Replies

Chet, it is not clear to me exactly what aspect of in-place upgrades you are defending. If your position is that in-place upgrades will complete successfully, I totally agree with you. However, I must disagree with your assertion that “there is little risk”.

There are a number of reasons an in-place upgrade is risky. With an in-place upgrade all the corrupt and orphaned entries and other nasties present in your current registry will transfer to your new installation. Any issues with installed applications will be inherited by the new system. Many of the ‘annoyances’ experienced on your old system will probably show up in your new one. And that’s just a few off the top of my head.

Will an in-place upgrade work? Of course it will. Should you make an in-place upgrade? Only you can make that determination based on your needs. I think of it as an opportunity for a fresh start free of any old baggage.

If any system has issues, annoyances, other nasties and Registry Orphans etc., they are not necessarily migrated with an in place upgrade nor does a clean install necessarily eliminate them – well, perhaps the Registry entries. Any significant problems with an in place upgrade are easily reduced to a level equal to those associated with a clean install simply by abandoning the initial in place upgrade and doing a clean install. There is no mechanism by which a clean install can be made as simple and straight forward as a successful in place upgrade.

I believe most of the reasons behind the practise of going to a clean install stems from practises that predate XP. The Upgrade Adviser does a good job of advising about potential problems and I don’t believe modern systems are plagued with those same problems of a decade ago. At least, I hope Microsoft has learned something and would no longer be so foolish as to turn out either flakey OSs or migration programs. They must be aware that the public will no longer tolerate that sort of nonsense and they have competing companies waiting in the wings just waiting for any stumble. Even when they don’t stumble there are a plethora of pundits more than willing to “create” some imaginary problem along with huge numbers of gullible willing to believe them and blindly repeat the the myths. If there is any company likely to produce flawless software these days it is likely Microsoft. They risk far too much to do otherwise. Yes, yes, I know, everybody has horror stories that they have heard from someone else but one rarely finds basis for these stories or the stories are so old as to be irrelevant.

Anyway, anyone who wants to start from a position that Microsoft software is so seriously flawed that it must be avoided during the installation of more Microsoft software, even though it is process known to require more work and problems than the presumed flawed software in likely to cause is welcome to do so. I just won’t make such a recommendation.

I presume you are talking about a desktop system and not a laptop so my comments are based on that.

There is no good reason to ever use Sleep Mode on a desktop and some good reasons for not doing so. In Sleep Mode, recovery is very fast but any power failure during Sleep will result in loss of data (in open files). Sleep is really designed for a laptop that has a battery so power failures are not a risk. After a Sleeping laptop drains much of its battery, the laptop will automatically save all data and Shutdown.

Desktop systems should use Hibernate Mode rather than Sleep. The system takes slightly longer to go into Hibernate but recovery is just about as fast as it would be from Sleep. The main advantage to Hibernate is that all data is first saved to disk in order to guard against power failures.

The front panel Power button on some/many/most systems (certainly mine) can be configured (in the Power Settings) to invoke any mode including Hibernate, which is how I configure mine. Pressing and holding the front panel Power button (for a few seconds) will override the Hibernate invocation and do a hard shutdown (power off). A brief press will invoke Hibernate if awake or Awaken from Hibernate if in Hibernate. The “Shut down” pseudo button on the Start menu still does a soft Shutdown.

My usual operating procedure is to press the front panel Power button to invoke Hibernate then walk away. On return, the system is in Hibernate and I simply push the Power again to bring the system out of Hibernate (about 30 seconds). During Hibernate, the system cannot be Awaken by any mouse movement or keyboard action.

A word of warning. A system that is in Hibernation can be woken by a special packet from the LAN. Unless you have a need to remotely wake your system, you should turn off the “Wake On LAN” option – especially if your system develops insomnia like mine did. I don’t believe using the front panel Power button is any harder on the system than any other way of invoking Hibernate (or Sleep) although, I guess you could wear out the switch since it is a physical switch.

The direction of this thread has shifted somewhat from partitioning strategy to drive strategy. Several people have mentioned the use of external drives, usually implying USB external drives.

When I was deciding to add external storage, I rejected USB as well as ESATA and 1394 Firewire in favour of NAS storage. The LAN based storage is likely slower but it offers several advantages as follows:
-Storage is always available to all systems on the LAN independent of any one system.
-The typical NAS enclosure provides an FTP server and Print server that operate independent of any one system.
-In the case of the FTP server, the data can be made available from anywhere on the Internet, greatly simplifying travel with a laptop or netbook.
-As a backup location, it provides an additional level of security by totally isolating the NAS drives from any system problems.

I don’t see any reason user directories could not be located on NAS drives but I don’t do so, in part because I think it would be slow, and because I otherwise only infrequently access the drives allowing them to spin down and remain in a standby state most of the time.

Chet,
Based on your years of experience you probably maintain your system in a very appropriate manner. I would suspect you keep apps up to date, run up to date virus scanners and malware scanners, clean out unwanted or unneeded files, etc. You are probably in the 5 % or so of users that do this. Unfortunately, the other 95 %, including many readers of this forum do not do, or perhaps know how to do the things you do without thinking about it. Because of this, an inplace update has many more chances of going awry. Yes, this allows the programs to continue, but see my last reply for my ideas on this.
I firmly believe that for the vast majority of users, going to a pristine system using clean install is the best choice. I have read at least 2 of you that the inplace upgrade worked fine, and I applaud you on doing what it takes to keep your system in top notch shape, but for those others, please take everything you read here with a grain of salt and do your research before running the in place upgrade instead of the clean install.

I seriously doubt my systems are anywhere a cleanly maintained as you would like to think. They aren’t even as cleanly maintained as I would like to think. I will concede that I prepared for the upgrades with steps that I am better able to appreciate than many common users. But, by the same token, I would also be better able to do any necessary post clean install work than would be most common users. I still elected to do in place upgrades because of the inherently simpler process, when it works. When it doesn’t work, the clean install is not significantly impacted.

I’d like to commend you for sticking to your guns. But I don’t believe anyone said you didn’t know what you were talking about. I only asked that you cite your source. Something you still have not done.

I would not expect someone on an internet forum to follow any advice given there unless it is backed it up with factual information from at least one other source. You don’t seem to keen on following the advice to clean install Window 7 for the best results, but only use your personal experience to verify your position.

You’ve been a member here for a month and have made 16 posts to date. How can anyone be expected to accept what you say at face value based on their experience with you here ???

To begin with, I have responded in post #6 with a citation, albeit a rather simple one. I certainly don’t have any additional statistical data to indicate the relative success rates of the two methods under discussion and I seriously doubt anyone has. I have provided the logic behind my reasoning. I have yet to hear any serious challenge to that reasoning. The logic is its own justification. Suggesting I should provide a citation to document research that is very likely non-existent is comparable to me demanding you provide a citation to support the idea that in place upgrades have been statistically shown to be unsatisfactory.

I am sure there are all kinds of opinion out there supporting the idea that Microsoft could not possibly produce a solid migration application but I attribute most of them to the popular fad of “Big Bashing”. In other words, many of these opinions are little more than some pundit pumping his own reputation by claiming he is so much smarter than Big Bad Microsoft. I can most definitely cite numerous examples of similar “Big Bashing” myths including many about Microsoft that, while popular, have proven to be complete fallacies.

My position has been, and continues to be, that there is little risk to attempting an in place upgrade and my experience is that every example that I personally know of supports the idea that the success rate is very good. A clean install starts from the position that there will be a significant post upgrade effort required.

To begin with, I have had experience with this issue. I have earned my living in the IT industry for many years and recently retired as a IT Consultant after many years with HP, most recently as a Performance Specialist on theit hp-ux systems. While my work was never devoted to PCs, they were the most common user interface to the systems I dealt with and I have used and abused them since the days when Bill Gates was producing an OS for IBM.

I have updated both my desktop (Acer M5620 with Q6600 CPU, 3 GB memory and 2 TB internal disk on a network that includes NAS drives, network printer as well as my other PC and even a Wii Console) and also laptop on the same network. Both updates were from Vista and were done as in place updates and there was only one hitch in the entire process that Microsoft quickly handled. While the laptop is a relatively simple system, the desktop has a wide range of somewhat unusual software on it. In addition to the two systems of my own, I have done two of my friends systems although one of those was also a very simple system. I did have (and still do have) Acronis and I had a complete image of my system drive on my NAS drive before starting both my updates along wiith a bootable CD.

So don’t try to tell me I don’t know what I am talking about. Sure there are people who have problems doing an in place upgrade. The ones with problems are the only ones you hear about. I am not looking for any kind of confirmation that an inplace upgrade works but I will point out that all the recommendations for a clean install are being offered from a position that lacks any significant experience with an in place upgrade. I am suggesting an in place upgrade has an excellent chance of working and worst case is a second attempt requiring a clean install. A clean install will always produce a significant amount of post clean install work.

Building cleanly will ONLY improve the chances of getting the OS installed. It will always complicate the process of migrating the remainder of the system.

There is little risk to an in place upgrade and much to be gained. I stand by my recommendation for the reasons stated.

I do agree that a clean install will almost always be successful in installing the OS. But, that is only part of a migration. All ‘upgrade’ users must continue with the additional job of installing, updating/patching and configuring all their applications along with reloading the application user data. All these operations present at least a lot of additional manual work and a worst, a lot of grief regarding re-registering the applications.

The Upgrade Adviser will warn of any potential software that may not be able to be migrated in place. An in place upgrade should result in a fully customized working system with all applications in place and working. If it doesn’t, the worst case is having to then do a clean install. An initial clean install will always require a lot of additional work beyond the simple OS installation.

Considering an in place upgrade is always coming from Vista, there is an excellent chance that it will go smoothly.

I too have heard many experts recommend a clean install. It is perhaps the most certain as far as the OS is concerned but certainly not the easiest in terms of the overall system migration. That is probably the biggest reason the ‘experts’ recommend a clean install. The experts are narrowly focused on an OS upgrade whereas the user is (or should be) concerned about the total migration process.

Can you site any sources to support your contention ??

Microsoft!

Separating your user data will allow you to make less frequent backups of your C: drive since there would be little on it that would change. You could, for example, backup the C: drive monthly while doing a backup of user files weekly or even daily.

In Win 7 Control Panel > Fonts > Font Settings, there is an option to allow fonts to be installed using a shortcut which installs a shortcut to the font rather than the font. This option should speed startup (I guess) but I also guess it might slow some applications.

Thanks Chet. I’m using a 27″ LCD Monitor with 1920 x 1200 resolution which is the highest this one will go (and the native setting).

I have set most of the applications up to suit this but would like to tweak this down by one point if it’s possible.

CK

I also have my monitor set to 1920 x 1200 although mine is a 24″ LCD. I found that I needed to set the Text Size to 150% in order to make the default reasonable. At 100% it appears quite small although readable but I cannot imagine making it smaller. For example, at the top of this window is a graphic of a lizard and thw words Windows Secrets Lounge Beta Formerly Woody’s Lounge. The lizard and words are a graphic and are unaffected by any Text Size setting. They are however affected by the IE8 magnification factor (bottom right hand corner of screen). The normal settings I use are as follows. The default Text Size set in the Display applet of the Control Panel is set to a scaling fgactor of 150%. The IE8 browser magnification size is set to 150%. The Text Size within the IE8 browse is usually set to medium. The magnification factor (bottom right hand corner) of 150% makes the lizard graphic and its words appear a reasonable size. The Default Text size for Windows of 150% makes the Windows text in locations like the Toolbars and Title Bar appear approximately equal to the Words “Formerly Woody’s Lounge” in the graphic. The IE8 browser application Text Size facility allows me to change the working text up and down from the that.

If you really want to make a fine adjustment to the Windows default as set in the Control Panel > Display > Set Custom Text Size then place the hand over the scaling ruler and click and drag the ruler to get any scaling factor you wish.

Before you try to reduce your text size, try increasing your monitor resolution. I suspect you have a relatively large monitor with the resolution set to a low resolution causing your text to appear abnormally large.

What is the make and model of your monitor or what is its size and type (CRT or LCD) and what resolution do you have your screen set to.

You should set the resolution to the native resolution of your monitor if it is a LCD. If it is a CRT, there are generally accepted resolutions for each screen size. Once you have the resolution set appropriately, you can make minor changes in the default text size that Windows uses by changing the Personalize settings in the Control Panel or it may be called something slightly different in various versions of Windows.

Keep in mind that the Personalization only sets the default text size used by the OS. Applications and even web pages within a browser can override the default. In addition, most applications have a mechanism for adjusting the text size in that application.

I believe Readyboost will help speed up the launch of applications (or, at least, that is where it is most noticeable) but the application launches that it helps most are the ones frequently used.

I had asked a question a couple days ago about the location of the stored results of the Readyboost test results. The OS only tests a flash chip to determine if it meets the minimum speeds for Readyboost. If you have several chips, as I have from my cameras, it is worthwhile looking up the Readyboost speed test results to determine the fastest chip. A faster chip will no doubt, deliver Readyboost advantages (to whatever extent they exist) better than a slower chip. The Readyboost test only performs the test once and, if the chip passes, it is never retested – unless you reformat it.

To see the test results and to be certain the results apply to the chip you are testing, format the chip then look at the results as described by Group Moderator Gerald in his response to my question.

“You can see the logs by going to Control Panel>Administrative Toosl>Event Viewer>Applications and Service Logs>Microsoft>Windows>ReadyBoost>Operational>

On the right panel click ‘find’ and type in ‘speed’ to see the speed test results. You can also look at all the logged ReadyBoost events in ‘Operational’.”

The results show both random read speed and write speed. By that they mean ‘random’ locations on the chip. Some chips have a high speed buffer that allows them to be classed at some very high speed but they may still be unacceptable for Readyboost. I believe their tests are done on relatively small file sizes since that is the type of data Readyboost is designed to speed up. Serial access to large files on disk is already very fast so Readyboost doesn’t attempt to speed up that kind of access.

First off, thanks very much for the location of the speed record of the Readyboost tests. I want the speed information mostly for a comparison between the SD chips I have because they are used as memory in my cameras and it is useful to know the relative speeds of each of the chips since that determines the speed of shooting. There isn’t any other (easy) way to determine the individual values for each chip.

I don’t think I agree regarding the usefulness (or lack thereof) of Readyboost however. As Readyboost, it (the extra memory) is a memory cache for disk IOs and as such relieves main memory from the need to supply memory for this task. In that function, it is as good as additional memory in that it effectively makes more of main memory available. The argument that it provides little benefit in systems with 3 or 4 GB of main memory is comparable to claiming that systems with 3 or 4 GB of main memory would benefit little from more memory (and the ability to use it). If this were true for all, or even most, cases, there would be little benefits to 64 bit systems and the applications that benefit from the larger memory potential.

I have 3 GB of main memory and the maximum I can install under a 32 bit OS is 4 GB (as is true for just about all systems) and of that maximum, only 3.1 GB can actually be used since the space above that is reserved. That makes it very unattractive to spend the money (any money) on anything above 3 GB. The cost of Readyboost is not only small but, considering the chip would otherwise sit around collecting dust if it weren’t employed as Readyboost, it is the most effective use I can put the chip to except when needed while actually taking pictures.

Linksys’s PC software is just for diagnostics, not drivers. It doesn’t check the Internet connection though.

If my suggestion of powering the cable modem off and back on doesn’t work, then yes, he needs to call Comcast.

I must agree with the concept of rebooting the modem. Remove the power for at least 30 seconds and that means actually unplug the power connector. While you are at it, I would do the same to your router. These are common recommendations from cable companies and I can say from personal experience that it frequently works.