Discussions

Jason, you're forgetting the adoption curve. The very people you're offending are the early adopters. Pragmatists will partially base their buying decisions on the opinions of early adopter types who they know. If your early adopters are offended, so offended that they themselves don't adopt, their advice to the pragmatists will be simple: don't buy it.

I'm trying to tell you that you've blown it, big time.

The best kind of advertising is supposed to be word-of-mouth. The only word on the streets at the moment is that Vista's licensing sucks.

PaoloM wrote:



Mike Dimmick wrote: A hint to Microsoft, though: don't do anything to force your XP customers to migrate, because they may not migrate to the platform you want them to. I have a feeling you may be supporting Windows XP for rather more years than in your current support lifecycle plans.

I am not going to say anything pro or con, but I distinctly remember those exact same words being said by many in 2001 regarding Windows 2000 and Windows XP...

I still see the Windows 2000 die-hards on CodeProject's forums almost every day, and at least a couple of my colleagues are running Windows XP, but doing so in the Classic theme. XP SP2 is enough of a jump over Windows 2000 that previous holdouts are starting to consider migrating, but a lot of them still consider that Windows 2000 was the best OS Microsoft have built so far.

Windows 98 to XP, no argument. Well, barring a few games that don't work on XP no matter how many compatibility bits you flip. I still have a couple of Windows 98 VM for this reason. (Psygnosis/Perfect Entertainment's "Discworld II" actually doesn't even work properly if you've got VM additions installed).

As for the MSDN Subscription argument, come off it. An Operating Systems subscription is £550 per year rather than a one-time purchase price of whatever edition is most suitable for the user, which is at most £260 for Vista Ultimate full packaged product (US price converted to GBP at ratio of 1.8, plus VAT at 17.5%). Your average homebrew enthusiast developer is going to balk at that, if they just want a clean image for testing software in. You give us Virtual PC for free, but take away with the other hand.

JasonOlson wrote:

Here's my opinion, there is NO doom and gloom ahead for Vista's licensing. In the end, you have to remember that us geeks are in the MINORITY of all computer users. The number of non-geek people I know that have actually _read_ the EULA I could count on a single hand (and, technically, I could count using my entire arms as digits (heck, I only need two bits to describe the number of people)).

Don't you think people should read what they're agreeing to? Should the license agreement not be intelligible, to be clear on what the terms actually are? If the customer believes they're entitled to one thing, but the activation service prevents them from doing so, you get confused and angry customers. A confused and angry customer will not be happy with paying for an additional licence at the point that you decide that they need one.

If we assume that users do not read the EULA, should not the terms of that licence be what a user would reasonably expect? Are we not trying to make 'intuitive' software, where a user can guess how to use it?

You should definitely not need a law degree and ten years' experience in copyright law to understand the licence agreement. It's too long, includes too many restrictions that are permitted in many jurisdictions (reverse engineering is explicitly permitted for compatibility in Europe, and this right cannot be restricted by the license agreement), and is not worded clearly.

I'm not telling you the geek opinion, I'm telling you what the general public believe they are entitled to do, and reasonably should be permitted to do. Forget tying 'OEM' installations to a specific machine; customers don't understand it and get upset if they can't transfer to a different machine. Geeks don't understand it.

I forgot to say what I'd be doing.

I'll be sticking with Windows XP. New hardware would be required for OS X - and new software, unless I put XP on as well. A desktop Unix would also require all new software - I'm perfectly happy with MS Office and will probably be upgrading to Office 2007.

Frankly, as a complete environment, I still like Windows XP the best.

A hint to Microsoft, though: don't do anything to force your XP customers to migrate, because they may not migrate to the platform you want them to. I have a feeling you may be supporting Windows XP for rather more years than in your current support lifecycle plans.

You're shooting yourselves in the foot.

Ever since these licences were announced, technical and enthusiast forums have been filled with posts arguing over what the licence means. Most of the posters have become highly disillusioned and announced an intention to switch to the Mac or to Linux.

The licensing is simply much more draconian than people will accept. It is of course your right to set whatever terms you want - that's copyright, that the copyright owner retains the sole right to make copies except as licensed by them. However, consumers have a reasonable expectation that they will be able to resell an item that they have purchased, and therefore that they can, without limit, trade the license that they have purchased.

For any interchangeable component, moreover, they have an expectation that they will be able to use that component with any other compatible component. They therefore expect to be able to move that component without limit. Limiting to a single transfer feels like you're abusing your customers. Some people will make additional copies without removing the previous copy, and you will have to catch this through activation.

Requiring additional licences for virtual machines is stupid. A virtual machine is a drain on the host machine's resources; it does not allow the customer to do more, rather to do less. If you do not require additional licences if the customer upgrades to a processor package with more cores, gaining a more performant machine, why require a new licence if the system is made less performant?

Saying something like this would go a long way to correcting the problem:

'You may at any time, subject to Microsoft's approval, install copies of the software, provided that you remove all other active copies from other hardware. A Microsoft online service will automatically approve requests except where Microsoft has reasonable grounds to suspect that the terms of this license have been broken.'

I'm no lawyer, so you'll probably want to reword it, but that's far more fair.

With the licence terms you have currently published, you will sell fewer licences of Vista, because people simply aren't accepting the terms. The people who should be your early adopters are refusing to bite. Take my terms, and you may sell more licences in the long run.

This licence change, or clarification, or whatever you want to call it, is the straw breaking the camel's back for me. I won't be upgrading to Vista. For me to adopt Vista, you'd have to fix this, you'd have to get compatibility working for eVC 3.0 and 4.0 (essential for my work, they currently both crash on loading a project), and you would have to commit to supporting Visual Studio .NET 2003 and SQL Server 2000, according to the dates outlined in your current support lifecycle pages (i.e. mainstream support to 14 October 2008 and 8 April 2008 respectively).

SQL 2005 Express creates databases with the 'autoclose' option switched ON by default, if you create them simply using the 'CREATE DATABASE' statement. When you first create a connection to a particular database, it runs through various checks to mount the database and roll forward the transaction log, and that takes a little time. By contrast the full versions of SQL Server mount all databases at startup.

If you use the SQL Server Management Studio [Express] graphical tools, Auto Close defaults to off. I'm not sure what happens if you use VS2005.

I've largely moved from VS 2003 to 2005 for Compact Framework 1.0 development. I'm finding it a lot quicker and more stable, especially when debugging. That's my experience, though.

At home, C++ Smart Device projects are broken - it simply says 'create failed' on the status bar. This may be due to IE7 Beta 2 Preview though (not installed the latest version yet) since the project wizard is an HTML dialog.

khgiese wrote:

I have the firewall on the exchange server truned of, i do not have any extra installed software such as groupshield, anti-virus or anything other then Windows 2003 server std sp2 and exchange 2003 sp1. I used netstat -an to verify that port 25 has nothing else using the port and the status is listening.
Can anyone help me here?

I'm going to assume that you've got those two service pack numbers the wrong way round, since the current Exchange Server 2003 service pack is SP2, while there isn't yet an SP2 for Windows Server 2003.

You must have some kind of firewall switched on. Did you run the Security Configuration Wizard from Windows Server 2003 SP1? See this Exchange team blog post for information on using SCW with Exchange Server.

The only other alternative I can think of is that DNS resolution between the client you tried and the DC is broken, which typically means that you don't have the right DNS configuration on the client. For AD to work, your clients - indeed all computers on your network including member servers and domain controllers - should only have internal DNS servers listed, so that all DNS queries go to your domain DNS servers. Your domain DNS servers should have forwarders set up (normally your ISP's DNS servers) to forward DNS queries for external sites.

If that's the case, the failure to connect to port 25 from a client computer is a red herring. Firstly, you need to tell the world where to send email. This is done by setting up MX records in your public DNS zone, to point to the appropriate server. Then you need to ensure that the SMTP server is exposed to the internet - if you have a hardware firewall you need to allow communication to port 25 on your SMTP server, and if you're using NAT you need to ensure port 25 on the external interface is mapped to port 25 on your server.

Normally I'm not a fan of Steve Gibson but he does have an end-user accessible port scanner. See ShieldsUp on his homepage. Enter port 25, then choose User Specified Custom Port Probe. The service is intended to tell you that ports are closed, therefore you want it to tell you that the probe failed and that port 25 is open.

Go and read the Winsock Programmer's FAQ. That'll answer a lot of questions about how to use sockets effectively.

The only way the client can tell that the server has disconnected (if disconnected gracefully) is by reading all the data until the receive function returns 0 bytes. Until that point it doesn't see the FIN flag. The server doesn't set the FIN flag in a transmitted packet until it's sent all buffered data.

In the case of an abnormal disconnection, eventually you'll either get a timeout, or you'll get an error response to sent data if the server sets the RST (Reset) flag in a response packet.

The Connected state basically just means that you've successfully performed the three-way handshake to set up a connection (client sends SYN, server sends SYN-ACK, client sends ACK). Beyond that point it doesn't really have any significance.

ScanIAm wrote:

I just made the switch in the opposite direction due to ATI's incompetance with drivers.

I just switched from nVidia to ATI for the exact same reason. Honestly, neither company is that hot on reliability right now. Although they're both very hot on radiated power

Actually a major part of my move was simply because nVidia wouldn't support my GeForce Ti 4800 SE on Windows Vista - it may of course not be capable of Glass. I didn't exactly push the boat out - I now have a Radeon 9550. I just wanted something to tide me over until buying a whole new machine some time later in the year (waiting for desktop version of Intel Core processors). Thankfully this card does not have a fan on it, so computer noise is reduced and IMO, reliability increased.

I'm not really a fan of Matrox any more either. The Millennium II I got with a machine back in 1998 was great (poor 3D acceleration, at the time that wasn't so important) but they really messed up the G200 and G400 lines.