Because apps derive their own paths from shell folder base locations
when they are installed, it's best to do these relocations before any
apps are installed or files are "opened". Else you can:
- clear MRUs, to "debulk" the reference load
- use Nirsoft's RegScanner to search for old location paths
- replace these with new paths to taste
Yep, apps are beginning to look up and use shell locations like
Documents, Pictures etc. (Vista drops the cutesy "My...", TG) but
that's better than mixing data with its own code files or dumping
things in the root directory, which is what they did before :-(
[/QUOTE]
See I just simply define a partion for Root and a partition for Home and
call it a friggin day. =) Takes about 5 seconds =P
And does this:
- automatically put large lumps Pics, Music etc. on larger volume?
- automatically put small lumps Pics, Music etc. on small volume?
- automatically update static, previously-derived paths?
If it doesn't, you'd have a similar amount of tap-dancing to do. Else
you'd just be moving the base of your account data subtree, which is a
lot shallower than what I'd like to do.
If the Linux architecture is elegent enough for all installed apps to
derive their paths through one-point-of-change variables, then the
last might be a yes, and that would be Very Nice. Vista supports that
sort of thing too (actually, so does XP) but as yet few developers are
using expandable strings for their internal paths.
Actually, expandable strings bring pain of their own, because it may
force you to conform to the OS's duhfault nesting strategy. There's
no "right" answer here, as it goes about what you want...
If you want strong user separation, as corporates and roaming-profile
fans do, then you'd want the present Windows approach of first forking
on user, then on content.
If you don't use user identities much, and would rather gather content
by type (e.g. small data to be backed up) irrespective of "whose" it
is, you'd want to fork first on content type (Documents, Downloads,
Pictures, Music etc.) and only then on user, if at all (i.e. you may
prefer a common store for all users).
Then you may have a piggy app you wish to exclude altogether, such as
The Sims and other games that dump 1G+ player data in the "Documents"
store, blowing out the ability to back this up on one CDR etc.
So what solves some problems, precipitates others.
I wish people would seriously get off this anti "piracy" trip.
Seriously...the only person it hurts is the consumer who runs into a
problem with the anti piracy measures. It does not actually *prevent* it.
It's getting to a point where the vendor's intent on securing their
rights is limiting or even crippling the value of their product -
something that is only possible to get away with if you are in a
monopoly situation. The only practical fix for that (other than
aimless ankle-biting by regulators) is a legal and effective WINE.
I use the word "vandor" (VANdal + venDOR) for vendors who are prepared
to presumptively payload your system on an automated suspicion of
wrong-doing, as self-defined in the EUL'A'.
Let's port that principle to the "real world".
You buy a car, and your car dealership thinks you are not keeping up
with payments, so they automate the disabling of the engine, or
perhaps they disable safety features like airbags (denial of access to
security patches). Or they break into your home to repossess it.
Then it turns out their record-keeping is faulty; you really were paid
up all along. And all they have to do is shrug and say "sorry"; the
small-print "agrreement" you signed to buy the car, gives them the
right to do as they please. This "agreement" was never negotiated
between them and any other entity; it's cooked up entirely within
their organization to meet their needs, and it overrides your local
government's legislation, the Internet being what it is.
And if you want to carry the particular goods you need to carry, you
have to buy that brand of car from that dealer, through that EULA.
This breaks a number of legal tenants common to many nations,
including the US; "innocent until proven guilty", "redress limited to
the goods in dispute", "privacy trumps credit status, only law
enforcement can search and seize", "national solereignty" etc.
MS may slowly overreach themselves in the rights they grant
themselves, with an entire industry greedily following its footsteps.
The eventual response may destroy or severely distort or cripple the
platform, or conditions may become so onerous that one is compelled to
switch to something else, even if that something else is an even worse
monopoly (Apple) or is yet to become fully-assed (Linux for dummies).
As an MVP, I'm prolly seen as committed to Windows, and I am in that
there is where the bulk of my experience, knowledge and value resides.
But my first loyalty is to the platform, not to that platform's
vendor. Where the vendor boosts the value of the platform, there's no
problem; where the vendor screws the platform up, we're in opposition.
I don't even find the Open Office UI to be that much different from MS
Office (before 2007 anyway) and in some ways more intuitive.
Yep. As I say, that UI change is one hell of a gamble, but I
understand why it was necessary... with a far richer feature set, MS
Office runs out of UI capacity far sooner than Open Office, and is
thus compelled to blink first
So I took a Vista + MS Office 2007 laptop out to a client who was
about to buy a new PC, and watched them dive into MS Office 2007's new
UI, cold, with no guidance provided. And they liked it; they picked
it up within a minute, and chose MS Office 2007 over 2003 and Open
Office alike, on the basis they liked the new UI better.
These weren't "computer people", but they weren't dummies, either.
They were a translator and a publisher, respectively, and they eat and
breathe word processing far more deeply than I do these days.
I rarely ever actually use Office...maybe twice a year...literally.
Same here, in these days of pervasive web and email. Before that it
was word process and fax or print, for everything. These days I use
spreadsheets more often, and that's only because all the "proper"
accounting packages I've tried have sucked beyond belief.
Took me seconds to find the margin settings in Open Office...
I spent about 10 minutes trying to find the stupid thing in MS Office...
Was that pre- or post-2007? That's the sort of hassle that 2007 aims
to address, where one is lost amongst a rich feature set.
Everyone says "so drop some features... just don't touch these crucial
ones". The problem is, everyone has a different idea of what the
crucial features are, and as I suspect most features have been
needs-driven, there's not much chaff to cut.
The fact that MS Office had "Page Setup" greyed out didn't help due to the
file being read only, but neither does the fact that it's grouped with the
other printer options in the file menu and not under formatting stuff where
I instinctively looked first (and where Open Office has it).
Yep - that is from the logic that these things are document-wide, and
in keeping with word processing's print orientation, seen as being
more closely linked to Print Preview and paper options than the rest
of formatting, which groups under sub-document style sheet scopes.
We'd think of it all as formatting, in that the actual activities are
so similar; margins, columns, etc.
I personally prefer the power of a commandline for many things. I am a
programmer though so I suppose I may think differently in that regard than
the average user. Even many applications I write, even under windows, if
they are just in-house utilities I generally make them command line apps.
Plus the development overhead is by magnitudes less for me.
Yup - that's why it's so much easier to develop as far as a CLI-driven
engine, then stop. Which is fine - just don't claim to have developed
as compete a product as one that provides a well-designed GUI.
The advantages of CLI over GUI is:
- if you learn to interact, you've already learned to automate
- it's much easier for phone support
The disadvantages are that it's far harder to get started, to get that
first 5% of ability that is often all that basic users need.
So, when I know I need to know something in-depth, I may well be happy
to CLI it. But when there are literally hundreds of different things
I may need to do once a year, there's no way I want to strive for a
50% ability level in all of them, as I may be obliged to do in Linux.
If Linux advocates are serious about <ahem> reaching out to Windows
users, then please, web up a quick how-to for these points of pain:
- how to "Ctl+Alt+Del" a task manager to kill tasks?
- how to write, save (where?) and run batch files?
- how to address drives and devices (hd0 etc.)?
- how to tell what type a file is?
Also UI Apps have the distinct disadvantage that they generally require
human input, making automation difficult if necessary. =)
Yup. This is an advantage, when it is malware that is doing the
automating
I can generally do things much faster with a powerful commandline than I can
with clicking around menus with the mouse.
Not if you have to do it once only, and the 5 secs keyboard vs. 2
minutes mouse clicks is offset by three hours of looking for and
reading manuals or man pages ;-)
Hmm never used Hoary, but I've used both the Dapper and Edgy installers
What vitage are those? I've not heard of either... have Ubuntu really
stayed on track with "new version every 6 months"?
Well yea, there are lots of distro's out there and I am sure there are
plenty of them that suck. =)
Picking the "right" distro (where "right" is YMMV) is an art in
itself... it's also the tribalism that bedevils compatibility in a
good way; by fracturing the target for malware.
If MacOS grew to (say) 40% market share, it would attract the same
sort of pounding that Windows "enjoys" today. With those
more-money-than-skills users, it would be a feed-fest.
OTOH, if the collection of Linuxen were to grow to 40% market share,
it would be a target of disparate parts, much as is the case with the
March 2007 av landscape in the Windows world.
Some by-design attacks would apply to all equally, but the deep and
nasty code exploits would often be limited to a particular distro.
Even if multiple distros shared the same defect, the offsets needed to
get code to run properly (as opposed to a simple DoS effect) would
likely be different - and that's a lot more work.
In fact, it's more likely that attackers would concentrate on
commonly-used edge-facing surfaces than the core OS itself.
I got pissed off earlier actually when I booted into XP to play some supreme
commander that is unfortunately not yet supported by Wine.
Games on WINE will be tough, as DirectX is large, complex, and has to
be really efficiently-coded to work acceptably.
OTOH, Vista itself poses "non-native OS" challenges to existing apps,
made worse by last-minute changes late in the final betas that
ambushed several vendors who were ready to go as late as mid-2006.
Log in...start SC....configure a little skirmish game...Windows decides to
pop up a dialog that it installed an update and gives me the 5 minutes
until it reboots dialog. I know the "reboot later" button is useless since
it will just bug me again in 15 minutes...
I strangle that at birth:
- Settings, Contral Panel, System, Advanced tab
- clear the "automatically restart on errors" checkbox, OK
It's an absurd duhfault, and I dunno what MS persisted this from XP
into Vista. The sort of thing that casts doubts on developer sanity.
--------------- ----- ---- --- -- - - -
Tech Support: The guys who follow the
'Parade of New Products' with a shovel.
