canopic jug writes:
Bryan Lunduke at Network World calls out what other mainstream media have been too timid, or bought out, to call out. He starts by pointing out that choosing Microsoft Windows for your organization should get you fired and that if you haven't already replaced Windows, across the board, you absolutely stink at your job.
There. Finally the topic is broached in mainstream media and a proper discussion can now start among decision makers who can arrange complete migrations to GNU/Linux, Chrome/Linux, one of the BSDs, or a combination of them.
As Microsoft security problems continue to escalate since even the pre-networked, MS-DOS days, managers and front-line grunts will find themselves increasingly culpable for selecting unviable software, such as Microsoft Windows. If they wish to pay big bucks for maintenance, there are plenty of companies around to participate in the money. Canonical, Red Hat, M:Tier are just a sampling.
[Ed. Note: I debated whether or not to run this story — in some respects it's just the Windows vs *nix argument all over again. Also, there are proprietary programs which are critical for certain industries which currently only run on Windows. On the other hand, gaining a mention like this in the more mainstream media, does that mean we are approaching an inflection point? Witness the increased displeasure with Windows 10's telemetry and the difficulty in completely blocking it. What programs do you use that are only available on Windows? What keeps you from moving to another OS? --martyb]
in some respects it's just the Windows vs *nix argument all over again.
There was an argument?
I thought the consensus was, Windows is used in businesses only for two reasons: 1) Windows-only critical software, and 2) suits prefer to buy software with support from a "real company" like Microsoft, since free software makes them uncomfortable even if you pay for support a la RedHat.
I have yet to meet someone who uses Windows because they like it; it's either lack of knowledge about alternatives, laziness to switch (<- that's me), or usage of Windows-only software (well, usually games).
The lack of business software on BSD/Linux is something that may put up obstacles.
Maybe Wine or VMs can mitigate it.
I don't know, Libre Office is a pretty solid program. Write is BETTER than Word, and Calc is almost as good as Excel.
And Powerpoint is a stupid waste of time and resources. Nobody really needs that piece of shit (they used it a lot where I worked).
MS Office is far, far more advanced than anything the open source world has. Well, advanced in terms of end-user features and usability (compared to LaTex being more advanced technically). Sure, LibreOffice is the same when considering student-level usage but it sucks at business scale usage. Where's the document encryption and security controls? Change traceability? 3rd party plug-in integration? Sharepoint support? Advanced graphing capabilities? Seamless embedded documents? Etc... Libre has similar features, but none of them are as good. Even the grammar and spelling support isn't as good let alone managing complex styles or imported data.
And you're ignoring the two biggest: Access and Outlook. Despite Access' shittyness, it's still best for non-programs to whip up their custom tools in the least amount of time. Forget about code quality, business users want tools which let them get their job done better and faster. Software always plays a support role, it is never an end onto itself.
Outlook/Exchange has nothing close to it. It is way more than just an email client. Encryption, email polls, secretary features (multiple people accessing the same account as one), calendars integrated with booking rooms and resources (you can buy wireless door displays which say who has booked that room and when), recalling emails the receiver hasn't viewed yet, seeing out of office messages before sending an email, etc...
And that's all ignoring the support community, tons of books, tons of tutorials, etc... of Microsoft products. If you want to replace Office, you need to be better on all those points and be better enough that it's worth the risk of changing. I don't like Microsoft, but despite their security concerns, the make very produce tools once you learn how to use them. The open source world just says "they suck" and never bothers to learn the features which makes putting up with the front facing suckiness worth it.
When Windows becomes seen as career suicide, all that software will quickly be ported to Linux (and possibly BSD).
Catch is that the ecosystem has to reach that point. Perhaps more M$-malware can fix that..
Yes there was. For about two decades (now almost three, if you still count the argument as not yet settled), since the very first day a Windows salesman first conned a prospect into taking Microsoft's toy operating system seriously.
...and 2) suits prefer to buy software with support from a "real company" like Microsoft, since free software makes them uncomfortable even if you pay for support a la RedHat.
For most of that time, *nix in business was something that came from major players like DEC, Sun, HP or IBM, often on support contracts with prices that make Microsoft's look cheap, but most of the time actually delivered real support.
And even leaving that aside, in the modern world there are plenty of independent companies around (not just OS vendors) who will gladly sell you robust support contracts for FOSS Unix or Unix-like systems such as *BSD or GNU/Linux. Most companies are too miserly to pay for real support. But those who do often get better support than they would from large software vendors at any price (and always better than what, if anything, they'd get from large software vendors at the same price).
That's an obsolete argument that, with respect to FOSS OSes, ceased to have any merit at around the turn of the century, and with respect to the various closed Unix flavours was never true back then (but might be now following more recent rounds of mergers & acquisitions...).
As for "Windows-only critical software", that's pretty much a contradiction in terms. If it's really critical, you audit it before deploying. If you can't audit the platform, you can't audit anything that runs on it either. If it doesn't warrant an audit, then it's not "critical software".
I agree with TFA and would go further -- if someone deploys unauditable software in a setting where a software error might cause loss of life, not only should that person lose his job, but he should also be charged with negligent manslaughter (if the risk eventuates), or at the very least reckless endangerment (if it doesn't).
For about two decades (now almost three, if you still count the argument as not yet settled), since the very first day a Windows salesman first conned a prospect into taking Microsoft's toy operating system seriously.
I'm grateful to that one salesman for one thing and one thing only: it helped made from IBM-PC an open (hardware) standard [wikipedia.org] thus an OEM friendly one, therefore cheap and affordable.
Without it, I doubt Mac or HP or DEC would have lowered the prices enough on their stuff. Without MS, perhaps other companies would have popped on the market to fill the same niche (eg Wang labs?), but I see nothing to indicate they'd have been more ethical.
Here's the IBM "portable computer" 5 years before PC [wikipedia.org] - introductory price of $9000 in the mid '70-ies - you'd need a small truck to "port" in between places.
There were other systems like Commodore Amiga of CPM.
The reasons why PC won domestic market are debatable, but I don't think Microsoft created the domestic market. Nevertheless, I will never deny y that Microsoft played a main role. On the other hand, it later destroyed competence and abused of users.
"I don't think Microsoft created the domestic market"
Debatable. There were all those others, creating and/or finding niche markets. Microsoft took advantage of a business relationship with IBM, if you'll recall, to get their Windows OS going. And, at the same time, mercilessly crushing most of the opposition. There came a point when Microsoft was calling all the shots with the OEM's. "You'll sell our product exclusively, or you won't sell our product at all!"
Microsoft probably had as much to do with creating the market for PC's as all other players combined.
Of course, their methods were despicable. Digital Research is still a sore spot, after all these years. MS knew that they couldn't their Windows very efficeintly without 32 bit disk access. That disk access was holding up the release of Windows 95. DR succeeded where everyone else faile - they got 32 bit disk access. And, MS leaned on DR, created a bit of code to inform Windows that it couldn't run on anything that wasn't Microsoft DOS, and the rest is history.
DYSEAC [wikipedia.org] predates the IBM-PC with 27 years (1954). And that really needed a truck. But it was portable ;-)
What got x86 IBM-PC going was that IBM had a solid standing with technologically clueless executive people. So IBM got the x86-ISA-IBM-PC sold. And Microsoft enhanced DOS so it could ride along. So executives decided to buy PC+Windows and once that happened they wanted to exchange files.. *slam-lockin*.
The problem stems from incompetent people in control of resources.
As a developer I prefer the win32 api over 5000+ individual libs.
As an end user, I hate every time a new program says I need a different .Net or VBA version installed. Programs should be fully self-contained, which is exactly what the original definition of object oriented meant. It meant if you exchanged a document, the program to view and edit the document were embedded within the document itself. It is one self contained object. Anyone who had a file had the software to view and edit that file. Sadly people applied OO at the code level rather than at the application level.
Phone apps and security concerns are slowly, slowly pushing us back towards that definition, but sadly it's more for vendor lock-in than end-user usability.