Are Widespread 'Microsoft-alike' Replacements Feasible?

Dr.Dubious DDQ asks: "With all the recent Microsoft(r) news, I see a lot of the usual complaining about Microsoft's unfair 'embrace and extend' practices. I do my own fair share of this, but I'd much rather actually *do* something about it.At the risk of prompting cries of 'No! That will only make them stronger!', I find myself asking: How possible is it to 'transparently' replace Microsoft-brand services with other (preferably, but not necessarily, Open Source) services (rather than flatly demanding migration away from all things MS)? Or put the other way around, what tweaks would have to be made to existing, standard services to make them 'bug-for-bug compatible' with MS versions, particularly OUTSIDE of the context of SMB/Samba, which is an already-obvious example?" While there are definite reasons why such an effort may be worthwhile, it is also possible that Microsoft could attempt to make legal attacks at such projects...even though they are designed with software interoperability in mind. Precedents in support of this idea do exist, such as: ReactOS and even standard Open Source openings like Gnumeric. "I've got two goals in mind here:

1. Ability to placate MS-platform applications that demand MS-brand services to connect to while ALSO allowing non-MS clients as close to 'full' functionality as possible with the same services
2. Naturally, ability to replace an MS-branded package would personally appeal to me as well for both technical and - yes, I'll admit it - philosophical reasons.

Ways of meeting either (or both) goal would be useful to me and, I suspect, a lot of other sysadmins.

For example:

• Is it possible (and feasible) to get OpenLDAP+Kerberos5 to fool Windows systems into believing they're talking to a "real" ActiveDirectory(r) server (without necessarily also having the entire Samba stack)?
• Can client programs that demand MS-SQL server generally use MySQL in MS SQL Compatibility mode instead, if MySQL is set to respond on the MS-SQL port (either directly or via ODBC?)
• How hard would it be to make a 'mod_dav_sharepoint type of module that spoofs Microsoft's special Sharepoint WebDAV behavior (which evidently also uses a 'special' non-standard SQL-like search mechanism - am I going to be kicked out of the club for thinking this looks, at least on the surface, like it might be a useful feature if usable by non-MS clients and implementable by non-MS servers)?
• Similarly, how feasible would it be to get non-MS DAV clients to be able to use Microsoft Sharepoint (or the hypothetical MS-alike drop-in replacement?)
• How good are the 'drop-in replacements' for MS Exchange?
• Are there issues with MS's implementation of IPP (are there any problems dropping Microsoft Printer Sharing entirely and using CUPS instead? It SEEMS that MS Windows 2000+ should support IPP directly, without resorting to Samba middleware - is this true?)
• Possibly risking heaps of derision for suggesting such an unlikely-sounding thing, but how about using mod_dav/Apache (as what Microsoft USED to refer to as 'Web Folders') as a replacement for SMB file sharing? Aside from possible performance issues, is this feasible, or are there too many incompatibilities in MS's DAV support for it to work?
• Are there any registry hacks or other tweaks that can be applied to Microsoft Windows-based systems to make them behave in more standards-compliant ways?
• ...etc?...

Are there other replacements people have investigated?"

Comments

It's a moving target.

[Feztaa]

Immitating microsoft is mostly pointless because it's a moving target; you'll always be playing catch up, and as soon as you've perfectly copied them, they'll release a new version that breaks everything and then you've got more copying to do. It's really wasted effort and I hate to see good programming talent used for such futile efforts.

Just make the existing stuff better, don't worry about immitating microsoft, be innovative in your own right, and slowly migrate all your stuff away from windows.

BeOS tried and got squashed

[Gary Destruction]

BeOS was an awesome OS, but Microsoft wasn't so eager to welcome it. BeOS is the victim of what other OS's suffer from and that's OEM licensing. Microsoft classifies it as a "trade secret". But in reality, it says that if you want Windows pre-installed, you can't provide the option to boot to a non-Microsoft partition. Scott Hacker wrote an article that details this licensing and it's impact on BeOS. He who controls the boot loader

The problem with immitating Microsoft is it can result in a Microsoft-oriented world without Microsoft. Look at Mozilla's XPI installer. It's not much safer than ActiveX because the amount of power it has. And no one wants to sign XPI packages. Even if they did, signatures could be spoofed. The more you try to immitate Microsoft, the more likely you are to make the same mistakes they've made.