Longhorn Server Scrapped

punkass writes "Microsoft announced Tuesday that plans for .Net Server, aka "Longhorn" have been scrapped and they will instead focus on the the release after that, code-named Blackcomb. NT4 came out in 96, 2k in 2000, and Longhorn was due out in 2005-06...Blackcomb seems to be a long time between releases."

Comments

This is good.

[Penguinoflight]

The longer Microsoft has between releases, the longer Linux has to come up with great releases. Just think how many security patches there will be between 2000 and blackcomb... that's not fun and sysadmins know it.

Re:This is good.

[4of12]

They seem to be desperately out of tune with their users in the server market

It's because they don't compete in a normal market anymore.

When you essentially own a marketplace, such as they do with desktop PC operating systems, then you can make a lot more decisions that your customers don't like but have to accept because the alternatives have disappeared or are considered too drastic (MacOS, Linux, etc.).

Actually, the competition they seem to be desperately out of tune with their users in the server markethey've endured trying to enter the server market has been good for them and their customers: each revision of NT was compared with UNIX. Early revs were laughable, but MS had a target to aim towards in terms of reliability and scalability. Finally, with Win2K they have something where they don't get laughed out of the room anymore. I doubt whether it would have been as a good product as it is without the competition.

They face a more serious threat in the future with their server operating systems. If they strongly leverage their desktop dominance in Windows and Office, then they can insure their servers are the only brand that works in a networked environment.

But if services are standardized and commoditized, which is what customers really like for their effects on price and quality (as in the PC hardware market), then open source flavors of UNIX will have already eaten their server marketplace for breakfast when they finally trot out some shiny chrome-plated Blackcomb .NET product that "does everything and more".

Re:This is good.

[ergo98]

Security patches are to be reckoned with. However, the long interval between new Windows server releases means that there will be large differences inbetween versions. If there's something that the IT departments don't like, it's large differences...So - I really have to wonder what crack Microsoft is smoking.

Obviously not as potent as the stuff you're smoking. Here are the facts as I see them. Please feel free to disagree.

• IT departments HATE non-stop tiny changes, each of which requires new testing, and likely breaks several of their applications. This can be seen in the reality that many shops still use NT 4.0 even with its successor 2000, and its successor XP, available. Hell, most IT departments shun at even installing service packs.
• Microsoft has oft been berated for doing what Apple now is the champion of, which is pushing minor changes as new versions (see 95, 98, 98SE, Me) and getting people to pay up. People don't like being "behind" when applications start using some esoteric feature of XP that adds little value, but suddenly renders obsolete their 2000 base.
• Because Microsoft has moved at such a rapid rate, many organizations have simple ceased moving with them: By the time they get a plan together and start to act, they're behind again. There are still a large number of organizations that aren't using Active Directory. There literally is such a thing as being too-rapid in your development (at least in areas where users and IT have to move with you) because the early-adopters will give up, and the late adopters will always feel that something better is just around the corner.

Re:This is good.

[peaworth]

Another aspect of this that is not necessarily releated to the annoyance / lack of annoyance in the technical side of the the IT department is the cost. Licensing 6.0 from MS went to annual subscription payments that are based on a 3 year break even rate. All those companies that signed up for licensing 6.0 this year on their server OS's will pay for the price of a new OS over the next 3 years but there will be no new product released in that time even if they wanted to "upgrade". They just got screwed. That ought to foster some more good will.

Not surprising, in the context of MS's new licensi

[Marx_Mrvelous]

So, let's look at some interesting facts:
1. MS Puts back the release of its latest Server OS.
2. MS is pushing a new licensing model where companies pay annual fees regardless of upgrades, but then get "discounts" on future upgrades.

So, does the new licensing plan allow them to basically, delay new technologies? It seems that, with their latest scheme, it reduces their motivation to release newer/better products.

Means essentially: nobody would buy one, anyway.

[thasmudyan]

One of the key paragraphs says:

Analysts expect the company's upcoming server software, .Net Server, to take off slowly because many businesses have either recently moved to Windows 2000 or are in the process of doing so. A majority of customers, then, would begin introducing .Net Server in late 2004--around the same time as the planned release of the Longhorn desktop and server software versions.

And that about covers my experience, too. Server overhauls take much longer intervals then changes in the desktop segment, where they install a new Windows every 3 years or so (doesn't matter, they are largely compatible versions, anyway... no admins, don't kill me, aaarrrghh).

So it actually makes sense to come out with a new server only if the changes are really signifcant and if the interval since the last major roll-out was more than, say, 5-6 years ago. Besides, nobody has money to throw at a new unproven technology right now (and in 2 years all the same), anyway.

Let's skip them, and do it ourselves.

[ka9dgx]

The goal of the next version (as I understand it) is to allow virtual folders, so that you can search everything with a common set of tools.

Let's build a virtual folder driver for Windows 98 and upward, to allow APPLICATIONS to virtualize the information they manage. It would be nice to have an email manager than presents emails as a list of files, or folders. Sending could be as easy as copying files to a folder, and then specifying an address. (To.txt?)

A virtualized database would present a list of folders in place of a table, with the fields being individual files, some read only (sequence numbers, keys, etc). To update the data, you just write to the file containing the appopriate field. If you wanted to add a field, you just copy a new file into the folder.

There is great power in letting an APPLICATION control the virutalization of the OS, this is why the idea of GNU/HURD is important for the future.

If APPLICATIONS can virtualize, then you get a freedom to innovate that would give Bill nightmares.

Virtualizing the address space for existing millions of users and applications could do more to help freedom to innovate than pretty much anything it's going to take Microsoft years to come up with.

Who's with me?

--Mike--

Has everyone missed this?

[drew_kime]

The remedy in the DoJ case mentioned Longhorn by name. Did anyone think Microsoft wouldn't just invent some reason to say of their next product, "This is not the Longhorn you are looking for"?