Slashdot Mirror

← Back to Stories (view on slashdot.org)

Are Widespread 'Microsoft-alike' Replacements Feasible?

Posted by Cliff on from the duplicating-the-entire-line-for-'compatibility' dept.

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?"

3 of 37 comments (clear)

  1. Re:obvious? by volteface · · Score: 2, Informative

    As long as you know the proper protocol (LDAP for example), making replacement client and/or server apps is only a matter of writing the implementation code. If it's a server, you have to make sure it duplicates functionality exactly, and that's where you could run into problems. It's basically like writing a new IRC client or web browser. I don't know if MS makes all its protocols public though, and, again, that's where you could run into trouble getting started.

    I'm not sure about the legality of it though. That is something that would need to be researched before any serious undertaking is made.

  2. Database choice by Saanvik · · Score: 3, Informative

    While MySQL is an okay database, it's not a drop-in replacement for MS*SQL.

    If you want to go for a closer one, try PostgreSQL. It's much more feature rich and stable than MySQL.

    Of course, the best bet is Oracle which runs great on Linux.

  3. Re:obvious? by Planesdragon · · Score: 2, Informative

    Are floating tables like a table inside a frame?

    Yes. And when I tell a table to "float" in MS Word and then open it in OOo, the table is wrapped in a text box ("frame"), which works rather well. But it's nowhere near as easy to get a table in OOo inside a text box.

    Sections, what the fubar are those?

    Many of the functions of sections--denoting page number styles and columns throughout a document--are done rather well and in some cases better with OOo's "page style" feature. However, there's no way to have a different "columns" setting at different areas of a page--something that I have done in about half a dozen instances over the past year.

    But wrt word-count, what's wrong with the one under File>Properties:Statistics ?

    The fact that it's three or four clicks to get, and it's automatically a count of the WHOLE document. Measuring what one has written by word count is not an unusual task--and by tripling the requirements to get a word count, OOo has made the task dramatically harder than it is in MS Office.

    What do you use floating tables and sections for, how do they convey information that can't be conveyed by OOo? Remember this is a word processor.

    I'm currently formatting the Prometheus Reference Document (http://www.thefga.com/), and the basic layout for readability reasons is going to be two-columns per page. However, a number of tables are far wider than the column space, and this means that the tables need to "float" so that they are not in the way of the columns.

    All of these are more annoyances than problems, and OOo has other features that do outweigh the problems--but the shortcomings do exist, and macros cannot fix all of them. OOo simply is not bug-for-bug or feature-for-feature a match for MS Word, and it'd doubtful that it ever will be.