Seriously, I think we have been giving this Pogson loon too much attention, and until one day he finds himself writing for a tech-bloid, I am not going to give him needless shout-outs here any more saving this one last time.
“Chuckle… After decades of telling people 'GUI good, CLI bad’, M$ is having to tell the world they were just joking.”
For someone who knows jack squat about MS products, he sure has plenty to say about them.
I must be getting old, but does anyone here remember this little beast called “Windows Scripting Host” from the late 90s, or is it just that Alzheimer’s has finally come knocking on my door? If the broad agenda of MS was to remove every trace of CLI’s existence, then why would they create such a thing in the first place?
Clue me in here, Pogson. By all means.
“I remember being at a place that switched from something to 2000 Server and one poor sap had to sit in the server room for many hours entering account information into a GUI, multiple clicks and lots of typing for each account.”
“net user”, anyone?
And, of course, here is the basic “For” usage for the absolute NT rookie.
——-
Go on, people – try post this FUD entry on his blog and see how fast he blocks you and then changes the subject entirely.
Typical Internet cretin.
——-
Addendum:
Comment poster “Ray” kind of reminds me of Ray Romano frantically flailing his arms in one of those hypochondriac moments in Everybody Loves Raymond:
“Personally, I believe registry cleaners to be a scam. They do absolutely nothing to speed up performance.”
That’s true, Ray. Every heard of this thing called “indexed storage”? The basic working of an indexed storage is like this – the storage keeps track of the location of every data entry in an optimized index (such as a balanced B-tree), and when you need to access an entry, the storage manager simply gets the entry for you by looking up the index and then jumping straight to the location of the entry itself. Every contemporary file system works like this. In fact, even the Windows registry works like this. Since we are talking about doing searches in logarithmic time i.e. O(log n), the mere act of erasing a few entries is most certainly not going to have much performance impact on the storage.
I wonder, though, if Pogson by any chance knows anything about logarithmic time. Or even time complexity in general, for that matter.
Comments
> Every contemporary file system works like this.
Unless it’s Linux.
Here is hilarious piece of reading from 2005: http://ext2.sourceforge.net/2005-ols/paper-html/node3.html
ext2/3 used to use linked lists exclusively
Than that NIH HTree was invented (because reusing existing implementation of B+Tree isn’t fun enough), which is basically same old linked lists, but splitted into buckets. Who cares about scalability, right? Linux is perfect. It has always been.
But… But… You have freedomz to use BTRFS/ZFS
Oh, and nevermind that HTree index is second-class citizen in extfs, so it could be inconsistent with what inodes tell or could be completely removed – nobody would see any performance degradation, because, as I’ve told before, Linux is perfect and will remain perfect no matter what.
“I remember being at a place that switched from something to 2000 Server and one poor sap had to sit in the server room for many hours entering account information into a GUI, multiple clicks and lots of typing for each account”
Because CSVDE, DSADD and LDIFDE don’t exist, right? Nor do all the third-party tools like AD Bulk Import.
RDP obviously doesn’t exist in Pogson-world either if the poor fecker had to freeze in the server room while he manually entered all these accounts.
Pogson’s experience of Windows server is three generations old, and doing it wrong at that.
“But… But… You have freedomz to use BTRFS/ZFS”
No, just use XFS. It can randomly mongle your data, but it’s a B+ tree! So basically, you have a tree and because it can allocate your data whenever it pleases, it sounds to me more like an A+ tree, because it’s so A+.
Also, I like this
“This feature was extremely well received, since for very large directories, performance improvements were often better by a factor of 50-100 or more.”
“Because CSVDE, DSADD and LDIFDE don’t exist, right?”
If you tell Pogson about it, he might think you speak of ADHD or some. Windows has no CLI, everybody knows this!
“But… But… You have freedomz to use BTRFS/ZFS”
No, ZFS is licensed under the evil and absolutely horrid slavery license CDDL (Sun bought the OSI approval for it, I’m sure) which is unfortunately GPL incompatible. So ZFS only if you FUSE. I wonder how FreeBSD and OSX got it?
“Personally, I believe registry cleaners to be a scam. They do absolutely nothing to speed up performance.”
LOL.
“But… But… You have freedomz to use BTRFS/ZFS”
When Btrfs is “ready”, I wonder how many people will avoid it simply because it’s from Oracle (never mind that it’s GPL, and GPL is supposed to “protect your Freedom from evil corporations”).
“ZFS is licensed under the evil and absolutely horrid slavery license CDDL”
Not true. Stallman says CDDL is okay, but you shouldn’t use it because it’s GPL-incompatible1.
“I wonder how FreeBSD and OSX got it?”
Because they’re under a BSD license2, which doesn’t have the draconian restricti^W Freedoms of GPL. And while CDDL is copyleft, it only operates at the source file level, not anything-that-touches-the-code like GPL. The problem here is GPL, not CDDL.
[1] http://www.gnu.org/licenses/license-list.html
[2] In the case of OS X, at least the kernel is BSD-licensed.
And there are also plain COM http://msdn.microsoft.com/en-us/library/ms676723 and WMI http://msdn.microsoft.com/en-us/library/aa389739 interfaces to AD, which are both activatable remotely (even “non-existent” RDP isn’t really necessary).
You must be signed in to leave comments.