Understanding .Net Framework 3.5
-
Well, Jim in Arizona I think if you read my blog posts you'll see that the new advancements are additive which is the whole point. You *can* stick with v2.0 and still use the latest IDE (VS2008) and only use v3 and v3.5 features when you are ready (in the same way you can use your own new libraries when you are ready). No runtime retesting since the CLR is the same and no compatibility issues since the framework bits are additive. Like I said, please read the posts for more info. I never thought I'd hear people complain for too much innovation :-)
May I take the opportunity to ask you about OS compatibility of .Net 3.0 and 3.5? I was told that both of them won't run on Win2k machines. As far as I can see there is no option in VS2005 to select the compiler, so what happens if I install .Net 3+? As I have some customers who still use Win2k, I'm a bit afraid they won't be able to use programs I compile afer updating. Thanks in advance Georg
-
Daniel Moth wrote:
I never thought I'd hear people complain for too much innovation
Hmmm..you haven't been around here for the last year or so then. :) While I'm not in this camp I hear from them all the time and I'll sum it up for you: There is actually quite a large percentage of professional developers here that are publicly expressing a *lot* of resentment towards Microsoft against the quantity and pace of new technology being released. Unusual and troubling for Microsoft (or it should be in any case) in that this board is populated by quite possibly the developers worldwide that are the *most* supportive of Microsoft development systems. They feel that they have just begun to really learn a new technology only to be bombarded by something completely new and they feel like Microsoft is employing a shotgun approach to new technology leaving them very wary of adopting anything new until the dust settles. It's not at all unusual to hear a *lot* of developers here express the opinion that they still won't touch .net because they feel like it's the flavour of the week at Microsoft who may very likely drop it effectively pulling the rug out from under them. I've maintained that I think Microsoft has bet the farm on .net, they respond that they just don't believe it and point to the fact that Microsoft doesn't use .net for any of their core business applications. Personally I'm no where near as skeptical as that and have fully embraced .net up to the 2.0 level and dabbled in the newest stuff, but come 3.5 I'm starting to feel the same level of burnout. I invested 2 years building up a rock solid business object framework using .net 1.1 have released major commercial apps based on it only to find that I could now take advantage of a lot of new technology that would improve it but it's far too late now that I have all this released code out there. Microsoft's timing and overall understanding of their development technology end users seems to have been slipping badly lately and I think it's likely most of the insiders who create this technology at Microsoft probably don't realize that at all because their main feedback comes from "keaners" who attend conferences and keep up with the technology for the sake of it, not working class in the trenches programmers who don't have the luxury of adopting every change that comes along because they've just got too damned much existing code that would have to be changed.
Thanks for the thoughtful reply John. Please note the smiley, as my comment was meant to be a bit tongue in cheek. IMO, quick subsequent releases are problematic only when they potentially break existing codebases. Even then, it is an opt-in model due to the side-by-side nature of the CLR. Like I said though, the new stuff is all additive (and multitargeting in VS2008 is a huge help). So you need only take advantage of v3 or v3.5 when you actually require it and not a day sooner. As a developer, I think that is a good thing. Happy coding :-)
-
I agree for the most part with the above reply. Added onto that is you have companies and HR people looking to see what the latest .Net is out there and then place job requirements on that. For example my current job required .Net 2.0, I came in and they weren't even using it yet! They had to order it. Add on to the mess is how unstable the 2.0 is. I have .Net1.1 at home and I am not switching due to the amount of frustertaions and time I have to wait for my windows apps to compile using .Net 2.0.
Please see my reply to John. Also please do not confuse the framework/CLR with the IDE. You are saying that .NET 2 is not stable and then you talk about compilation issues. If you have issues with the IDE/compilers (irrelevant to the framework) then maybe VS2008 will sort them out for you :-)
-
May I take the opportunity to ask you about OS compatibility of .Net 3.0 and 3.5? I was told that both of them won't run on Win2k machines. As far as I can see there is no option in VS2005 to select the compiler, so what happens if I install .Net 3+? As I have some customers who still use Win2k, I'm a bit afraid they won't be able to use programs I compile afer updating. Thanks in advance Georg
I am not sure that this is the best place for support questions (I don't want to upset the moderators), so please use the free online forums for that: http://forums.microsoft.com/MSDN/default.aspx?SiteID=1 Generally though, your understanding is correct. Remember that NetFx3 was going to be Vista-only and after customer feedback it was made to work on XP and WS03 as well... HTH
-
Daniel Moth wrote:
I never thought I'd hear people complain for too much innovation
Hmmm..you haven't been around here for the last year or so then. :) While I'm not in this camp I hear from them all the time and I'll sum it up for you: There is actually quite a large percentage of professional developers here that are publicly expressing a *lot* of resentment towards Microsoft against the quantity and pace of new technology being released. Unusual and troubling for Microsoft (or it should be in any case) in that this board is populated by quite possibly the developers worldwide that are the *most* supportive of Microsoft development systems. They feel that they have just begun to really learn a new technology only to be bombarded by something completely new and they feel like Microsoft is employing a shotgun approach to new technology leaving them very wary of adopting anything new until the dust settles. It's not at all unusual to hear a *lot* of developers here express the opinion that they still won't touch .net because they feel like it's the flavour of the week at Microsoft who may very likely drop it effectively pulling the rug out from under them. I've maintained that I think Microsoft has bet the farm on .net, they respond that they just don't believe it and point to the fact that Microsoft doesn't use .net for any of their core business applications. Personally I'm no where near as skeptical as that and have fully embraced .net up to the 2.0 level and dabbled in the newest stuff, but come 3.5 I'm starting to feel the same level of burnout. I invested 2 years building up a rock solid business object framework using .net 1.1 have released major commercial apps based on it only to find that I could now take advantage of a lot of new technology that would improve it but it's far too late now that I have all this released code out there. Microsoft's timing and overall understanding of their development technology end users seems to have been slipping badly lately and I think it's likely most of the insiders who create this technology at Microsoft probably don't realize that at all because their main feedback comes from "keaners" who attend conferences and keep up with the technology for the sake of it, not working class in the trenches programmers who don't have the luxury of adopting every change that comes along because they've just got too damned much existing code that would have to be changed.
The endless stream of new software, whether by Microsoft or others, is a pet peeve of mine. The primary benificieries, when push comes to shove, are the book publishers. In my little work-group, I am generally the last to consider the study of the newest technology. whilst some look for successors for Ruby,* (perhas Cubic Zirconia or Diamel?). Right next to me is, coincidentally, the one in our midst who is most constantly buying new "Programmer's Books". While he talks a good talk, brimming with the latest jargon and correct techniques, I've yet to see much of this be put into any use. When push comes to shove, knowing what you know, and knowing it well, is the source of good, solid, and robust code with real performance and . . . a short development cycle. And quit often, when a solution to some difficult SQL is required, 'they' come to the problem solver (my unabashed primary asset). Instead of spreading beyond, I explored the ODBC drivers for Text and Excel. Wouldn't you know it . . . expanding current knowledge was coincidentally of value to my employer (and cohorts) within a month. Now, I admit I moved on some time ago from C to C++. In fact, though not totally of my own free will, I started using C#, though admittedly with a C++ mentality. The single set of methods in .NET, though a pain to learn at first, is something I bought into. I'm not the complete Luddite. But it is perhaps a weakness that we, as a group, are so easily whip-sawed by the technology-of-the-week. I program for .NET 1.1 because that's what several hundred workstations on this site are currently guaranteed to have installed. I need to be sure that the apps work the first time. I'll jump to 2.0 as soon as I've a customer base. So long as Microsoft maintains reliable backward compatability (and I thus exercise due diligence), all is well. So long, and no longer. Just stop allowing yourself to become a victim - changing your platforms and frameworks as often as your socks. Maybe I am a Luddite. 8086 assembly still has a romantic attraction. It's not the ego-pop of using the assembler, though, but a time when work spent learning something new was time well spent - not a fart into the wind. At the risk of alienating any support I have left - aren't some of us really starting to be looked upon as Jack-asses of all trades, master of none? *Ruby picked at random - no actual aspersion (or support) is intended. Really. So don't flame that particular point. Feel free to flame the r
-
I am not sure that this is the best place for support questions (I don't want to upset the moderators), so please use the free online forums for that: http://forums.microsoft.com/MSDN/default.aspx?SiteID=1 Generally though, your understanding is correct. Remember that NetFx3 was going to be Vista-only and after customer feedback it was made to work on XP and WS03 as well... HTH
Thanks for the hint!
-
Thanks for the thoughtful reply John. Please note the smiley, as my comment was meant to be a bit tongue in cheek. IMO, quick subsequent releases are problematic only when they potentially break existing codebases. Even then, it is an opt-in model due to the side-by-side nature of the CLR. Like I said though, the new stuff is all additive (and multitargeting in VS2008 is a huge help). So you need only take advantage of v3 or v3.5 when you actually require it and not a day sooner. As a developer, I think that is a good thing. Happy coding :-)
Daniel Moth wrote:
IMO, quick subsequent releases are problematic only when they potentially break existing codebases.
Well certainly that would be a nightmare of epic proportions, but even if they don't break existing code they break something in the developers frame of mind. What you're doing with all these new releases, many of which contain some better way of doing something fundamental than the previous releases, is creating a mindset in the developer of fuck it, I'll wait until they figure out what they're going to do. In my opinion it would be far better to lay down a moratorium on new .net releases for at least a period of 2 to 3 years. Plan in advance all the features that will go into the next release carefully, ***Consult with the developers who would be using it in real world applications, not hobbyists or students or people with computer science degrees who have never written a commercial application in their life ***, and then give developers time to prepare for it and release it all at once on a longer cycle. In the trenches, real world developers with commercial products like myself need one thing above all else: stability. I have easily over a million lines of .net 1.1 level code with a tiny smattering of .net 2.0 in existing commercial apps. Because of your rapid release cycle I have very large commercial apps that are in use world wide that will *never* benefit from anything beyond .net 1.1 in any meaningful way. You're giving us a moving target to code against. I mastered winforms only to discover that Microsoft wasn't really planning on *that* being the ultimate way of doing UI when WPF came along. WPF is early days yet and with the way things have been going with .net I'm certain somethign else is just around the corner that will make much of what I code in WPF pointlessly outdated so if I'm starting a new app today I'm inevitably going to choose winforms still because I'm betting my business on the technology that I use, not writing little samples for developers and in many cases investing a year or two of time in that project. From my outsider perspective Microsoft seems (for some reason that just doesn't make any sense) trying to cram out release after release relentlessly without stepping back and thinking "are we doing the right thing for our customers the developers who use this code in the real world?". I realize your going to say that it's all opt-in but you're not taking into consideration the psychology
-
The endless stream of new software, whether by Microsoft or others, is a pet peeve of mine. The primary benificieries, when push comes to shove, are the book publishers. In my little work-group, I am generally the last to consider the study of the newest technology. whilst some look for successors for Ruby,* (perhas Cubic Zirconia or Diamel?). Right next to me is, coincidentally, the one in our midst who is most constantly buying new "Programmer's Books". While he talks a good talk, brimming with the latest jargon and correct techniques, I've yet to see much of this be put into any use. When push comes to shove, knowing what you know, and knowing it well, is the source of good, solid, and robust code with real performance and . . . a short development cycle. And quit often, when a solution to some difficult SQL is required, 'they' come to the problem solver (my unabashed primary asset). Instead of spreading beyond, I explored the ODBC drivers for Text and Excel. Wouldn't you know it . . . expanding current knowledge was coincidentally of value to my employer (and cohorts) within a month. Now, I admit I moved on some time ago from C to C++. In fact, though not totally of my own free will, I started using C#, though admittedly with a C++ mentality. The single set of methods in .NET, though a pain to learn at first, is something I bought into. I'm not the complete Luddite. But it is perhaps a weakness that we, as a group, are so easily whip-sawed by the technology-of-the-week. I program for .NET 1.1 because that's what several hundred workstations on this site are currently guaranteed to have installed. I need to be sure that the apps work the first time. I'll jump to 2.0 as soon as I've a customer base. So long as Microsoft maintains reliable backward compatability (and I thus exercise due diligence), all is well. So long, and no longer. Just stop allowing yourself to become a victim - changing your platforms and frameworks as often as your socks. Maybe I am a Luddite. 8086 assembly still has a romantic attraction. It's not the ego-pop of using the assembler, though, but a time when work spent learning something new was time well spent - not a fart into the wind. At the risk of alienating any support I have left - aren't some of us really starting to be looked upon as Jack-asses of all trades, master of none? *Ruby picked at random - no actual aspersion (or support) is intended. Really. So don't flame that particular point. Feel free to flame the r
Balboos wrote:
The endless stream of new software, whether by Microsoft or others, is a pet peeve of mine
You're not alone. The only reason about 50 other people with the same opinion haven't jumped into this discussion is because of the title of the original post. I'm starting to come around to your way of thinking with .net 3.5 coming along.
"110%" - it's the new 70%
-
Daniel Moth wrote:
IMO, quick subsequent releases are problematic only when they potentially break existing codebases.
Well certainly that would be a nightmare of epic proportions, but even if they don't break existing code they break something in the developers frame of mind. What you're doing with all these new releases, many of which contain some better way of doing something fundamental than the previous releases, is creating a mindset in the developer of fuck it, I'll wait until they figure out what they're going to do. In my opinion it would be far better to lay down a moratorium on new .net releases for at least a period of 2 to 3 years. Plan in advance all the features that will go into the next release carefully, ***Consult with the developers who would be using it in real world applications, not hobbyists or students or people with computer science degrees who have never written a commercial application in their life ***, and then give developers time to prepare for it and release it all at once on a longer cycle. In the trenches, real world developers with commercial products like myself need one thing above all else: stability. I have easily over a million lines of .net 1.1 level code with a tiny smattering of .net 2.0 in existing commercial apps. Because of your rapid release cycle I have very large commercial apps that are in use world wide that will *never* benefit from anything beyond .net 1.1 in any meaningful way. You're giving us a moving target to code against. I mastered winforms only to discover that Microsoft wasn't really planning on *that* being the ultimate way of doing UI when WPF came along. WPF is early days yet and with the way things have been going with .net I'm certain somethign else is just around the corner that will make much of what I code in WPF pointlessly outdated so if I'm starting a new app today I'm inevitably going to choose winforms still because I'm betting my business on the technology that I use, not writing little samples for developers and in many cases investing a year or two of time in that project. From my outsider perspective Microsoft seems (for some reason that just doesn't make any sense) trying to cram out release after release relentlessly without stepping back and thinking "are we doing the right thing for our customers the developers who use this code in the real world?". I realize your going to say that it's all opt-in but you're not taking into consideration the psychology
"Opt In" - I love that phrase in this context. Except that, when managers (read: pointy haired guy from Dilbert) hear about a new technology from Microsoft, you will be pushed to use it. It's about as much an opt-in as were eating and sleeping when you were born. The targeted of sales, via slick jargon, of the latest and greatest of what they (management and buyers) neither user nor understand is no less shameful than the Pharaceutical industries snake-oil level ads for their latest (and most expensive) drugs on TV. In a previous position, my employers used as a reason for letting me go is that I wasn't using .NET and their competitors were AND THE CUSTOMERS WERE ASKING FOR IT! Even though they, themselves admitted 'we' had the best and easiest user-interface, they needed to sell the NAME not the PRODUCT. What - you're still not using .NET 6.02E23 ? How will my resturant's customers ever eat knowing their bill was generated with out-of-date software?
"The difference between genius and stupidity is that genius has its limits." - Albert Einstein
-
"Opt In" - I love that phrase in this context. Except that, when managers (read: pointy haired guy from Dilbert) hear about a new technology from Microsoft, you will be pushed to use it. It's about as much an opt-in as were eating and sleeping when you were born. The targeted of sales, via slick jargon, of the latest and greatest of what they (management and buyers) neither user nor understand is no less shameful than the Pharaceutical industries snake-oil level ads for their latest (and most expensive) drugs on TV. In a previous position, my employers used as a reason for letting me go is that I wasn't using .NET and their competitors were AND THE CUSTOMERS WERE ASKING FOR IT! Even though they, themselves admitted 'we' had the best and easiest user-interface, they needed to sell the NAME not the PRODUCT. What - you're still not using .NET 6.02E23 ? How will my resturant's customers ever eat knowing their bill was generated with out-of-date software?
"The difference between genius and stupidity is that genius has its limits." - Albert Einstein
I tend to agree: it sounds like not being with that company was the right result for you - whether by choice or otherwise :-)
-
I tend to agree: it sounds like not being with that company was the right result for you - whether by choice or otherwise :-)
>> I tend to agree: it sounds like not being with that company was the right result for you - whether by choice or otherwise Better than right result - a true stroke of good fortune. 1) I was only unemployed for the summer, a vacation type of which I'd not had for decades. 2) I now work in an environment with several other programmer: stimulates learning and debate. 3) At one point, they had to hire a second programmer to assist my replacement in trying to get a new version of their software completed. They complained it took me 2 years to complete their (non-.NET) Windows version, and my replacement had already done the same in one year (elsewhere). Now, roughly four years later, they're still selling their only product: the version I wrote, with a few minor enhancements. Two programmers unable to replace me - a major laugh, and one of the rare ego-boosts - but mainly a major laugh. The extra cost was directly out of the owner's pockets (a very small company) - but getting a big 2 weeks severence pay after nine years? My inner devil was delighted. 4) They finally had to move from New York to South Carolina to cut expenses - selling the new houses they recently bought (as a result of selling their only product: the POS package I wrote). From my point of view, it's more than the right result . . . good fortune helped me duck a bullet.
"The difference between genius and stupidity is that genius has its limits." - Albert Einstein
-
I agree for the most part with the above reply. Added onto that is you have companies and HR people looking to see what the latest .Net is out there and then place job requirements on that. For example my current job required .Net 2.0, I came in and they weren't even using it yet! They had to order it. Add on to the mess is how unstable the 2.0 is. I have .Net1.1 at home and I am not switching due to the amount of frustertaions and time I have to wait for my windows apps to compile using .Net 2.0.
This is exactly why I wrap absolutely everything. That way I can switch versions without affecting my app code.