“These days, the ‘revolution’ is all about Linux. The word alone has become a catchcry for everything anti-establishment, anti-Bill, and anti-licensing fees. If you listen to the hype, it’s being used everywhere, in businesses of all sizes, to do everything but make the coffee. Just because everybody’s using Linux, however, doesn’t mean everybody’s happy for that fact to be known, as I found recently while looking for potential candidates for this special report about companies that had made the switch from Windows and Linux, and vice versa.”
Planning makes a big difference on the outcome of any project. It doesn’t seem Austereo made very wise decisions. Why did they not test the Blackberrys before commiting to them? There are PDA’s that work perfectly with linux. And although I don’t have any experience with CRM software, that surely can’t mean the whole company needs to use the app. And I can’t get email back after I delete it either and I use Exchange. So I don’t see why the big push to switch to MS. Also, I’d like to see those 11 people at Coffey fix their stuff the next time their systems get infested by the latest plague. Thats almost 160 machines per person to fix.
Why did they not test the Blackberrys before commiting to them?
And they didn’t do their research as there is open source software that will push to blackberries:
http://www.funambol.com/
Hey TechGeek i love to see the way you try to justify Linux by placing blame on the system managers. Love your attitude buddy, keep it up. You MUST be the most knowledgeable person since you so easily found that all Austereo employees are dumbass. Nice…
…isn’t really about Linux vs. Windows, but rather Planning vs. No Planning.
If you don’t have any stated requirements, no specifications, no idea what you want, and no idea what the chosen solution can give, then you ARE screwed – no matter the chosen solution.
The problem is that many zealous Linux advocates would have you think that switching to Linux will solve *all* of your problems right away.
That’s when the IT managers buy in, and switch over the whole environment without any deep planning.
I’ve never heard of such claims from “zealous Linux advocates”. Nor have I ever heard such claims from “zealous Windows advocates”.
However, that doesn’t mean such claims have never been brought forth, but if IT managers are stupid enough to believe such statements, then they are not suited to their positions. They are simply incompetent.
No matter what solution you choose, some planning is necessary. Not so much for the home user, where it matters little, but for corporations it’s very important. These IT-managers cannot have any qualifications if they don’t plan ahead.
The problem is that many zealous Linux advocates would have you think that switching to Linux will solve *all* of your problems right away.
I’ve never heard of such claims from “zealous Linux advocates”. Nor have I ever heard such claims from “zealous Windows advocates”.
However, that doesn’t mean such claims have never been brought forth, but if IT managers are stupid enough to believe such statements, then they are not suited to their positions. They are simply incompetent.
Well, I guess it’s a good thing that a lot of people who post on these forums and others (such as ZDNet) are not IT managers
I can see it now …
IT manager: “We MUST roll out open source everywhere NOW!!!’
CEO: “Why?”
IT manager: “Because Richard Stallman is GOD!!!!”
Edited 2006-04-24 22:10
The problem is that many zealous Linux advocates would have you think that switching to Linux will solve *all* of your problems right away.
That’s when the IT managers buy in, and switch over the whole environment without any deep planning.
IT managers that don’t plan are caused by zealots who want you to think [fill in the blank]? Above them or below them in the hierarchy? And who listens to managers who jump into untested plans proposed to them by over-enthusiastic fanbois? Does that mean that poor executives are also caused by the influence of these “other camp” cheerleaders because they listen to poor managers that have been subverted by unknowing or, worse yet, active cultic agents? Didn’t anyone learn anything from 70’s TV sci-fi? Have we all forgotten to “Ignore alien orders?”
The intersection of social interaction and business strategy seems to be much larger in your diagram of the world than mine. Let’s test and report back with results! Set your watchpoints …
Interesting stories…
… but these are just samples from a much bigger, and more complex, picture.
Jumping from one OS to another is not something that should be taken lightly and without due thought. Crossing a street without thinking can cause a LOT of problems, too.
The business goals must first be documented and reviewed for missing details. The small stuff can kill you, as Austereo found out. Find out what platform can meet your needs. Most likely, both Windows and Linux will work. In some cases you have a critical application or need that will require a particular direction. In cases where it seems that your choices are open, that is where the real work at planning comes in.
And, by the way, I could not find a Linux coffee maker, but I did find Linux powered ice cream (which is far more tasty than coffee, anyway)
http://www.linuxdevices.com/articles/AT9296154631.html
/end discussion
There really isn’t much more to say, so I sense a Windows v. Linux flamewar coming.
I’d rather want a coffee maker
I know there were coffee makers for the C64. Well, there were circuit designs for such systems, at the very least. You had to build them yourself though.
At least now I can tell every Linux fanboys to STFU when they tell me that Linux is the solution to all the problems. It is right tool for the right job and Windows excel when it comes to delivering a total integrated solution.
Linux has a place like earlier Unixes but not a Windows killer by any distance.
The reverse is also true.
Windows is not superior in any way, however Linux excels when it comes to delivering a Total Integrated Solution.
Why not just say:
X is not superior in any way, however Y excels when it comes to delivering a Total Integrated Solution.
I think what you should have said was
“I believe Windows is not superior in any way, however Linux excels when it comes to delivering a Total Integrated Solution. ” as this is an opinion and not fact. They both have thier uses, and they both excel at certain tasks. The best tool for the job is the tool to use.
Yeah, whatever
I just wanted to show how little value there was in the comment from CuriosityKills.
Kerberos, RADIUS, Certificate Servers, NIS, NFS, CIFS, CUPS, LPR, LDAP, Groupware, SMTP, DNS, IPSec, and PAM; How do I, as a small shop of 1000 employees, configure each one of these. If I am able to configure them all, how do I then proceed to integrate them all together? And once I do, how do I troubleshoot all these disparate systems? And if I have 5+ experts just taking care of these basic systems, who will take care of the network, IDS, firewall, SQL database, the CRM, the PBX, the ERP, the backups, desktops, applications, etc? Microsoft solution might not make sense for a Google, a large institution like a university, for banks, etc., but for the majority of businesses, it does something very well, it lets your basic infrastructure get out of your way and lets you focus on business, financial, and customer service applications.
Yes Microsoft makes it easy to ignore all those piddling little things. But when you let the wizards setup your system how do you its right? How many exploits are default? If you are the admin you need to know this stuff.
Well, there are wizards available for Linux as well.
RHEL contains many wizards, though you don’t have to use them.
“But when you let the wizards setup your system how do you its right?”
The same way you know whoever coded the Linux kernel, postfix, BIND, coded it right. The same way you know who ever put your distribution together put it together right. I can assure you, whoever puts these systems together, knows a lot more about the subject matter than you will ever know. I don’t understand how people can’t trust Microsoft putting together their OS but can somehow trust all the behind the scene voodoo of any operating system such as Linux. At least be consistent, and audit the entire system, all 30 million plus lines of code.
“How many exploits are default? If you are the admin you need to know this stuff.”
Again, how many exploits does Sendmail get? OpenLDAP? Linux? The fact is that when there are exploits, you are reliant on a company to provide patches and you use these patches to secure your system. Windows 2003 already has pretty good defaults. If you design your network infrastructure properly, these become darn good defaults. Finally, if you use the Microsoft Lockdown wizard, you get a system that becomes excellent.
Compare that to Unix/Linux. I’ve seen one too many Unix companies that use NIS. I’ve seen one too many Unix companies that use NFSv3 and lower. These protocols by default are not secure. You cannot do something to secure them. Many Linux companies do not take advantage of LDAP, of Kerberos, OTP, etc. because they are too difficult to integrate.
Edited 2006-04-24 22:29
>Compare that to Unix/Linux. I’ve seen one too many Unix
>companies that use NIS.
NIS is as dead as… dead.
> I’ve seen one too many Unix companies that use NFSv3
> and lower. These protocols by default are not secure.
But they widely support encryption. So why use the default? This is a bit like saying “I put the new lightbulb in… but dammit, it won’t light without this switch on the wall here!” Granted, not all NFS daemons support encryption themselves… most of those old.
>You cannot do something to secure them.
Cue grating game-show buzzer! Wrong! There are many things you can do. You just don’t know how.
To whit:
— cut
sec=mode
Set the security mode for NFS transactions. If
sec= is not specified, then the default action
is to use AUTH_SYS over NFS Version 2 mounts, or
to negotiate a mode over NFS Version 3 mounts.
NFS Version 3 mounts negotiate a security mode
when the server returns an array of security
modes. The client picks the first mode in the
array that is supported on the client. Only one
mode can be specified with the sec= option. See
nfssec(5) for the available mode options.
— cut
…and…
— cut
The mount_nfs(1M) and share_nfs(1M) commands each provide a way to specify the security mode to be used on an NFS file system through the sec=mode option. mode can be either sys, dh, krb5, krb5i, krb5p, or none. These security modes may also be added to the automount maps.
— cut
>Many Linux companies do not take advantage of LDAP, of
>Kerberos, OTP, etc. because they are too difficult to
>integrate.
So pick a distribution that does. I.E. RedHat Enterprise. Besides, why should they automatically go right to LDAP when most systems won’t be using it?
Kerberos for NFSv3 is has been hacked in as an afterthought (Secure NFSv3.) So what you have in the marketplace are clients and servers that might not necesarily support Kerberos. If you look at the marketplace, many implementations for NFSv3 and less do not take advantage of Kerberos (standards are important to some people.) The first version of NFS that has Kerberos support in its specification is NFSv4. So while I did not realize someone actually took the time to hack in Kerberos into NFSv3, I stand by my original assesment that NFSv4 is the only secure version of NFS.
>The first version of NFS that has Kerberos support in
>its specification is NFSv4. So while I did not realize
>someone actually took the time to hack in Kerberos into
>NFSv3, I stand by my original assesment that NFSv4 is
>the only secure version of NFS.
You talk a good game, but it doesn’t wash. Those man pages are from Solaris version 9, which is… what… at least five years old now? Older than XP for sure, and lots older than W2003 Server.
Please.
Edited 2006-04-25 07:59
Kerberos, RADIUS, Certificate Servers, NIS, NFS, CIFS, CUPS, LPR, LDAP, Groupware, SMTP, DNS, IPSec, and PAM; How do I, as a small shop of 1000 employees, configure each one of these
With one lone engineer. I have most of that automated on Linux, and can tailor and integrate it for a business in one week, testing included.
And I’m not the brightest admin/architect out there.
And for 1000 employees, more than 1 engineer sure enough won’t kill you.
Actually, do like they’ve done in the article for Windows, hire some Linux consultant to do it, and support it.
If I am able to configure them all, how do I then proceed to integrate them all together? And once I do, how do I troubleshoot all these disparate systems? And if I have 5+ experts just taking care of these basic systems, who will take care of the network, IDS, firewall, SQL database, the CRM, the PBX, the ERP, the backups, desktops, applications, etc?
BS 1 : you don’t need 5+ experts to do the things you described in the first line : you need 1. I’m not an expert, I’m a specialist, and I can do it alone. An expert will do it fingers in the nose. But doing all that in one week is unwise, even if possible. Now, to take care of the IDS, firewall, etc., of course you need some other people.
BS 2 : you complain about being a small shop and wrongly needing 5+ experts, and you want to have ERP and CRM ? Go back to planning board please : this is nonsense, you’re doomed to fail.
Microsoft solution might not make sense for a Google, a large institution like a university, for banks, etc., but for the majority of businesses, it does something very well, it lets your basic infrastructure get out of your way and lets you focus on business, financial, and customer service applications
That’s wishful thinking. Like hinted in the article, MS solutions need as much support, if not more, than anything.
Where did this BS entered your head ? The security of MS OS alone prevents you from only focusing on business, financial, and customer service applications.
Even MS can’t rely solely on their software, you know !
And I’m not talking about the rest : supervision, backups, …
“With one lone engineer. I have most of that automated on Linux, and can tailor and integrate it for a business in one week, testing included.
And I’m not the brightest admin/architect out there.
And for 1000 employees, more than 1 engineer sure enough won’t kill you.
Actually, do like they’ve done in the article for Windows, hire some Linux consultant to do it, and support it. ”
But how many have you actually mastered? Are you fully utilizing their entire functionality or just using the bare minimal functionality? If you aren’t fully utilizing them, what is wrong with just using the Windows solution?
“BS 1 : you don’t need 5+ experts to do the things you described in the first line : you need 1. I’m not an expert, I’m a specialist, and I can do it alone. An expert will do it fingers in the nose. But doing all that in one week is unwise, even if possible. Now, to take care of the IDS, firewall, etc., of course you need some other people.
BS 2 : you complain about being a small shop and wrongly needing 5+ experts, and you want to have ERP and CRM ? Go back to planning board please : this is nonsense, you’re doomed to fail. ”
A small shop of 1000 people will usually have one system administrator, one help desk person, maybe one network administrator, and about 5-10 people taking care of all the business applications. Optimizing DNS performance, configuring a mailer that can send 10 million emails a day, or trying to figure out how to integrate the LDAP server from their two sites (one in New York, the other in San Francisco) and get their open source groupware solution to work is not in their interest. They need a basic infrastructure to do only what the business needs and then get out of the way.
The problem with this discussion and indeed this way of thinking is that Linux is portrayed as a monolithic, single-application system.
For example, remote users struggled to grapple with a virtual private network (VPN) login system that required three different passwords to establish a connection.
This is an example. What VPN software were they using? I’ve been using OpenVPN with Linux for maybe two years and I have had no such problem. So the problem may have been that they selected the wrong software, not the wrong OS.
>Microsoft solution might not make sense for a
>Google, a large institution like a university, for
>banks, etc., but for the majority of businesses, it
>does something very well, it lets your basic
>infrastructure get out of your way and lets you
>focus on business, financial, and customer service
>applications.
You’re perpetuating the Big Lie: no matter that you can click-n-drool through Windows wizards, they only expose you to the simple configurations. “Basic infrastructure” usually means a simple SMB server or a SMB peering network, and not much more.
To approach the kind of functionality that you describe in Windows, you can’t be a clueless newbie, you will need to know your stuff. When you are throwing in certificate servers, RADIUS, Kerberos, LDAP, Groupware, SMTP, etc. etc. then you will *NOT* be a candidate for simple point-n-drool wizards to set up your Windows network. You’ll be on par in terms of OS/application familiarity with your average *nix admin to implement all of that.
Here’s an interesting snippet from the article:
After four months, Microsoft Active Directory and Exchange Server 2003 had replaced now-discontinued Linux servers to provide a consistently managed, centralised messaging infrastructure across 20 Coffey offices. “Previously, all the e-mails were effectively stored on the desktops and there was no central location of the data,” Parsons explains. “That’s a nightmare both because of litigation, and because of duplication across the company and all the problems that duplication brings.”
It’s called IMAP. If they are using POP to deliver mail to desktops in a large business, then that is their own stupid decision. I’m quite confident that Exchange was a better solution than this, but that doesn’t mean that their previous solution was the best a Linux based infrastructure can provide.