Over a year ago, I bought a new computer. This was a carefully considered decision, as I don't change computers that often and so try to get something I will be satisfied with for as long as possible.
The machine itself is a marvel: A silent computer, completely fanless (from http://www.endpcnoise.com). It weighs a lot because it's basically a big finned heat sink, with heat pipes connecting to the motherboard. Future upgrades will definitely not be replacing this case (A TNN500AF), but just swapping out processors or boards.
Years ago, in order to run concurrency experiments, I got a dual-processor machine that had so many fans it was like being at the aft end of a jet engine; I even took to wearing ear defenders while programming. Since then I have been much more aware of noise in my computers. My father even built me a special desk with an insulated cubbyhole for the computer, but the ventilation fans for that box turned out to be no quieter (the cubbyhole is now being retooled for conventional drawers). Subsequently, I discovered the fanless machine (which also has two processors, but as dual cores rather than separate boards).
Whereas the agile approach would be to try out one small experiment at a time, I am prone to overreaching. I change these things so infrequently, I think, that I might as well move to the latest and greatest MS OS in the process. This was not so rash a decision as it might seem; the process of installing new software is consuming enough that not doing it twice (once for a new machine, and again later for a new OS) constitutes significant time savings.
Since the old XP system was still working, the Vista experiment on the new box became a side project in which I occasionally dabbled. There were definite incompatibilities: I had to upgrade the machine's BIOS by downloading and flashing new code into an EEPROM, which was fairly straightforward but not really designed for the average computer user. Some drivers had to be chased down. There may have been a hardware incompatibility in the sound card, or the sound card may have just failed at the same time I was installing Vista; in either case the company that built the machine sent me a new one.
Eventually, Vista booted. Over the past six months I've used it effectively for exactly one thing: pulling video from my camera and making DVDs. The program that does this came with my version of Vista (I don't know if it comes with all versions) and it does it quite well; in fact, the only flaw I found in the Macbook was that its video creation program didn't quite recognize my video camera (it saw the camera, just couldn't properly download the video). Leopard will probably work correctly, and in any event XP also has a DVD creation program, possibly the same one.
Also about a year ago, I was setting my dad up with some new software that involved a Java applet for printing postage. His machine, which was ancient but still quite functional, took so long to load and start this applet that we thought it had hung. I decided it was time for a new one, went to Costco and bought what they had, which was a Gateway dual-core Vista machine. It was all preconfigured and preinstalled so it seemed to work fine -- but kind of sluggish considering what a huge improvement in hardware it is over his previous machine. Dad complained about the new UI for awhile but eventually got used to it. One thing I notice is that his disk seems to be working all the time -- Vista performing mysterious tasks in the background. So yes, it works for him. It's tolerable. But it also seems like we're going in the wrong direction; progress either slowed or going slightly backwards. You don't really get a sense of any big improvement in the OS, which isn't surprising since they ripped all the interesting stuff out (the new file system, etc.). The chrome is prettier, but that wears thin fast. And in exchange, you get software compatibility problems.
There will be the inevitable comments to this story saying "why don't you just quit Windows altogether?" Indeed, my experiment with the Macbook has been quite successful and I have taken to traveling with it quite happily (at first I thought I'd need to carry a Windows notebook as well, but that has yet to be an issue). And I've just installed Parallels and XP within Parallels, and that seems to work fine as much as I've tried it. So there's a lot of promise there; I could even imagine just having a desktop Mac (one of those 8 processor ones) and doing all my Windows work inside Parallels.
Linux is also attractive. At Javapolis I heard talk that Acer is releasing a $300 Linux subnotebook with a Flash disk (no hard drive), and it sounds like it will have no hiccups when suspending and resuming. That will be worth buying just to play with.
I'm still tied to Windows because certain software tools only work on Windows. Although OpenOffice is great, its word processor can't handle documents the size of a book (initially it would crash, now it doesn't but it runs so slowly it's unusable). MS Word is still the only word processor that can produce camera-ready documents that are book-sized (Yes, I do know about all the two-step approaches: "just write it using this one tool, and lay it out using this other one." It's only efficient when you say it like that, not in reality).
Camtasia is another tool that only seems to run on Windows, and it has recently become important to me. Then there are all the programs around the edges, both payware and freeware, that have been written for the largest market and not yet ported to the other machines. I use enough of those that it's still useful to have a Windows box running. They may in fact run on a Mac, but now I have this big heavy silent machine (applying the "sunk cost" economic model).
But back to Vista. I kept putting off the transition, partly because I kept hearing about problems people were having. I didn't want to try moving over to the new machine only to discover that several key pieces of software didn't work on it right in the midst of an important project. So I put it off, and eventually came up with the idea of using a Network Attached Storage (NAS) disk as a transition device. All my data files would go on the NAS disk, and if I discovered I couldn't do something on the Vista machine, I could just fire up the old XP box and work on the same files.
Although I got the NAS disk, I (fortunately) didn't go through with the plan.
My brother Todd is a sole proprietor who keeps small business systems up and running in San Diego and the surrounding area. He's on the ground all the time and knows what's happening with both hardware and software. He's a target for canvassing by Microsoft in their attempts to figure out how to get people to "accept" Vista. (They went as far as to try to educate him about how to "market against Linux." Todd shrugs and says "Why do I care about marketing against Linux? If Linux is better I'll install that."). So if there are problems, Todd knows about them.
Vista has problems. A number of business programs require a (paid) upgrade to work on Vista; you can't just reinstall your existing programs. This is after paying for the heftier hardware and for the OS that slows it down without producing a discernible benefit. After that you don't know if you'll have to pay more to upgrade other software. Or worse, whether you'll even be able to get a version of the program at all that will work on Vista.
Years of FUD (Fear, Uncertainty, and Doubt, generally spread by larger companies in an attempt to prevent positive changes fomented by smaller companies or sometimes just groups of people) in the computer industry have come home to roost, this time on Microsoft's doorstep. Microsoft's traditional approach has been to create something buggy in order to hold the market space, then to fix it with subsequent service packs. It's possible that Vista Service Pack 1 will arrive in time to save the day, and if that happens it will certainly be an amazing recovery. But in the meantime, Vista has become a disaster: both as a product, and even more as the once-legendary Microsoft marketing team goes down in flames.
But really, what could those marketers do when the majority of consumers can easily discover that there's "no there, there?" Consider:
http://blip.tv/file/340692 [It must be pretty horrible to be a marketing person when people are producing videos about how bad your product is].
http://news.zdnet.com/2100-9590_22-6220201.html?tag=nl.e540 "Vista, both with and without SP1, performed notably slower than XP with SP3 in the test, taking over 80 seconds to complete the test, compared to the beta SP3-enhanced XP's 35 seconds." [So it looks like Vista SP1 isn't really going to do anything about the speed problem, and we can expect less than half the performance of XP for at least another year.]
http://crave.cnet.co.uk/gadgets/0,39029552,49293700-10,00.htm "Any operating system that provokes a campaign for its predecessor's reintroduction deserves to be classed as terrible technology. Any operating system that quietly has a downgrade-to- previous-edition option introduced for PC makers deserves to be classed as terrible technology. Any operating system that takes six years of development but is instantly hated by hordes of PC professionals and enthusiasts deserves to be classed as terrible technology."
The torrent of stories expressing disappointment with Vista only seems to grow.
Operating systems account for 1/3 of Microsoft's income, and Office accounts for another third. Although I'm (so far) still stuck using Word, I've become very reluctant to upgrade to new versions. In the past, I would upgrade when Word started crashing too badly, but at some point it became reasonably stable. But it's as if there's no more work for the Office group to do, so instead they rearrange features in the products. One version, style sheets are fairly quick to access, the next version they completely change it around so you have to learn and remember an entirely new sequence which is not particularly intuitive. These "improvements" come from their so-called "usability lab" where they are apparently testing out new features using golden retrievers (who, in case you don't know one, are enthusiastic about everything). As a result, I resist upgrading to new versions of Office for as long as possible, because I know there will undoubtedly be a lot of hassle involved. Better to stick with what you know than risk having your time sucked up by another Microsoft experiment.
What's amazing is that the only really innovative operating system that has appeared in at least 25 years was crushed, probably with a lot of help from Microsoft while at the same time they were claiming to be innovative. This was BeOS, and life would have been much better right about now if Microsoft had bought them instead, because it was designed around creating lots of processes, which would be awfully useful as multicores come online. (BeOS has been resurrected in the open-source world as Haiku so there's still hope, although I don't have any sense of the state of that project).
So does this mean all the innovation has gone out of Microsoft? No, I don't think so. For one thing, I think C# and .NET are significant improvements in programming, and what I've seen of Silverlight is quite impressive (Silverlight is included in the upcoming RIA Jam in Ann Arbor, January 14-16); if I knew that I was only going to live in a Microsoft world I would definitely be moving in that direction. In fact, I can only see hope for Microsoft in the OS world if they create an operating system which is basically .NET sitting on a thin layer of hardware glue. That would be a big step forward; innovative, even.
And I recently made a fast pass through cell phone technology, which was largely disappointing when it comes to getting the phone to interact with your computer. In that area, it looks like Windows Mobile is a strong contender.
But as far as my desktop OS, yesterday I reinstalled XP on the silent machine, and it doesn't look like I'm going to be trying Vista again until I hear some really, really good things about Service Pack 2. Basically, I'm done experimenting with Microsoft OSes for at least a year and a half.
> I'm still tied to Windows because certain software > tools only work on Windows. Although OpenOffice is great, > its word processor can't handle documents the size of a > book (initially it would crash, now it doesn't but it runs > so slowly it's unusable). MS Word is still the only word > processor that can produce camera-ready documents that are > book-sized (Yes, I do know about all the two-step > approaches: "just write it using this one tool, and > lay it out using this other one." It's only efficient > when you say it like that, not in reality). > > Camtasia is another tool that only seems to run on > Windows, and it has recently become important to me. Then > there are all the programs around the edges, both payware > and freeware, that have been written for the largest > market and not yet ported to the other machines. I use > enough of those that it's still useful to have a Windows > box running.
If you virtualize your Windows (XP) box with VMWare and let Linux run the processor, you get much better stability overall. The virtual Windows rarely crashes, because it doesn't take much stress just running a few selected apps now and then, and if it does, so what? You just reboot the virtual machine. And Linux just keeps on truckin', doing everything else.
Then there's Wine and its proprietary relatives. They may provide just what you need; the only way to find out is to try.
> In fact, I can only see hope for Microsoft in the OS world if > they create an operating system which is basically .NET > sitting on a thin layer of hardware glue. That would be a > big step forward; innovative, even.
I'm not sure if you're slyly alluding to MS's Singularity project or not, but MS Research has been working on something very similar to what you've described for quite a while now: http://research.microsoft.com/os/singularity/
It's a research OS written in a dialect of C# (mostly) that uses type safety and software isolation techniques. You've got a microkernel (your thin layer of glue), and all "user" processes are in written in managed code (in Sing#, an extended dialect of C#).
Maybe they've added new features, but in the past FrameMaker has been a page layout program, not a word processor. MS Word allows me to create and lay out pages, do automatic index and TOC generation etc. all in one place. Otherwise you're going back and forth every time you make a change.
If FrameMaker is now a full-fledged word processor then it's interesting; otherwise it adds work to the process.
I was once too an anti-vista person. I hated it and hated how most apps would crash and my reliability meter thingie would be at 4-5 I installed SP1 RC and my reliability is now 8-9 programs don't crash as much copying is now faster upgrading the intel drivers for USB I now copy/paste to and from my 250 USB drive at 20-30mb an increase of 15mb before SP1 RC. I use Linux also but sadly wine can't run most programs and you lose speed when using a VM on big things like VS or Cad. This is my 5 cents. Criticism is welcome.
I think Vista is a good indication of MS fall from dominance in the OS arena. (I know this has been predicted many times before) The reason is not because of Vista's issues its because MS is blind to the fact that Vista solves NOTHING.
What do I get for 'upgrading' to Vista? Desktop search? Aero? Supposedly better security? (We have definitely heard that before).
All for 100-300 dollars? Gee thanks. Throw DRM in for free too.
The very idea that MS believes that doubling my RAM and upgrading processor and HDD for Vista is a benefit proves they are losing touch with their customer base.
On a quick sidenote regarding word processing and layout: I've found learning LaTeX to be far and away worth the initial investment to learn it. It produces gorgeous layout, by and large obviating the need for direct intervention. Despite this, it gives you very fine grained control when needed. I use it for everything longer than a couple sentences, now. . .On Windows, MikTeX is easy to get up and running, and on Mac, MacTeX.
I used Latex for my very first book (Computer Interfacing with Pascal and C). The problem with any system of indirection, where you write code and then process it, is that you must constantly verify that everything is correct, and there's no way to perform that test other than visually.
The reason I moved to a WYSIWYG word-processing system where a single program does everything is that when you make any additions or changes, that's when you're most focused on it and that's when you're most likely to catch errors. Any multistep process provides more room for errors.
I, and my coauthors, build all kinds of testing into books, to constantly verify that the code you see is correct. But when the only way you can test for proper layout is visually, anything other than "change it/see it/verify it" adds errors.
> The reason I moved to a WYSIWYG word-processing system > where a single program does everything is that when you > make any additions or changes, that's when you're most > focused on it and that's when you're most likely to catch > errors. Any multistep process provides more room for > errors.
I'm not writing books but I write a lot of technical documentation for work and open source projects (the latter is more of a 'need to' than a 'have done'.) What I would really like is a tool that combines a good WYSISYG document editor with a good display-agnostic document format. In other words something that lets me write the document and render it based on my preferences immediately without tying me to that format and layout.
The big benefit of this for me is that a lot of technical documentation is basically a form document. Being able to specify a basic schema for this allows the technical information to be stored in a way that can be easily transformed into many different forms and searched in a very direct way.
For the time being I've been using XML and transforming it into DocBook and then to PDF or HTML. This works but it's a little ad hoc. It's not something I can hand to someone that doesn't know a lot about XML, XSLT and a little about FO, DocBook and HTML and expect them to use it. I've tried using XLMMind in the past and it's OK but didn't quite meet my expectations.
I can appreciate what you say about LATEX. I found the constant recompilation bothersome, even for smaller papers and reports. There was a brilliant word processor that worked around a similar principle of adding content to structure rather than trying to beat content into structure as an afterthought - but it ran on OS/2 and I forget it's name.
And that brings me to my one point of agreement with you and my one gripe with Microsoft: Their total domination of the market combined with their insistence on re-selling the same product for 20 years, means that we are still looking at the same inferior OS and the same terrible word processor, that we where 20 years ago.
I myself can not run Vista - make fun of my tin foil hat if you must, but Vista coming out after MS and NSA had chats, combined with all that "activity" on the hard drive...basically you don't have a clue what your machine is doing and who is on it.
I am using preinstalled Vista on Dell laptop. So maybe because I got computer built for Vista I didn't have any problems (I must note that I don't use printers, scanners etc. with this laptop). In my opinion Vista is first Microsoft OS built with security in mind, and Microsoft should be praised for that. Yes it's slower than XP and some programs like Windows Media Player run slow for no apparent reason (although I suppose it is fault of built DRM in Vista), but it is more secure than previous versions of Windows and that's why I recommend it to all ordinary home computer users who ask me for advice.
"MS Word is still the only word processor that can produce camera-ready documents that are book-sized"
??? Last time I had a problem with either editor was when I edited a document with a 70 pages long table. Word (I think it was 2000) was not able to open it - it simply hang on opening the doc. OOo opened it without a glitch. Indeed, if I used page layout, it sometimes lagged big time before updating the display, especially when scrolling a lot, but I found out that I can edit it without problems if I'm not using page layout, then switch to page layout only for a final check.
> I used Latex for my very first book (Computer Interfacing > with Pascal and C). The problem with any system of > indirection, where you write code and then process it, is > that you must constantly verify that everything is > correct, and there's no way to perform that test other > than visually. > > The reason I moved to a WYSIWYG word-processing system > where a single program does everything is that when you > make any additions or changes, that's when you're most > focused on it and that's when you're most likely to catch > errors. Any multistep process provides more room for > errors.
I beg to differ. IMO, the more structured, organized a process is, i.e. the more small, quite well defined and isolated steps you do, the less errors are likely to survive. Especially in large documents, I often fix something at page x, then find that the fix introduced a layout problem at page y, fix it there, only to find that a new problem occured at page z. I surely would like a tool which does all these fixes for me, instead of me always having to check manually - and LaTeX does this in an excellent way.
> I, and my coauthors, build all kinds of testing into > books, to constantly verify that the code you see is > correct. But when the only way you can test for proper > layout is visually, anything other than "change it/see > it/verify it" adds errors.
Basically, you just said that programming in anything other than a scripting language (i.e. without compilation) adds errors :-)
IMO, parsing the LaTeX input files for code, and performing automated tests on LaTeX source should be at least as easy as doing this with Word. In fact, LaTeX has been used before for literate programming (does CWEB ring a bell?). Which should be something quite useful for the type of books you are writing, if I can judge this correctly.
LaTeX was purposefully developed to implement the wisdom and science of many many typographers. I won't bvelieve you if you tell me you have acquired this wisdom and science (nor do I think you will pretend to have done so). This wisdom and science is not pure convention, nor is it just hot air, it's knowledge that many generations of typographers have built up, and which helps lay out a document on paper so that it is most easily read and comprehended, no matter what the content of the document is.
As such, IMO it is quite likely that a document laid out by you contains at least some layout errors which make it harder to read than if you'd simply let LaTeX do the layout for you. Even things which may look sub-optimal to you may actually be right, but if you don't ask professionals, you will never know.
You know many people and travel a lot. You might want to ask an old professional typographers (there are very few younger typographers left which actually do know their craft) about the layout of documents you create in Word.
Alternatively, you may want to have a look at http://www.webmonkey.com/design/fonts/tutorials/tutorial3.html, so you can understand why neither OOo nor Word are proper tools for doing layout. Using LaTeX, however, is likely to allow you to produce high quality layout even if you don't involve a professional typographer.
Flat View: This topic has 28 replies
on 2 pages