How long(horn) blues
How should we allow for all this from the point of view of hardware purchasing? I think the answer to that is simple; namely, ignore it for the time being. For most companies, Longhorn is going to be at least 24 months away in both their thinking and purchasing, and no-one should be making hardware purchase decisions today based on what Longhorn might or might not require in such a long timeframe. Indeed, anything you buy today, even at a reasonably high-end level, will doubtless be completely passé in performance terms by the time the shrinkwrap comes out.
Will the new user interface provide any real business benefits? This is hard to say. I’m somewhat disappointed that the SideBar technology appears to be dead. This would have allowed you to write applications or applets that located themselves in a strip of Desktop space running down the side of the display. I particularly liked the idea because it allowed you to bring together, in an orderly fashion, various information feeds that are important to your work. It was a much bigger space than the fiddly and awkward System Tray in Windows XP and harked back to the modular Desktop concept first presented for Cairo. Apparently, it is been deemed not worth pursuing and has gone to vapourware heaven. I only hope that Microsoft does not view the scattering of widgets all over the Desktop as an acceptable alternative, as this is being tried out in the new OS X 10.4 Tiger version on the Macintosh and I’m far from convinced of its value. Having a fake extra layer does not work when the basic Desktop is 2D. If real work can be done on making the Desktop into a tangible and navigable 3D space, where Z-order actually matters, the Tiger solution might gain some legs.
Oh, and why no screenshots of the new Longhorn build in this column? Microsoft has banned the press – both print and online – from showing any screenshots of this build. Seems a little overcautious to me, but it is a build aimed solely at the device driver community, so pretty visuals are somewhat lacking. Nevertheless…
Executive Software took considerable umbrage at my comments regarding its latest defragmentation product Diskeeper 9. In that column, I criticised it heavily for the scaremongering words used to describe disk-fragmentation problems and their possible effect on system reliability. Executive Software’s letter to me says: ‘I would like to acknowledge the point you made concerning Diskeeper’s warning messages. Your view on this is understood and this section of your article was forwarded to our head office and plans are already in place to have the text modified for the next release of Diskeeper, version 10. Such messages were not intended to scare the user, though I can see how you may have arrived at that conclusion. All that was intended was to simply emphasise the importance of running a defragmenter and therefore improve system performance.’ I’m pleased to see that Executive Software has decided to tone down those dialog box warnings and make them less scary to the less technically informed user.
The letter goes on to complain about my other comment, that fragmentation does not and has not ever been the cause of system problems. It pointed me at a number of Microsoft Knowledge Base articles, mostly from the NT 4 era, in which Microsoft appears to be recommending defragmenting hard disks to improve system reliability. An example is 259421, which reads to me as a powerful example of why you just shouldn’t let a Windows machine run out of disk space, and how to go about recovering a barely workable system if you do. Another Knowledge Base article, 274318, referred to severe fragmentation causing problems when attempting to install Office 2000 onto Windows ME, which of course uses the FAT32 file system. Knowledge Base number 306524 covers how burning a CD can fail from a badly fragmented hard disk.