16-bit ADDRESS bus and 8-bit DATA bus to be exact. And yes more than 64K is doable on such a system (172K seems a bit odd-sized though--192K would make more sense as that is exactly 3 64K banks). The CPU just needs a bit of help is all. It's called BANK SWITCHING and it was common in those days. Essentially multiple banks of memory would be wired to the same address lines, then the output enable pins would be wired to be active at different times depending on a register. That's how my Z-80 based Coleco ADAM could work with up to 4096K--in assembler you use an OUT instruction to set a register to set which little piece of that you want to use (I think it was actually 2 32K banks or 4 16K banks you could select).
Bank-switching was a bit awkward because you couldn't "see" all the memory at once (you certainly couldn't do anything like calling outside of your current memory segment reliably) but you could do a lot with all that memory nonetheless (mmmm...ram disk)
Hand tuned a floppy drive with an oscilloscope?... how and what the fuck for?
Well, specifically with a system like the II+ such a thing was quite possible. I used II+ machines at school as a youngster but I am not that familiar with that hardware. However I know that Apple peripherals were "dumb" (not in a bad way--just very elegant and simple from a hardware standpoint but very reliant on programming). IIRC the timing of the floppy was very closely related to the system clock of the computer, so if you could get the drive to spin EXACTLY the right speed it could read a track in the correct number of CPU cycles all the time. If it was "out of sync" you'd get read errors (kind of like "buffer overruns" and "underruns"). I'm guessing that is where the 'scope came in--comparing the pulses from the rotating drive with system clock signals so they matched up right.
My systems of choice were Atari XL and Coleco ADAM which had serially-connected "intelligent" peripherals that were rather less sensitive to timing. Less hacker-friendly however. Anyways, compared to today's machines even they were blissfully simple. Even though those machines couldn't do as much it sure was nice to actually know HOW they worked. Oh yeah--and even though the ADAM had a reputation for being "very buggy" when it was first released, the final "R80" revision machines were quite fine--and even the original "very buggy" machines crashed far less often than Windows--and I never had a virus on any of the 8-bit machines I ever owned...ever.
...but I won't. Rother I'll explain why you have no clue.
Insightful, how about idiotic. What can you program in Unix that you can't in Windows.
That wasn't the point the original poster was trying to make. The point is HOW you program in Un*x vs Windows. Nobody will argue that you can do anything you want with either platform. However, a great many people would argue that the "UNIX way" is FAR mor elegant.
In Windows you have C and C++ just like Unix. Java, Perl they are all there as well.
This statement really demonstrates your inability to comprehend the differences. To extend the "building toys" analogy, C/C++, Java, Perl et al are NOT the pieces, they are the plastic/wood/metal with which the pieces are made. You could make lego bricks out of the latest space-age carbon fibre composites, but they would be useless if the "bumps and holes" on each brick were different sizes and wouldn't lock together.
Now the platforms may be different but largerly they are more similar than not from a progammers ability to make a program perform a required task.
There I'd really have to disagree with you. There are things that Un*x style architectures do easily that are arduous to perform in the Windows environment. Similarly, there are things Windows excels at. IPC was really much more refined under UN*X--some might say Windows works with threads so well because it has to since its IPC abilities have historically sucked--really in UN*X it is much easier to get various components to play nicely with each other yet keep their resources separate and protected. OTOH, there are reasons Windows-based games are so far ahead besides simple market share--graphics interfaces are one of those "funny shaped blocks" in Windows that is very well suited to its task.
Really that Lego analogy is very apt indeed. UN*X is very uniform in how it works, just like a bucket of classic Lego bricks. You have a library of pipes, sockets, shared memory etc. that is very standard across all programs that extends all the way to the user interface (you can pipe all manner of programs input and output together right on the command line to a degree not yet seen in production releases of Windows). Once you get the hang of the UN*X Way you can snap these blocks together esily to suit your needs.
For all the "object orientedness" of Windows, there is not that level of uniformity in interfacing to make those reusabel objects work together. Instead, you have an overly complex framework in the form of DDE/OLE1/OLE2/COM/DCOM that was largely designed to accomodate disjointed, inconsistent interfaces between various components/applications. This is something like the "licensed from the movie" Technics sets with all the little odd-sized rods/axles, funny-shaped blocks, special wheels and so on. There many little sets where the pieces fit together very nicely in a few commonly required configurations, but when the time comes where you want to make your own creation not in the instruction booklet you become frustrated with the useless pieces. For many kids, the six or so really cool things you can build are good enough, for the 10% "most geeky" kids it would bore you quickly.
I can't say I really know for sure what a "mainframe toy" would be--mainframes don't seem like fun at all. I think "mainframers" may have forgotten what childhood was like, or perhaps hatched from a pod fully grown, who knows. I do not have a lot of exposure to that philospohy/culture. If I HAD to pick a toy that was most mainframe-like I might say Mecanno, because like UN*X they are fery uniform in structure, however you have tediously fiddle with those little screws to put anything together, just like a mainframe--you have your "special screwdrivers" (arcane knowledge) and have to follow tedious processes to get things done. Or, perhaps it is like building a birdhous with popsicle sticks, where you have to tediously glue the pieces together with Elmers glue, wait for it to dry bef
Simply stating there is "nothing wrong" with the security model of DCOM with the only supporting argument being that a single one of a large number of DCOM attacks was related to a buffer overrun bug and not a design flaw is pretty weak. If it was simply a matter of buffer overruns and memory leaks then these nasty problems would all go away once the implementation was fixed with MS03-039. They are not going away. Years after blaster there are still critical fixes being is
Sorry, I know ALL TOO MUCH of what I speak. The software my employer maintains and uses historically relies very heavily on DCOM and they are dedicating a LOT of resources to completely eliminate depencency on DCOM over the next couple of years. The hard truth is that DCOM security is unintuitive and overly complex. I'm not just talking about the protocols or low-level models around security. I am saying this from a top-to-bottom view. For example, the tools MS provides with the OS to administer DCOM are very lame (DCOMCNFG.EXE really really sucks--it is usually not clear what account should be allowed to do what thing with what component from the crappy interface). Even though the need to use DCOMCNFG is limited to quite a narrow segment of users there sure are a lot of questions, discussion and confusion around it.
Also, default security behaviours are DESIGN characteristics of a system rather than IMPLEMENTATIONs. Those default behaviours in DCOM (implied access to all components by any connection by default) happened to be so flawed that MS altered them in XP sp2--it was so important that we all had to deal with the broken apps that resulted from the changes ourselves.
In my experience it is also quite tricky to effectively secure DCOM servers without relying on domain security, meaning you need to establish a dedicated machine for a domain controller if none exists.
All in all, actually, it seems DCOM in general is very convoluted to use for no good reason whatever--I don't profess to understand it well at all myself but I'm forced to use it. That in itself is a design flaw because 3rd party software developers and end users who do not understand DCOM well enough can very easily open security holes.
DCOM simply gives me headaches. It gives everyone headaches, and I've never...ever...heard anyone say anything along the lines of "DCOM totally rocks--it is the best!". Usually "supportive" comments are along the lines of "it works much better than it used to" or "it works fine it's just that no-one uses it right". Mostly I hear "almost no one needs it and if you totally disable it your system will be safer". Hardly a ringing endorsement and hardly a solution to corporate/enterprise users who must live with it.
Uh, I thought there was public consensus on this that packaging software is the whole point of the "distribution" concept! What is it you think distributions are supposed to do, if not this?!
The problem is that there is no standard so the whole point is completely lost! It's fine and dandy to say RPM is a standard packaging method but I can't just say (for example) I don't like SuSE's Apache RPMS so I'll just use Mandrake's instead--they are both RPMs sure, but if I try that you'll get mired in "dependency hell" and you have to know what you are doing and override dependencies to get it to work. The same potential for hangups presently exists around Debian-derived distros as well--you can't always count on stock Debian packages working in derivative distros or vice versa.
I have no problem at all with distro makers maintaining its own repository of packages. You're right--that is what those organisations exist to do. The problem is that you have to depend on that ONE SINGLE REPOSITORY FOR ALL THE SOFTWARE that you want to manage properly in the system. If Linux devotees want to knock Windows off its pedestel it'll have to work on that.
I don't think compiling from source is a hack at all.
Well, that is your problem--from a user's perspective it is VERY MUCH a hack. Compilation is a normal part of the development and release process--it is a normal part of a PROGEAMMER'S use of a computer. It is NOT something an end-user/sysadmin does or should be expected to do. Even "source-based" distros like Gentoo or Linux from Scratch have tools like Emerge or ALFS to automate the compilation process with a low-level "packaging system"--essentially making the compile process the same as a slower version of a binary package install from the users viewpoint. Even then LFS concedes it is a distro NOT meant for everyone.
Even if you cannot see how compilation from source can be viewed as "hackerish", you certainly cannot dispute that doing so to install software on a distro based on RPM or DEB binaries leads to a very "hacked" situation--you install something from source either because you cannot find the version you want or it isn't available at all as a binary package. Later it IS packaged and other packages are dependent on it. You either have to remove your custom-compiled version and replace it with the RPM or you start down the dangerous road of --nodeps.
It's all more work than it needs to be. If I don't like the job SuSE is doing I want to be able to replace their packages with Fedora equivalents seamlessly for example. If I find some nifty niche-app on SorceForge it would be really cool if they could maintain a binary package themselves and not have to worry that it is too specialised for my distro publisher to bother maintaining.
Not using IE and using Firefox instead almost completely secures an up-to-date Windows box.
Dumping IE is a good step to take in improving the security of your Windows PC, but to say that one action "almost completely secures an up-to-date Windows box" is a dangerous oversimplification of the problem.
Firstly, despite the legal disputes surrounding the strategy, Microsoft has deliberately engineered IE into its OS. Windows 95 and NT4 and previous versions had no dependency on IE at all. On an "up-to-date Windows box" it is LITERALLY IMPOSSIBLE to completely "get rid of IE" on your system. You can remove the icon, make Firefox your default browser and so on, but IE remains in place. You cannot remove the IE rendering engine or any other "meat" of IE contained in system DLLs or you break a multitude of applications and important parts of Windows itself. No matter how hard you try to avoid it, you still need IE to do something as basic as keeping your system up-to-date and run your basic applications properly.
All that needs to happen is for an ambitious group cyber-terrorists to commandeer and taint one single system of servers--the Windows Update site--and the world's IT infrastructure can be brought to its knees. I know saying "all that needs to happen" understates the difficulty of pulling such a thing off, but it IS possible--and the point is that Windows Update is a very serious potential single point-of-failure. Even though Windows update is a huge site run by many many computers, it is still accessible through a single network address and maintained by a single company and operates the same way for everyone. The fundamental concepts behind Microsoft's Windows Update are seriously flawed and without constant vigilance on the part of Microsoft it could be the most serious vector of attack in the history of computer security.
Another fundamental design flaw of Windows from a security standpoint is OLE/COM/DCOM/Microsoft's RPC implementation. Microsoft themselves have acknowledged this with its efforts in Longhorn to create a new service-oriented programming framework for distributed applications (Indigo). Even in post-blaster 2005 there are still reported flaws around DCOM. Ever since OLE was introduced with Windows 3.1 it has been more convoluted than required to do its task, and even with this added complexity it was not designed with the highly connected world of today in mind. Eventaully COM would come out with the still klunky OLE2 interface built on top of that, and distributed application functionality would be tacked on with DCOM. Holding onto a foundation that had quickly become rickety for this long was a grave mistake. MS should've started pushing everything and everyone away from this whole kludgy mess ten years ago when it started becoming clear that the network would be central to computing.
Thankfully, there are limiting factors to the whole DCOM disaster in that home users don't really need the "D" part of it at all, so you can disable it in the registry and/or block its ports with a firewall. Unfortunaely, that doesn't fly in the corporate world as there are a lot of client/server products that rely on it to function (think ERP, industrial automation, custom integration systems and so on). This is why corporate adoption XP sp2 and 2k3 Server sp1 were not at all rapid (so much for the "up-to-date" part of your argument). Those service packs close up much of DCOM and break corporate apps. Thus, those updates are not rolled out until procedures are available to make updated PCs work with existing systems. Guess what? Those procedures generallyinvolve UNDOING some of the changes made in those SPs to secure systems!
I'm sorry, the headline of this article putting Windows at par with ANYTHING in terms of security is unconvincing to me. While it is true that there are some Windows systems out there that are better secured than some Linux or BSD and it might even be true that overall the implemented systems out there are equally secur
That these guys advertise this compatibility speaks very ill about their distro, IMO
Well in *my* opinion the majority of people do NOT share your opinion about compatibility. I don't think I've ever heard an average end user of PCs (or any other product for that matter) say "hmm...this product interoperates with multiple standards...that means it sucks. I'll pick this other one that's only half as compatible with other people's stuff--since they didn't think of interoperability it must be better at other stuff anyways".
The fact that multiple methods of package management are supported by a Linux distro doesn't mean it is "hackerish"--it doesn't HAVE to anyways. It means the distributors recognise that there are two prominent standards and that end users demands require both. Being debian based this distro's repository will be DEB only. There is nothing "hackerish" about wanting to have 3rd party programs distributed as RPMs "just work" even if the system is natively Debian-based.
Like it or not, RPM is the leading format in the enterprise space--it is the native format of Red Hat, SuSE and Mandrake. Support for RPM is also a requirement for Linux Standards Base (LSB) certification. I and many others consider LSB compliance to be a requirement. RPM is not going away and people just have to learn to accept that. Similarly, the Debian format has its own merits and with so many distros being Debian variants it is a defacto standard that RPM adherents have to accept.
It might not be the most "idealistic" or technically elegant thing to do, but supporting both major package management formats is the most professional and business-sound decision to make. Linux will never, ever achieve dominanace if every distributor has to maintain its own packages of every popular application demended by its users. People will not switch to Linux if you either have to live with the packages maintained by a single organisation or be TRULY "hackerish" ahd compile parts of the system from source or shoehorn 3rd party packages into the system with things like "rpm --nodeps" and crap.
Think about it...one of the biggest complaints about Microsoft is "vendor lock-in", where nothing works quite right unless you buy into MS end-to-end. Well, if Linux distributors followed your mentality then they risk falling into that same trap. I want a solid, stable, easy to manage OS on which to base my systems. I actually DON'T want the distro comapny to put so much time and effort into its package repositories. The folks responsible for Apache, Sendmail, Postfix, MySQL, PostgreSQL etc should be able to make LSB compliant packages that I can just click on and install. I really don't care if I can't get every package I need from the same people who put together the distro--I just want to be able to get it SOMEHWERE and have it work!
Anyways, I wish the good people behind this new enterprise linux all the best, and hope that their talk of "compatibility" is substantial.
On the Amiga, data types were handled as plug-ins. If you needed the capability to open a jpeg you added the data type.
That would be great if it worked that way on Windows. MS could go ahead and include all the plug-ins it wanted, and if you didn't like IE or its rendering engine you could "unplug" it and "plug in" Gecko. OR, if you aren't processing JPEG or MPEG data at all you could just unplug those extensions.
That isn't the way Windows works however--there is absolutely no consistent interface for handling data and no logical delineation between components. What stemas me is that much of this wasn't done for legacy/compatibility/technical limitations--it was done DELIBERATELY. IE USED to be a distinct component where Windows would run without it just fine. MS DELIBERATELY mixed it right in so its dlls sat right along Kernel, User, GDI and so on and encouraged its own developers and others to add as many hooks and ties as they could. Then they went and did that with WMP. Now there is an astonishing amount of software that ties into these libraries for different reasons and so on, and there is no published/advocated interface that makes IE and WMP libraries "pluggable" with replacements.
This is analgous to hardware systems that have these super-integrated motherboards with sound, video, IDE, network, kitchen sink etc all soldered together. If you think the sound is crappy you can add a card, but you still have the built in stuff sucking power, generating heat and hanging around to be a potential nusiance or point of failure. Such a thing is tolerable for a disposable home or office PC but there are legitimate reasons for avoiding that sore of integration.
If MS wants to sell its Windows platform as suitable for anything you could use the hardware for, then it MUST re-architect it so that its components can be decoupled, much like it is still possible to buy hardware (with thinkgs like video and sound on separate cards).
MS can throw whatever they want onto the CD and into the installer if it wants to--hell, that is what Linux distros do too. The point is that the components are SEPARABLE. You can do a custom/"expert" install of Linux and choose not to install a web browser or media players or even a GUI if you are savvy and have special needs, or you can just hit "default workstation" or whatever and it will put a GUI, web browser, media player and even the office suite (or several of each of those!).
That is not what I have a beef about--I'd be perfectly happy if MS had a "one click" install that puts all the toys on--hell they could even put MS Office into the Windows install for all I care. What I do NOT like is that there is NO WAY to avoid installing all or parts of the GUI, IE, WMP etc etc..even if I don't NEED or WANT those things. I'm NEVER gonna play.WMF files on my mail server thank you. Also, what if I like alternatives and will never use the "default" components? If I use another media player, another browser or GUI environment all the factory MS cruft just wastes processor cycles, memory and hard drive space.
And your comment about Apple starting it all...well you're right...and that is a big reason I never got a Mac. There are times when I LIKED the command line and scripts and stuff and it frustrate me to no end not to at least have that as an OPTION. I DID have an Atari ST and that was similar, but even with the Atari you could shut down GEM and use it with a CP/M68K command line environment. And now even Apple has seen the light, at least partly. MacOS X follows the UNIX philosophy underneath. I still get the impression the GUI is welded on a bit too tightly but I think with PCs being more complex and interconnected that a hyper-monolithic architecture is outdated and unsuited to todays needs.
The actual Media Player program is just a front end for Windows' media layer. So programs can place calls to it to play media, without having to implement their own codecs and playback mechanism.
THAT is what the whole beef is about! THAT is the reason the American DOJ had the brouhaha over IE "bundling" as well!
The problem really has NOTHING AT ALL to do with "including" WMP or IE with Windows--MS should be allowed to do so. "Bundling" is a really misleading term when you think about it because what MS has done is not "bundled"--it has engineered what were once distinct applets or applications into libraries and integrated/hooked/merged the "meat" of these applications into the OS. Thus, when you open an IE browser window or the WMP applicaiton the.exe is just a thin shell with not much more substance than a desktop icon.
This "melding" into the OS, or transforming "applications" into "OS components" is how MS gets things over the "Chinese Wall" they supposedly keep between applications groups and the OS development groups. THAT is what damages competitors.
RealPlayer and FireFox have to load up their own libraries and so a lot more under-the-hood work to launch because the REAL parts of their applications are not "system libraries". Ether they have to to pre-loading tricks at startup, which makes the system startup time longer, or the user has to wait for startup longer.
Furthermore, users cannot DISABLE the loading of IE or WMP components as they are "system components". Other apps depend on IE's rendering engine, you need it for Windows Update, and WMP libraries are used by all sorts of games, etc. to play video. You cannot even disable pre-loading of them on normal startup to my knowledge.
The whole end result is that MS's own bundled "applications" (really just thin shells) seem to work just that much smoother than the competitor, and for all practical purposed are not removable without altering functionality. Thus you have big huge monolithic bloated OS. And now MS is paying the price because they made strategic decisions to tie in all these features for convenience and market domination all at the expense of efficiency, stability and security.
The MS way of things is the antithesis of the UNIX/Linux way of things. I do not like that a certain amount of web browser and media player and GUI CRAP just HAS to be installed on domain controllers, file/print servers etc etc. With a UNIX like OS components are much more discrete--I can more easily configure the install to include just exactly what I need/want and no more or less. With Windows that is absolutely impossible right now.
The result of MS "bundling" is more complex and far reaching than RealPlayer and Netscape whining about having to push their products harder than MS to get people to use them. After years of designing around this philosophy MS has spread sickness throughout the global computer ecosystem.
The website is not standards compliant. In fact it just as broken if not moreso than Slashdot:
* There is no DOCTYPE--it just starts with an HTML tag
* There are UPPERCASE tags mixed with lowercase tags. This is bad practise and strictly speaking all tags in HTML4 and XHTML are supposed to be lowercase.
* There are proprietary attributes in the document (ms_positioning)
Besides spweing forth really crappy HTML, the techniques the authors used to do the layout are exteremely poor. The site is the exact OPPOSITE of semantic web design. Remember 1996 when websites were all becoming GIF Jigsaw puzzles? That is almost what bungie did. I see empty DIVs, no-break-space characters and some javascript malarky.
and as far as position:fixed goes, at least FireFox supports it even if it is slow...IE DOESN'T SUPPORT IT AT ALL. What you see on Bungie is the aforementioned javascript fakery. Don't give me static about FF bugs--it is LIGHT YEARS ahead of IE in terms of correct support of standards.
Bungie very obviously didn't get an award for good coding practise. The site may be architecturally correct but the coding as abysmal. Furthermore, the site COULD have been made to work far better on FF than it does not WITHOUT FANCY WORKAROUNDS. As a matter of fact, if they DID use position:fixed and other CSS positioning and layout techniques properly it would work extremely well. It would not, however, look as sharp in IE6. Understandably, since Bungie is a Microsoft property, they'll use Microsoft platforms and development tools, and test much more extensivley with IE6.
As for the case study touted by MS, it makes some very important points about how to build the architecture of a site. Any suggestion that you need.NET technology to make it all possible is bunk though. You build just as scalable sites using Apache and mod_perl on a Linux or BSD box (and it might not even be as processor intensive either). You might need to be more disciplined, but you can employ page caching, separation of logic and presentation, etc. with a Free softweare stack quite readily.
Where the ASP.NET solution excels is in its IDE and library of components--Visual Studio.NET really does kick ass there. It's just too bad that the output to the browser tends to suffer when developers get caught up in a fancy design (not only does the HTML look ugly and break standards, it results in inconsistent behaviour). I hope the next version of VS.NET has addressed this.
Apple wants to sell pretty, uncrackable, all-in-one, gold standard computers. They've been doing that since the 1970's and they will continue to do so.
They haven't been like that since the 70s--from 1977 to 1984 Apple was opposite in most ways: Its computer was prettier than everyone elses but it was NOT all-in-one and it was VERY crackable/hackable: Pre-Mac Apples had a whole bunch of slots (with well-known specs) and the early ones even came with all the schematics!
I'd say Apple wasn't so proprietary until Woz left, and it might not stay that way if someone like Woz were to rise to a position of influence in the future. It'll be snowing in Hell before that happens though.
Apple knows that there is competition out there, they don't ignore it like other companies
Of course Apple is painfully aware of the competition. However, Apple has historically been king of the "not invented here" syndrome. They in fact DO ignore the competition as much as possible. That is why Apple is not the company fighting anti-trust lawsuits today. As you alluded to, Apple is not an idealist champion of Free software--Jobs is just smart enough to recognise that it is the best way to tap into a huge pool of development talent and to be the most responsive to the user community.
Apple wants you to run whatever software you want, on their PC's
Not really...they TOLERATE you running whatever software you want on your Mac. They WANT you to run THEIR OS, THEIR productivity software and THEIR hardware. To achieve Jobs' nirvana experience you need to buy into the whole package. However, being a niche player in the PC industry it HAS to tolerate users who want to tinker with Linux or Windows on their hardware. However, those options "weren't invented here" so expect to be ignored by all official Apple support resources if you inquire about ANY third party hardware or software.
Don't you think linux development and customer support is going to skyrocket when there is just a few configurations to develop for?
No I don't. Apple's strategy is to keep variation of both the hardware AND THE SOFTWARE configurations to a minumum. Supporting Linux would dilute their resources and hinder the success and progress of OSX.
Maybe Apple will be the first large computer manufacturer to offer a choice
It can't be the first because some other hardware vendors already offer a choice between Windows, Linux and no OS installed. Given its history since 1984, Apple is likely to be the LAST to "offer a choice". The original Mac had no slots (despite the engineers having tried to "sneak" one in) and was not expandable and there was not even any developent tools/languages available to the general public when the Mac was first released (BASIC was an extra-cost option released some time after the Mac itself--unheard of in the 80s when BASIC was burned into the ROM of every other PC out there).
From the start, the Mac was NOT about choice--you did what it let you do...no more and no less. The Mac was about usability--about making computers more humane. Think appliance: When you buy a stove you can't get dozens of different types of burners, racks, knobs and lights to customise or upgrade it. You get your Kenmore ZX123 and it cooks your food and just IS the way it is. If it breaks you get replacement parts that are just like the old ones and except for the colour and trim that match your decor it is just like every other Kenmore ZX123. Not only is there NO choice, you don't ever care. It cooks your food and that is it--you don't even remember or care it is a ZX123 model unless you need to fix it.
That is what Jobs' and Raskin's vision was with the Mac--make it pretty, elegant, useful and humane. Expansion slots and OS configuration choices and build-to-order and source code just doesn't fit anywhere in that puzzle. All that stuff is the domain of their developers and engineers. For A
False, Microsoft tries everyting that may remotely work, then toss the things that don't
While this is certainly true right now, MS did not have the resources to employ such a scattershot strategy for most of the first decade of its existence. It was only after MS/PC-DOS entered wide distribution in 1981 that MS could make any big high-risk/questionalble bets without causing undue harm to the company. I'd argue that it wasn't until the mid 80's that MS had a venture fail badly enough to be abandoned: That would be the MSX platform.
In any case that is actually a reason for MS business success. It goes to great lengths to diversify and knows when to drop unsuccessful ventures.
As I write this, the post I am responding to is modded "Troll". Seems it is more "Insightful" than anything else.
BillG is a smart guy who surrounds himself with smart guys. MS started out of a motel room in New Mexico and didn't become a near monopoly solely out of luck any more than it did out of sheer creativity. MS is huge for a few important reasons:
1. They recognise opportunities and make maximum use of their resources and connections. MS became the "king of programming languages" for micros inthe 70s and 80s because BillG immediately saw the potential of the Altair and the desperate need for a friendlier method of programming the system. They also used their connections and networking skills to arrange a meeting with IBM re. DOS as well as to locate and purchase QDOS.
2. They are a bit sneaky--they will sell something they don't even have yet (DOS) and create demoware/vapourware to stall and kill competition in a field where they are lacking (GO/pen computing/etc--"they might have it now, but big ol' MS is gonna have it REAL SOON NOW"--yeah right).
3. THEY RESEARCH THEIR COMPETITON--MS has historically been very paranoid. Even with their position today they view EVERY competitior as one who could destroy them. When MS plays in a market they research EVERY LITTLE THING about that market and EVERY COMPETITOR. BillG himself operates that way. If he meets someone who has something interesting to talk about but BillG knows nothing about it, BillG will spend every waking moment for a couple days learning about the subject. The next time he meets that person he can talk with that person like he is a seasoned expert.
It is for that third reason why MS has a whole department of Open Source Specialists in its employ and has had for years. It is also how the Halloween Memos came to be. It doesn't matter how badly MS slags Linux or how much it scoffs at Free software--it has ALWAYS apporached it as competition with the potential to destroy Microsoft. I'm willing to bet it's been on BillG's personal radar for a decade already (when MS was just starting to realise the Internet was a game they had to play).
So the parent to this post is exactly right: MS is essentially "stockpiling ammunition" for the battle with Free software. How they will use the knowledge and people they acquire could go many ways:
1. They could use it to make their FUD sound more credible--for example, some weakness in Linux to exploit in the "get the facts" campaign or items to avoid or downplay where Linux has the advantage.
2. MS operates by acquisition, not innovation. They might have to avoid GPL code to keep its code secret, but it can at least steal IDEAS from GNU apps. It is also already well known they've lifted BSD code many times. This is OK though--at least MS software gets better as a result.
3. If they CAN'T beat Free software, they'll be prepared to "join" it. It may be a cold day in hell before MS Office is open, but if Linux meets or beats Apple's market share and all indications are that it won't go away, MS will be prepared to form a "Linux business unit" to port Office and other apps to the platform. It won't be "Free/Libre", but if MS dominates application software for Linux it can steer the platform and continue to be the industry's biggest player. This is what they have done with Apple--Microsoft is the biggest vendor of MacOS applications, and stunning industrial design aside, Macs are becoming more and more PC-like.
Yep it would be nice if MS was more cooperative, but it is just too far from their business model. MS NEEDS software to be closed or else it would have to completely re-invent itself. It is simply easier for it to try and make the competition work to its advantage or simply go away.
Now that's the dumbest thing I've heard today, especially for software that commands less than 10% of the market. Give me a break.
Actually it isn't dumb--it is completely true. You alluded to one of the main reasons some sites are still broser-specific--they have decided that the cost of changing their legacy, non-compliant code is more than the cost of lost potential business from ten percent of the market. In essence, they in fact do NOT want your business because they have deemed you do not have enough to offer them as a customer for the perceived effort it takes to cater to your needs.
Weather.com has worked fine in all production versions of Firefox. Of course your theory that sites (other than MS-owned) would "deliberatly" write code to break a browser is downright asinine. WTF would they do that?
I too am puzzled about the original poster's experience with weather.com--perhaps it is becasue he uses an old release, or maybe it is the presence or lack of certain plug-ins, or an overall configuration problem. It doesn't mean you have to be insulting. As far as his "asinine" theory goes, it has been proven to happen. In some cases it is sneaky, and in other cases it is very deliberate (a stupid move IMHO, where a site sniffs your broswer and displays a "sorry you aren't running IE--buhbye" page).
Broser sabotage is done to compel users to another browser, either because (as in the case I refence here) said company has a vested interest in the success of a specific browser, or (in the case of a lot of intranet/private sites) tech support wants to ensure a homogenous client operating environment.
Seriously, remove your lips from the crack pipe every once in a while and get a grip on reality.
Seriously, remove the cactus from your rectum and learn to relax...you might find a little insight in what people say, even when it doesn't seem to make sense.
And word to the shortsighted designers out there who STILL make public sites that break standards and rely on platform-specific behaviour: you are doing a disservice to all who use the internet. It doesn't matter if a new PC is $3000, $300 or $30, if it requires more care and feeding than an exotic pet and spftware does not behave with a reasonable degree of consistency from one machine to the next, then PCs will NEVER reach their portential. Consumers do not need a monopoly, but they DO need software developers of all stripes to play by the same rules (USE THE STANDARDS PEOPLE!).
Also, don't count on MS and IE being dominant forever--the machines on the leading edge of the ultra-low-cost market have Linux factory-installed, and even if some form of IE remains the dominant browser, MS has shifted efforts to security at the expense of compatibility (it had to becasue its platform has been flawed from the start). I expect that IE7 or possibly whatever the equivalent of IE8 will be will break more than a few IE6 sites just like XP SP2 broke a number of applications.
It's one thing to charge for an fork...but without food, that cutlery is just worthless hunks of metal and plastic.
That argument makes no sense at all. For one thing, people CAN and SHOULD be able to offer free and "Free" food. My mother obtained strawberry plants--at no cost--form a fellow parishoner many years ago. They have since thrived and MULTIPLIED and now mum's strawberry patch is too large for her to manage. She put many plants in the compost but gave some to me so I could grow them in the back yard. The ladies out there also sell fresh strawberries, pies and preserves at farmer's markets in addition to giving away stuff to friends.
There you have it--the biological version of "peer-to-peer file sharing". There is no law against the "unauthorised duplication and/or distribution of strawberries" that a large corporation can use to restrict individuals or potential competitors from making strawberries. International dumping laws are the closest thing, and they only apply to governments and large companies who sell at a loss to deliberately eliminate a competitor.
Microsoft is trying to be like Monsanto--who has new breeds of oilseeds and grains and is trying to use the patent system to restrict and control a natural process. Neither is right in doing that. Microsoft's solution around CIFS is like Monsanto saying "we'll publish the information on our roundup-ready Canola, but that co-op of Saskatchewan farmers or that university research group cannot ever make a royalty-free version to compete with our variety".
In both situations I think this is damaging to society--especially in ther case of agriculture. Food is much too important to be the proprietary domain of a molopolistic enterprise. While software is not an important basic need for people, it is nonetheless a vital part of the global economy, so I believe Micorosft's proposed remedy is entirely insufficient. It holds a monopoly over a protocol that enables information exchange important to the functioning of most business systems. MS has demonstrated in the past it cannot be trusted with SMB/CIFS as it has made changes to deliberately break reverse-engineered systems. Microsoft should've been ordered to relinquish control of the protocol to a standards body like ECMA/IEC/IEEE/ISO...and prohibited from applying restrictions of use which are obviously targeted at preventing their main competitor from operating under the terms of its own choosing.
Actually, web sites that cannot stand Slashdotting are badly designed.
That isn't always the case. Even after/. started bringing linked sites to their knees it was run on relatively modest machines...and other very busy sites like rpmfind.net ran on servers built with leftover parts to handle amazingly big loads (most easily done with Apache running on Linux or BSD but even MS IIS isn't that bad--there is just a big OS overhead).
To handle/. loads of traffic requires mainly intelligent caching and generous use of RAM, but MOST IMPORTANTLY you need a big fat pipe to the 'net. Those sites that fall over often would've been fine even on an entry level PC, except that they are hosted on a single machine with two dozen other dynamic sites sharing a 100MB connection, or perhaps they are hosted of a DSL or cable connection out of a home or small office. No amount of smart designing will prevent a small connection from saturating.
Although the metre is an SI unit, m^3 per minute is not considered "proper" becasue SECONDS is the standard base of measurement for time for all SI engineering units. If one second is not a reasonable amount of time for the application then it is prefixed like any other SI unit (microseconds, milliseconds, kiloseconds....). Stuff like "kilometres per hour" is commonly used metric but scientists wouldn't use that for the same reason--for them velocity is always metres per second.
However you measure it though a "15 centimetre" fan that draws 1500 watts of power to cool a CPU is just stupid. Not sure if it is a slow news week for nerds but it seems the stories are getting silly (still a good laugh though). This CPU cooler would be good for the guy who buys that dumb "type R" PSU that was featured earlier. It seems the PC-modder crowd is seeing more and more "rice boxes" out there.
I can picture it now..."cool" nerd going a LAN party, driving a brightly painted Civic with that screaming 115HP motor, windows tinted almost opaque with 18" wheels on a goofy camber due to the chopped springs/hack lowering job and glowing neon and red "R" stickers all over the place.
In the passenger seat sits his 'leet gamer PC (the trunk is full of amps and speakers). It is an amazing construction of aluminum, plexiglass, neon and silver and gold Tremclad housing a type R power supply (as advertised on/.!), the biggest heat sink that'll fit inside and six fans to coll that screamin' $80 mobo and Celeron CPU. Good thing that mobo has built in audio, video and LAN becasue the slots are blocked by all the other gear...
Just a word of advice...if you "pimp your ride", then "pimp your PC" to match your ride, it is time to seek therapy.
Although CGI is in its decline, it is still used plenty.
Using this argument makes little sense becasue CGI is quite irrelevant today. It is the kind of back-handed comment that people use about languages like COBOL.
The "death of CGI" rather than the fact it is still twitching is the real argument to use mod_perl over many of the alternatives. mod_perl is probably the most mature of all the non-CGI web application platforms--it pre-dates everything except perhaps PHP (which looks to have emerged at around the same time). J2EE, mod_python, ruby on rails and so on are great performers I'm sure, but mod_perl and PHP are much more established, and literally nothing beats the vast library offered through CPAN.
I totally agree with you about the fun factor though--for all the hype Java solutins get, Java is probably among the LEAST fun languages to use--perhaps it is better than straight C or FORTRAN bt I'd even rather program in VB. Perl does lack a mandatory structure, letting the coder create ghastly code, but it DOES give the disciplined coder the ability to right extremely maintainable code as well (my Perl code tends make Pascal look free-wheeling).
The basic truth is this--it doesn't matter what the language does to make code maintainable, a language needs to be "fun" to make developers the most productive. That is why there are millions of essential, popular scriptsout there written in Perl. I think if Apache2/mod_perl2 had a nice, refined graphical IDE it could kick major ass on Visual Studio/ASP.NET...
The likes of Atari ST / Amiga /... "could" have ended the MS monopoly
That was 1985--a different time. Microsoft was far from being a monopoly. Wordstar was battling upstart WordPerfect to maintain supremacy in the Word Processor market. Accounting was still commonly done on CP/M and Apple II computers running VisiCALC spreadsheets. Although it was the biggest player in business PCs, IBM still commanded less than 50% of the market. IBM was almost non-existant in the home market--that market was dominated by the "big three": Apple, Commodore and Atari. The PCJr was laughed out of the market because it was complete garbage compared to the superior offerngs from Commodore and Atari.
It was IBM that was shunned by the home market--Apple, Commodore and Atari were not only accepted but very much embraced by the market. IBM was very successful in business because they presented a logical argument to businessmen--one big trusted vendor to hadle all your computing needs. If you spent megabucks to get that mainframe from them it only makes sense to get the PC from them too. In the home IBM was NEVER dominant--by the time the platform became popular in homes it was compatibles that took that market (Tandy/Radio shack probably sold more home PCs than IBM at first).
MS was also very successful, but mostly through licensing technology to IBM and others. Until the 1990s MS (like Intel) had no brand image at all--people who owned Commodores or TRS80s might've vaguely remembered the name from the copyright notice on the bootup screen. Their OS was nothing remarkably different than CP/M, their word processor (PCWord) sucked and was a bit player against WordStar and WordPerfect. Their spreadsheet (Multiplan) equally sucked against the kings Lotus and VisiCalc, both in terms of functionality and market share. They had no GUI except some plan to introduce the pile of crap Windows 1.0 "real soon now".
So why did MS succeed? Because they learned from their mistakes (even if it took a vew years and three versions) and Commodore and Atari didn't--those ones got much more right at the start but not only didn't learn from their mistakes, they repeated them with magnified intensity. While Microsoft was gradually building its brand and improving its products, Commodore and Atari were wandering aimlessly. The other reason MS is now dominant is that they "innovated" faster--I put quotations there becasue they borrowed or acquired innovation and their products were not tied to the success of one hardware vendor (even Commodore Tandy and Apple were licensees). The others suffered from "not invented here" syndrome and tried to engineer everything in-house and ket their designs closed and very proprietary. Commodore would've died much quicker if it stuck to what it was doing and didn't purchase Amiga--and when they did they bungled everything about it except the hardware itself.
Just because MS is big doesn't mean it will be there forever--it used to be smaller than Commodore. It'll just be a bit harder to knock them off--not because they are big but becasue they are smart/savvy. It'll also most likely take a collaberative effort and an open software/hardware design to do it.
IBM/Microsoft DOS was based on CPM, an open source/free OS
Gary Kildall must be spinning in his grave right now. CP/M was a PROPRIETARY operating system by Digital Research. Maybe there is an open or public domain incarnation today, but it was very much proprietary when DR still extisted. The STANDARDS were open in that the BDOS calls were pubically available, and CP/M variants ran on multiple platforms (8080, 8085, Z80, 8086, 68000) and CP/M machines were usually open architecture S100 machines. You could definitely not obtain a copy of CP/M legally for free nor could you see the source code without a special agreement and extra cost.
The BIOS for the IBM PC was also open
ummm...no it wasn't. Even the BIOS calls weren't 100% fully published. Phoenix and Compaq developed a compatible BIOS against the wished of IBM (it was the one and only part of the original PC that wasn't an off-the-shelf component in a design a small group of hoppyists could easily replicate). The way it went was like this: a group of people disassembled the IBM bios and wrote a detailed specification of all the entry and exit points of all the calls and what effect they had on the system. Then a separate group of developers at different company (Phoenix) who had sworn a legal oath that they had never examined an IBM PC used that specification to create the first IBM compatible BIOS.
It wasn't really Microsoft or IBM that created the advantage of which you speak at all--they merely took good advantage of "open architecture" and the co-operative efforts of others. When it comes to the creation of the industry, others did all the work and IBM and MS used their marketing savvy to take maximum advantage and profit (the ones who did the work were not marketers obviously).
When IBM finally realised that a little firmware was not enough to keep a lock on the market it was too late--they no longer steered the direction of that market. The MCA bus was technically superior to EISA, but it was closed and incompatible and IBMs share of the market they created was less than 50% or at least fast heading that way.
Don't confuse open architecture hardware platforms with Free/open software--they both have an advantage in that information is more free to move about, however control oof the design and direction of the former is still firmly in the grip of a select few hardware vendors: Intel controls the bus and motherboard dimensions, Intel and AMD the CPU and chipset, ATI and NVidia video and so on.
Microsoft has one of the best track records in the business for backwards compatibility....except they don't. My employer guarantees 100% compatibility with 20 years of future products as a condition of sale. We have fully supported software that communicates with hardware that was made in the 1970s. IBM has that sort of track record of long-time compatibility and support for its mainframes too. That kind of track record for Microsoft would mean being obliged to offer support for stuff like BASIC on the TRS80.
That being said, MS DOES put a great deal of effort into backwards compatibility--to the point of including a DOS emulator in NT4/2K/XP (WoWExec) that is so seamless most people would never think that the aforementioned OSes are no more compatible with DOS than Linux is (it just happens to have really good emulation). There is a blog by a microsoftie called something like "the old new thing" that explains the lengths MS goes to to maintain compatibility with popular legacy apps.
There are two problems with the efforts MS has put into legacy support: Firstly, it has done a lot to make their codebase cryptic, nearly unmanageable and sub-optimal. This is a problem the likes of IBM and my employer have to contend with as well, except that DOS and NT were not engineered with then intention of being the core of a product for decades. As a result, you get a massive blocks of code,.ini files and registry settings specific to legacy apps. You might never run the DOS version of Simcity from 12 years ago on your new system, but there is code in the current windows that was placed there specifically to make that one app run. All of that legacy support is quite a hodgepodge at times.
The second problem with MS Legacy support is that it tends to be rather selective. In the past, when there was a very popular 3rd party app that sold a lot of copies of Windows (certain desktop publishing packages come to mind) legacy support was done without question. when MS Office sales are slumping...well it looks like time to add a few more features that 0.001% of users asked for and use them as an exuse to break file format compatibilities. The thing about IE7 beig "too advanced" for anything older than XPsp2 is another one of those cop-outs. A little company and a non-profit foundation managed to make more secure browsers with innovative features that runs on multiple platforms and MS can't use their billions to engineer something that works with multiple versions of a SINGLE PLATFORM? Bullshit. They are trying to accelerate the elimination of Win2k because it is limiting their revenue potential.
I understand that legacy support is expensive and that MS is beter than a lot of SW companies like Red Hat (not that that is totally Red Hat's fault--they just don't have the resources). The difference is that Free software often continues to work on anything it'll compile on, and if you do have to upgrade you don't often have to pay through the nose for a highly disruptive upgrade. The IE7 compatibility issue is artificial--MS could EASILY make it run on win2K with its resources and say "there is no official support--use at your own risk". They just made design decisions to deliberately create critical dependencies on XPsp2. Even more than concerns about support costs, MS wants to boost stagnant OS sales.
Problem is, that makes IE7 an expensive browser for someone like me, whose only MS OS is win2K. Firefox is free in all senses of the word, so IE7 makes for a pretty weak justification for an OS upgrade when Firefox is much more convenient to get and I don't need to re-install my OS.
The feature article is from the mainstream press. The author has basically come to the realisation that there is no such thing as 100% secure/safe web browsing. This is something all competent IT professionals know--in fact the only "safe" computer in terms of security is one that is never connected to a network and is not physically accessible to anyone but yourself (pretty much impossible to do of course). Any IT professional who believes otherwise is not competent in computer security.
There are degrees of "safeness" however. Rationally explaining this does not pull in readers of the mainstream press of course--the local news knows being a bit more alarmist than required will boost ratings. The fact is that IE and Microsoft Windows are architecturally flawed. ActiveX object, BHOs and the like in IE grew out of ancestors like COM, OLE etc, which have roots in the Windows 3.x era when the Internet was not even on MS' radar (remember as late as the end of 1995 BillG thought an MSN based on its own proprietary infrastructure could compete with the open standards of the Internet). As a result, the components that make IE powerful are wholly unsuited for a networked environment. Windows XP/Server 2003 have stable VMS-inspired underpinnings that make them acceptably stable, however for compatibility reasons and due to MS culture it retained higher-level interfaces and deplorable security model of DOS/Win3.x/9x/Me.
By contrast, the Mozilla team threw out the unmanageable legacy Netscape spaghetti-code and re-architected from the ground up at a time when ther was already a good deal of awareness of internet security. Linux and MacOS X are rooted in UNIX heritage. Although it has a longer history than MSDOS, UNIX was designed form the start for a networked environment. The combination of these browsers and OSes are thus inherently superior regardless of their marketshare because their very foundations are better.
Microsoft can build the biggest, deepest moat around its house, put bars on the windows and doors, and add a layer of brick to the walls, but the creaky foundation will still crack, shift and leak and allow toxic mold to creep in. The F/OSS house may need repair from time to time, but it is much less likely to be condemned for sitting on crumbling footings.
Yes, Intel and AMD publish info on their processors--this is exacly why they are successful. There have been proprietary processors in the past as well. They are not common knowledge because they were big flops. The IBM PC model 5150 that spawned the architecture we use today was NOT the first PC that IBM made--that honour goes to the IBM model 5100. Why was it a complete flop? Well, IBM did NOT release detailed specs for the processor of the 5100, or any of the hardware for that matter. They kept the instruction set and everything secret! Non-IBMers could only program it in high-level languages!
FYI open SOFTWARE isn't very new either--in the early days where computing was a more purely scientific/academic endeavour coders tended to share their code with relatively few restrictions. When computers became popular and commercial software got proprietary--the GNU movement was actually reactionary in a sense--a desire to return to the early days of true sharing and collaberation.
There appears to be an analogue situation happening with PC hardware. In the 70s there was little effort put into protecting designs--Intel and AMD are just following that legacy with their CPUs. The Homebrew club members proudly showed off their clever designs. Computers even came with schematic drawings (even pre assembled ones)! By the 80s hardware vendors started going down the proprietary path (encrypted Atari7800 cartriges, The Macintosh, IBM's MCA bus, etc) until we have absurdities like ATI and NVidia GPUs, "soft" modems and wireless chipsets and so on that you cannot make work without proprietary information.
Open ANYTHING isn't new...it's just the relisation that success of an idea is ultimately limited or impossible in an environment that is too closed.
Oh, and you might not want to talk about FPGAs becasue they're "uncompetitive"--that is not the point--I personally HAVE obtained HDL code and configured an FPGA and "RAN" that code for far less than $1000. Could a pre-fabbed ASIC be obtained cheaper? absolutely...probably for one tenth the cost and far less trouble...but it is unmodifiable. I can take a RISC CPU core and other bits and pieces and make my own customised design for next to nothing...almost as easily as making custom software apps. The arguments about TCO of MS vs GNU are continually shot down for software--the same can be said about hardware...there are far more important benefits to open-ness than just the cost factor once you've achieved a certain critical barrier-to-entry.
Nice to see someone as important as IBM realising the importnace of open HARDWARE. I've found that until recently the concept has been overlooked or even derided. Even open software advocates didn't "get it" and said it could never work, becasue hardware is different--the argumant was that hardware isn't something individuals or small companies could influence becasue of the high cost of entry, and big companies needed to make money off licensing closed IP to fund development and production of new hardware.
This day and age, such an argument is complete BUNK. Hardware design is done on computers and chip specifications are more often than not specified in VHDL or Verilog--the "source code" of hardware if you will. Not only is design and simulation within the reach of even hobbyists, the end result is very similar to software in characteristics. While IBM is not completely opening things up to the point of showing the "source code" of the Cell processor, it is a great step to see all the specifications etc. without encumbrances.
Quite frankly I'm surprised the open source movement hasn't advocated open hardware much more vigourously. After experiences around NVidia and ATI and Intel you'd have to be a fool not to realise that open hardware isn't just an interesing idea, it is NEEDED for the success of open software.
How are things on the militia base?
on
Tinfoil Hat House
·
· Score: 2, Interesting
Must get lonely out there in the middle of Montana or North Dakota or wherever you are, hours away from any other civilisation. You should get into town more--despite being "fascist" the good citizens of those towns really are nice people.
Seriously though I'm guessing you have a bit of a warped sense of humour or you are a masterfull troll. Besides the fact that there ARE health and safety concerns associsted with prostitution and scrap vehicles, a typical community code is quite reasonable. If all people were reasonable and considerate there would be no need for such codes.
Alas, a small but significant minority of people are complete jackasses. Given the chance these people will paint their houses hot pink with purple, orange and yellow trim and park a rusty old truck with two flat tires on the front lawn under the shade of their 12-foot C-Band dish (the one that pipes hard-core porn into their living room which they view with the volume on so high everyone on the block can count the orgasms).
These codes might not be about health and safety, but they are about the right of law abiding citizens to live in realtive peace. Unfortunately, there are other types of jackasses who are the opposite--these are the ones that insist on hearing their favourite pins drop after 9PM, mow their lawns in a checkerboard pattern and wish everyone would do the same, and hold up city business with requests for more traffic-calming measures, more noise bylaws and more regulation of every little thing that slightly annoys them.
If you find you are living in a redneck hellhole, or nazi-Stepford-suburbia there is a way to stop the madness. It is called "civic duty". Get involved in politics--especially at the local level. Attend the odd community planning meeting, endorse non-wacky candidates for office and vote for them...or at least vote for cryin' out loud. The latter form of jackass knows this already, and that is why California is so whacked that a government led by Conan the Barbarian is actually a vast improvement over the previous rock-bottom situation with Davis.
Slashdot Mirror
172K of ram... in a system with a 16-bit bus?
... how and what the fuck for?
16-bit ADDRESS bus and 8-bit DATA bus to be exact. And yes more than 64K is doable on such a system (172K seems a bit odd-sized though--192K would make more sense as that is exactly 3 64K banks). The CPU just needs a bit of help is all. It's called BANK SWITCHING and it was common in those days. Essentially multiple banks of memory would be wired to the same address lines, then the output enable pins would be wired to be active at different times depending on a register. That's how my Z-80 based Coleco ADAM could work with up to 4096K--in assembler you use an OUT instruction to set a register to set which little piece of that you want to use (I think it was actually 2 32K banks or 4 16K banks you could select).
Bank-switching was a bit awkward because you couldn't "see" all the memory at once (you certainly couldn't do anything like calling outside of your current memory segment reliably) but you could do a lot with all that memory nonetheless (mmmm...ram disk)
Hand tuned a floppy drive with an oscilloscope?
Well, specifically with a system like the II+ such a thing was quite possible. I used II+ machines at school as a youngster but I am not that familiar with that hardware. However I know that Apple peripherals were "dumb" (not in a bad way--just very elegant and simple from a hardware standpoint but very reliant on programming). IIRC the timing of the floppy was very closely related to the system clock of the computer, so if you could get the drive to spin EXACTLY the right speed it could read a track in the correct number of CPU cycles all the time. If it was "out of sync" you'd get read errors (kind of like "buffer overruns" and "underruns"). I'm guessing that is where the 'scope came in--comparing the pulses from the rotating drive with system clock signals so they matched up right.
My systems of choice were Atari XL and Coleco ADAM which had serially-connected "intelligent" peripherals that were rather less sensitive to timing. Less hacker-friendly however. Anyways, compared to today's machines even they were blissfully simple. Even though those machines couldn't do as much it sure was nice to actually know HOW they worked. Oh yeah--and even though the ADAM had a reputation for being "very buggy" when it was first released, the final "R80" revision machines were quite fine--and even the original "very buggy" machines crashed far less often than Windows--and I never had a virus on any of the 8-bit machines I ever owned...ever.
those were the days...
...but I won't. Rother I'll explain why you have no clue.
Insightful, how about idiotic. What can you program in Unix that you can't in Windows.
That wasn't the point the original poster was trying to make. The point is HOW you program in Un*x vs Windows. Nobody will argue that you can do anything you want with either platform. However, a great many people would argue that the "UNIX way" is FAR mor elegant.
In Windows you have C and C++ just like Unix. Java, Perl they are all there as well.
This statement really demonstrates your inability to comprehend the differences. To extend the "building toys" analogy, C/C++, Java, Perl et al are NOT the pieces, they are the plastic/wood/metal with which the pieces are made. You could make lego bricks out of the latest space-age carbon fibre composites, but they would be useless if the "bumps and holes" on each brick were different sizes and wouldn't lock together.
Now the platforms may be different but largerly they are more similar than not from a progammers ability to make a program perform a required task.
There I'd really have to disagree with you. There are things that Un*x style architectures do easily that are arduous to perform in the Windows environment. Similarly, there are things Windows excels at. IPC was really much more refined under UN*X--some might say Windows works with threads so well because it has to since its IPC abilities have historically sucked--really in UN*X it is much easier to get various components to play nicely with each other yet keep their resources separate and protected. OTOH, there are reasons Windows-based games are so far ahead besides simple market share--graphics interfaces are one of those "funny shaped blocks" in Windows that is very well suited to its task.
Really that Lego analogy is very apt indeed. UN*X is very uniform in how it works, just like a bucket of classic Lego bricks. You have a library of pipes, sockets, shared memory etc. that is very standard across all programs that extends all the way to the user interface (you can pipe all manner of programs input and output together right on the command line to a degree not yet seen in production releases of Windows). Once you get the hang of the UN*X Way you can snap these blocks together esily to suit your needs.
For all the "object orientedness" of Windows, there is not that level of uniformity in interfacing to make those reusabel objects work together. Instead, you have an overly complex framework in the form of DDE/OLE1/OLE2/COM/DCOM that was largely designed to accomodate disjointed, inconsistent interfaces between various components/applications. This is something like the "licensed from the movie" Technics sets with all the little odd-sized rods/axles, funny-shaped blocks, special wheels and so on. There many little sets where the pieces fit together very nicely in a few commonly required configurations, but when the time comes where you want to make your own creation not in the instruction booklet you become frustrated with the useless pieces. For many kids, the six or so really cool things you can build are good enough, for the 10% "most geeky" kids it would bore you quickly.
I can't say I really know for sure what a "mainframe toy" would be--mainframes don't seem like fun at all. I think "mainframers" may have forgotten what childhood was like, or perhaps hatched from a pod fully grown, who knows. I do not have a lot of exposure to that philospohy/culture. If I HAD to pick a toy that was most mainframe-like I might say Mecanno, because like UN*X they are fery uniform in structure, however you have tediously fiddle with those little screws to put anything together, just like a mainframe--you have your "special screwdrivers" (arcane knowledge) and have to follow tedious processes to get things done. Or, perhaps it is like building a birdhous with popsicle sticks, where you have to tediously glue the pieces together with Elmers glue, wait for it to dry bef
there's nothing wrong with DCOM's security model
Simply stating there is "nothing wrong" with the security model of DCOM with the only supporting argument being that a single one of a large number of DCOM attacks was related to a buffer overrun bug and not a design flaw is pretty weak. If it was simply a matter of buffer overruns and memory leaks then these nasty problems would all go away once the implementation was fixed with MS03-039. They are not going away. Years after blaster there are still critical fixes being is
Sorry, I know ALL TOO MUCH of what I speak. The software my employer maintains and uses historically relies very heavily on DCOM and they are dedicating a LOT of resources to completely eliminate depencency on DCOM over the next couple of years. The hard truth is that DCOM security is unintuitive and overly complex. I'm not just talking about the protocols or low-level models around security. I am saying this from a top-to-bottom view. For example, the tools MS provides with the OS to administer DCOM are very lame (DCOMCNFG.EXE really really sucks--it is usually not clear what account should be allowed to do what thing with what component from the crappy interface). Even though the need to use DCOMCNFG is limited to quite a narrow segment of users there sure are a lot of questions, discussion and confusion around it.
Also, default security behaviours are DESIGN characteristics of a system rather than IMPLEMENTATIONs. Those default behaviours in DCOM (implied access to all components by any connection by default) happened to be so flawed that MS altered them in XP sp2--it was so important that we all had to deal with the broken apps that resulted from the changes ourselves.
In my experience it is also quite tricky to effectively secure DCOM servers without relying on domain security, meaning you need to establish a dedicated machine for a domain controller if none exists.
All in all, actually, it seems DCOM in general is very convoluted to use for no good reason whatever--I don't profess to understand it well at all myself but I'm forced to use it. That in itself is a design flaw because 3rd party software developers and end users who do not understand DCOM well enough can very easily open security holes.
DCOM simply gives me headaches. It gives everyone headaches, and I've never...ever...heard anyone say anything along the lines of "DCOM totally rocks--it is the best!". Usually "supportive" comments are along the lines of "it works much better than it used to" or "it works fine it's just that no-one uses it right". Mostly I hear "almost no one needs it and if you totally disable it your system will be safer". Hardly a ringing endorsement and hardly a solution to corporate/enterprise users who must live with it.
Uh, I thought there was public consensus on this that packaging software is the whole point of the "distribution" concept! What is it you think distributions are supposed to do, if not this?!
The problem is that there is no standard so the whole point is completely lost! It's fine and dandy to say RPM is a standard packaging method but I can't just say (for example) I don't like SuSE's Apache RPMS so I'll just use Mandrake's instead--they are both RPMs sure, but if I try that you'll get mired in "dependency hell" and you have to know what you are doing and override dependencies to get it to work. The same potential for hangups presently exists around Debian-derived distros as well--you can't always count on stock Debian packages working in derivative distros or vice versa.
I have no problem at all with distro makers maintaining its own repository of packages. You're right--that is what those organisations exist to do. The problem is that you have to depend on that ONE SINGLE REPOSITORY FOR ALL THE SOFTWARE that you want to manage properly in the system. If Linux devotees want to knock Windows off its pedestel it'll have to work on that.
I don't think compiling from source is a hack at all.
Well, that is your problem--from a user's perspective it is VERY MUCH a hack. Compilation is a normal part of the development and release process--it is a normal part of a PROGEAMMER'S use of a computer. It is NOT something an end-user/sysadmin does or should be expected to do. Even "source-based" distros like Gentoo or Linux from Scratch have tools like Emerge or ALFS to automate the compilation process with a low-level "packaging system"--essentially making the compile process the same as a slower version of a binary package install from the users viewpoint. Even then LFS concedes it is a distro NOT meant for everyone.
Even if you cannot see how compilation from source can be viewed as "hackerish", you certainly cannot dispute that doing so to install software on a distro based on RPM or DEB binaries leads to a very "hacked" situation--you install something from source either because you cannot find the version you want or it isn't available at all as a binary package. Later it IS packaged and other packages are dependent on it. You either have to remove your custom-compiled version and replace it with the RPM or you start down the dangerous road of --nodeps.
It's all more work than it needs to be. If I don't like the job SuSE is doing I want to be able to replace their packages with Fedora equivalents seamlessly for example. If I find some nifty niche-app on SorceForge it would be really cool if they could maintain a binary package themselves and not have to worry that it is too specialised for my distro publisher to bother maintaining.
Not using IE and using Firefox instead almost completely secures an up-to-date Windows box.
Dumping IE is a good step to take in improving the security of your Windows PC, but to say that one action "almost completely secures an up-to-date Windows box" is a dangerous oversimplification of the problem.
Firstly, despite the legal disputes surrounding the strategy, Microsoft has deliberately engineered IE into its OS. Windows 95 and NT4 and previous versions had no dependency on IE at all. On an "up-to-date Windows box" it is LITERALLY IMPOSSIBLE to completely "get rid of IE" on your system. You can remove the icon, make Firefox your default browser and so on, but IE remains in place. You cannot remove the IE rendering engine or any other "meat" of IE contained in system DLLs or you break a multitude of applications and important parts of Windows itself. No matter how hard you try to avoid it, you still need IE to do something as basic as keeping your system up-to-date and run your basic applications properly.
All that needs to happen is for an ambitious group cyber-terrorists to commandeer and taint one single system of servers--the Windows Update site--and the world's IT infrastructure can be brought to its knees. I know saying "all that needs to happen" understates the difficulty of pulling such a thing off, but it IS possible--and the point is that Windows Update is a very serious potential single point-of-failure. Even though Windows update is a huge site run by many many computers, it is still accessible through a single network address and maintained by a single company and operates the same way for everyone. The fundamental concepts behind Microsoft's Windows Update are seriously flawed and without constant vigilance on the part of Microsoft it could be the most serious vector of attack in the history of computer security.
Another fundamental design flaw of Windows from a security standpoint is OLE/COM/DCOM/Microsoft's RPC implementation. Microsoft themselves have acknowledged this with its efforts in Longhorn to create a new service-oriented programming framework for distributed applications (Indigo). Even in post-blaster 2005 there are still reported flaws around DCOM. Ever since OLE was introduced with Windows 3.1 it has been more convoluted than required to do its task, and even with this added complexity it was not designed with the highly connected world of today in mind. Eventaully COM would come out with the still klunky OLE2 interface built on top of that, and distributed application functionality would be tacked on with DCOM. Holding onto a foundation that had quickly become rickety for this long was a grave mistake. MS should've started pushing everything and everyone away from this whole kludgy mess ten years ago when it started becoming clear that the network would be central to computing.
Thankfully, there are limiting factors to the whole DCOM disaster in that home users don't really need the "D" part of it at all, so you can disable it in the registry and/or block its ports with a firewall. Unfortunaely, that doesn't fly in the corporate world as there are a lot of client/server products that rely on it to function (think ERP, industrial automation, custom integration systems and so on). This is why corporate adoption XP sp2 and 2k3 Server sp1 were not at all rapid (so much for the "up-to-date" part of your argument). Those service packs close up much of DCOM and break corporate apps. Thus, those updates are not rolled out until procedures are available to make updated PCs work with existing systems. Guess what? Those procedures generallyinvolve UNDOING some of the changes made in those SPs to secure systems!
I'm sorry, the headline of this article putting Windows at par with ANYTHING in terms of security is unconvincing to me. While it is true that there are some Windows systems out there that are better secured than some Linux or BSD and it might even be true that overall the implemented systems out there are equally secur
That these guys advertise this compatibility speaks very ill about their distro, IMO
Well in *my* opinion the majority of people do NOT share your opinion about compatibility. I don't think I've ever heard an average end user of PCs (or any other product for that matter) say "hmm...this product interoperates with multiple standards...that means it sucks. I'll pick this other one that's only half as compatible with other people's stuff--since they didn't think of interoperability it must be better at other stuff anyways".
The fact that multiple methods of package management are supported by a Linux distro doesn't mean it is "hackerish"--it doesn't HAVE to anyways. It means the distributors recognise that there are two prominent standards and that end users demands require both. Being debian based this distro's repository will be DEB only. There is nothing "hackerish" about wanting to have 3rd party programs distributed as RPMs "just work" even if the system is natively Debian-based.
Like it or not, RPM is the leading format in the enterprise space--it is the native format of Red Hat, SuSE and Mandrake. Support for RPM is also a requirement for Linux Standards Base (LSB) certification. I and many others consider LSB compliance to be a requirement. RPM is not going away and people just have to learn to accept that. Similarly, the Debian format has its own merits and with so many distros being Debian variants it is a defacto standard that RPM adherents have to accept.
It might not be the most "idealistic" or technically elegant thing to do, but supporting both major package management formats is the most professional and business-sound decision to make. Linux will never, ever achieve dominanace if every distributor has to maintain its own packages of every popular application demended by its users. People will not switch to Linux if you either have to live with the packages maintained by a single organisation or be TRULY "hackerish" ahd compile parts of the system from source or shoehorn 3rd party packages into the system with things like "rpm --nodeps" and crap.
Think about it...one of the biggest complaints about Microsoft is "vendor lock-in", where nothing works quite right unless you buy into MS end-to-end. Well, if Linux distributors followed your mentality then they risk falling into that same trap. I want a solid, stable, easy to manage OS on which to base my systems. I actually DON'T want the distro comapny to put so much time and effort into its package repositories. The folks responsible for Apache, Sendmail, Postfix, MySQL, PostgreSQL etc should be able to make LSB compliant packages that I can just click on and install. I really don't care if I can't get every package I need from the same people who put together the distro--I just want to be able to get it SOMEHWERE and have it work!
Anyways, I wish the good people behind this new enterprise linux all the best, and hope that their talk of "compatibility" is substantial.
On the Amiga, data types were handled as plug-ins. If you needed the capability to open a jpeg you added the data type.
That would be great if it worked that way on Windows. MS could go ahead and include all the plug-ins it wanted, and if you didn't like IE or its rendering engine you could "unplug" it and "plug in" Gecko. OR, if you aren't processing JPEG or MPEG data at all you could just unplug those extensions.
That isn't the way Windows works however--there is absolutely no consistent interface for handling data and no logical delineation between components. What stemas me is that much of this wasn't done for legacy/compatibility/technical limitations--it was done DELIBERATELY. IE USED to be a distinct component where Windows would run without it just fine. MS DELIBERATELY mixed it right in so its dlls sat right along Kernel, User, GDI and so on and encouraged its own developers and others to add as many hooks and ties as they could. Then they went and did that with WMP. Now there is an astonishing amount of software that ties into these libraries for different reasons and so on, and there is no published/advocated interface that makes IE and WMP libraries "pluggable" with replacements.
This is analgous to hardware systems that have these super-integrated motherboards with sound, video, IDE, network, kitchen sink etc all soldered together. If you think the sound is crappy you can add a card, but you still have the built in stuff sucking power, generating heat and hanging around to be a potential nusiance or point of failure. Such a thing is tolerable for a disposable home or office PC but there are legitimate reasons for avoiding that sore of integration.
If MS wants to sell its Windows platform as suitable for anything you could use the hardware for, then it MUST re-architect it so that its components can be decoupled, much like it is still possible to buy hardware (with thinkgs like video and sound on separate cards).
MS can throw whatever they want onto the CD and into the installer if it wants to--hell, that is what Linux distros do too. The point is that the components are SEPARABLE. You can do a custom/"expert" install of Linux and choose not to install a web browser or media players or even a GUI if you are savvy and have special needs, or you can just hit "default workstation" or whatever and it will put a GUI, web browser, media player and even the office suite (or several of each of those!).
.WMF files on my mail server thank you. Also, what if I like alternatives and will never use the "default" components? If I use another media player, another browser or GUI environment all the factory MS cruft just wastes processor cycles, memory and hard drive space.
That is not what I have a beef about--I'd be perfectly happy if MS had a "one click" install that puts all the toys on--hell they could even put MS Office into the Windows install for all I care. What I do NOT like is that there is NO WAY to avoid installing all or parts of the GUI, IE, WMP etc etc..even if I don't NEED or WANT those things. I'm NEVER gonna play
And your comment about Apple starting it all...well you're right...and that is a big reason I never got a Mac. There are times when I LIKED the command line and scripts and stuff and it frustrate me to no end not to at least have that as an OPTION. I DID have an Atari ST and that was similar, but even with the Atari you could shut down GEM and use it with a CP/M68K command line environment. And now even Apple has seen the light, at least partly. MacOS X follows the UNIX philosophy underneath. I still get the impression the GUI is welded on a bit too tightly but I think with PCs being more complex and interconnected that a hyper-monolithic architecture is outdated and unsuited to todays needs.
The actual Media Player program is just a front end for Windows' media layer. So programs can place calls to it to play media, without having to implement their own codecs and playback mechanism.
.exe is just a thin shell with not much more substance than a desktop icon.
THAT is what the whole beef is about! THAT is the reason the American DOJ had the brouhaha over IE "bundling" as well!
The problem really has NOTHING AT ALL to do with "including" WMP or IE with Windows--MS should be allowed to do so. "Bundling" is a really misleading term when you think about it because what MS has done is not "bundled"--it has engineered what were once distinct applets or applications into libraries and integrated/hooked/merged the "meat" of these applications into the OS. Thus, when you open an IE browser window or the WMP applicaiton the
This "melding" into the OS, or transforming "applications" into "OS components" is how MS gets things over the "Chinese Wall" they supposedly keep between applications groups and the OS development groups. THAT is what damages competitors.
RealPlayer and FireFox have to load up their own libraries and so a lot more under-the-hood work to launch because the REAL parts of their applications are not "system libraries". Ether they have to to pre-loading tricks at startup, which makes the system startup time longer, or the user has to wait for startup longer.
Furthermore, users cannot DISABLE the loading of IE or WMP components as they are "system components". Other apps depend on IE's rendering engine, you need it for Windows Update, and WMP libraries are used by all sorts of games, etc. to play video. You cannot even disable pre-loading of them on normal startup to my knowledge.
The whole end result is that MS's own bundled "applications" (really just thin shells) seem to work just that much smoother than the competitor, and for all practical purposed are not removable without altering functionality. Thus you have big huge monolithic bloated OS. And now MS is paying the price because they made strategic decisions to tie in all these features for convenience and market domination all at the expense of efficiency, stability and security.
The MS way of things is the antithesis of the UNIX/Linux way of things. I do not like that a certain amount of web browser and media player and GUI CRAP just HAS to be installed on domain controllers, file/print servers etc etc. With a UNIX like OS components are much more discrete--I can more easily configure the install to include just exactly what I need/want and no more or less. With Windows that is absolutely impossible right now.
The result of MS "bundling" is more complex and far reaching than RealPlayer and Netscape whining about having to push their products harder than MS to get people to use them. After years of designing around this philosophy MS has spread sickness throughout the global computer ecosystem.
The website is not standards compliant. In fact it just as broken if not moreso than Slashdot:
.NET technology to make it all possible is bunk though. You build just as scalable sites using Apache and mod_perl on a Linux or BSD box (and it might not even be as processor intensive either). You might need to be more disciplined, but you can employ page caching, separation of logic and presentation, etc. with a Free softweare stack quite readily.
* There is no DOCTYPE--it just starts with an HTML tag
* There are UPPERCASE tags mixed with lowercase tags. This is bad practise and strictly speaking all tags in HTML4 and XHTML are supposed to be lowercase.
* There are proprietary attributes in the document (ms_positioning)
Besides spweing forth really crappy HTML, the techniques the authors used to do the layout are exteremely poor. The site is the exact OPPOSITE of semantic web design. Remember 1996 when websites were all becoming GIF Jigsaw puzzles? That is almost what bungie did. I see empty DIVs, no-break-space characters and some javascript malarky.
and as far as position:fixed goes, at least FireFox supports it even if it is slow...IE DOESN'T SUPPORT IT AT ALL. What you see on Bungie is the aforementioned javascript fakery. Don't give me static about FF bugs--it is LIGHT YEARS ahead of IE in terms of correct support of standards.
Bungie very obviously didn't get an award for good coding practise. The site may be architecturally correct but the coding as abysmal. Furthermore, the site COULD have been made to work far better on FF than it does not WITHOUT FANCY WORKAROUNDS. As a matter of fact, if they DID use position:fixed and other CSS positioning and layout techniques properly it would work extremely well. It would not, however, look as sharp in IE6. Understandably, since Bungie is a Microsoft property, they'll use Microsoft platforms and development tools, and test much more extensivley with IE6.
As for the case study touted by MS, it makes some very important points about how to build the architecture of a site. Any suggestion that you need
Where the ASP.NET solution excels is in its IDE and library of components--Visual Studio.NET really does kick ass there. It's just too bad that the output to the browser tends to suffer when developers get caught up in a fancy design (not only does the HTML look ugly and break standards, it results in inconsistent behaviour). I hope the next version of VS.NET has addressed this.
... not with your reasons.
Apple wants to sell pretty, uncrackable, all-in-one, gold standard computers. They've been doing that since the 1970's and they will continue to do so.
They haven't been like that since the 70s--from 1977 to 1984 Apple was opposite in most ways: Its computer was prettier than everyone elses but it was NOT all-in-one and it was VERY crackable/hackable: Pre-Mac Apples had a whole bunch of slots (with well-known specs) and the early ones even came with all the schematics!
I'd say Apple wasn't so proprietary until Woz left, and it might not stay that way if someone like Woz were to rise to a position of influence in the future. It'll be snowing in Hell before that happens though.
Apple knows that there is competition out there, they don't ignore it like other companies
Of course Apple is painfully aware of the competition. However, Apple has historically been king of the "not invented here" syndrome. They in fact DO ignore the competition as much as possible. That is why Apple is not the company fighting anti-trust lawsuits today. As you alluded to, Apple is not an idealist champion of Free software--Jobs is just smart enough to recognise that it is the best way to tap into a huge pool of development talent and to be the most responsive to the user community.
Apple wants you to run whatever software you want, on their PC's
Not really...they TOLERATE you running whatever software you want on your Mac. They WANT you to run THEIR OS, THEIR productivity software and THEIR hardware. To achieve Jobs' nirvana experience you need to buy into the whole package. However, being a niche player in the PC industry it HAS to tolerate users who want to tinker with Linux or Windows on their hardware. However, those options "weren't invented here" so expect to be ignored by all official Apple support resources if you inquire about ANY third party hardware or software.
Don't you think linux development and customer support is going to skyrocket when there is just a few configurations to develop for?
No I don't. Apple's strategy is to keep variation of both the hardware AND THE SOFTWARE configurations to a minumum. Supporting Linux would dilute their resources and hinder the success and progress of OSX.
Maybe Apple will be the first large computer manufacturer to offer a choice
It can't be the first because some other hardware vendors already offer a choice between Windows, Linux and no OS installed. Given its history since 1984, Apple is likely to be the LAST to "offer a choice". The original Mac had no slots (despite the engineers having tried to "sneak" one in) and was not expandable and there was not even any developent tools/languages available to the general public when the Mac was first released (BASIC was an extra-cost option released some time after the Mac itself--unheard of in the 80s when BASIC was burned into the ROM of every other PC out there).
From the start, the Mac was NOT about choice--you did what it let you do...no more and no less. The Mac was about usability--about making computers more humane. Think appliance: When you buy a stove you can't get dozens of different types of burners, racks, knobs and lights to customise or upgrade it. You get your Kenmore ZX123 and it cooks your food and just IS the way it is. If it breaks you get replacement parts that are just like the old ones and except for the colour and trim that match your decor it is just like every other Kenmore ZX123. Not only is there NO choice, you don't ever care. It cooks your food and that is it--you don't even remember or care it is a ZX123 model unless you need to fix it.
That is what Jobs' and Raskin's vision was with the Mac--make it pretty, elegant, useful and humane. Expansion slots and OS configuration choices and build-to-order and source code just doesn't fit anywhere in that puzzle. All that stuff is the domain of their developers and engineers. For A
False, Microsoft tries everyting that may remotely work, then toss the things that don't
While this is certainly true right now, MS did not have the resources to employ such a scattershot strategy for most of the first decade of its existence. It was only after MS/PC-DOS entered wide distribution in 1981 that MS could make any big high-risk/questionalble bets without causing undue harm to the company. I'd argue that it wasn't until the mid 80's that MS had a venture fail badly enough to be abandoned: That would be the MSX platform.
In any case that is actually a reason for MS business success. It goes to great lengths to diversify and knows when to drop unsuccessful ventures.
As I write this, the post I am responding to is modded "Troll". Seems it is more "Insightful" than anything else.
BillG is a smart guy who surrounds himself with smart guys. MS started out of a motel room in New Mexico and didn't become a near monopoly solely out of luck any more than it did out of sheer creativity. MS is huge for a few important reasons:
1. They recognise opportunities and make maximum use of their resources and connections. MS became the "king of programming languages" for micros inthe 70s and 80s because BillG immediately saw the potential of the Altair and the desperate need for a friendlier method of programming the system. They also used their connections and networking skills to arrange a meeting with IBM re. DOS as well as to locate and purchase QDOS.
2. They are a bit sneaky--they will sell something they don't even have yet (DOS) and create demoware/vapourware to stall and kill competition in a field where they are lacking (GO/pen computing/etc--"they might have it now, but big ol' MS is gonna have it REAL SOON NOW"--yeah right).
3. THEY RESEARCH THEIR COMPETITON--MS has historically been very paranoid. Even with their position today they view EVERY competitior as one who could destroy them. When MS plays in a market they research EVERY LITTLE THING about that market and EVERY COMPETITOR. BillG himself operates that way. If he meets someone who has something interesting to talk about but BillG knows nothing about it, BillG will spend every waking moment for a couple days learning about the subject. The next time he meets that person he can talk with that person like he is a seasoned expert.
It is for that third reason why MS has a whole department of Open Source Specialists in its employ and has had for years. It is also how the Halloween Memos came to be. It doesn't matter how badly MS slags Linux or how much it scoffs at Free software--it has ALWAYS apporached it as competition with the potential to destroy Microsoft. I'm willing to bet it's been on BillG's personal radar for a decade already (when MS was just starting to realise the Internet was a game they had to play).
So the parent to this post is exactly right: MS is essentially "stockpiling ammunition" for the battle with Free software. How they will use the knowledge and people they acquire could go many ways:
1. They could use it to make their FUD sound more credible--for example, some weakness in Linux to exploit in the "get the facts" campaign or items to avoid or downplay where Linux has the advantage.
2. MS operates by acquisition, not innovation. They might have to avoid GPL code to keep its code secret, but it can at least steal IDEAS from GNU apps. It is also already well known they've lifted BSD code many times. This is OK though--at least MS software gets better as a result.
3. If they CAN'T beat Free software, they'll be prepared to "join" it. It may be a cold day in hell before MS Office is open, but if Linux meets or beats Apple's market share and all indications are that it won't go away, MS will be prepared to form a "Linux business unit" to port Office and other apps to the platform. It won't be "Free/Libre", but if MS dominates application software for Linux it can steer the platform and continue to be the industry's biggest player. This is what they have done with Apple--Microsoft is the biggest vendor of MacOS applications, and stunning industrial design aside, Macs are becoming more and more PC-like.
Yep it would be nice if MS was more cooperative, but it is just too far from their business model. MS NEEDS software to be closed or else it would have to completely re-invent itself. It is simply easier for it to try and make the competition work to its advantage or simply go away.
Now that's the dumbest thing I've heard today, especially for software that commands less than 10% of the market. Give me a break.
Actually it isn't dumb--it is completely true. You alluded to one of the main reasons some sites are still broser-specific--they have decided that the cost of changing their legacy, non-compliant code is more than the cost of lost potential business from ten percent of the market. In essence, they in fact do NOT want your business because they have deemed you do not have enough to offer them as a customer for the perceived effort it takes to cater to your needs.
Weather.com has worked fine in all production versions of Firefox. Of course your theory that sites (other than MS-owned) would "deliberatly" write code to break a browser is downright asinine. WTF would they do that?
I too am puzzled about the original poster's experience with weather.com--perhaps it is becasue he uses an old release, or maybe it is the presence or lack of certain plug-ins, or an overall configuration problem. It doesn't mean you have to be insulting. As far as his "asinine" theory goes, it has been proven to happen. In some cases it is sneaky, and in other cases it is very deliberate (a stupid move IMHO, where a site sniffs your broswer and displays a "sorry you aren't running IE--buhbye" page).
Broser sabotage is done to compel users to another browser, either because (as in the case I refence here) said company has a vested interest in the success of a specific browser, or (in the case of a lot of intranet/private sites) tech support wants to ensure a homogenous client operating environment.
Seriously, remove your lips from the crack pipe every once in a while and get a grip on reality.
Seriously, remove the cactus from your rectum and learn to relax...you might find a little insight in what people say, even when it doesn't seem to make sense.
And word to the shortsighted designers out there who STILL make public sites that break standards and rely on platform-specific behaviour: you are doing a disservice to all who use the internet. It doesn't matter if a new PC is $3000, $300 or $30, if it requires more care and feeding than an exotic pet and spftware does not behave with a reasonable degree of consistency from one machine to the next, then PCs will NEVER reach their portential. Consumers do not need a monopoly, but they DO need software developers of all stripes to play by the same rules (USE THE STANDARDS PEOPLE!).
Also, don't count on MS and IE being dominant forever--the machines on the leading edge of the ultra-low-cost market have Linux factory-installed, and even if some form of IE remains the dominant browser, MS has shifted efforts to security at the expense of compatibility (it had to becasue its platform has been flawed from the start). I expect that IE7 or possibly whatever the equivalent of IE8 will be will break more than a few IE6 sites just like XP SP2 broke a number of applications.
It's one thing to charge for an fork...but without food, that cutlery is just worthless hunks of metal and plastic.
That argument makes no sense at all. For one thing, people CAN and SHOULD be able to offer free and "Free" food. My mother obtained strawberry plants--at no cost--form a fellow parishoner many years ago. They have since thrived and MULTIPLIED and now mum's strawberry patch is too large for her to manage. She put many plants in the compost but gave some to me so I could grow them in the back yard. The ladies out there also sell fresh strawberries, pies and preserves at farmer's markets in addition to giving away stuff to friends.
There you have it--the biological version of "peer-to-peer file sharing". There is no law against the "unauthorised duplication and/or distribution of strawberries" that a large corporation can use to restrict individuals or potential competitors from making strawberries. International dumping laws are the closest thing, and they only apply to governments and large companies who sell at a loss to deliberately eliminate a competitor.
Microsoft is trying to be like Monsanto--who has new breeds of oilseeds and grains and is trying to use the patent system to restrict and control a natural process. Neither is right in doing that. Microsoft's solution around CIFS is like Monsanto saying "we'll publish the information on our roundup-ready Canola, but that co-op of Saskatchewan farmers or that university research group cannot ever make a royalty-free version to compete with our variety".
In both situations I think this is damaging to society--especially in ther case of agriculture. Food is much too important to be the proprietary domain of a molopolistic enterprise. While software is not an important basic need for people, it is nonetheless a vital part of the global economy, so I believe Micorosft's proposed remedy is entirely insufficient. It holds a monopoly over a protocol that enables information exchange important to the functioning of most business systems. MS has demonstrated in the past it cannot be trusted with SMB/CIFS as it has made changes to deliberately break reverse-engineered systems. Microsoft should've been ordered to relinquish control of the protocol to a standards body like ECMA/IEC/IEEE/ISO...and prohibited from applying restrictions of use which are obviously targeted at preventing their main competitor from operating under the terms of its own choosing.
Actually, web sites that cannot stand Slashdotting are badly designed.
/. started bringing linked sites to their knees it was run on relatively modest machines...and other very busy sites like rpmfind.net ran on servers built with leftover parts to handle amazingly big loads (most easily done with Apache running on Linux or BSD but even MS IIS isn't that bad--there is just a big OS overhead).
/. loads of traffic requires mainly intelligent caching and generous use of RAM, but MOST IMPORTANTLY you need a big fat pipe to the 'net. Those sites that fall over often would've been fine even on an entry level PC, except that they are hosted on a single machine with two dozen other dynamic sites sharing a 100MB connection, or perhaps they are hosted of a DSL or cable connection out of a home or small office. No amount of smart designing will prevent a small connection from saturating.
That isn't always the case. Even after
To handle
Although the metre is an SI unit, m^3 per minute is not considered "proper" becasue SECONDS is the standard base of measurement for time for all SI engineering units. If one second is not a reasonable amount of time for the application then it is prefixed like any other SI unit (microseconds, milliseconds, kiloseconds....). Stuff like "kilometres per hour" is commonly used metric but scientists wouldn't use that for the same reason--for them velocity is always metres per second.
/.!), the biggest heat sink that'll fit inside and six fans to coll that screamin' $80 mobo and Celeron CPU. Good thing that mobo has built in audio, video and LAN becasue the slots are blocked by all the other gear...
However you measure it though a "15 centimetre" fan that draws 1500 watts of power to cool a CPU is just stupid. Not sure if it is a slow news week for nerds but it seems the stories are getting silly (still a good laugh though). This CPU cooler would be good for the guy who buys that dumb "type R" PSU that was featured earlier. It seems the PC-modder crowd is seeing more and more "rice boxes" out there.
I can picture it now..."cool" nerd going a LAN party, driving a brightly painted Civic with that screaming 115HP motor, windows tinted almost opaque with 18" wheels on a goofy camber due to the chopped springs/hack lowering job and glowing neon and red "R" stickers all over the place.
In the passenger seat sits his 'leet gamer PC (the trunk is full of amps and speakers). It is an amazing construction of aluminum, plexiglass, neon and silver and gold Tremclad housing a type R power supply (as advertised on
Just a word of advice...if you "pimp your ride", then "pimp your PC" to match your ride, it is time to seek therapy.
Although CGI is in its decline, it is still used plenty.
Using this argument makes little sense becasue CGI is quite irrelevant today. It is the kind of back-handed comment that people use about languages like COBOL.
The "death of CGI" rather than the fact it is still twitching is the real argument to use mod_perl over many of the alternatives. mod_perl is probably the most mature of all the non-CGI web application platforms--it pre-dates everything except perhaps PHP (which looks to have emerged at around the same time). J2EE, mod_python, ruby on rails and so on are great performers I'm sure, but mod_perl and PHP are much more established, and literally nothing beats the vast library offered through CPAN.
I totally agree with you about the fun factor though--for all the hype Java solutins get, Java is probably among the LEAST fun languages to use--perhaps it is better than straight C or FORTRAN bt I'd even rather program in VB. Perl does lack a mandatory structure, letting the coder create ghastly code, but it DOES give the disciplined coder the ability to right extremely maintainable code as well (my Perl code tends make Pascal look free-wheeling).
The basic truth is this--it doesn't matter what the language does to make code maintainable, a language needs to be "fun" to make developers the most productive. That is why there are millions of essential, popular scriptsout there written in Perl. I think if Apache2/mod_perl2 had a nice, refined graphical IDE it could kick major ass on Visual Studio/ASP.NET...
The likes of Atari ST / Amiga / ... "could" have ended the MS monopoly
That was 1985--a different time. Microsoft was far from being a monopoly. Wordstar was battling upstart WordPerfect to maintain supremacy in the Word Processor market. Accounting was still commonly done on CP/M and Apple II computers running VisiCALC spreadsheets. Although it was the biggest player in business PCs, IBM still commanded less than 50% of the market. IBM was almost non-existant in the home market--that market was dominated by the "big three": Apple, Commodore and Atari. The PCJr was laughed out of the market because it was complete garbage compared to the superior offerngs from Commodore and Atari.
It was IBM that was shunned by the home market--Apple, Commodore and Atari were not only accepted but very much embraced by the market. IBM was very successful in business because they presented a logical argument to businessmen--one big trusted vendor to hadle all your computing needs. If you spent megabucks to get that mainframe from them it only makes sense to get the PC from them too. In the home IBM was NEVER dominant--by the time the platform became popular in homes it was compatibles that took that market (Tandy/Radio shack probably sold more home PCs than IBM at first).
MS was also very successful, but mostly through licensing technology to IBM and others. Until the 1990s MS (like Intel) had no brand image at all--people who owned Commodores or TRS80s might've vaguely remembered the name from the copyright notice on the bootup screen. Their OS was nothing remarkably different than CP/M, their word processor (PCWord) sucked and was a bit player against WordStar and WordPerfect. Their spreadsheet (Multiplan) equally sucked against the kings Lotus and VisiCalc, both in terms of functionality and market share. They had no GUI except some plan to introduce the pile of crap Windows 1.0 "real soon now".
So why did MS succeed? Because they learned from their mistakes (even if it took a vew years and three versions) and Commodore and Atari didn't--those ones got much more right at the start but not only didn't learn from their mistakes, they repeated them with magnified intensity. While Microsoft was gradually building its brand and improving its products, Commodore and Atari were wandering aimlessly. The other reason MS is now dominant is that they "innovated" faster--I put quotations there becasue they borrowed or acquired innovation and their products were not tied to the success of one hardware vendor (even Commodore Tandy and Apple were licensees). The others suffered from "not invented here" syndrome and tried to engineer everything in-house and ket their designs closed and very proprietary. Commodore would've died much quicker if it stuck to what it was doing and didn't purchase Amiga--and when they did they bungled everything about it except the hardware itself.
Just because MS is big doesn't mean it will be there forever--it used to be smaller than Commodore. It'll just be a bit harder to knock them off--not because they are big but becasue they are smart/savvy. It'll also most likely take a collaberative effort and an open software/hardware design to do it.
IBM/Microsoft DOS was based on CPM, an open source/free OS
Gary Kildall must be spinning in his grave right now. CP/M was a PROPRIETARY operating system by Digital Research. Maybe there is an open or public domain incarnation today, but it was very much proprietary when DR still extisted. The STANDARDS were open in that the BDOS calls were pubically available, and CP/M variants ran on multiple platforms (8080, 8085, Z80, 8086, 68000) and CP/M machines were usually open architecture S100 machines. You could definitely not obtain a copy of CP/M legally for free nor could you see the source code without a special agreement and extra cost.
The BIOS for the IBM PC was also open
ummm...no it wasn't. Even the BIOS calls weren't 100% fully published. Phoenix and Compaq developed a compatible BIOS against the wished of IBM (it was the one and only part of the original PC that wasn't an off-the-shelf component in a design a small group of hoppyists could easily replicate). The way it went was like this: a group of people disassembled the IBM bios and wrote a detailed specification of all the entry and exit points of all the calls and what effect they had on the system. Then a separate group of developers at different company (Phoenix) who had sworn a legal oath that they had never examined an IBM PC used that specification to create the first IBM compatible BIOS.
It wasn't really Microsoft or IBM that created the advantage of which you speak at all--they merely took good advantage of "open architecture" and the co-operative efforts of others. When it comes to the creation of the industry, others did all the work and IBM and MS used their marketing savvy to take maximum advantage and profit (the ones who did the work were not marketers obviously).
When IBM finally realised that a little firmware was not enough to keep a lock on the market it was too late--they no longer steered the direction of that market. The MCA bus was technically superior to EISA, but it was closed and incompatible and IBMs share of the market they created was less than 50% or at least fast heading that way.
Don't confuse open architecture hardware platforms with Free/open software--they both have an advantage in that information is more free to move about, however control oof the design and direction of the former is still firmly in the grip of a select few hardware vendors: Intel controls the bus and motherboard dimensions, Intel and AMD the CPU and chipset, ATI and NVidia video and so on.
Microsoft has one of the best track records in the business for backwards compatibility. ...except they don't. My employer guarantees 100% compatibility with 20 years of future products as a condition of sale. We have fully supported software that communicates with hardware that was made in the 1970s. IBM has that sort of track record of long-time compatibility and support for its mainframes too. That kind of track record for Microsoft would mean being obliged to offer support for stuff like BASIC on the TRS80.
.ini files and registry settings specific to legacy apps. You might never run the DOS version of Simcity from 12 years ago on your new system, but there is code in the current windows that was placed there specifically to make that one app run. All of that legacy support is quite a hodgepodge at times.
That being said, MS DOES put a great deal of effort into backwards compatibility--to the point of including a DOS emulator in NT4/2K/XP (WoWExec) that is so seamless most people would never think that the aforementioned OSes are no more compatible with DOS than Linux is (it just happens to have really good emulation). There is a blog by a microsoftie called something like "the old new thing" that explains the lengths MS goes to to maintain compatibility with popular legacy apps.
There are two problems with the efforts MS has put into legacy support: Firstly, it has done a lot to make their codebase cryptic, nearly unmanageable and sub-optimal. This is a problem the likes of IBM and my employer have to contend with as well, except that DOS and NT were not engineered with then intention of being the core of a product for decades. As a result, you get a massive blocks of code,
The second problem with MS Legacy support is that it tends to be rather selective. In the past, when there was a very popular 3rd party app that sold a lot of copies of Windows (certain desktop publishing packages come to mind) legacy support was done without question. when MS Office sales are slumping...well it looks like time to add a few more features that 0.001% of users asked for and use them as an exuse to break file format compatibilities. The thing about IE7 beig "too advanced" for anything older than XPsp2 is another one of those cop-outs. A little company and a non-profit foundation managed to make more secure browsers with innovative features that runs on multiple platforms and MS can't use their billions to engineer something that works with multiple versions of a SINGLE PLATFORM? Bullshit. They are trying to accelerate the elimination of Win2k because it is limiting their revenue potential.
I understand that legacy support is expensive and that MS is beter than a lot of SW companies like Red Hat (not that that is totally Red Hat's fault--they just don't have the resources). The difference is that Free software often continues to work on anything it'll compile on, and if you do have to upgrade you don't often have to pay through the nose for a highly disruptive upgrade. The IE7 compatibility issue is artificial--MS could EASILY make it run on win2K with its resources and say "there is no official support--use at your own risk". They just made design decisions to deliberately create critical dependencies on XPsp2. Even more than concerns about support costs, MS wants to boost stagnant OS sales.
Problem is, that makes IE7 an expensive browser for someone like me, whose only MS OS is win2K. Firefox is free in all senses of the word, so IE7 makes for a pretty weak justification for an OS upgrade when Firefox is much more convenient to get and I don't need to re-install my OS.
The feature article is from the mainstream press. The author has basically come to the realisation that there is no such thing as 100% secure/safe web browsing. This is something all competent IT professionals know--in fact the only "safe" computer in terms of security is one that is never connected to a network and is not physically accessible to anyone but yourself (pretty much impossible to do of course). Any IT professional who believes otherwise is not competent in computer security.
There are degrees of "safeness" however. Rationally explaining this does not pull in readers of the mainstream press of course--the local news knows being a bit more alarmist than required will boost ratings. The fact is that IE and Microsoft Windows are architecturally flawed. ActiveX object, BHOs and the like in IE grew out of ancestors like COM, OLE etc, which have roots in the Windows 3.x era when the Internet was not even on MS' radar (remember as late as the end of 1995 BillG thought an MSN based on its own proprietary infrastructure could compete with the open standards of the Internet). As a result, the components that make IE powerful are wholly unsuited for a networked environment. Windows XP/Server 2003 have stable VMS-inspired underpinnings that make them acceptably stable, however for compatibility reasons and due to MS culture it retained higher-level interfaces and deplorable security model of DOS/Win3.x/9x/Me.
By contrast, the Mozilla team threw out the unmanageable legacy Netscape spaghetti-code and re-architected from the ground up at a time when ther was already a good deal of awareness of internet security. Linux and MacOS X are rooted in UNIX heritage. Although it has a longer history than MSDOS, UNIX was designed form the start for a networked environment. The combination of these browsers and OSes are thus inherently superior regardless of their marketshare because their very foundations are better.
Microsoft can build the biggest, deepest moat around its house, put bars on the windows and doors, and add a layer of brick to the walls, but the creaky foundation will still crack, shift and leak and allow toxic mold to creep in. The F/OSS house may need repair from time to time, but it is much less likely to be condemned for sitting on crumbling footings.
Yes, Intel and AMD publish info on their processors--this is exacly why they are successful. There have been proprietary processors in the past as well. They are not common knowledge because they were big flops. The IBM PC model 5150 that spawned the architecture we use today was NOT the first PC that IBM made--that honour goes to the IBM model 5100. Why was it a complete flop? Well, IBM did NOT release detailed specs for the processor of the 5100, or any of the hardware for that matter. They kept the instruction set and everything secret! Non-IBMers could only program it in high-level languages!
FYI open SOFTWARE isn't very new either--in the early days where computing was a more purely scientific/academic endeavour coders tended to share their code with relatively few restrictions. When computers became popular and commercial software got proprietary--the GNU movement was actually reactionary in a sense--a desire to return to the early days of true sharing and collaberation.
There appears to be an analogue situation happening with PC hardware. In the 70s there was little effort put into protecting designs--Intel and AMD are just following that legacy with their CPUs. The Homebrew club members proudly showed off their clever designs. Computers even came with schematic drawings (even pre assembled ones)! By the 80s hardware vendors started going down the proprietary path (encrypted Atari7800 cartriges, The Macintosh, IBM's MCA bus, etc) until we have absurdities like ATI and NVidia GPUs, "soft" modems and wireless chipsets and so on that you cannot make work without proprietary information.
Open ANYTHING isn't new...it's just the relisation that success of an idea is ultimately limited or impossible in an environment that is too closed.
Oh, and you might not want to talk about FPGAs becasue they're "uncompetitive"--that is not the point--I personally HAVE obtained HDL code and configured an FPGA and "RAN" that code for far less than $1000. Could a pre-fabbed ASIC be obtained cheaper? absolutely...probably for one tenth the cost and far less trouble...but it is unmodifiable. I can take a RISC CPU core and other bits and pieces and make my own customised design for next to nothing...almost as easily as making custom software apps. The arguments about TCO of MS vs GNU are continually shot down for software--the same can be said about hardware...there are far more important benefits to open-ness than just the cost factor once you've achieved a certain critical barrier-to-entry.
Nice to see someone as important as IBM realising the importnace of open HARDWARE. I've found that until recently the concept has been overlooked or even derided. Even open software advocates didn't "get it" and said it could never work, becasue hardware is different--the argumant was that hardware isn't something individuals or small companies could influence becasue of the high cost of entry, and big companies needed to make money off licensing closed IP to fund development and production of new hardware.
This day and age, such an argument is complete BUNK. Hardware design is done on computers and chip specifications are more often than not specified in VHDL or Verilog--the "source code" of hardware if you will. Not only is design and simulation within the reach of even hobbyists, the end result is very similar to software in characteristics. While IBM is not completely opening things up to the point of showing the "source code" of the Cell processor, it is a great step to see all the specifications etc. without encumbrances.
Quite frankly I'm surprised the open source movement hasn't advocated open hardware much more vigourously. After experiences around NVidia and ATI and Intel you'd have to be a fool not to realise that open hardware isn't just an interesing idea, it is NEEDED for the success of open software.
Must get lonely out there in the middle of Montana or North Dakota or wherever you are, hours away from any other civilisation. You should get into town more--despite being "fascist" the good citizens of those towns really are nice people.
Seriously though I'm guessing you have a bit of a warped sense of humour or you are a masterfull troll. Besides the fact that there ARE health and safety concerns associsted with prostitution and scrap vehicles, a typical community code is quite reasonable. If all people were reasonable and considerate there would be no need for such codes.
Alas, a small but significant minority of people are complete jackasses. Given the chance these people will paint their houses hot pink with purple, orange and yellow trim and park a rusty old truck with two flat tires on the front lawn under the shade of their 12-foot C-Band dish (the one that pipes hard-core porn into their living room which they view with the volume on so high everyone on the block can count the orgasms).
These codes might not be about health and safety, but they are about the right of law abiding citizens to live in realtive peace. Unfortunately, there are other types of jackasses who are the opposite--these are the ones that insist on hearing their favourite pins drop after 9PM, mow their lawns in a checkerboard pattern and wish everyone would do the same, and hold up city business with requests for more traffic-calming measures, more noise bylaws and more regulation of every little thing that slightly annoys them.
If you find you are living in a redneck hellhole, or nazi-Stepford-suburbia there is a way to stop the madness. It is called "civic duty". Get involved in politics--especially at the local level. Attend the odd community planning meeting, endorse non-wacky candidates for office and vote for them...or at least vote for cryin' out loud. The latter form of jackass knows this already, and that is why California is so whacked that a government led by Conan the Barbarian is actually a vast improvement over the previous rock-bottom situation with Davis.