted other companies for every complete copy of Open Desktop that they sold -- fees for the TCP/IP stack, the compiler, the AT&T license, the Motif skin, CDE, the IXI desktop ... even purchasing IXI didn't reduce it by much. They tried to go head-to-head with Windows NT as a workstation OS on the desktop and nearly broke the company, then backed off to focus on enterprise sales and support contracts. There was no way they could compete with a solid Linux as a desktop option after the late 1990s. And SCO was a much larger and more powerful company than Mark Williams (by 1991 they had about 1200 employees worldwide and $200M/year revenue -- going by memory).
- justin66 6y agoThat is wild. I didn't know SCO was ever losing money in quite that way. I worked tech support for an early PC point of sale credit card payment processing company and Linux was actually one of the last *nixes we supported. I don't believe they had even rolled it out in 1998 when I quit. We supported DOS, Windows, SCO, and maybe one other commercial Unix with just a few users. It was excruciatingly obvious to the more knowledgeable rank and file where things were going, and I even wrote an internal document once explaining how to run the SCO version of our software on Linux using its iBCS support. The internal discussion I remember about which operating system we would support next as online payment processing took off was mostly about BSDI vs Linux (but with our upper management focused very strongly on NT). Perhaps oddly, both options were treated seriously.