Eadon, I have had enough of you derailing our forums. You are zapped.
FFS grow up.
Server 2003 has been a good friend for the past decade. I have built a career on this operating system, I know its personality and its tics, and quirks have become second nature to me. In 2015, we will see the official end of support for Server 2003, so the time has come to start polishing the business case for the migration …
As articles those are by definition on topic - an article cannot derail itself.
However, the religious fundamentalists who invariably pop up and tell Lewis that climate isn't weather but rather a magical thing that happens when people are naughty and don't pay money to green charities and eat lentils; those people can and do derail.
This post has been deleted by its author
This post has been deleted by its author
It's true though, if one is replacing a 2003 server anyway, and given Server 2012 is so different (it is, just 2003 to 2008 is pretty much entirely different and 2012 looks pretty dfferent from that...), then it makes sense to at least take a look at other options rather than just automatically buying 2012. You could save serious cash on software and server hardware too (I found Server 2008 straight-up bloated, hopefully 2012 has gone on a diet.)
Sadly that link encapsulates my dilemma as well.i When I first started in computing, my PC's occupied the first floors of whole buildings and I got my real exposure to the disciplines involved courtesy of IBM system engineers. I've risen all the way to CIO along the way, not being beholdened to any one firm, just picking the best tools to do the job. Too bad I missed the free perks but there were absolute (go to Ft. Leavenworth) rules against that where I worked. My personal machines were anything but PC-compatible.
I still like the tech it's just way beyond the reach of the SOHO/SMB's and they can forget "The Cloud." That doesn't work at all by any measure for those markets. What to do?
Aside: Like you, they will pry my WinXP VM from my cold, dead fingers. And I'm a pretty deadly customer ;-).
Yes, I upgraded from Windows 2000 to Linux. It's been a pain though, even though the Win2K box had Apache & IIS as well as MS SQL and MySQL. The "common" LAMP CMS did work on the "WAMP" though I may add one 2003 Server next month (and upgrade it to something Linux flavoured before 2015), or not if the SAMBA shares stay stable.
Fortunately I don't have to run Exchange or Sharepoint.
But what to replace the XP with?
"The hardest question to answer is the simplest: why make the move from Server 2003 to Server 2012?"
Well, I think I can top this one with an even better question: "Why make the move to Server 2012 with its dinkey-toy interface, while you can still get Server 2008R2?".
Not only will the price of purchase be lower, you'll also get a product which was designed to be the direct upgrade for 2003, thus making the upgrade process a whole lot easier per definition. Another thing to keep in mind is the product itself: 2008 has been around for quite some time now, so you can be sure that a lot of the "out of the box" bugs have been addressed by now. Server 2012? Considering the very poor reception it gets I have my doubts there.
Finally there's an even bigger issue to keep in mind: the dreaded End Of Life issue which we're addressing right now. Windows Server 2008R2 extended support ends in 2020, as can be seen here (link to lifecycle page of Microsoft support).
The Windows Server 2012 EOL date on the other end only adds a meagre 3 years, it's support ends in 2023 as you can see here.
Sure, time is money. But not having to deal with the dinkey toy interface that was Metro is also worth a lot of money to a lot of systems administrators. There is much more to this story then "merely upgrading".
In my opinion companies are much better off picking 2008 over 2012.
"Why make the move to Server 2012 with its dinkey-toy interface, while you can still get Server 2008R2?"
This is a damned good question, sir, and one I still don't have a satisfying answer to. Server 2012 R2 is worth the jump, despite the absolute shite interface. Server 2012 versus Server 2008 R2...
...well, the only big reasons I have for that are storage related. SMB 3 is way better than SMB 2, but this sort of presumes you are using Windows 8. Also: the iSCSI target in Server 2012 is way better than the shitemobile in 2008 R2. Oh, and DirectAccess; if you do IPv6 things then you don't want to faff about with Server 2008 R2. Oh, and the whole virtualisation-aware AD controllers. Those are damned cool.
I'd say "IIS finally stopped sucking monkey dong in Server 2012" but honestly here, who uses IIS?
So it's really a toss up. If you use iSCSI off of Windows Server then 2012 is a no-brainer. I have an article coming up on why you really should be upgrading your AD controllers to 2012. DirectAccess is a bit of a niche still, but if you use it, go 2012.
Otherwise? 2008 R2. Until you pry it from my cold, dead hands. Just like Windows 7.
Reasons to use 2012 not 2008:
a) If you use Remote Desktop, it's miles better. At everything.
b) If you use Remote Access, it scales further.
c) If you use the centralised deployment, it saves you lots of time
d) If you use Hyper-V, there is literally no comparison between 2008 and 2012.
e) If you use SMB3, try telling me that's not better.
There are LOADS of reasons.
a) Unless you keep the number of systems you RDP from to a very narrow cone - or you're made out of money - RDP/VDI is too costly to implement. So what good is that?
b) Remote Access is asstastic if your client is Windows 7 because Microsoft refused to backport the client.
c) Centralised deployment is awesome! If you're building Azure. We'll have arguments about the usefulness of that particular religion when it comes to servicing downmarket customers.
d) Damn rights: Server 2012 is finally at parity (well, almost) with VMware, and Server 2012 R2 is so good that you'd swear it can't be the same team that coded Hyper-V 2.0
3) SMB3 is eleventy squillion times better than SMB 2. That said, SMB 3 requires Windows 8, so what good is that?
Most of the reasons for Server 2012 over 2008 are only applicable in three scenarios:
1) You have so much "rolling around in it" money that you can ignore being pecked to death by licensing ducks.
2) You are willing and able to block move your entire infrastructure to Windows 8/Server 2012 all in one go.
3) You operate at datacenter scale.
If you are going from Server 2003 you might as well go Server 2012 in the hope that Windows 9 will be "not ass" enough to make the leap on your clients from Windows 7. At that point you've at least upgraded your backend to be better suited to that hopeful (but bloody unlikely) future.
If you are trying to build the case for Server 2008 --> Server 2012, I might be able to stomach the arguments. 2008 was pretty bad.
If you are trying to make the argument for 2008 R2 --> 2012 you've lost me completely. If you were using Hyper-V on 2008 R2 then by now you're already 2012 .
As to the rest of the market on Server 2008 R2...
Server 2008 R2 + Windows 7 ain't broke and I have yet to be presented with even a marginal business case to attempt to fix it. The greatest value Server 2012 brought to the extant 2008 R2 customer was the ability to drive down VMware licensing costs come renewal time.
Actually, Windows 8's & Windows Server 2012's GUI has been the first from Microsoft where I haven't done a rip and replace on the entire "experience/" Where I've been putting my mouse pointers to for a couple of decades is exactly the same place I'm doing normally in "Metro." My only complaints are that the ribbon is for crap, no matter where in app or on desktop. Give me a pop-up menu with those choices. Oh, if I can tear off menus and park them wherever convenient, that's be nice too. [I was already doing that on mi Amigas, but NeXT sealed that deal.]
If you don't like the desktop, break out the scalpels! In the case of WS2012, don't even install the GUI.
... include Server 2008 R2 for those still infected with the "OH BOG IT'S BASED ON WINDOWS 8" hatred- 2008 and WIndows 7 are based on the same code base, and win7 is pretty damn stable IMHO. Admittedly, the shiny hasn't quite worn off the majority of the 2008 VMs I help build, but we are looking at 2012 at this point in time, probably starting with our AD infrastructure. (I'm *not* looking forward to that and the inevitable Exchange upgrade, but hey, that's why I get paid the semi-big bucks)
I can certainly understand why businesses are reluctant to move to 2008 and newer- we have one vendor at $work that would probably be delighted if we still ran 2000 on both the servers and desktops- they are the only reason why we have server 2003 boxes left in our environment, and even then they are starting to realize that like it or not, the gaming industry need to move at the same pace as the rest of the IT industry, even as they are being dragged, kicking and screaming, in that direction. (At least they didn't scream too much when we presented them with virtual instances on it running on our exceedingly grunty production cluster.)
Anon to protect those semi-big bucks that comprises my paycheck, largely because there are not too many big name vendors in the gaming industry and one can probably figure out just who I'm talking about. :)
"There are capital expense discussions to be had concerning end of support. For instance, upfront costs of required tools – intrusion detection systems, more advanced firewalls, network segmentation and so forth – are such that buying new Server 2012 licences is almost guaranteed to be cheaper"
I don't understand won't you require such tools with Server 2012 ..
"A large and expensive piece of industrial equipment incorporating custom hardware and applications that absolutely require Server 2003 is a mirror of what keeps me running Windows 2000".
If you had access to the source code then you could recompile the old app to run on Server 2012.
"Server 2012 contains a vast array of new technologies"
?
I don't understand won't you require such tools with Server 2012.
Nope, you won't need to buy all (or most) of those tools if you run an OS under support. Server 2012 is quite capable of taking care of itself, thank you. This isn't the Windows XP era; you don't need to piss yourself in terror at the thought of running a Windows system with an external IP address. You can pretty much do an "upgrade in place" to your existing badly-designed network without having to rearchitect the whole damned thin and carrying on with a shite security model for the next 10 years or so.
Should you be tearing up your network and redesigning everything so that there is massive amounts of segregation, multiple firewalls from independent vendors, IDS systems to detect everything, etc? Yes. Will 99%+ of companies unless you have a gun to their heads? No.
Remember that most companies are SMEs. "Eggshell security" (a hardened edge offering a single point of defence protecting a relatively wide-open and "squishy" internal network) is the best they can afford. Not merely for CapEx but for OpEx reasons.
As soon as Server 2003 moves out of support there is no rational way to keep on doing eggshell computing. The first trojan that happens along with annihilate your entire network. Server 2012 is secure enough and keep up to date with patches, etc, that you can keep on this path with about the same level of risk (probably lower) than you had before Server 2003 went out of support.
It isn't the proper way to do things, but it is the common way people do things. Server 2012 basically buys you another decade or so during which you don't have to redeisgn the entire network and retrain all your staff.
If you had access to the source code then you could recompile the old app to run on Server 2012.
Maybe. Maybe not. I'd still need someone who could deal with library changes between the two and so forth. Either way, most people don't have access to the source code. Welcome to capitalism. It sucks. Mind who you vote for next time and maybe we can slowly start to change this, eh?
We pretty much moved over to Server 2008 R2 (and Windows 7 for most of our IT department) within six months of the OS being released. Never looked back since...especially since you can certainly feel Server 2003's age by now. Heck we could feel it years ago.
And I'm still deploying new Server 2008 R2 systems. Server 2012 doesn't exist in my world. Much in the same way how vanilla Server 2008 doesn't exist either. Will Server 2012 R2 be better? Only time will tell.
I do admit that not everyone will find it necessary to move away from Server 2003. Thing is though, these people tend to be the ones who don't fiddle around with their servers much in the first place. You know, screw Windows Updates right. To these people Server 2003 going into End-of-Support is the least of their concern... especially if one of the pending updates is "Windows Server 2003 Service Pack 2".
To a full-time administrator there's lots to love about Server 2008 R2. Lots of tiny little improvements here and there to make life less of a misery. I for one welcomed WSUS becoming an MMC snap-in. Managed Service Accounts being another feature I like since managing the dozens of passwords for the accounts of all our various services has become quite the headache in the past. PowerShell 3.0... and et cetera.
Oh, and did I mention PowerShell 3.0 yet? Guess I did.
Again, these are things which really benefit the administrator.
But if there isn't an administrator... or if it's someone who comes into the office all of five minutes of his week just to visit the loo... then there really are bigger issues at hand which need looking into.
With the sole exception of domain controllers - seriously guys, virtualisation aware DCs in Server 2012 are the shit - I can't make the case for moving from Server 2008 R2 to Server 2012. The case for Server 2003 --> Server 2012 is easy. But Server 2008 R2 is a beloved old friend that is more than good enough to get the job done.
It used to be that I was perfectly comfortable with Server 2003 R2. DFSR was really all I needed to make most of what I do work fine. Then I got a Server 2008 R2 licence and set it up as a DC with DNS and DHCP. Now I can't go back. I just can't do it; Server 2003 is just too old.
You know what won me over? The ability to right click on a system in DHCP and add a reservation. That's it; right there. I wasn't won over by a firewall or a protocol, I wasn't won over by encryption or the power of Greyskull. I was won over because someone put an improvement into the operating system that does what computers are supposed to do in the first place: make boring repetitive tasks easier.
I find it interesting to note, however, that I am not nearly so rah-rah about Server 2012? Why? Because Microsoft threw ease of use out the Window. They became obsessed with the technology itself and fuck the people who have to actually use it.
If I wanted to live in a world like that, I'd use Linux. Oh wait, I do! Though it would send our resident hypertroll into paroxysms of rage, I pick Linux not because of nerdly masturbation or ethical handwaving. I choose it because in very specific circumstances it is actually *way* easier to use than Server 2012.
If you go the Server 2012 route you're stuck with the same damned things as you are on the "commercial Linux" route: mind-bogglingly shitty UIs or the shell. The shell is the aformentionned "rote memorization route" and we've wound the argument back 'round to "this isn't going to work for SMEs."
The real question is "what's going to come after Server 2008 R2 for the small business world?" It sure as shit isn't the cloud; well, not for anyone that cares about their data or not getting sued into a lump of coal. (Hi, Echelon!) The truth is that I don't really know the answer. I think there's a gap in the market here that simply isn't being filled.
That basically leaves me with hoping Synology decides to build a rockstar ecosystem around the DSM. They seem to be the only play that gives anything close to a damn any more...and it's not really all that close to a damn at all.
I personally think that the era of installed operating systems is simply over for the SME. Virtual and physical appliances are the future. The overwhelming majority of these will be Linux based, with the off BSD and Windows units making appearances for colour.
Microsoft and Oracle (via Solaris/ZFS) are sitting on the technologies required to make great SME gear. They won't do so because they fear cannibalising their cloud and enterprise licenceing markets.
Too many Linux types are Eadon-class zealots. They can't see beyond their own neuroses long enough to solve the UI problems. My recent interactions with Microsoft make it clear that under no circumstances do the give any fucks whatsoever about addressing usability issues either.
The closed source giants say "fuck users and SMEs, they aren't worth the money." Open source giants say "fuck users and SMEs, it's their own damned fault if they are too stupid to see the perfection of our glorious design." The next-generation SaaS vendors are all about the users and SMEs, slaving over designs until they are intuitive, but demand vendor lock-in, or your privacy in exchange for that usability. Worse, they're mostly based in the US, so the other 6.3 Billion of us can't use them!
It's starting to feel lonely here down at the bottom. No love for the SMEs or end users from any of the players out there. You know you've hit rock bottom when your hopes for the endpoint boil down to "maybe Tizen won't suck too bad" and your hopes for user-grokkable servers are a black box like Synology.
Beer, because this is damned depressing already.
> You know what won me over? The ability to right click on a system in DHCP and add a reservation.
That was a lovely addition indeed as I've always been fond of running Windows DHCP servers. And yup, I'm also a pretty heavy user of DHCP reservations as I like to be in absolute control of all devices on my network... all the way down to IP addresses.
This does prove my point though; that Server 2008 R2 isn't about one big change but rather that it's about a million of tiny little improvements put into an awesome package to make life a lot less of a misery.
Unfortunately however a lot of SMB's couldn't care less.
I don't know man, my SMB stable is pretty hard core Small Business Server 2011. You'll pry that out of their cold, dead hands. When you do, I'm almost certain it will be replaces with a Synology, not "Microsoft Azure NSA Edition with +1 to yearly subscription $$$." I wholly expect them to keep clinging to that OS until it goes out of support.
"Server App-V is probably the best path forward if you can get your applications packaged up for it. It enables you to stream applications to your users and offers a very good chance of successfully supporting applications that would otherwise be a real pain to migrate."
<end user alert>
I'm at home and logged into a remote desktop session to access my work desktop. It is 2008/R2 and you have now explained what that little orange cube is in the system tray is doing. As a result, I can run various 'business applications' of some vintage (I'm guessing from the interface 'design').
On log-in, scripts run to set up my remote desktop but the same scripts run to set up my desktop on an actual staff machine, and on the student machines when we log into the 'staff' domain.
I conclude that the IT guys have a remote access/roving desktop system that supports 600 teaching staff with about 300 staff machines and 500+ local student access machines with 15k accounts. We appear to be using a Windows 7 actual desktop session. Most of the hardware is dual core/2Gb with some 1Gb machines knocking around and some thin clients that used to be Citrix but now run the same as the others.
I suspect this estate and setup will be operated for as long as it can be as dosh is hard to find right now in the public sector, and it works rather well.
My personal BYOD solution is a laptop running CentOS with an rdp session running over the open wifi in college. Fast screen updating and lets me work anywhere in the building and I imagine is reasonably secure.
</end user alert>
"The ability to right click on a system in DHCP and add a reservation."
Congrats. You need to hire a DNS/DHCP admin. Nothing, in my opinion, make DHCP harder to manage than widely scattered DHCP reservations.
IMNSHO, if you need to give a machine a reserved address, power off the machine, create a reservation in the right place, and power the machine back on.
This is the main reason to upgrade for many small businesses, preferably before it happens.
Like many (as mentioned in the article), I prefer if it ain't broke etc. But it will break eventually, so you might as well pick your own time to fail. My 6 year old HP ML350 now runs 4 x virtual desktops instead of server2003 and until it breaks it can crunch some numbers for me. It never worked this hard before.
Zentyal has a $deity long way to go before it's actually usable. Not the least of which is birthing a remote access protocol that isn't made out of slow and horrible. It has an almost usable web UI - but still nowhere near as comprehensive or well supported as Virtualmin/Webmin/Usermin - and about the only nice thing I can say for the desktop UI is "at least it isn't Unity."
Zentyal needs more UX TLC and to start putting R&D into places in the Linux ecosystem (like the aforementioned remote access protocols and UX) that other distros won't. Until then it's Just Another Distro that doesn't net much over *min. For the cost, I expect more...or at the very least a commitment to a roadmap that will get those who bet their businesses on Zentyal where we need to go.
Wake me when Zentyal has taken over development of the FreeRDP server (now merged into Weyland, but still needing lots of dev support to move forward). When they've made a commitment like that - with some real, tangible benefits to the SME user that will ultimately place Zentyal ahead of the myriad other competing Linux distros in the area - I will start to believe.
That said, they're not a bad distro if you can't afford a Synology.
This post has been deleted by its author
It's a little cheeky but you could just grab an HP Microserver (around £100 including cashback) a couple of 4GB DIMMS (around £30) and a Technet subscription (£99~ish per year) which gives you everything that isn't labelled "Enterprise Edition" or a dev environment.
Job done.
@Dogged - I think it would fall on the "outright fraud" side of "a little cheeky". Personally I have a couple of microserver and technet for my home lab, but this is what it's for. People using Technet for upgrading their workstations are the reason that the workstation licences were cut back. People using Technet to feel like they're using software legitimately do great harm to other companies. If people rip off Microsoft's software, it stifles development in competitors and FOSS.
Damn straight. Microsoft kicked SMBs in the nuts, let them fall the the ground, then demanded those same SMBs subscribe to be kicked in the nuts over and over again each year.
Small Business Server 2011 was bloody brilliant and I won't soon forgive them for murdering it. The bastards.
Exchange in the cloud is not a solution unless you're flush with too much cash and American. The rest of us don't upgrade every refresh (so Microsoft's cloud offerings are TERRIBLE value) and live in countries where we as business owners can be sued if and when the PATRIOT act is used to get access to our customers' data (which in SMBs lives in e-mail far more often than any other form of storage.)
Sorry, but for all that Office 365 is awesome - well, when it's working, which isn't a lot lately - it is emphatically *not* an appropriate solution for a non-American SMB..or any SMB that doesn't upgrade with each release. (Which would be most of them.)
What's more, upgrading with each release is not a good thing. Microsoft has a nasty habit of releasing a steaming bucket of gorilla shit and then taking 3 versions to get the thing mostly usable. Then they crap out another bucket of shit and we're on the train again.
SBS 2011 had a user limit of 75. Essentials 2012 has a user limit of 25. Microsoft got so much shit for that they had to come up with a horrible workaround that rightfully earned them scorn and enmity from virtually their entire SME partner base.
In addition to killing off Exchange, Sharepoint is gone.
Essentials 2012 does not have WSUS
And that's just off the top of my head. If you are honestly peddling Server 2012 Essentials as a replacement for SBS 2011 you're either a complete idiot or an MS PR flunky that is truly abominable at their job.
Microsoft has abandoned the SME and kicked them in the nuts on the way out. SME is my turf, bub; don't teach your grandmother to suck eggs.
We've been running around 1500 Server 2003's across our estate for years, the inevitable ahs happened though, and our IBM servers are about to start shipping without support for 2003 - the storage array drivers haven't even been written for the old O/S, so it's got no way to address the target volumes. Add to that the fact we run local DB's on Oracle 9i, there are compatibility issues around moving beyond 2003.
Queue Server 2012 - where I'm hoping to install it in Hyper-V mode (GUI obviously...I am a Windows guy after all!) and build a Server 2003 VM to run on it.
Can anyone think of any gotchya's or potential hurdles I'm about to face?
Ta muchly.
Hmm, I see what you're saying, but wouldn't Hyper-V present a legacy driver to a VM running an older O/S? It's akin to VMware allowing an XP VM to run on newer hardware, because it can present an IDE interface to the virtual machine which allows to it see the raw VMDK's you've added to it?
No, not Small and Medium Business (my day job allows me to ignore those, though my "evening job" does not), but SMBv3. For file shares, the kind that need to be accessed over high-latency WAN links, the DFS/SMB3 combo seems a huge improvement. Not based on real world usage though, just some firsts tests. Actual deployment : we'll wait for R2, though we have a couple of 2012 servers running for file/print, AD and DHCP/DNS.
I agree with a previous commentard, UI is not really an issue on the server side. If you have multiple servers, using RDP to manage them seems odd.
As far as the Eadon thing goes: he'll be happy to know that, after cleaning out the storage department and finding 50 or so still working desktops/laptops, we installed Mint on them and will be donating them to schools in the area. Which will probably have a bigger effect on Linux adoption among "users" than his comments here ever did.
RDPing into the server is not odd.
A) when you work remotely you have to RDP into something (unless you use VPNs, which many/most SMBs don't have set up.)
B) The remote administration tools only work on "like" versions of the client operating system. You can't administer Server 2012 from Windows 7 and fuck you very much if you suggest I use Windows 8.
If I have two servers in from of me and each one does what I need it to do with the only difference in at the end of the day functionality being UI then I will pick the server with the UI that works best for me.
After all, I am the guy that has to use it. A UI is there to make my life as an administrator easier. That is not an irrelevant aspect of the operating system design. In fact, it's a goddamned critical one. Somehow a whole chunk of nerds seem to have forgot the reason we invented these things in the first place: to make our lives easier.
If the new version doesn't make my life easier than the old version why in the flaming monkey fuck would I spend money it?
"Newness" has no value. "New for the sake of new" is not a relevant argument. The item must deliver tangible benefits to me or my client for money to be spent on it. If it doesn't, then there is no incentive to splash the cash on that mouldy metro stash.
Ok, valid point. I was referring to the guys that rdp into a dc to manage ad, into another to manage dhcp and so on. I would assume though that you rdp in over vpn ? Genuinely interested, I would think that you don't just set up 3389 forwarding on a firewall.... allthough, sadly, many SMB's still do this (combined with no-ip or dyndns).
I could argue that a lot of mmc add-ins to manage 2012 still work on Win7, but I won't since there is AFAIK no official support. I could also argue that, if you bite the bullet, server manager on Win8 isn't half bad to manage both 2008R2 and 2012, and might even make your life easier. I won't, since my colleague on Win7 has the same issues with TIFKAM as you do.
I do agree that it's a damn shame that Microsoft does not make the 2012 RSAT available on Win7. I'm hoping against hope that they'll do an "xbox one" on that.
That being said, I was hoping that you could enlighten me regarding performance of SMBv3 vs SMBv2 (in a DFS scenario) over high latency WAN links. My tests show improvement in both access speed and bandwidth, but I'm curious about real-world usage scenarios. And my tests are just that : a 2012 guest vm/dfs host/rodc (forest trust) on each side, with guesstimates on performance. After all, I have limited time since I also have to deal with licensing for MS/VMWare/Adobe. On 2 continents. In separate contracts. I've read your articles here and on your blog, so I assume you can feel my pain. To be fair : I have quite a good relationship with both Adobe and MS where I live, and they do help out where they can. But unfortunately the good MS guys (even from sales) are limited to their region, cross any border and you're back to square one.
What's wrong with 3389 forwarding? Port 18354 --> 3389 on a target. There has been exactly one RDP bug in the past decade that could allow a user to spoof the protocol and log in. They patched it ASAP. RDP is reasonably secure, especially if local administrator-level and domain-controller level users have been denied access. I know there are security paranoids out there that would prefer I use RDP inside an SSH tunnel inside a VPN through TOR, but this isn't the DoD. It's a guy wanting convenience to get to his data. Ease of use, you know, that thing about which no fucks are given by asbergers types?
I'm getting a certain sense of "I like Windows 8, so anyone who doesn't needs to be secretly condescended to because my preferences really should apply universally. If you don't like the same things I like, there is quite obviously something wrong with you, you short-bus loser." It's rarely a conscious thing, but when people have preferences we don't understand we all do it. To some extent at least. Computer nerds are the worst. You'd think they'd be beyond these sorts of petty ostracisation, but they enjoy finding reasons to isolate and belittle more than any group I've encountered. (VIM! EMACS! FIGHT!)
SMB3 versus SMB2: SMB3 goes faster in almost every scenario. Does it go fast enough to justify having to switch to Windows 8? Fuck no.
Well, you are right : I do like Windows 8. I also like Brussels sprouts. In both cases I would never consider anyone not liking what I like to be a "short-bus loser".
I have exactly 3 Windows 8 clients on the network (and no plans whatsoever to roll out), so I wasn't really considering SMB3 for that. I was considering it for traffic between servers over a high latency (around 200 ms roundtrip) WAN link. I'm guessing the upgrade from 2003/2008 in that case would be a Good Thing here.
and fuck you very much if you suggest I use Windows 8.
Yeah, because everyone knows that Windows 8 doesn't have a desktop. It's all tiles, all the way down, right Trevor? ALL TILES ALL THE TIME. THE TILES ARE COMING FOR YOU TREVOR. RUN AND HIDE TREVOR.
Or in the universe the rest of us live in, make your desktop tile first on the Start Screen and click "enter" when you log in.
The desktop enhancements are worth it.
I tried to persuade a customer who is having issues with his apple products accessing / crashing Ex2003 that Server 2012 / Exchange 2013 would be the best way to go forward (and that support will finish soon).
After the long chats and quotes, he came back to me with "Will accessing the files be any faster?" (as in the speed of them coming from the server to the desktop). I said, no as they have a gigabit network and that's the bottleneck.
He's now more than happy to reboot the server every couple of weeks rather than spending thousands on a new server.
Actually, if by "apple products" you don't mean iOS devices but actual OSX clients, and if your customer would be willing to also dip in his/her wallet (to a lesser extent) for OSX 10.9 later this year, file access might be significantly faster since 10.9 supposedly will use SMB2 by default for file share access. But then again, in that case 2008R2/Exchange2010 will do just fine.
My personal log system is an old Sanyo 386 notebook with the screen sheared off (I am using a 10" orange IBM external monitor instead) with 1GB of RAM running DOS, WP 5.1 and Lotus 123. I use it every single day. There is a working TRS 80 in my basement used to monitor the Lizard's cage via a sensor suite.
"Old" doesn't mean "useless."
I think I am like a lot of folks with TechNet licenses. I still want to believe Microsoft is relevant, and that what they build matters. My employer however doesn't want me investing in Microsoft skills. So I have the TechNet subscription (which I pay for personally) for this purpose.
Has it been handy? Not very often, but having it meant there was always an opportunity to work at home on MS-focused technical content, and I have used it on occasion. For example, I had planned to use this license to work on my hyper-V skills, to see if I could utilize this to justify purchasing an upgrade to my one Windows Server 2008 license (EE). Now I will not be doing this; but I am working hard on building my VMware skills needed for my other 10 lab servers.
In the "old days" (the 1990's - and wow, that seems so silly sounding) I remember TechNet as being marketed for Admins, Technical Architects and similar folks - not those doing "development" but technically testing and deploying MS core products. MSDN on the other hand was for serious development (MS SQL, MS Visual Studio, etc.), and my organization had one of the enterprise MSDN agreements then. So I don't agree with the "it was never for testing" BS; the _type_ of testing is the issue, and I always viewed TechNet testing as the "install, upgrade, integrate, deploy" type.
So I am asking MS to reconsider this move. I am not personally going to spend $700 per year for a MSDN license, and the DRM hassle of moving stuff around is quite a pain. (For example, I have a licensed copy of MS Office 2010, and every time I change my laptop hard disk I have to do the song-and-dance with MS about the license key). I can see reason in not giving me access to EVERYTHING, but the various versions of Server and Office and perhaps a limited number of other applications would be sufficient. Asking me to pay much more however is going to get me to vote with my feet.
I have plenty of corporate support for building Linux skills. I am about to permanently convert my Windows Desktop to Linux. I had removed OpenOffice (actually its kind-of-twin-clone LibreOffice) from all my personal machines (Windows desktop, Mac Mini, Windows laptop) but will now instead uninstall my TechNet Office 2010 licenses and <sigh> really learn how to use OpenOffice (or whatever).
In my experience at work, this is how it really begins in this business. When those at the top forget how they got there - because they had the support of those on the ground - it becomes very difficult to prevent this kind of decline. Last week BlackBerry pulled the plug on the Playbook (something they had been hinting at for a few months, after they had reduced inventory sufficiently). It would seem so straight-forward: the Playbook was a "distraction", not part of their future. And I know they need to focus on the Z10, Q10, and the other phones still coming, along with getting corporate customers to upgrade to BES10. Fortunately, I bought mine at $150 (32GB model), which was a great buy what what it does. It's just a shame BB can't see the value in at least giving the Playbook a new home as a full Android tablet. It may not be able to run BB10, but it could probably run Android 4.2 just fine thank you.
My point is BlackBerry also did the same thing as Microsoft seems to be doing - pissing off a whole load of enthusiasts who have supported and defended them for years in this area. Not only will I not be interested in any future BB tablets (something no mobile company can be without), but I am going to move away from BB devices for my phone as well, as soon as I am able (8 months and counting). I can't count Microsoft out in the same way, but I won't spend any time (or my own money) building skills favorable to them. How much "good will" is going to be lost by this move? I will bet the asset value of this was not properly calculated, versus the supposed improvement in revenue and costs by consolidating that fraction of users from the TechNet category into MSDN that are willing to make the transition.
All of the above to state one thing - "penny-wise, and pound-foolish!"