I wish I could mod you funny :)

A trend I've noticed in many Windows SW is that many commercial titles have stalled, and their offerings look pretty much as they were in 2001. Many commercial developers have been going after web products for so long that have let their desktop counterparts wither and look antiquated (and their web offerings are still not fully fleshed out or even feature complete to compete with the desktop version).

This combined with VDI (and cloud services in general), makes a case in some organizations to ditch the windows client, and just run the legacy apps on a RemoteApp window through VDI on the cloud. Then it doesn't matter what OS you run, you can always reach your legacy apps.

What MS needs is to give developers a reason to develop native apps again for its platform (in contrast for iOS or Android), so it can extend its reach past the desktop in a meaningful way. Otherwise, the platform will be relegated to back end legacy apps that can be run remotely.

Win32s API and the Windows compatibility layer in OS/2 were a serious threat to MS dominance at the time. They offered app developers a measure of compatibility with the present (OS/2 and NT) and the yet unreleased Win95, and it could have stopped MS in its tracks to get Windows to 32 bits. If Win95 would have taken longer, it would have made sense for more apps to migrate over to OS/2 Warp (or to Win32s and run on all 3 operating systems).

What MS lacks right now is an unifying development environment that spawns both form factors. Many Windows apps never migrated to WinMo (and WinPhone) because the API was too crippled and it was too difficult. And now the .Net framework hasn't taken the place of the full Win API for most commercial development and it is becoming too fragmented and it is too confusing what is supported on which platform. With Win8, MS is looking to reset all the versions of the framework into a single version: WinRT. With it, it might finally achieve some compatibility across mobile and desktop, so it is easier for developers to target both platforms (even if with different binaries, most of the WinRT code will run on both types of environments with just a recompile). This might make it more palatable for existing apps to migrate to .Net or all the way to WinRT in the near future (especially if there is .Net framework that is backwards compatible that allows WinRT apps to run on Win7 and WIn2008, but that's a bit far as speculations go).

The DOS 4 flop was pretty bad (most users stayed with DOS 3.3 for the longest time), but it also made DOS 5 and 6 look like gold when they came out, and made it harder to make the case for OS/2, which seemed like too much bloat and closer to DOS 4.

I think of current globalization as -nearly- a return of slavery. Right now, outsourcing to India and China seems cheaper than automation because the initial investment is low. Even if in the long term automation within USA and EU for their markets would be cheaper and lead to sustained growth and better quality: business tend to only make short term decisions and won't even consider automation.

It is also reminiscent of the aborted Rome/Greece industrial age (No Industrial Revolution in Ancient Greece?): it was cheaper to keep slaves than to invest in building machines that would work using steam power. Right now, even though we have the technology, it is considered a good business practice to outsource labor to low tech markets because their startup cost is low, and there is no regard for quality and long term viability.

That's the competitive advantage that business are giving up when they replace their local, hard earned and paid for local systems in favor of the new and shiny Cloud.

Then companies become shells, that are geared towards sales. The only thing these business will do is figure out ways to sale the re-branded generic products.

As customers realize that most providers are pretty much the same, there will be more and more pressure to lower prices, and will force these same business to figure out ways to be competitive (cheaper and distinctive). Some will figure out ways to combine cloud offerings and/or their own proprietary tech, and might come up with new distinctive offerings. Most will be bypassed and the market will consolidate (ie, going out of business, since the only one doing the work is the Cloud provider).

MS is just jumping on the bandwagon. Windows is perceived as "cloud late", so it is fighting the perception.

The Cloud concepts can useful tools in the IT arsenal. But we have to remember that not everything is a nail, and right now we're still in the phase where managers thing that their cloud hammer is good for everything.

Cloud services make a lot of sense for retail and some manufacturing and store fronts. It makes no sense for specialized service and office workers. Trying to use it for everything is the proverbial square peg into the round hole. Users haven't realized the discrepancy yet. The question is, how long it will be before they begin the reaction and correction?

Indeed, having Cloud services complement and augment a local infrastructure can be a good idea. However, most public cloud providers aren't doing this, and are bypassing the local infrastructure completely (assuming one exists in the first place). Hybrid and private clouds can make a lot of sense, but managers don't make the distinction.

And the question of productivity isn't being addressed. Many cloud systems are severely limited in performance, to the point that they can't compete with local systems. It isn't a 1:1 equivalence between LAN and Cloud. But business managers just hear Cloud and can't wait to jump. And then they become trapped in low performance closed systems, and it takes a major event for them to roll back and rebuild their local systems.

Against

Second parragraph (in essence, putting numbers to the "amount" of climate change):

Yes, MS offers discounted versions of Windows and Office, but they're not cheap enough when competing with free alternatives, and many times not include all the functionality you want the users to learn (or there is a pirated old version available w/o DRM restrictions and enough functionality, which might be obsolete, but gets the job done, but also means that the student is not really being trained in a current product, so MS loses in both fronts: no license purchased, and the user won't promote its products when he/she becomes an employee of a company that can afford it). Also, prices didn't use to be as affordable as they're are now. If I remember correctly (it's been a while), the savings use to be in the 10% to 20% range, and there wasn't that much of a distinction between the commercial and academic version. Now the academic version is heavily discounted (seems about 50% or more) and you get extras like Encarta and other resources, but it might be too little too late.

And in their rush to make "affordable" editions of Windows and Office, MS has fragmented it and made it too confusing (to the point that users don't know what they need or have). Before, you had Windows and Office, and you pretty much knew the capabilities of your setup, and knew that you had everything you needed, you just had to click around the help files and figure it out. It makes sense to have up to 2 editions (maybe 3 stretching it), with minimal differences between them (ie advanced features, not complete applications missing). You could go for the low end to economize, knowing you could unlock more features, but now with all the editions and variations, you can't tell what you have. And then they introduce competing apps and suites (Outlook Express vs Outlook, and Works vs Office), further fragmenting the name.

You have corporate users doing diagrams in Excel or doing graphics designs in Powerpoint (instead of using Visio o Publisher), partly because of ignorance (ie, not knowing that there is a better program for that), and because the more suited application is not included with their edition of office, or what is worse, because someone in their team actually put thought into it, and opted to use the less optimal app in order to make it easier to distribute the documents even when their team had the budget for the full Office Suite (or their IT could install an Office viewer for the unsupported format). Instead of making it easier, it is just more complexity.

Maybe if MS were to give away started editions, with more features (maybe also forego activation on starter editions, and on older versions - if a user is stuck with XP or Office XP, let them, not even bother them with activation, even offer the old version for free on the web site). Getting students and home users hooked on an old version is better (for MS) than letting them go to a free alternative on another OS.

I don't really have a solution for MS: it just might be too late. The Vista debacle (and other MS missteps, like repeatedly failing with Tablets and Phones) and Apple and Google push for moving everything tot he network have cut short MS window to remain relevant by years (decades?). It's only option now is to hit a home run with Windows 8, and maybe remain relevant for a few more years, but it will be difficult to regain momentum. In a best case scenario, it might actually get what it wants: with thin clients (and W8 on arm), and most everybody will end up running its apps remotely (using RDP RemoteApp) on the cloud, and it will be sure that most anybody running its SW will be a legal user (since its on the cloud, there will be a stronger control on illegal copies). However, it will still only have a shrinking user base, with more and more users opting for less restricting options w/o DRM being recommended by their younger relatives or newly hired coworkers.

Complex applications require that people know how to use them, and it takes time and investment for people to get trained.An growing expert user base is the best advertising that you can get. Having your SW out there, in the hands of students and young people trying to figure out how to use it helps it remain relevant as they go to work for companies that end up purchasing the SW.

IMO, more than open source and the Internet and hosting (paradigm shift), this is what is actually killing off Microsoft. It used to frown on piracy, and fight it mostly to scare up business that could afford to pay, but more or less allowed for the general population, since ensured that new users would have an easier time finding its SW, and that would encourage them to remain on the Windows platform. With XP and its activation scheme, MS didn't stop piracy (ie, determined users that aren't going to pay you anyway will either break it, or use alternatives), but made it harder for new users (students and home users) to get into its products, and with he rise of alternatives, and the Vista fiasco, it is relegating itself to oblivion ("the harder you hold on, the more you lose").

There is also the logic that these companies see new users as a source of revenue, not only as licenses, but as requiring training. So instead of giving away their SW to people that would self-train, they expect them to pay to get trained. With companies not wanting to send employees for training, and with motivated individuals unable to pay for it themselves, this IMO is a losing strategy (it generates short term revenue if your product is an industry standard that most be learned, but you lose out on dedicated people, and your user base tends to erode and eventually your product becomes irrelevant).

I seem to recall Windows 2003 was branded "Windows .Net" for a short while.

Indeed, legacy business apps are living far beyond the lifecycle of the underlying technologies, many times thanks to virtualization, and because business are reticent to re-engineer old apps (virtualization is allowing them to stay with old tech a lot longer).

I dislike profoundly the concept of desktop WOA, but it might make sense for some. MS is looking to drastically simplify the platform. If they get rid of most legacy APIs (and old ways to do things) and make it easy for consumers and enterprise users and IT to just install a light client (ie, WOA), they might have something. Having WOA for the endpoint might make sense because you can always provide the good old legacy apps using RDP and RemoteApp, getting the best of both worlds (as long as you have a connection, disconnected laptop users be dammed). MS will be charging for 2 licenses per user: one for the WOA device, one -or many- for the RDP server client licenses.

As for the list of tools: Delphi might make it. Embarcadero is pushing cross platform development with their FireMonkey framework in the XE2 line of products (ie, Delphi and C++). They already support WIndows, OSX and iOS, and there are plans to support Android (blog posts hint at a beta sometime this quarter or next).

FireMonkey is not a direct Delphi VCL (or even OWL) replacement, so it's not like you'll be able to just recompile your old Delphi apps, but you at least get a starting point, and Delphi component vendors seem to be taking notice and creating components for the framework. It remains to be seen how much vapor and how good it is, and how many organizations are willing to port their Delphi apps.

So for native cross development for iOS and Android, there might be a few options in the horizon:
* Xamarin monoTouch and monoDroid, at $400 each (for single developer license)
* Embarcadero FireMonkey XE2 products (when it supports Android)
* Qt

And these support WIndows x86, and will likely support WOA.

Yes, MS logging sometimes is overly obtuse and opaque. There seems to be a generalized aversion of putting data into the log, maybe trying not to disclose sensitive information with the error, but it leads to useless error messages in general.