My Internet is flaky, but will try to get this uploaded…
Images from Thurrott.com (thanks)
Where can I find great software for Windows?
Browse your file system like in the 90s!
Win32 and misconceptions
There has been a fear campaign targeting Win32. Microsoft is responsible for it. This site is adding fuel to the fire that Win32 doesn’t belong in a “modern” world. Let’s look at common misconceptions surrounding Win32.
One can argue that downloading programs from the Internet isn’t the safest way to get software. I can agree with that. So let’s put Win32 programs in controlled environment, a Store. This is what I always thought would happen. I’m still surprised it hasn’t happened.
If you want “modern” Win32 program management outside Enterprise you are limited to “hobby” solutions such as Chocolatey. Chocolatey gets the job done, but it isn’t something a regular Windows user installs. Steam exists for games so the lack of a Win32 application store is surprising. After all Win32 is the world’s greatest software platform for any form of productivity, at home or at work.
What makes people afraid of downloading Win32 programs from a controlled environment like Windows Store? I don’t get it. Well-written Win32 programs don’t slow down your system. In a store they would be checked for malware and annoying installers would be a thing of the past.
Let’s address sweeping accusations of “Windows rot”. I have never experienced Windows-rot, simply because it isn’t real. What is real are identifiable problems related to programs, drivers, Windows updates or Windows itself. Malware also exists.
Many “Windows rot issues” stem from OEM images. Often junk-filled with lots of unnecessary programs and processes. Starting with a clean Windows iso is the best way to never get caught by “OEM rot”. If you lose special drivers from the OEM image you can download them manually. Just make sure you have the network driver before you wipe your disk clean. OEMs have been a disaster for Windows’ reputation. Some machines are barely usable out of the box. Some Lenovo consumer machines can’t be cleaned, because Lenovo will reinstall crapware through UEFI when you reboot. I’m not making this up, but business ThinkPads seem spared.
So say you experience “Windows rot”. What to do? Well, start by looking in Windows Task Manager or use the excellent program Process Explorer. What’s eating your CPU cycles? Also look in Task Manager for start-up programs which may be legit programs, but don’t need to run all the time. Ccleaner gives you a better view of start-up programs, scheduled tasks and even browser extensions.
If a program is problematic, upgrade/downgrade/reinstall or uninstall it. Same goes for Windows updates (and drivers). If it is a System process causing trouble, use services.msc or a GUI tweak tool to disable it or set it to manual start. Black Viper (blackviper.com) has a list of services for every Windows version which are safe to disable.
I have rarely seen a Windows problem (in 20+ years) that couldn’t be identified (and solved) by basic troubleshooting. Google is there to help you. When you have identified what is causing “Windows rot”, just search for it and Internet will help you solve it. My Windows machines tend to become faster – not slower – because I optimize them more and more the longer I use them. Windows Update can be a problem, but waiting a week or two gives Microsoft the time to pull a bad update. That’s why manual control of updates is crucial, not only for Enterprise.
The problems we have discussed are related to Windows, but not particularly to Win32. Some people like to blame Win32 for “Windows rot”, when in fact Win32 is just an API for writing Windows programs. Sure, programs may cause problems, but bad programs, drivers and Windows updates will always cause trouble. Win32 programs installed from a Store would be just as tested as Windows updates, wouldn’t they…?
This may be one reason for Microsoft’s Win32 resistance. MS doesn’t want to test every Win32 program so instead they write a “UWP-sandbox” that isolates the program from the system. It saves MS a lot of work (developers have to do that work), but also enforces handicapped apps. This is the “lazy approach”. Bad UWP apps can find their way into the Store because not much testing is done. MS is confident that these apps can’t cause harm anyway.
Compare this “lazy approach” with small teams of Linux distro maintainers who build and test 40K – 50K packages every release. There is no reason MS can’t have 500K well tested Win32 programs in the Store including automatic testing of all the updates. MS could even put a few restrictions on the programs like Apple does in the Mac App Store. For example no auto-start, all installed files go into a specific folder, even not touching the registry… Portable programs would fulfill those requirements from day one, just needing the Store to put an icon in the start menu. I think developers would support Windows Store if Windows Store supported their Win32 programs with minor modifications.
Conclusion: There are no sane technical arguments against Win32 which can’t be solved with minor modifications and a Store. It’s just politics… and unwillingness to actively curate a Store. Steam can do it and so can Microsoft.
skane2600
<p>"even not touching the registry"</p><p><br></p><p>I think the idea of the registry as a big bad is just another misconception. It plays an important role in Windows and one that can't be achieved through using a per-app repository instead.</p>
skane2600
<blockquote><a href="#254956"><em>In reply to hrlngrv:</em></a></blockquote><p>The question is whether the leftover data in the average person's registry is actually significantly harming performance or just irritating the delicate sensibilities of developers. </p>
skane2600
<blockquote><a href="#255238"><em>In reply to hrlngrv:</em></a></blockquote><p>But I assume that registry "cleaners" are scanning the entire registry, that's not what normal programs that access the registry do.</p>
skane2600
<blockquote><a href="#255298"><em>In reply to hrlngrv:</em></a></blockquote><p>It sounds like neither one of us really knows how it's done, but obviously it would depend on how many stubs are there and whether the resulting time delay is human perceptible or not. When people complain of Windows slowing down, I doubt that it's because of how long it takes to do an insert object in Excel (not a very common activity, IMO). Of course, it's not as if every uninstalled program leaves stuff behind in the registry anyway.</p>
skane2600
<blockquote><a href="#255350"><em>In reply to hrlngrv:</em></a></blockquote><p>I think this is perhaps a Raymond Chen question. We don't know how any of this is implemented internally in Windows. </p>
skane2600
<blockquote><a href="#255701"><em>In reply to hrlngrv:</em></a></blockquote><p>When it's impractical or impossible to gather the data required to come to a conclusion, you simply can't. </p>
skane2600
<blockquote><a href="#254677"><em>In reply to paul-thurrott:</em></a></blockquote><p>The problem with the topic of Windows rot is that there seems to be little if any technical verification of its existence. Just anecdotal statements that "my computer slowed down after awhile". It shouldn't be that hard to do a technical investigation that identifies precisely what the causes are and how they affect system performance. Unless the slow-down has some specific cause unique to Windows on Intel, Windows on ARM or any other approach won't "solve" the problem.</p>
skane2600
<blockquote><a href="#256100"><em>In reply to hrlngrv:</em></a></blockquote><p>Windows S went a long way toward being the leisure version but the devices that included it were overpriced and sales (apparently) were dismal. Now Microsoft, going back to their Windows 8 playbook, want to force it on all Windows 10 devices, at least on installation.</p><p><br></p><p>But as others have noted, an OS not called Windows would be a better basis for a scaled-down OS.</p>
skane2600
<blockquote><a href="#254734"><em>In reply to TheJoeFin:</em></a></blockquote><p>I don't think a car analogy is particularly apt, but what you suggest would be like a car maker who doesn't allow the car to be driven faster than 40 miles per hour and limits where the car is allowed to drive. </p><p><br></p><p>If MS wanted to create a brand-new OS without a misleading name like "Windows" that would be fine. There are already plenty of computing options like Chromebooks for those folks that have trouble seeing over the steering wheel.</p>
skane2600
<blockquote><a href="#254857"><em>In reply to TheJoeFin:</em></a></blockquote><p>I don't see UWP as a platform that has "lots of benefits". </p>
skane2600
<blockquote><a href="#254961"><em>In reply to TheJoeFin:</em></a></blockquote><p>I don't see how UWP apps are easier to distribute and monetize than Win32 apps.</p><p><br></p><p> "lightweight and fast" is marketing talk, IMO. All that matters is the relative speed and power between an UWP app and the equivalent Win32 program. I can easily believe that programs with scaled-down capabilities (regardless of platform) take less resources than their more capable counterparts. </p><p><br></p><p>The vast majority of Win32 programs don't have updaters and the few that do often can control the update mechanism. </p><p><br></p><p>Win32 programs can be installed just as easily as UWP apps. As simple as downloading a single EXE file and running it. Of course some programs have more complicated installers often to allow a default setup that would otherwise have to be done by the user by browsing through menus. </p><p><br></p><p>Touch and other controllers are handled properly in Win32 programs as well. As a practical matter there's really only one UWP target, the PC.</p><p><br></p><p>Yes, I'm a developer, but only since 1981.</p>
skane2600
<blockquote><a href="#255085"><em>In reply to TheJoeFin:</em></a></blockquote><p>If you don't use the Store it's not as if you have go through an elaborate procedure every time you want to sell a new program. Anyone sophisticated enough to write a non-trivial program isn't going to have a problem creating a simple website and there are many payment systems out there that aren't any harder to setup than setting one up with Microsoft. How many serious programs use in-app-purchases? You see that most often with smartphone games, and UWP has little or no presence on smartphones.</p>
skane2600
<blockquote><a href="#255208"><em>In reply to TheJoeFin:</em></a></blockquote><p>Not all Win32 programs require an installer/uninstaller. For Some, install: download the single EXE, done. uninstall: delete the single EXE, done.</p><p><br></p><p>Obviously some Win32 programs require more setup but often the setup is required to accomplish something that UWP apps can't do anyway.</p>
shameermulji
<blockquote><a href="#254817"><em>In reply to skane2600:</em></a></blockquote><p>"If MS wanted to create a brand-new OS without a misleading name like "Windows" that would be fine."</p><p><br></p><p>What they should have done and what I still think they should do.</p>
skane2600
<blockquote><a href="#254794"><em>In reply to doofus2:</em></a></blockquote><p>The problem is that even these more modest changes would still impact compatibility of existing programs. Most likely anyone still using Windows 7 isn't really worried about the alleged problems that later versions of Windows (and UWP in particular) are supposed to solve.</p>
skane2600
<blockquote><a href="#254982"><em>In reply to doofus2:</em></a></blockquote><p>It's hard to determine how easy it would be to update to a new system that's just a vague hypothetical idea. For, example, I have a program that requires being able to automate Office through scripting which doesn't use macros within Office or Office documents. It also requires access to information in a central Registry. I'd think any "safe" version of Windows would make it impossible. Besides many programs work fine but will never be updated.</p>
skane2600
<blockquote><a href="#254949"><em>In reply to hrlngrv:</em></a></blockquote><p>I mostly agree, but MS was already on the road to UWP before 2015. If you look at their development timeline they started working on their "One Windows" approach at least back as far as the Windows 7 mobile days (whatever they were calling it). I think they have been and continue to be following Newton's First Law with respect to their strategy and so far no external force has acted on it.</p>
skane2600
<blockquote><a href="#255429"><em>In reply to wright_is:</em></a></blockquote><p>The future viability of any technology product is based on its market and business relevance not its technical relevance. The latter is as much about fashion as it is advancement.</p>
skane2600
<blockquote><a href="#255679"><em>In reply to wright_is:</em></a></blockquote><p>Seat belts were no great tech advance and airbags are flawed both in design (they inflate too fast to protect people not wearing seat belts at the expense of those who do) and often execution (metal fragments that injure people).</p><p><br></p><p>But to compare the danger to people in a car accident to the "danger" of having an insecure OS (which in the case of Windows, you exaggerate) is absurd unless that OS is the basis of some potentially dangerous device. </p><p><br></p><p>I've spent years developing medical devices and we certainly don't just throw in an OS or other code and just assume that it would be OK. We test s**t out of it.</p><p><br></p><p>If you are concerned about computer safety you should look at self-driving cars that are on the streets with no safety or testing standards (at least in the US). Apparently states are so excited about the financial aspects, they don't think safety is that important.</p>
skane2600
<blockquote><a href="#255779"><em>In reply to Jules_Wombat:</em></a></blockquote><p>It's tech people who are fixated on things like the ill-defined "Windows Rot", not normal users. </p>
skane2600
<blockquote><a href="#255975"><em>In reply to jimchamplin:</em></a></blockquote><p>"The rot is real, and it's part of how Win32 software installs and runs."</p><p><br></p><p>Care to elaborate? For example, please explain how Win32 programs on Intel processors "run" differently than UWP apps or MacOS apps run on the same CPU and how that difference translates into so-called "rot".</p>
skane2600
<blockquote><a href="#256057"><em>In reply to Jules_Wombat:</em></a></blockquote><p>"Most consumers are and have left Windows for general computing needs."</p><p><br></p><p>Not sure what you mean by "left Windows". You mean most non-business Windows users have thrown away their PC or never turn it on anymore? The fact that people use smartphones doesn't mean they have "left" Windows. I watch a lot of things online but it doesn't mean I've "left" TV.</p>
skane2600
<blockquote><a href="#256053"><em>In reply to longhorn:</em></a></blockquote><p>I think a lot of the time people talk about "Windows rot" as just another way of saying "Windows Bad". Not only don't they want to bother to figure out what's going on, finding a specific problem with a third-party process undermines the idea that something is inherently wrong with the OS.</p>
skane2600
<blockquote><a href="#256084"><em>In reply to hrlngrv:</em></a></blockquote><p>I think another reason driving initiatives like UWP is that programmers in general and less experienced programmers in particular really hate maintaining old code. They want to create something new and that desire is not necessarily associated with business acumen. </p><p><br></p><p>Combine that with the lack of historical perspective on the failure of "universal" platforms and you end up embracing the latest and greatest attempt (currently PWA, unless the next one is going to be proposed tomorrow). It seems the futility of the pursuit has to be painfully learned by each generation of developers.</p>
skane2600
<blockquote><a href="#256140"><em>In reply to hrlngrv:</em></a></blockquote><p><br></p><p>"I doubt maintaining UWP code would be any more appealing than maintaining Win32 code."</p><p><br></p><p>You're probably right in the long run but the UWP code was "new" for those who created the platform and it's a lot "fresher" than the code that implemented Win32. As I'm sure you know, the longer a code base is actively modified the harder it is to maintain.</p><p><br></p><p><br></p>
skane2600
<blockquote><a href="#256147"><em>In reply to hrlngrv:</em></a></blockquote><p>Got it. I probably didn't make it clear what I was referring to originally.</p>
skane2600
<blockquote><a href="#256281"><em>In reply to Jules_Wombat:</em></a></blockquote><p>But those same "normal" people don't think about services and the registry as a problem that needs solving, so they won't see moving to Store apps as a solution (whether it is or isn't). It's not as if Microsoft is promoting store apps by saying they get rid of those nasty services and that nasty registry. </p><p><br></p><p>What you don't seem to realize is that the world of Win32 software development isn't limited to familiar major applications by major brands and new Win32 programs have been written and will continue to be written. </p><p><br></p><p>Naturally new platforms have more new development than old ones because they have greater gaps in coverage. The "age of complex software" can end only when the need for powerful software no longer exists. Simplicity and power have always been traded-off in software development and nothing can change that. It's fundamental.</p><p><br></p>
funnyjokes
<p>I regularly visit your site and find a lot of interesting information. </p><p>Not only good posts but also great comments. </p><p>Thank you and look forward to your page growing stronger.</p><p><a href="http://flipdiving.online" target="_blank">flip diving</a></p>