Microsoft causing lost productivity. A rant.
-
Erik Funkenbusch wrote:
but at the time many people thought it sucked.
No doubt they did, but I didn't. However, it's not just a performance thing but a usability thing and that's got nothing to do with hardware. Help has objectively gotten progressively worse with each new version just based on usability.
Kevin
Not that I'm calling you a liar, but let's just say i'm skeptical because so many people seem to have selective memories. Maybe you're not one of them, and great, but overall I think most people seem to forget the past, or misremember it. I've watched so many people claim "Windows 9x is so much better than Windows NT", then those same people claim "Windows 2000 is so much better than XP", then those same people claime "XP was the best MS release, and Vista sucks", seperated by a number of years of course... but for some reason they just don't remember how they felt back in those days. Even my own mother does this. When I tell her stories as a kid, she'll claim they never happened.
-- Where are we going? And why am I in this handbasket?
-
Not that I'm calling you a liar, but let's just say i'm skeptical because so many people seem to have selective memories. Maybe you're not one of them, and great, but overall I think most people seem to forget the past, or misremember it. I've watched so many people claim "Windows 9x is so much better than Windows NT", then those same people claim "Windows 2000 is so much better than XP", then those same people claime "XP was the best MS release, and Vista sucks", seperated by a number of years of course... but for some reason they just don't remember how they felt back in those days. Even my own mother does this. When I tell her stories as a kid, she'll claim they never happened.
-- Where are we going? And why am I in this handbasket?
I have to agree there. Windows 9x made me want to scream with the crashes, NT and 2000 were continuously plagued by the BSOD. XP was the first stable version I have used (going all the way back to windows 2.1 I believe or what ever just preceeded 3.11) but it requires a little tweaking to get it just right. Now Vista comes out and tries to address some serious security issues, and has to move a little closer to the Unix side of the world and the same people who scream for resolution of their devestating security issues do nothing but complain about the fix. You can't have security without some form of intrusiveness. It just is not possible.
-
Erik Funkenbusch wrote:
If you press the wrong button, how is that Visual Studio's fault?
There are operations that, no matter what, are going to take a long, long time to complete. Hitting F1 accidentally, invoking the refactoring system etc. What would be nice would be a small "Loading..." or "Initialising..." dialog with a big bright "Cancel" button.
cheers, Chris Maunder
CodeProject.com : C++ MVP
Chris Maunder wrote:
There are operations that, no matter what, are going to take a long, long time to complete. Hitting F1 accidentally, invoking the refactoring system etc. What would be nice would be a small "Loading..." or "Initialising..." dialog with a big bright "Cancel" button.
But would that not then require multi-threaded programming, sucking up resources that we are already ranting about? If you click the close button on the help dialog and it doesnt close, just hit alt+F4 and end program now. Works every time.
-
I have to agree there. Windows 9x made me want to scream with the crashes, NT and 2000 were continuously plagued by the BSOD. XP was the first stable version I have used (going all the way back to windows 2.1 I believe or what ever just preceeded 3.11) but it requires a little tweaking to get it just right. Now Vista comes out and tries to address some serious security issues, and has to move a little closer to the Unix side of the world and the same people who scream for resolution of their devestating security issues do nothing but complain about the fix. You can't have security without some form of intrusiveness. It just is not possible.
You have to remember, that each version of Windows was better in most ways than the version before it. In some ways, some considered each new version worse... largely because they required more resources than the last version. Windows 95 was much better than Windows 3.1, but there were people that complained that Windows 95 took 4-8MB when 3.1 worked great in 1 or 2MB. Windows 98 was more stable and better than Windows 95 in many ways, but it needed 32MB which many people complained about. Of course 32MB at the time costed about $500. And so on, and so on... The most complaints have always come from those that "upgraded" without upgrading their platform. Or maybe they did upgrade their platform, but they felt resentful about it. Computing will continue to use up all available resources. As hardware gets faster or bigger, new software will figure out ways to utilize it. I think this is a *GOOD* thing. I mean, back in the P200 days, the idea of continuous voice recognition from almost any speaker without training was all but impossible. Today, Vista comes with such functionality for free. And it works pretty damn well too. Software will always push the bounds of the available hardware. Sometimes it's because of lazy coding, but sometimes it's a tradeoff between better productivity and using more resources. We don't program in assembly today because we don't need to. Maybe the human genome is so complex because god used a framework library ;)
-- Where are we going? And why am I in this handbasket?
-
Not that I'm calling you a liar, but let's just say i'm skeptical because so many people seem to have selective memories. Maybe you're not one of them, and great, but overall I think most people seem to forget the past, or misremember it. I've watched so many people claim "Windows 9x is so much better than Windows NT", then those same people claim "Windows 2000 is so much better than XP", then those same people claime "XP was the best MS release, and Vista sucks", seperated by a number of years of course... but for some reason they just don't remember how they felt back in those days. Even my own mother does this. When I tell her stories as a kid, she'll claim they never happened.
-- Where are we going? And why am I in this handbasket?
What you say is true in general but there are some genuine issues too. With OSes I preferred XP to 2000 from day one - though I certainly recall lots of people complaining about XP's "Fisher Price" UI. No doubt many still do. It never bothered me. Re: Visual Studio, I do think each new version is better overall, especially for .NET developers, despite performance problems. And even then there tend to be a few usability regressions. However, with VC++ in particular the consensus seems to be that it falls short as an IDE. Hence why MS is promising a major revamp for version 10.
Kevin
-
Chris Maunder wrote:
There are operations that, no matter what, are going to take a long, long time to complete. Hitting F1 accidentally, invoking the refactoring system etc. What would be nice would be a small "Loading..." or "Initialising..." dialog with a big bright "Cancel" button.
But would that not then require multi-threaded programming, sucking up resources that we are already ranting about? If you click the close button on the help dialog and it doesnt close, just hit alt+F4 and end program now. Works every time.
No. Plenty of hang-time in outlook is caused by network waits. The CPU is doing nothing.
Dave Buhl wrote:
just hit alt+F4 and end program now
Not much of an option if it's Visual Studio or Outlook.
cheers, Chris Maunder
CodeProject.com : C++ MVP
-
Cristian Amarie wrote:
Armies of developers still uses VC6 (me included). Why? As you pointed out - is easy to use.
Actually, i think it's more because there's billions of lines of legacy code that nobody wants to port to the newer, standard conforming compilers. I also think there's armies of lazy programmers that don't want to learn the "new" C++, complete with templates and STL, which often means that those that DO want to use those "newfangled" features are restricted from doing so because the lazy programmers won't know how to maintain it. Also, of course, VC6 was the last version to adequately support MFC with the wizards. Personally, I haven't used the MFC wizards in almost a decade. I found I was far more productive by learning how MFC worked internally and just hand-writing my code. Get "MFC Internals" by Scot Wingo and George Shepherd. Treat it like the bible. You'll be 100% more productive.
-- Where are we going? And why am I in this handbasket?
Erik Funkenbusch wrote:
Actually, i think it's more because there's billions of lines of legacy code that nobody wants to port to the newer, standard conforming compilers.
Not for me. And the discussion was about *productivity* - specifically about the IDE which is ok, but consumes far more resources (most important, my time). The compiler is indeed far better - but one can also use the 2005 (or Platform SDKs) compiler with 6.0 IDE as well. About the legacy code - it happens to be in such a project for 10 years, with legacy code which was COMPLETELY ported to Visual C++ 2005 while it also compiles on VC6. But when you want to do a simple task such as editing a resource, VC6 comes to rescue. Heck, the development cycle is now like - fix/change/debug quickly in VC6 - add/change resources in VC6 - do the final rebuild and test on VS2005. And this goes on and on.
Erik Funkenbusch wrote:
I also think there's armies of lazy programmers that don't want to learn the "new" C++, complete with templates and STL, which often means that those that DO want to use those "newfangled" features are restricted from doing so because the lazy programmers won't know how to maintain it.
Touché. There were templates and STL also in 6.0 (ok, I know these from 2005 and - again - the compiler is better). C++, new or old, isn't for lazy programmers anyway. Even 6.0 was too much for them. You pointed exactly on how you can identify these "ticks": templates.
Erik Funkenbusch wrote:
I found I was far more productive by learning how MFC worked internally and just hand-writing my code.
You're not using hand-built EVENTSINK maps with ActiveX's (and remember each and every VTS_xxx arguments for every dispatched event handler), do you?
Erik Funkenbusch wrote:
Get "MFC Internals" by Scot Wingo and George Shepherd. Treat it like the bible. You'll be 100% more productive.
I think reading a book will usually make you smarter/quick/capable on dealing with new problems. Productivity - I think - is mainly related not on what you don't know, but on what you know but can't solve easier. Also me I discovered years ago that a corrupted/malfunctioning class wizard can be improved by manually deleting the .clw file. Unfortunately, that's the kind of thing that consumes most of the time. At least for me.
-
Chris Maunder wrote:
IE spends an inordinate amount of time with the little "Connecting" icon when you open a tab.
This one does NOT make sense to me. Whenever I use IE7, I get the feeling I'm waiting for a boat to sink. It's a strange feeling. Incidentally, my original bias against Firefox was the time it took to load compared with IE6. Once I finally started using it, I decided that 2-3 second difference wasn't really important to me. Now with IE7, I can gloat that I don't have to wait 20 seconds for the first connection to work, even if it's the about:blank page.
Chris Maunder wrote:
Outlook is unable to do anything without an obligatory 10 second pause.
This has ALWAYS infuriated me. Why in the WORLD is this application (apparently) single-threaded when EVERYTHING it does can fail or take 2 minutes to time out? WHY??
Chris Maunder wrote:
Visual Studio has lots of nasty little "oops - I pressed the wrong button and will now wait 2 mins
My favorite is accidentally hitting F1 when I mean to hit ESC. Great. All I wanted to do was close the Find dialog box. Now, I have to wait 2 minutes for you to FINALLY open Help, only for me to instantly close it again. Yeah. Those are fun. X|
Chris Maunder wrote:
It's just dawned on me how much of my time I spend waiting for apps to unfreeze.
One of the reasons I spend (possibly unnecessarily extreme) amounts of time ensuring every app I write is ridiculously responsive. I HATE that non-responsiveness and it confuses users who think the app is broken.
It has become appallingly obvious that our technology has exceeded our humanity. - Albert Einstein
You spend in your life: A of your time sleeping B of your time working C of your time watching TV . . . and 1 year waiting for Microsoft products to complete a task :laugh: Why don't we get Microsoft source code and put Linux developers to improve it? Life would be good ;P
_class MySignature _{ __public override void toString() __{ ____return "hi ;)"; __} _}
-
Chris Maunder wrote:
Come on Microsft. How about we have a year where you spend your time making what you ask us to pay $400 for faster, leaner and more usable. Usability doesn't mean more features and stuff like making it impossible to tell the difference between an active window caption and non-active window caption. Usability means it's simple and easy to use and makes us more productive.
Can't agree with you more! I just threw some rather expensive hardware at it, and the shit still freezes up. It's not even connecting to the net or anything important. Just bringing up a context menu in windows explorer is a dog sometimes (but the icons sure look pretty :) ).
- S 50 cups of coffee and you know it's on!
-
So what you are saying is that the developers at Microsoft should anticipate the setup desires and skill level of every user worldwide, write their code to automatically configure every possible setting to satisify the needs of every purpose, include it all in a single package under a single title, operate in every written/spoken language under every culture, all while consuming no resources or requiring the user to think about a single aspect of their own use cases? Users need to decide if they want pure ease of use, security, or whatever and choose the operating system the works for them. Stop bitching, especially if you are a developer. Make your own operating system the meets all your requirements and then market it to the masses. It is not difficult to research how to do the configuration changes if you do not already know and yes you can turn off search indexer and just about every other resource hog that you don't care for in Vista. I have been running Vista since beta1 which barely worked at all and now have Vista Ultimate running on my laptop and never go above 25% ram usage (2gb) and rarely hit the 100% cpu usage with a P4 3.2 machine. Oh and have not had a single issue with VS05 except the initial install which only required service pack being applied and now works like a champ. Take some personal responsibility and initiative and you might find dramatic improvement. Meanwhile, barely computer literate users will have all the functionality they need (and more) and as they learn can change the things that need to be to refine their user experience. I used to hate MS and still prefer Linux sometimes but if you sit back and consider the sheer scope of what Windows is, you have to give credit to MS or you are simply looking for something to rant about.
Or else, get a Mac.:cool:
-
No. Plenty of hang-time in outlook is caused by network waits. The CPU is doing nothing.
Dave Buhl wrote:
just hit alt+F4 and end program now
Not much of an option if it's Visual Studio or Outlook.
cheers, Chris Maunder
CodeProject.com : C++ MVP
Chris Maunder wrote:
Not much of an option if it's Visual Studio or Outlook
Nope but if you are complaining about having to wait for something like "Help" to load after accidentally hitting F1 then it works quick and easy. But a lot of time in Outlook the hang time can be traced back to personal folders bulging at the seams with junk that you will never really read again. Each time you open outlook it reads the index of each attached folder. fwiw
