OSNews

One key point I want to throw out, (it will apply most to #14):
Windows to Linux is oranges to apple-seeds. We need to either compare the NT Kernel to Linux, Win32 to GNU/Linux, or compare Windows XP to Ubuntu, Redhat, Suse, etc. When comparing to Windows, distros are so unique and varied that you cannot just refer to them as "Linux". There is no Linux "OS" in the Windows sense of the words "OS". And we should not *expect* Ubuntu, SuSE, and Redhat to be any more similar than Apple and Windows, as they are different companies. It is purely well-planned convenience that there *is* software that can be used on all Linux distros.

13) can I trust open source drivers for my critical business work? even seldom corruption or process deadlock could cause big problems. And when it's time to upgrade hardware will the O.S. drivers be ready, even after they've had to reverse engineer the spec?

Can you trust closed source drivers? What if you are a US government contractor or big US company oft targetted for corporate espionage? Can you run those closed source Taiwanese motherboard drivers?

The driver readiness for your hardware upgrade has little to do with the OS. Microsoft does not write the drivers in the Windows-space; the vendor does. More and more vendors are releasing Linux drivers with their new hardware, and the FOSS community writes the rest, with much larger variety. Try finding drivers to make Windows run on your Sparc or PowerPC hardware!

14) where the heck is X installed? well does the dist use GNU /usr/local/... or the LSB standard, or more of a debian deriv or slack or bsd?

Why does it matter? We are using software for either Windows, or Ubuntu, or Redhat, or ..., and therefore it is packaged appropriately for that environment. It doesn't matter to me, the normal user, as the packaging takes care of it.

As a power-user, it might matter, as I may want to step outside the box, but as a power-user I know enough to control that - even if it is Windows freeware that wants to install to "c:", or c:[company name][software].


15) boot problems but don't want to reinstal whole system? win = /fixmbr, lin = (lilo, grub, hda0, hd0, /boot, dd if=/bkup/hd0_boot_part.img of=... bs=...)
How do you run fixmbr if you can't boot? You must be using a very limited, command-line only, DOS based, user-unfriendly Windows boot disk. Or you are using the ASCII graphics based Windows Install CD Rescue mode. I'd much rather use the GUI repair modes of your distros LiveCD/Install CD.

I will admit, users do not want to change from what they know. But if Linux distros simply become what is already known, we will lose all of the possible innovation. It will take time, but if some Linux flavor is truly the best, it will be accepted. But mark this - Linux will not gain acceptance by simply mimicking what already exists. Why change at all, if it will just be more of the same? Developers - don't waste your time striving for compatibility with product X (e.g., Exchange) - strive for compatibility with innovative standard Y (e.g., CalDAV), supported by N other products. As these products get better and more accepted, those that are not compatibile will fade away.

<delete duplicate>

Edited 2006-07-23 03:43

One key point I want to throw out, (it will apply most to #14):
Windows to Linux is oranges to apple-seeds. We need to either compare the NT Kernel to Linux, Win32 to GNU/Linux, or compare Windows XP to Ubuntu, Redhat, Suse, etc. When comparing to Windows, distros are so unique and varied that you cannot just refer to them as "Linux". There is no Linux "OS" in the Windows sense of the words "OS". And we should not *expect* Ubuntu, SuSE, and Redhat to be any more similar than Apple and Windows, as they are different companies. It is purely well-planned convenience that there *is* software that can be used on all Linux distros.

You are correct about that; lets take it even further; what about distros like gnusolaris hence I can understand in regards to the reason why it should be called GNU/Linux - because GNU userland can be transplanted into any environment; or in the case of gnusolaris, IIRC, they use the gnu compilers, Debian packaging system and other opensource components.

Regarding the 'planned' part; one could compare the distros to Apple; and how Apple bundle all that you need in the box; you get a system, iLife etc. In the case of GNU/Linux, you get the operating system and all the stuff you need to get on with work.

Can you trust closed source drivers? What if you are a US government contractor or big US company oft targetted for corporate espionage? Can you run those closed source Taiwanese motherboard drivers?

Now how about how ISPs worked with the US government to tap peoples emails and correspondance; considering how compliant the US public is, when ther is the 'if you don't comply and worship the government, you're obvious a terrorist and anti-American' theme in the air (and been like that since the formation of the US), I wouldn't be surprised if the chipset manufacturers don't start adding 'features' in accordance to 'government directives' for 'counter terrorism and protecting the democracy'.

The driver readiness for your hardware upgrade has little to do with the OS. Microsoft does not write the drivers in the Windows-space; the vendor does. More and more vendors are releasing Linux drivers with their new hardware, and the FOSS community writes the rest, with much larger variety. Try finding drivers to make Windows run on your Sparc or PowerPC hardware!

What I hope is this; Intel and AMD purchase Ati and Nvidia respectively, they open up all the specs and allow prices, supply and demand, and volume to sort out things. Screw secrecy when you have the production facilities to provide the whole kit to the OEM vendors; video card, chipset, processor etc. all under the same roof; they'll compete supplying the whole thing, not just the graphics card, thus, if they do opensource their drivers, they won't actually lose any competitive edge.

inux will not gain acceptance by simply mimicking what already exists. Why change at all, if it will just be more of the same? Developers - don't waste your time striving for compatibility with product X (e.g., Exchange) - strive for compatibility with innovative standard Y (e.g., CalDAV), supported by N other products. As these products get better and more accepted, those that are not compatibile will fade away.

True; create a product, end to end; server and client, which utilises open standards protocols and provides the same level of functionality as the commercial products.

These exist NOW; Sun sells the Solaris Enterprise System, which includes EVERYTHING which is included with Exchange; you can use Outlook with it, or Evolution - and yet, why the hell don't people use it? I mean, I'd love to know why a company chooses expensive proprietary software from Lotus or Microsoft when Sun can provide the SAME level of funcationality at a LOWER price.

Same goes for the Kolab project as well; why are these pushed off by the high flying CIO's - the fact that there aren't multibillion dollar marketing budgets behind them? the fact that they don't include cheap gimicy crap; the fact that they don't have smooth talking, silver toungued sales represenatives with ethics so low, they rival Karl Rove in in the "who can be the most unethical".

Edited 2006-07-23 03:47

I mostly agree with your replies. But I still think it's easier for a non-comp-savvy user to get this with a default windows instal, no? Keeping that configuration however... like you said, "...one of the constant IT troubles".

You have to agree because they are all true, as are these ones.

13) can I trust open source drivers for my critical business work? even seldom corruption or process deadlock could cause big problems. And when it's time to upgrade hardware will the O.S. drivers be ready, even after they've had to reverse engineer the spec?

I'm speaking from my viewpoint only, not trying to represent major point of view

If you give HCL a spin. Then carefully decide on HW. Yes, OSS drivers are more reliable. Problem in the closed source drivers is that you can't realy do much about the state they are in.

14) where the heck is X installed? well does the dist use GNU /usr/local/... or the LSB standard, or more of a debian deriv or slack or bsd?

Now, this comment is simply stupid.

I don't need to know where X is installed. As long as it just works, why would I?

If you meant any piece of software instead of X, then answer is simple. Why would I need to know where it is? As long as it is in any point which is defined by PATH you can access it from anywhere. As long as I can access this in menu, why would I bother. As long as clicking associated file calls that software, why would I care?

btw. Windows softwares are installed in much more messier fashion. Distro at least follows its way of doing, while people on windows constantly reinvent the wheel.

15) boot problems but don't want to reinstal whole system? win = /fixmbr, lin = (lilo, grub, hda0, hd0, /boot, dd if=/bkup/hd0_boot_part.img of=... bs=...)

fixmbr will fix something? Either I don't know anything about computing or the fact that you're pulling that last and the most insane remarks.

While the solution with dd would fix your problem in any case. That is if you copy mbr.

Fixing one at the cost of others is not a solution or a fix. And even that it would be fixing the case when you want to get rid of lilo or grub, name some other case. fixmbr won't help you when your system goes into bluscreen during boot (even in safe mode).

I'm with ya though... I totally 100% grew up on unix. I breathe unix. But I gotta admit, the average "PC Enthusiast" can fix my moms computer. The average "Linux Enthusiast" has 2+ years of CompSci and can admin or find the documentation they need for their own system.

No, you aren't. We are completely different league. If you try to represent a contra point at least show a decency and present non-stupid and valid remarks. You putting your self into the same league as me, well... I even kinda feel insulted.

There are a lot of reasons why Windows is primary desktop. But, sorry... you didn't name one. All you've said so far is that you don't know what you talk about.

Why Windows still rules the majority of desktops?
- Comming preinstalled with machines
- Applications (and saying outlook or exchange client is just insane, it is more than that) people are used to, and application which simply don't exist on llinux
- People are used to it
- People are using it at their workplace
- games
- loads of crap downloadable from anywhere, and people just thrive to do that, no matter that this is the most stupid thing they could do with their computer
- linux not yet fully representing viable solution for their workplace

For now the state of enthusiasts is like like you said, but you don't seem to be smart enough to resolve why. Linux is mostly used in higher purpose (aka. servers), desktop purpose is not spread enough to pose the need for typical enthusiasts to be needed. It might even get to that point, but you can believe that you'll (or any enthusiast) have all the tools you need by hand at that time. Most of the current efforts goes into servers and the first mistake of the computing world is providing servers where enthusiasts can thrive. At the first real problem, they will still be just enthusiasts and nothing more.

shane

I'm glad you are

Edited 2006-07-23 19:22

While I appreciate your comments, can we agree to share some professional respect? It's not my intention to insult you -- whatever league you're in. I'm well schooled, fairly well experienced, and a darn good Linux S.E. There will always be others with better insite and more experience. You may be one. By "I'm with ya", I meant, "I agree with you". No insult intended. The points I made were to bring some ideas to the table regarding why linux isn't mainstream... and I think most of them are valid considerations or hurdles for a predominantly Windows IT world. Yes, there are solutions or alternatives, and yes, some of them shouldn't be show-stoppers for Linux platform integration or even migration. I'm not standing behind these with a pitchfork as arguments against GNU/Linux platform. Did it really come across that way?

Regards,
shane