NoelC

Live and learn. I figured for technical things I can't figure out I can ask for help, but for licensing things that's naturally got to be the manufacturer. It cost me more than it should have, but I have to say it's been working perfectly. -Noel

With the vertical RGB bands I actually sense more color fringing on characters rendered with pure grayscale, as well as a bit more of a "broken up" look to them. These are good reasons for having configurability. Configurability that actually works, I mean. EVEN IF it were system-wide, I could set it to use less color (there are three choices) and have a happier medium. And why is it considered impossible that the software could automatically enable the color when the device is oriented a certain way, and disable it if turned up sideways. Oh, that's right, computers are getting less powerful and have smaller storage. -Noel

Thanks. I had forgotten I had posted the other thread on this forum. It's gratifying to know that a few folks actually understand the problem. At least I've gotten one response now that actually answers the question, "does it happen for you?" I realize the subject and detail of font rendering are subtle, but I really expected more folks to contribute to this discussion. Didn't Microsoft announce that Windows 8 was going to be all about perfection in typography? Is that all anything is any more? Say it, don't bother to do it? -Noel

Well, I've done a bit of experimentation... The OpenType font I had on hand to try was Adobe Myriad Pro. Not good for a UI font, but good enough for some testing. Unfortunately, your suggestion really isn't a viable workaround. 1. NO color is used in rendering the font on any monitor, regardless of the ClearType Tuner setting. The use of color is actually good on my center monitor, which is normally oriented and has vertical RGB stripes. 2. Choosing the UI font does not affect all the fonts in the system. For example, it does nothing good for the font Notepad uses by default. Notepad in particular is not important; it's the concept. The ClearType Tuner needs to affect all the fonts that are rendered with color, in dialogs/windows put up by any application. While it may be too late for Windows 8.1, a new version of Windows is being developed. That's where an actual fix needs to be. For me, now, this is only a minor issue because the fonts rendered with inappropriate color on my side monitors are readable even if they're a bit sub-optimal, and I probably use my side monitors more often than anything else for displaying reference info from the web and IE doesn't use color in rendering. This is a subtle issue - one that a lot of folks don't even know exists. But that doesn't mean it shouldn't be done right. -Noel

VERY interesting. Thank you. I will try and see what changing the font can do. Who would think that Microsoft's own font wouldn't exercise the system capabilities to their fullest, especially when IE turns off all the color on them. -Noel

Or just log off and on again. -Noel

What do you mean by that? Please be specific. -Noel

I thought I saw in passing that the size was not small. I don't recall the exact size now, and I'm not sure how to see it after the fact. Looking for further info online, the two updates appear to be a specific fix for blue screens, and a security update for Flash Player in IE. Perhaps they were trivial after all. -Noel

Yes, I've seen that glitch too. Glad you reported it, My1, I had forgotten to do so. BigMuscle, do you have an updated version of your GUI tool? Even with Win 10 I've found the old version you published a very handy thing for helping get the colors and transparencies of everything just right. By the way, there are a couple of new Windows Updates for Win 10, one of which is non-trivial. As usual, the "More Info" link doesn't go anywhere useful, though they may fix that as they often do. The good news is that these updates don't break the existing symbol dependencies of Aero Glass. -Noel

LOL, you sure do have your finger on the pulse of today's high tech. Remember when "heuristics" was a bad thing? -Noel

Your assessment is probably right, BigMuscle. I've personally seen only failures to load new symbols for a short time (as much as a few days) after a new OS release, but always (if I wasn't the first to find out) when someone has said the symbols have become available the next run has auto-loaded them. With KB3000850 I installed it pretty early - hours after it came out - and the first bootup (on two different systems) loaded the symbols correctly. But my troubleshooter mind always asks the questions. -Noel

Big SSDs don't cost much any more... Example, Crucial MX100 - 128 GB for $59.99 US. How much is your time worth? -Noel

Without the update installed, does your system pass a check using the following? SFC /VERIFYONLY I ran across this page, published by Microsoft, which describes ways to reset your Windows Update components, and provides a Microsoft Fixit troubleshooter. https://support.microsoft.com/kb/971058?wa=wsignin1.0 I realize that in your case the update has been applied without error and the problem is in running the system afterward, so I don't know if it's directly applicable. But it at least gives you an inroad to start troubleshooting problems with your servicing database. -Noel

I have heard, from an old Digital Equipment Corporation acquaintance, that someone, somewhere is resurrecting VMS (the OS that used to run on Vax computers). As you may know, the architecture for Windows NT was derived from VMS, which in turn was derived from earlier DEC operating systems. Digital had its technical head screwed on straight. It's a shame they fell by the wayside. Re-developing VMS could actually lead to good things. But again, they'll have to create a subsystem for running Windows applications pretty seamlessly. The chance of that working well is greater than, say WINE, due to the underlying architectural similarities. By the way, to this day I use a Digital LK250 keyboard. The one I'm typing on right now is literally 30 years old, and as wonderful to use as it was the day it was built. No "Win" key and I rather like that. I only mention this to prove that some things done in the past were actually good, and haven't been improved on yet. -Noel

If you're at all serious about that comment, the specific things it has been finding [susp] (suspicious) and blocking were off-mainstream tools, e.g., from Microsoft's old .chm help file creation system or Adobe's Photoshop SDK. Avast's claim to fame is that they've invented some kind of iterative "virus signature minimization" system that they run on super fast computers (possibly even on GPUs) in order to - supposedly - make the virus scanning process as efficient as possible. In real terms, my guess is that they have a huge database of malware, and a possibly even huger database of legitimate software. Their system strives to identify all the malware and exclude all the goodware. Obviously, when the numbers of things checked is in the millions (or more), this leads to the possibility of generating virus signature databases that could inadvertently match legitimate software that they don't have in their database. Notably, upon using their "false positive report" system, the false positive detections go away very quickly - like within a day or two. Human involvement is minimal. It's a pretty cool idea, conceptually, and they obviously pulled it off for years, but the actual implementation seems to be starting to fall apart at the seams. -Noel

I have been using Avast since around 2005. Back then one NEEDED a 3rd party AV solution, and they were one of the first who really embraced 64 bit computing. I tested it against Windows Defender in late 2013, when I first installed Windows 8.1 (at that time I moved up from Windows 7), and I found Avast to have a slight edge in performance (i.e., Avast was better), so I stuck with it. Up to early 2014 Avast has been an efficient system, so I stayed with what worked and I knew to be stable. You get to know a product and its characteristics, and you build up a comfort level with it. However, during 2014 the product has clearly changed (see some of my bullet items above). I don't know if the company went through upheaval or they lost a particularly smart engineer or what, but toward later 2014 the product began to cause problems (e.g., more false positives) and just cost more overhead by being in the system. As a long-time career engineer, I perceive failure inside the company through subtle cues from the changing nature of the product. That Avast completely borked my test system on the release of KB3000850 was the last straw, and that pushed me over the edge to take it out (the only recourse at the time; it took Avast several days to acknowledge and fix the problem). Based on completely restored performance, I'm glad I've made that decision. Funny thing is, because of all the other things I do to keep away from malware, and the nature of how I use my systems, Avast really hasn't blocked any real threats in a long, long time, so my risk in removing it is minimal. Whatever differences in protection between the two (and it may well be that Windows Defender is actually better now) are mostly inconsequential. As a side note: The first full scan I did with Defender after removing Avast turned up a threat on a backup drive. Some Java Runtime 6 component that had never been identified by Avast as a problem. That component is long gone from my system - everything's on version 7 now. It's really too bad for them... I've published two books on configuring Windows, and in each is a whole page where I recommend Avast enthusiastically. That page is going to come out of my updated Windows 8 book. -Noel

On several occasions people have reported trouble with Aero Glass failing to automatically download symbols, yet folks able to succeed at doing it other ways. So I guess the real question is this: Is there a bug in the Aero Glass auto-download process that could be fixed, or different servers that could be accessed, that would make the auto-download process more reliable / robust? -Noel

Sounds like something fun to experiment with, thanks. -Noel

The idea behind ReactOS is cool, but XP-level compatibility really isn't good enough any more. They'll need to take it further. I've heard WINE runs a lot of things directly under Linux/OSX. I have had only the briefest exposure to it myself, some years ago. -Noel

Mostly it's the degrading performance of Avast over the past year (it now slows my systems down about 50% just by being present), but also a lot of other little things that added up to me concluding that the quality of their engineering is slipping. I wrote this up on the Avast forum a few days ago... FYI, on my main workstation I HAVE ditched Avast, as of 2 days ago. I'm happy about the decision as it has returned all my lost performance. Then subsequently... -Noel

I guess I'm guilty as an incurable "early adopter', but no one's invented an essential Metro/Modern App yet (and it may never happen) so it's kind of academic anyway. -Noel

Thanks, but that's unfortunately not an acceptable approach for me for another reason - I have multiple systems networked, and expect my own username to be privileged. Yes, I know I could change everything over to use Administrator, but it doesn't buy me the ability to run Metro/Modern apps seamlessly anyway directly with the one login, so it's a moot point. I'm a bit surprised no one has taken it upon themselves to figure out the hack required to circumvent Microsoft's "Is UAC On" check for running Metro/Modern apps. Maybe no one serious enough about computing to know how things are put together has wanted to run such apps badly enough to figure it out. -Noel

For what it's worth I was trying to use DISM a little while ago on a Win 10 system (which honestly I don't know if even makes sense, but if the servicing database is corrupted I figured I'll try). I was seeing timeouts. It's possible Microsoft's servers aren't working properly or are overwhelmed. They're not running a tight ship over there. -Noel

​Did you try the commands I listed in post 30? -Noel ​ ​ ​ ​ ​

Nice piece of info in the comments... -Noel