Forgot your password?
It's funny.  Laugh.

All Hail Bloatware 290

Posted by Hemos
from the hehe-bwahah dept.
Zarn writes "In Tuesday's Slate edition Andrew Shuman, in his article The Love Bloat, argues that the problem with bloated software is that it isn't bloated enough and that we, the customers, are the ones demanding bloat! " Heh. I'm wiping a tear off of my check from laughing so hard - Jonathan Swift, here we come.
This discussion has been archived. No new comments can be posted.

All Hail Bloatware

Comments Filter:
  • Dancing paperclips... Actually I do believe people demanded this. And this is one of the reason why I am afraid of the masification of GNU/Linux: Perhaps we will begin to see dancing paperclips and similar appear everywhere.

    It's just like the GUI-based Caldera installator. I am not saying it is a bad thing, I suppose it is nice... I don't know. Anyway, I can't see a *real* reason why having that GUI would be better over having a menu-driven text installer. But the problem is there are persons, lots of them, who believe the GUI is a real improvement over the menu-based interface. Can I say bloat? The depressing thing is many persons actually think that having a GUI makes it easier to install the OS.

    I don't have problems with user friendliness, but I just hope massification won't turn GNU/Linux into just another OS designed for idio, erm, umm, user-friendly, expert-hostile. Please, please, please, please, don't add a dancing paperclip to GNU/Linux installation process in the name of user friendliness. The dancing paperclip pretty much shows how extremely annoying `user friendliness' can be to expert users.

    And no, the GPL and the fact that we can change the code won't help in this case.

  • (2) Monolithic design, which is NOT a feature. MS Word has features targeted at lawyers (and useless for everybody else), at accountants, at writers, etc., etc. You don't need most of them, but get all of them anyway. Pluggable modules would have been a much cleaner solution (you are a lawyer? plug in the "Lawyer" module...)

    Actually this is one of the things that they were trying to do with COM/OLE/etc. Maybe they'll get there yet but I rather doubt it. It looks like the usual couple of big company problems: (1) too much legacy code to rewrite, and (2) a lot of internal groups with different goals and not all that much desire to work together.

    The potential is still there but some of the Linux component platforms may have a better chance simply because the component support will exist before the applications are to bulky to change. Of course there's the eternal question of which component platform is today's favorite...

  • > Bloat is bad in that it adds complexity which is the enemy. Insofar it consumes computer resources it is tolerable.

    Exactly. You can get a 20GB disk for around $300, so 200MB of storage costs about $3. Anyone know what 5-10MB of disk space cost in 1988? I bet it was a lot more than $3.

  • That sounds more like a problem with Visual Basic and your 'windows weenie' to me. I wrote a Pong game that uses the Keyboard and Direct Input, and it's only 100k including 78k of bitmaps. it loads in less then one second, and only 'nonstandard' DLL it uses is dinput.dll, witch is part of directX.

    It is *posible* to write non-bloated win32 apps, if you feel like it
    "Subtle mind control? Why do all these HTML buttons say 'Submit' ?"
  • cpt kangarooski writes:
    If I may ask, why don't you use a DTP program? Quark is a good choice, and lots of people like PageMaker (although it's being replaced by the new InDesign program) and they're both a jillion times better at this than Netscape, a spreadsheet or WordPerfect.
    A wonderful idea! Please tell me where we may purchase the Linux versions of Quark and PageMaker.

  • Excuse me, but why is bloat in a competitor bad ???

    If it weren't for MSbloatware, we wouldn't have $0.20/MHz CPU, $1/MB DRAM, and $20/GB disk. As someone (esr?) pointed out, Linux owes MS a great debt for making the fast hardware market cheap by its' volume.

    Where would we be if MS-Win9* ran just fine on a 386/16, 4MB RAM, and 100 MB HD? It'd still cost $800.

    -- Robert
  • and I used DirectInput for the mouse, windows messaging for the keyboard.
    "Subtle mind control? Why do all these HTML buttons say 'Submit' ?"
  • No offense intended, but you missed the point. If the paperclip were simply a graphic, there would be no problem. I've changed many of the icons I use in WordPerfect. But the paperclip is further work that MS is doing to bury the system so that users have little to no ability to configure systems to their own preferences. I tried to use Word97 at my old school but couldn't figure out how to make the paperclip stop appearing. All I wanted was a help system that waited for me to ask for help. WordPerfect let me do that, so I installed it (at a cost of $39 for the full suite--something my department chair was very happy about.)

    As for "hotdog users", well, I think that you're referring to savvy users--those who take the time to learn about their tools and prefer craftsmanship to heavyhandedness. I've used keyboard commands in WordPerfect for years not because I feel cool doing it, but because I leave my hands on one input device rather than two. It's the same reason that I use one remote for my television and vcr--I don't want to be bothered with two of them. There are times when a mouse is useful, but when I can get things done faster on the keyboard, that's the way I will go.

    No one at WordPerfect made anyone memorize keyboard commands. When you need one you can choose to remember it, look at the keyboard template, or go to the help system. I just don't want to deal with a program that forces me to use one method--this is what I've experienced when forced to use ms programming. It just doesn't work for me.

    Last point: If you don't like bloat, don't buy it. There is only one language that commercial companies of any sort understand and that is the language of dollars and cents.
  • As I was reading this page, I suddenly, without any action on my part, got a Blue Screen of Death. (I am, unfortunately, on a Win95 machine right now.) What was the culprit? Microsoft's fat, lazy Java VM - a Java app was running in the background.

    Time to trim the fat (or VFAT, perhaps?) off of M$ products.
  • These are the same people who use MS Excel (or Word) to store record-and-field (read database) type info because they don't know what a database is or does.

    What, you mean that you are not supposed to use a spreadsheet to store record and field information? Why not? I mean, what else is it for?
  • Emacs is just an editor. All it does is accept input from the user. Well, that and run Lisp programs.

    Don't all programs just accept input from the user? To tell you the truth, I've never seen "just an editor" applied to Emacs. Usually it's the "kitchen sink" that comes to mind.

    But anyway, speaking about bloat. I just started a new XEmacs process on a Solaris Sun box with 196Mb of RAM. Did a 'ps' which in my case evaluates to 'ps -e -o user,pid,pcpu,pmem,vsz,tty,comm | sort -r -k 3,3 | more' and lo and behold: my new XEmacs process (no files open except for scratch) takes 4.3% of my memory (pmem: that's the resident set and is equal to 0.043*196 = 8.4Mb) or 10072Kb (vsz: allocated space). I don't know about you, but from my point of view these are pretty high numbers for "just an editor".

    Note that I am not complaining -- I have enough memory and XEmacs is one of the more useful things I've run across -- but it is not lean-and-mean by any count.

    Also, the initial point was to contrast the one-program-does-all philosophy and the many-small-tools approach. Emacs does use other programs, sure, but the design goal of Emacs was that you never have to leave its environment -- shell, mail, compile, etc. are all available from within Emacs. Contrast this to the design of Unix: for example the 'ps' alias above uses three Unix programs to achieve the result I want.

    Anyway Emacs is just a front end to other programs.

    No. Shell can be thought of as a front-end to other programs, but I don't see it as useful to think of Emacs in these terms. If all you want to do is to call other programs, use shell scripts -- no need for a Lisp interpreter to be involved.

  • It will.

    I've heard they're moving to XML.

    (A bell should have gone off, warning you that Microsoft is intent on making yet another proprietary perversion of a standard.)
  • I would like to see a graphics program that can do it all. I have PhotoShop5, CorelDraw7 with Kai's Power Tools, LViewPro, MS Paint (natch), and NetSketch (which is awesome for lettering). My home computer is now low end - 133 MHz, 1.6GB HD.

    I have all these programs, and not one does it all
    (and I'm by no means a newbie at these programs).
    LViewPro eats jpegs for breakfast, and loads very fast. I use it primarily for compression, converting and cropping. PhotoShop5 handles layers beautifully - it's how I made plates for a Shockwave movie - while in CorelDraw, the object and layers is more complex and even mystifying.
    Neither PS or CD lay lettering down cleaner than
    NetSketch, and MS Paint is what I use to put pictures inside the hollowed out letters from Netsketch. If I could have all these things rolled into one graphics package, I would be so happy! But then, it would probably take forever to load - filters, plug-ins, there you have it...

    I also tried PSP and found it disappointing...but it was a trim little program nonetheless!

  • I'm not as sure how correct I am, I was just postulating and guessing, mostly.

    > The "feature" of not being bloated doesn't sell
    > because it has much less "features" compared to

    Could it be that the world has gone crazy? Why won't people use the ordinary $20 screwdriver that 's made out of solid stainless steel by a respected company instead of the multi-faceted piece of junk from wallmart costing $30? I can't answer to that. Perhaps this is why there is natural selection, and the windows users, in a short few millenias, will fall by the wayside to more rational thinkers.

    No, if rational thinking was a desirable trait, humans would have developed it already. Back to my NT workstation (which I use to telnet to a Unix.)

  • "As for "hotdog users", well, I think that you're referring to savvy users--those who take the time to learn about their tools and prefer craftsmanship to heavyhandedness. I've used keyboard commands in WordPerfect for years not because I feel cool doing it, but because I leave my hands on one input device rather than two."

    I was a WordPerfect user and I, too, used to know the keys by heart. (I still remember F11 - View codes!). I'm not sure what you mean by "hotdog users" though.
  • I think the only program I've seen that kept old versions for sale was Corel, with CorelDraw. When CorelDraw7 came out, you could still buy 5 and 6, and 5 was bundled with Lexmark printers (at least here in Canada). Why not check out a college or university computer store? By the time Netscape Communicator came out, the UofC Microstore still had 2 versions on the shelf.
  • As for fast lean web browsers, chimera, arena, and amaya all come to mind, except they've all been in beta-state with no development since 1996.

    Forget about Amaya as a browser. When I used it a few months ago it seemed unable to display any pages at all correctly. It makes a decent editor, however.
  • > I'm much happier with the UNIX way of having small applications that do just enough.

    Bingo!. You prefer the "many small utils that do one thing and do it well" approach over the "one big app that does everything" approach.

    Realize that not everyone prefers things this way. I also like this way of doing things, but my mother (for example) does not.

    I use her as an example often, because she is a "typical" (is there such a thing?) PC user. She teaches chemistry at college level, and uses MS-Word to make up her exams, MS-Excel to keep track of her students' marks, etc... and Netscape to surf the web. She logs in to her AIX account and uses Pine to get her mail, and is quite happy.

    She has no interest in using any apps other than those 5 I just mentioned (Word, Excel, Netscape, Pine, (VT-Terminal)).

    Even just using MS-Word to make up an exam, if she has to ermbed a molecular diagram or something in the document, she will use Word's built-in "MS-Draw" thingie (even though it sucks rocks); she will use the equation editor extensively and all the rest (actually, she uses a lot of the built-in features).

    And you know what happens if she needs to embed an image that is more complex than the MS-Draw sub-applet can handle?

    I head for the hills -- fast!

    The amount of complaining and grousing she will do at having to use a different program to draw the image, then import it back into her document is more than I, for one, care to handle.

    She's not stupid, she is quite capable of doing it. But she doesn't want to. She wants to be able to open a single application (MS-Word), and create everything she needs -- from scratch, if needs be, then print it out, without ever having to use another program.

    And most of the time, Word allows her to do this. Which is why she's happy with it. And why I wouldn't even try to convince her to use something like vi, that can't even "format text to a given width" in and of itself.

    It's a different philosophy, because it suits her needs better.

    You (and I) prefer many small utilities that work together. She prefers a single large app. Neither approach is necessarily "better", just different.

    So, perhaps you "cannot accept that many of the "features" in (say) MS Word belong in a word processor."

    I can.
    - Sean
  • The always-informative Andrew Leonard has whipped up a quick response to this article. Read it here. :) []
  • Aside from that, I have yet to see a useful application of java or javascript on a webpage.

    I went to and played their games. They use a java application which takes a long time to download. However, once it is downloaded it works great. I played chess with an 11-year-old. Isn't that useful?
  • Well, I am a newbie at graphics programs, or at least use them almost never. Does the GIMP not do the sort of things you're looking for? When did you try it (if you have)? It might be more feature-ful now -- from what I understand, the GIMP is very modular, and the features you need might have been developed since you tried it out.
  • by eponymous cohort (8637) on Wednesday July 07, 1999 @06:05AM (#1815042)
    Every program expands until it can read mail!
  • "Err, uhh, last I checked, stuff like commctl.dll was considered a system *library*. Just a resource for programmers. Using this thing doesn't rely on tight connections between the OS and Apps folks."

    No, but the commctl.dll IS tightly coupled to the OS version. Which means that the same code that is included in the Win98 system library was duplicated in the software in order to get the gimmicky menus and widgets to work on all OSs, instead of just linking dynamically to the version that was out there. If I'm a Win95 user, I think I can deal with not having the Win98ish menus and gui items. MS threw in the kitchen sink for the "neat-o" effect (which has no effect on me whatsoever).
  • "In Excel 7.0 open a blank sheet and use your arrow keys to get to the 95'th row, highlight the entire row by clicking to the left of columnA and then hit TAB to col. B, then go to help about Excel and hold down Shift+Ctrl and click on tech supt. This should give you a doom like window. arrow keys move you around and d,c look up and down. if you turn away from the stairs and type EXCELKFA the wall will disappear and you can go into a room with pictures of the developers."
    - Sean
  • Consumers only want to buy an upgrade if it contains new features. But they also want backwards compatibility. Hence, bloat.

    Once upon a time anything that ran in more than 512K of memory was bloated. Anything that required more than a 14.4K modem was bloated.

    Love bloat. Bloat is your friend. Accept your bloatedness.
  • Yes MS Office is bloated. So is Visio, so is Corel Draw, so is IE4 and so is Netscape Communicator.... so is Just any popular non-shareware package you care to name.
    There are lots of marketing driven reasons that this is the case. M$ Office is only a 'prime' example because it has beaten its competitors by becoming more bloated. Yes, I am one of those consumers who when faced with a choice between word processors (10 years ago) chose the one with more whistles and bells. M$ did a better job at consistently coming up with more of them and it 'won' the war of the office suites.
    [ But you all can go back to bashing M$ now ]
  • Uh...I think bloat is driven by Marketing departments. Marketing can also take it's fair share of the blame for the poor quality of some of todays software. Today you have to pre-announce a product (MS is already talking about the Neptune system that is slated for release in 2001 :) and load it up with more *features* than the other guys stuff if you want to survive. Sad but true.

    Bloat goes away when big software companies figure out that they can make more money by selling a base package for $x and then offering a buffet of other options for $y. Sorta like buying a car. I would love to have a lean mean version of MS Word (I need a little more than wordpad). But noooooooo...I've got to sit and wait everyday while NT plods through loading up an amazingly fat application of which I use about %10 of it's *features* Quite frankly Xemacs isn't much better.

    Bloat is driven by Wall Street..not by bad coders.
  • by haro (34457) on Wednesday July 07, 1999 @06:13AM (#1815048)
    In Andrew Leonard has an article [] about it.
  • Ford Expedition -vs- MS 2000 ?? yeah right.

    Ford Expedition: Sure, it's too damn big, guzzles gas, and has annoyingly high headlights that bother other drivers at night. BUT I can still jump in the driver's seat, start it up, and go pick up a 6-pack without so much as an extra thought.

    MS 2000: Similarly, too damn big (200 MB), guzzles CPU time, and has the famous dancing paper clip. BUT to write a letter, I have to fumble through the menus, uncheck every box that has the word "auto" in it, kill the paperclip, and do an ungodly amount of work before I can just type text.

    The point is, Micros~1 does not understand the far-out concept of ease-of-use. They prefer to TELL me the easiest way to type a letter (anyone else had that f@#king paperclip say "It looks like you're writing a letter...". He's worse than Jar Jar for chrissake), and tell me which 'features' to use, instead of just letting me get my work done.

  • The development tools that are used contribute to bloat as well. In the early 80's, when PC's and other computers didn't have much memory, developers often used Assembly Language, resulting in very lean, fast programs.

    Of course writing a big app in assembly will drive you insane. As PCs got bigger, C and C++ were used, and developers increaingly linked in bigger and bigger libraries. This helped contribute to the bloat.
  • Speaking as professional web developer(must fix user info here) I've seen a couple of uses for Javascript. Mostly for when you want to change a form element and have other form elements change, without sending the whole damned thing back to the server again. Another use was for surfing through a file structure that the client wanted to set up, more gracefully than normal directory indexing(and more securly as well).

    However, Javascript has one major bad thing against it: most of our clients demand that we support Internet Explorer 3.0 And, as any programmers out there know, IE3's DOM *sucked*. It sucked hard and it sucked a lot. So until IE3 goes away, we can't use a lot of Javascript. Which means we have to, at times, come up with work-arounds that feel like kludges.

    --John Christensen
    Applications Developer
    Auragen Communications
  • Bloat happens because you hire cheap inexperienced
    programmers that you can treat like indentured
    servants. Bloat goes away when you have a global
    network of experienced talent working on the
    code, looking out for each other.
  • >"Bloat is the American dream: bigger, better, and everywhere all at once. Supersize it!"

    >Sadly enough, this is quite true, and quite disgusting.

    Couldn't agree more. If you look at almost every aspect of American culture, that's what you see.
    Las Vegas is the epitany of it, I think :-)that's why Hunter Thompson subtitled FLLV "A savage journey to the heart of the American Dream".

    Funny tho, that supersize it attitude can have very negative effects. The quest for "more" and "faster" in the world of agriculture provided the factory farm, which killed family farms, forces animals to live in an unnatural and stressed out way, introduced hormone and antibiotic abuse, not to mention a rampant spread in bacteria (pig farmers interestingly enuf have a higher level of e. coli in their intestines than average people).

    Bigger is certainly not better, esp when you're running a slow computer like mine at home. It cranks along and pauses like a faithful old dog! But if something is perceived to be bigger and shinier, than it certainly "must" be better!

  • I didn't have to much trouble finding out how to get rid of that thing IIRC, I just asked it 'how do I kill the paperclip?' It seemed to work pretty well.


    I totally never thought of that. I spent literally hours, as I used 5 different RegEdit sessions, and it took forever.

    I'm kinda surprised the little monster helped you. If you were to ask "How do you kill SamIIs?" I probably wouldn't be the first to help you out.
  • Or the complainers single out features that they never use, such as AutoSummarize in Word or the Journal feature in Outlook (that can slow even the fastest computer to a crawl). My advice to these complainers: Turn these features off or ignore them.
    I have a better idea. Give me the source code and let me ./configure those stupid features right out of the executable!
  • ---Quote---
    The day that Microsoft fails to convince you to upgrade--i.e., to buy a product that the malcontents call bloated--is the day that Redmond becomes a ghost town.

    That day happened for us about 2 years ago when we decided not to upgrade to Win98 on one of our boxen (others running one of several Linux distros, currently Caldera 2.2 and RedHat 5.2). Now, where can I load up my truck with all the worthless (to them) hardware?

  • This article is scary, since I don't think the author (a self-confessed Micros~1 programmer) is kidding.

    >Sadly, it is you, the customer, who demands bloat, forever clamoring for new features.
    > The day that Microsoft fails to convince you to upgrade--i.e., to buy a product that the malcontents call bloated--is the day that Redmond becomes a ghost town.

    In other words, "the customer demands" that Micros~1 stays in business and keeps hauling in money. Yeah, right.

    > Most bloatware complaints come from users who own 2-to 3-year-old machines. They don't understand that the new (bloated) versions of software are meant for the new 400-megahertz machines [...] not their Pentium 133 doorstops

    This would be OK, _IFF_ there was any form of document compatibility between versions. Otherwise, it's just a forced-upgrade circle jerk with the CPU manufacturers. "Sorry, your 1997 car doesn't work with the 1999 gasoline". AutoCAD is another program which regularly pulls this scam, and I think it's about time for customers to stop accepting this philosophy.

    > The elegance of the Windows 98 operating system is that it runs practically every application from the DOS days and all those goofy Windows 3.1 programs.

    Insert your own sarcastic reply here.

    >Software companies take your wish lists seriously, and then make them happen.

    So in closing, which customer asked for dancing paperclips, and can somebody please hurt him/her???

  • The point is not making a nice small hello world program... if you want one, go ahead and write one in perl, but nobody will care...

    The reason everyone uses "hello world" is that it has essentially no user code, which makes it a good benchmark on the amount of code a certain compiler will link in just to write to the screen.

    The question is how small can you make your program if necessary. If hello world compiles to 50k, then if you need a program that fits in 40K of memory and actually does something, you need to find a different compiler or language.

  • Not so much. Typesetting is not the same as layout. Typesetters would create blocks of text, while compositors shuffle that text around in paste-up, before everything goes to film.

    Of course, computers have changed all that.

    Now they've both sort of merged into a bigger and better version of layout. Text is actually editable while you're compositing, so we have the freedom to change things ourselves. Typesetting has not done too well though, because with so many documents created on computers from day one now, there's not a whole lot for them to do. Mostly typesetting now means making sure that styles are all properly in place.

    TeX however, from all that I've seen of it, is not a layout program. It's a typesetting program, and very good at it. They're not the same thing, even though we're heading towards that.

    Typesetting deals with the appearance of TEXT, while layout deals with the appearance and arrangement of all of the elements on a page. Text comes in big blocks like everything else and is not really treated as special.

    If you think that the typsetters have had a rough time of it with the whole DTP revolution, just wait and see what's happening now. New techniques are being developed for going from computers directly to plates, and even from computers to the actual presses. This does not bode well for the strippers or plate-makers.
  • Not quite. In AR-style static libraries (the good old .a files) the linking granularity is object file level. (eg. the .o file level) with traditional linkers.

    There are linkers which work at much finer granularity these days. Some work by dealing with code in a not-fully-compiled intermediate format. The intermediate code is linked, inlined, and generally mucked about. Then, after linking, a final code generation step produces the final output code from the linked intermediate code. Such an approach is a huge win for languages like C++, because you can inline methods from a library into your program and other nice things.

    Not that I'm switching to C++ anytime soon.


  • How does clean/tight code result in bloat? Sounds mutually exclusive to me. What you seem to be saying is that clean/tight code results in more bugs than bloated code, which is something else altogether and I wouldn't want to get involved if you decide to start telling coders that their nice slim programs must be full of bugs while MS Word is relatively bug-free by comparison. What you're saying is probably irrelevant. Not only would I not agree with you, but the whole thing would get real ugly.

  • is sounds like he is useing windose and GIMP on win32 is buggy and doesnt do everthing the linux one does. for example i can't get script-fu to werk at all, on win32 THE best reason IMHO to use GIMP on win32, cause i just havent seen scripting done in anyother program.

    #include "standard_disclaimer.h"
  • Hooray for ad hominem arguments. Without them, the Anonymous Cowards around here wouldn't have much to say.

    If you think about what he said, he has a pretty decent point. Without adding new features, how can Microsoft justify selling a new version? Answer: They can't. It's not that farfetched to believe that they might start taking a lot of suggestions from a few consumers here and there that would otherwise have never even been considered. It might be something that a couple of people want, but Microsoft has hundreds of millions of customers that don't want or need that feature. It gets tossed in anyway to add another bullet to the feature list on the back of the box. Looks impressive, no?

  • I've become so numb to sarcasm I don't know it any more. Was this sarcasm or just idiocy?
  • My theory behind bloat is as follows:

    Back in the "good" ol' days of DOS, that boot loader trying to pass itself off as an operating system, any application that needed to do something had to do it itself. Want to talk to the modem? Include code to do it. Want a spell checker? Thow that in. What ever you need, throw it in. And the users learnt that everything should be integrated into one application. Hence,
    every application must now contain an integrated
    spell checker ( how 'bout it Mr. Taco? ), a drawing program, a program development envrionment, a file manager... or people get confused...
    Try explaining the unix theory of small compatible programs, do one job well, to the average person on the street, and see what happens. Retraining a person in computer usage is nearly impossible.

  • Since we are on this topic, I'm trying to get people to test a program [] I made. This is Windows only right now, but it compresses a Win32 Executable and adds a stub program to it that decompresses the program in memory when it runs. This turns out to be a very tricky thing to do because of dll imports, resources, thread local storage, and so on. Anyway, if you have any programs that don't work with it, mail me...
  • ask a stupid question: where are these easter eggs?
  • I didn't have to much trouble finding out how to get rid of that thing IIRC, I just asked it 'how do I kill the paperclip?' It seemed to work pretty well

    of couse, you need to *use* the paperclip at least once to get rid of it
    "Subtle mind control? Why do all these HTML buttons say 'Submit' ?"
  • You do realize, of course, that your hundred and some megs of ram are why it loads fast? Load up sysmon some time and take a look at the allocated memory. It's probably sitting around 130 meg. That DRAGS a computer with less ram to its knees. Sure, Word loads fast on my machine too ... it also crashes if I edit a 6 meg document. Why shouldn't it edit a 6 meg document, it can insert 6 meg pictures, can't it? Microsoft adds a lot of toy value to its software, but power-hungry users are left begging for REAL software. Unfortunately, we (the power people) are the minority ... that is, we're NOT Microsoft's market. That, and most power-hungry users just go out and buy more hardware to make the software run. I'll pay an extra $100 for a better video card to make my games go from 8 fps to 50 ... but to get a real-time spell checker? hello?
  • >libc5 weighs in at 600k

    >libc6 weighs in at 3mb!

    #ls -l /lib/
    -rwxr-xr-x 1 root root 1170636 Jun 15 10:30 /lib/*

    Huh? 3 meg? Looks like about 1.1 meg to me.
  • Unfortunately, windows.h in Windows references every possible function ... so including it (which you almost must) will link your program against all library functions -- some of which are NOT dynamic. Also, since Windows does DLLs so poorly (not versioned in the filename), most big apps are linked statically or come with their own copies of the same DLLs to put in their local path to make sure nothing gets broken. Ever searched a Windows machine for MFC*.DLL or winsock*.dll ? ... have a good one!
  • This one had me rolling, really. Customers are asking for bloat? I've yet to meet one. Frankly, if it's really expected of us all to have a PIII/500 and 256MB of RAM just to run a word processor, a spreadsheet, a half-baked database and a broken contact manager/mail client, then I for one am content to fall behind - I do as well just installing something like StarOffice that does 95% of the same in useful functions and has a quarter of the disk footprint.

    More frightening yet, though, is that best as I can tell, the author believes everything he's said. He doesn't realize that he's going to have to put on an asbestos suit to open his mail for the next several days (I for one won't be a part of this - he's below my response threshold, too brainwashed to bother). -Drayke
  • >Stupid features for stupid people I guess.

    Gosh, who buys all the software? Linux still doesn't have a majority market share by any stretch of the imagination. Not to mention when anyone proposes you guys pay for software, you act like it's a crime and break out Molotov cocktails...

    I of course don't mean you in particular, but I still mean to make a point. Slashdot is not the real world.
  • I have a K6-200, got 192MB ram (but had 64) and Word/Excel etc load basically as fast as IE5. eg. less than a second.

    The first time you load it, it might take about 2 seconds, but subsequent loads will take less than a second.
    And Word is pluginable (It's called COM/ActiveX)- and it is. It doesn't load up modules (spell checker, macro engine etc) until it's used for the first time per session.
  • I don't think he's supporting the point that he thinks he's making. Usually, consumers don't initiate featuritis, it comes from the marketroids. In order to survive financially, the company has to keep creating new reasons for people to buy their product (remember the yearly model changes from Detroit?). Fortunately (for the companies, at least), the average consumer will swallow the marketing drivel once the features are there. Then the consumers buy the product, and the company is happy.

    Fortunately, I am not a typical consumer, and I'm guessing many others here at Slashdot aren't either. I prefer functionality, simplicity, and elegance. That's why I drive an Acura Integra rather than a Lincoln LandMass, a Buick SnootyStreet, a Cadillac UrbanSprawl, or a Ford BunkHouse. I don't have, nor will I pay for, auto-dimming rearview mirrors, automatic headlights, auto-on windshield wipers, auto-adjusting seats, massaging seats, or even auto-shift transmissions.

    Every time I turn around, there's another marketing term that fabricates a difference in the consumer's mind -- "Intel Inside" -- "Where do you want to go today?" -- "Duratec V6" -- "Vortech V8" -- "StabiliTrak" -- "You may already be a winner" ...

    I'm inclined to think that the US consumer is inordinately susceptible to this marketing propaganda. I'm not really sure why, though. We're all supposed to be individuals here, and think for ourselves, right? Then why do so many people believe such transparent marketing scams?

    Or maybe I should ask the opposite question: why don't I always believe every claim or slogan from a salesman or commercial? Is there something wrong with me???

  • his is entirely offset however, by certain operating systems which require about half a gig to install comfortably. Moores law should not be used as a crutch or excuse for sloppy coding. I sincerely doubt that there is a single piece of microsoft software out there that could not be made in half the size with twice the speed.

    Quite unprovable; I'm not convinced that MS programmers are either sloppy or stupid.

    I did not say that Microsoft programmers are sloppy or stupid. I simply said that just about every Microsoft program could probably be made with twice the speed and half the size. That has to do with optimization. If they used assembly in key parts of the program, they could probably achieve both of these features, though it would take considerable time and effort, not to mention that it would make modifying the code a pain in the ass.

    Standards evolve when vendors provide more features; this is normal and healthy, and the way the industry is supposed to work. HTML mail is good because it's a simple, portable, non-patented extension; in contrast, Word document attachments are evil because they're complicated, non-portable and proprietary. Given that people actually *want* more structure to their mail than ASCII can povide, this is a natural direction for the industry to take. (Remember that we can't (and shouldn't) standardise before we have implementation experience!)

    HTML is far from the best choice for this though. HTML is, in many ways, quite limited in the degree of control it gives you over how your text is formatted. Not to mention that different browsers will display similar html differently. Attachments on the other hand, do not suffer this failure. Note that I did not say word attachments. You could just as easily use PDF, RTF, or any other format you please. You use the ascii text in the email to give a description of the attached files and a brief message, and leave the bulk of the data in seperate, attached files, from where it can be transfered to specialized programs (.cpp files to MSVC, .doc to word, etc).

    Most people don't know, at install time, which set of features they'll need; their requirements do change over time. Besides, would you really
    want to spend hours figuring out which of the thousand features you want, specially when the resource you seem to be optimising for (disk/memory space?) is in fact quite plentiful? (A gigabyte of disk is worth ~$30; how many man-hours is that worth?)

    First off, features can be packaged into components. You could have you "Basic Documents" component, with the base editor, spell checker, and grammar checker. Take on a "Label Editor" component, "Letter Editor" component, and so forth. Components could be installed from a CDROM, or over the net from a server (with proper authentication and license keys this would work). As for relying on large harddisks . . . remember that the more shit you dump on the disk, the longer the harddisk will have to seek to find data. And . . . let me put it this way. Would you rather have a 200 meg word processor and 800 megs of MP3s, or would you prefer one fat 1000 meg word processor? Harddisks are getting bigger, but so are data files. Use your resources wisely.
  • I think the paperclip is there because they got probably got a ton of people in focus groups who just sat at a blank screen for half an hour and then complained about "not knowing what to do." Its called the lowest common denominator.
  • I didn't have to much trouble finding out how to get rid of that thing IIRC, I just asked it 'how do I kill the paperclip?' It seemed to work pretty well

    I first saw the paper clip when a (then) cow-orker was taking a look at Office 95. We both loathed it at first sight. He told it to "F*ck off and die." It did its search for things we might want help with, and getting rid of the "Office Assistant" was right at the top of the list.

    I guess you don't even have to be polite...

  • Sorry, someone had to use that word in here. But at any rate, part of the issue is Unix-minded people looking at Windows-minded people. I know people who honestly believe that Outlook is a good program because it does everything that it does. Mind you, they run Pentium-II 300+ machines with 8gig or more of drive space and twice as much RAM as necessary (in my mind).

    Unix has always aimed at having dozens (or hundreds) of tiny single-purpose programs to do individual groups of tasks. Shell scripting is a very powerful tool in this environment because combining grep, awk, sed and find you've got more tools than Windows has already (using ANY built-in tools). The issue is ease-of use.

    Bill Gates wants "a computer on every desktop" and what he understands is that people don't want to write shell scripts. People here is defined as those who spend money on computers on average. We the *nix community do not fit into that category. We are not the vast majority of users. Therefore, bloatware gets built, not out of laziness or pressure tactics, but because of a philosophy of ease-of-use. Making a program intuitive and easy means adding hundreds of features to make it intuitive and easy to all the people who might come across it. emacs is intuitive to some people (I don't know who :) whereas vi makes sense to others. Some prefer the feel of jed or joe. In the MS world, people all use Word and Notepad. That's it (most don't even use Notepad).

    Most of the people I've trained in Windows have told me they didn't even know Outlook (which they may use daily) had sticky-notes built in. I like those sticky notes ... but to make them built-in, Outlook's code stays in memory even if you just want one stupid sticky note on the screen. People don't know the feature exists because the software's intuitive enough to get them working without having to explore. No exploration = no novelties.

    As people become more and more computer savvy, they start buying toolkits, like Norton or Nuts&Bolts to do things to their computer ... they want stuff to work with. *nix has those things built-in, the difference is that we, the anti-bloat community, don't appeal to the masses because they don't want to type

    /sbin/ifconfig | head -2 | tail -1 | sed -f 2 ' '

    to get their IP address. Heck, they don't care that they have an IP address.

    Should we fix this? I don't think so. I don't think that the Linux community will benefit from appealing to the general public because Linux will then become bloated as well. Linux is not bloated because it is limited in scope and tries to do what it does well. It's that different paradigm thing :).

    Enough of my rambling ...

    - Windows on the desktop, sure ... but can we keep the server alive for an hour, please?
  • The dancing paperclip is just a bitmap. It's there to provide a link to online help. That's what it provides me, anyway. In Office 2000 it can be changed to a number of different 'themes' if you (like me) find the paperclip an annoyance.

    Hotdog users are always going to resent online help. It threatens their guru status when people stop begging them for assistance, and start doing it themselves.

    For all the people I've talked to about it, the paper clip itself isn't an annoyance. We don't care for the notion that the help system is somehow made more useful by appointing an animated character as its gatekeeper.

    To put it another way:

    • I don't want to show off how studly I am by memorizing every feature of an application.
    • I don't want people "begging [me] for assistance". (I prefer that they have a good, complete, accurate help system so they don't bother me with their questions.)
    • I don't want 'themes' for the Office Assistant.
    • I don't want to be bothered by someone's idea of a 'friendly face' popping up to offer me advice on a program I'm already familiar with.
    • I want to press F1 and get the help index.
    • I want help to show up when I ask for it and at no other time.
    I don't hate "the paper clip" because I want a bad help system. I hate it because (for me, at least) it's a very poor interface to the help system.
  • Speaking of bloated easter eggs, I seem to recall that the one in Windows 95 had lots of images, and played about a 30 second WAV file.. that sucker would have at least taken 5 - 10MB of disk space and was probably installed on every machine no matter how 'lean and trim' you made your custom installation.

    I guess that would suck if you were trying to fit as much as you could on a smaller hard drive.

  • Er, if you have gcc you should have g++. They come in the same package, so I don't know why someone would build it without the C++ language.

    A more valid test would be:
    int main(void)
    printf("hellow world");
    return 0;

    This compiled with gcc (egcs) turns out to be 3012 bytes when stripped on Linux-ix86 with -Os and 2976 with g++ (after changing filename to hello.cpp).

    A more realistic test is to compile with -O6 -march=pentiumpro, which are the options I use all the time. In this test, a stripped C compile is 3012 (oops, must no be much to optimize there!) and a stripped C++ compile is 4028 (probably due to alignment for pentiumpro optimizations or something).

    Damn, I wish &lt and &gt worked!!!
  • by delmoi (26744)
    say what you want about netscape, but IE *is* really good. its modular, and a lot faster at loading pages then netscape, witch is sad I guess... I use them both, usualy just netscape, but if a page dosn't load right, I just go to IE. it relly does work a lot better :(
    "Subtle mind control? Why do all these HTML buttons say 'Submit' ?"
  • Actually, What I'd like with netscape would be for it to stuff a lot of its functionality into dynamic libraries so it doesn't even bother to load the JVM or JScript (which I leave disabled), the mail/news client (which I never use), CSS (which in NS 4.08 requires JScript, which I disable), the bookmarks editor (which I only use occasionally, the history viewer (which I use once in a while).

    Being a plugin architecture isn't enough if you automatically load in all available plugins into memory immediately.

    Maybe we need to throw mozilla through gnu ropes to get its paging and function orginazation coherent.

    Speaking of this, can anyone offer an explanation why the preferences dialog is SO BLOODY SLOW?

  • [to tune of "With a Backpack On My Back"]

    I love to sit and stuff my apps
    with great huge chunks of code
    and that is why performance is
    beginning to implode.

    Feature-eee, feature-iii,
    feature-eee, feature-i-i-i-i-i-i,
    feature-eee, feature-iii,
    my programs are obese.

    [idea stolen largely from a song sung by Garrison Keillor]
  • why I haven't used MS products in about 5 years.

    What is bloat? Bloat is the American dream: bigger, better, and everywhere all at once. Supersize it! From VCRs to food processors to Ford Expeditions, industry has historically provided consumers with features to have, not necessarily to use. How many of you have programmed your VCR? Minced carrots with your Cuisinart? Or gone off-roading in your SUV? Why should software be any different?

    Well, let's see. If that is the American dream, I must be an accidental American. I hate SUVs, I drive a small sports car. I don't own a Cuisinart, and yes, I have programmed my VCR (it's really not that hard, god bless Sony's industrial designers).

    Maybe that's why the software I use is different?

  • by bogado (25959)
    Who said that? I had a 386, 16m and it ran winword 6.0 and it had all the features that 90% of the people use today of their word 97/2000. With maybe one exception that it would not keep underlining words that it don't find in it's dictionary (3 out of 4 persons that uses word that I know HATE this, I actualy like but this is because I tend to make a lot of mistakes).

    I now use linux, and I hear my boss saying that he likes linux but it will not catch untill it has a office suite that is 100% compatible with M$ office. He says that people need that, and WP, SO and other word processor for linux don't render the word docs the same way that word does.

    The only problem I see in this line of argument is that not even word does render 100% the same in every computer, people usualy config diferent page size then your printer, have diferent fonts then the ones installed in your computer (and shure they will use theirs "mangalo extra bold italic 13pt" font in their email).

    "take the red pill and you stay in wonderland and I'll show you how deep the rabitt hole goes"
  • The problem with easter eggs is that they obviously didn't go through any formal review process. Flight simulators and the like may be an exception, but *some* easter eggs are nasty lawsuit bait that *no* company would ever let out the door.

    Anything that doesn't go through a formal review process is an unknown. Maybe it's bug free, or maybe it has a nasty bug that will make the entire product look bad. Run "flight simulator" and corrupt your disk. Or maybe it has actively malicious code embedded inside. Run "flight simulator" and have a copy of all "encrypted" Office documents ftp'd to a remote site, if you also have network connectivity.

    We simply don't know.

    The software vendor simply doesn't know.

    And that uncertainty, in itself, is enough to cast a dark shadow on the product. It brings to mind angry line workers who weld bolts into car body cavities so the victim will be plagued with an uncorrectable rattle. Or the angry construction worker who seals a carton of milk, with a small hole, behind the drywall so the annoying owner who comes by with his annoying requests will suffer from an unlocatable stench for years.

    I don't mind a simple group photo or list of contributors; in fact I think it's very reasonable to put that information under the "About" button so the programmers can show pride in their work. But anything beyond that raises serious questions about the quality of the software produced in that shop.
  • last time I checkd, (yesterday) 128Megs of PC100 sdram was only $58. very nice. that's about 50cents a meg. when I baught my computer in 1995, it was $50 a meg... hehe
    "Subtle mind control? Why do all these HTML buttons say 'Submit' ?"
  • Well...I don't use "mail", but I do sometimes telnet to the SMTP port ;)
  • I would suggest that writing code that is both clean and tight is possible, but extremely difficult - as it hinges upon getting the design just right. A great design will practically write clean and tight code for itself, assuming a competent coder.

    Of course, getting the design right almost never happens the first time, and is nearly impossible while specifications and requirements are changing underfoot. In most real-world situations one almost never gets a chance to fix the design.

    Given an imperfect design, one can write sneaky, impenetrable, yet tight code, or can keep things relatively clean -- and pay the price in bloat.

    I would much rather maintain the latter...

  • The article was not clearly written. The author refused to separate bloatware from overly-feature-laden ware. Two big differences. Bloatware is badly written code that sucks memory ad HD space due to bend-over-backwards-compatibility, overuse of legacy code, etc. Feature-laden (something I never connected with MS...) is having a lot of features.

    I'm all for features, but since when was MS more feature-rich than previous options or current alternatives?? Frontpage-- lots of buttons, but if you're making a serious webpage, where is the power to customize it to do exactly what is needed? GUIs in generall are feature-poor or at best overly complicated and hard to get to the important features compared to command-line versions.

    As for bloatware...ugh. It scares me. IT becomes harder to deal with and requiring bigger and better computers to do the same (...or less!!) work as before.

    PS: As per the autocorrect, a quick backspace will undo the autocorrect without deleting your typing.
  • The aging, ever expanding behemouth software product I make my living on suffers from bloat big time. Many of the reasons for this have been correctly identified already by Shuman, Kaa, and others here: featureitis, swiss army knife disease, designer drivel, acres of templates and sample datas, etc.

    But the probllem is compounded -- and the customer is screwed -- by process, schedule, and market imperatives.

    There is no time to do it right and meet the schedules. There is often no chance to coordinate and integrate with concurrent development efforts. Identifying opportunities for code sharing and consolidation requires dedicated time and energy that is usually allocated elsewhere.

    And once its been slapped together with spit and glue, there is no chance to backtrack and revisit the implementation due to fears of destabilization -- especially if the feature was adopted by the customer base. If you do get a chance to fix it, it has to be timed carefully at the benginning of a development cycle. And even then you get the attitude that "It works (sort of) goddamit and if you redo it, what'll we get besides bugs?" Management has learned the hard way to adopt a very dim view of "elegance".

    The problem is compounded again by consumer loyalty to market dominating products. New, better products cannot compete with industry standard gear on merit alone. Everyone knows M$Word, all documents are in Word format, all the corporate standards are built atop it,... The fact that product X can blow the doors off Word for this or that use means nothing. So everyone continues to buy Word and M$ continues to jam more stuff in the box.

    PS: Everyone needs a clue about the 'backward compatability' issue. Typcial developer mindset. Its not the user being stupid and using older, out of date gear and expecting the software to save their asses. The fact is that the end result of using software tends to be data: documents, databases, etc. The wealth accrued to the customer is the data. The software, installing it, living with it, upgrading it,... is just an ephemeral nuisance.

    The fact that until very recently the industry standard was to orphan old data had more to do with softweare vendor arrogance than anything else. One of the primary reasons our product has dominated its market is because of our consistent comittment to protecting the customer investment in their data. And document written by any version of our software on any platform can be read by any latter version of our product -- and no funky translation steps either. Just open the damn document. So blame my employer, among others, for the fact that customer are now saavy enough to demand backward compatibility -- and just deal with it.

  • Excuse me Mr. Shuman, but not all of us can afford a nice new system everytime a new Intel product comes out. Many of us have to make due with what we have until we can afford a new system. Becasue of Micro$oft and a few others, the typical lifespan of a home computer is greatly reduced, thus in the Micro$oft world your system becomes obsolete quicker

    I think his point is that you don't HAVE to use the latest versions of software. No one's forcing you to run Office 2000 on your 486!
  • and if it weren't for Microsoft, we wouldn't NEED them. It's like how the fellow with Claris on a floppy said:

    I don't think that guy had Clarisworks on a floppy... I think he had AppleWorks. As in, for the Apple II. One of the greatest programs ever written. Iit lacks functionality that I would want, but it's so simple and quick and dirty...

    Like, if I wanted to catalog all my comic books or records or something, or keep a mailing list or something, it would be so good for that. Now, for making a resume, it'd look like crap by today's standards. Maybe I should go buy a dman Apple IIe and be done with it and get Appleworks and get organized. It seems like it'd be even easier than learning EZbase and the other stuff I downloaded.
  • Look, Navigator is bloated on every platform, not just Linux. What an advertisment for cross-platform-icity: "Navigator! Bloated on over forty platforms!"


  • I hate bloatware, but, at least with M$ the problem is not the features, the problem is usually easter eggs. Why does Excel need a 3D flight simulator? why does word need a pinball game? why does outlook need a picture of the developement team. These are not features and I doubt they were on "consumer wish lists" if they are features they should be advertised as such. but they won't be because its just wasted code.

    anyhow that's my $0.02
  • Do you really think software developers add features just for fun, like some cackling tormentor?

    Yes, that's exactly what programmers do! This is especially true for people working on projects like MS Office - the more features the program has, the easier it is to sell as an upgrade. I fully expect many programmers on the Office team to sit around talking about what more junk they can throw into their suite. In fact, it would not surprise me in the least if there was some kind of monetary reward program for people who come up with the most ideas. Certainly anything that can be patented is rewarded.
    Timur Tabi
    Remove "nospam_" from email address

  • Bloat is not our friend.

    Yes, it is true that people are demanding more and more new features, and they also demand backwards compatibility, and these things do take more space.

    But it is mostly the fault of the software companies.

    Any large system becomes bloated. Just look any large burocracy. The problem is communication between all of the people who are writing the software. They don't coordinate, and they don't care about the system for the system's sake. They do thier job with the least amount of effort on thier part. If it's easier for them to make a new file format rather than stick with the old one, then we have a new format, and one more bit of old code for 'backwards compatibility'. Why bother coordinating this fancy feature with the one the guy down the hall is writing so that we don't suck up all of the processor. Why don't we write the same routine as ten other people because we didn't know somebody else had written it already for their own work? Why not assume that everybody who wants to run this program has a computer under 2 years old? Why not add in hard-coded limits to data sizes and whatnot just for the sake of convienience?

    Why doesn't anybody put any effort into making their software elegant, internally coordinated, optimized, and expandible for the future?
  • Actually, while I could be wrong, it's my understanding that emacs is pretty much just a lisp interpreter. It's the lisp programs that actually do everything. I think that things are a bit different in Xemacs. I'm not sure how much this is valid, but I do believe it someone, a plain xemacs binary with no elisp packages can't do all that much. But with those elisp packages, damn. I just recently discovered etags (man, meta-. binring you to a variable's declaration or a function definition really rocks). I wonder how much more there is in Xemacs.

    Well, anyhow, I think that the guy that you were responding to is right, though, a good portion of xemacs isn't xemacs itself but the lisp packages. If you were to remove all of the xemacs lisp packages, I bet you that it would have a somewhat smaller memory footprint. :-)

    Either way, the extremely modular design means that once the core is done, you get a stable program. I can't remember Xemacs ever crashing on me, and I use it for all of my programming. And if I was low on diskspace, I could get rid of all of the lisp packages that I don't use. That and there are very few bugs in Xemacs. Probably because it's a mature program. Funny, I remmeber someone talking about linux being based on 30-year-old technology. emacs is something like 20-30 years old, I think, maybe 10-20, and damn does it rock. I guess the Petreley was right when he said that programs shouldn't be measured by the yardstick of windows NT - normally programs do get better as they get older.
  • One was to register for a conference, and was used to check that the data was complete (not that this justified it). The other used it ust as gratuitiously, to get at the realaudio broadcast of foxnews.

    Aside from that, I have yet to see a useful application of java or javascript on a webpage. But then, I also beleive in jail time for blinking text or gifs . . .

  • The thing is, some linux, has gotten bloated also. Anyone every use netscape!? It runs slow on a AMD 400 w/ 256 megs of ram.
    Is there any open source browser that are faster, leaner, meaner, not as bloated or buggy as netscape? Just pictures and text, no java, no java scripts, no nothing , runs under X with pictures and text, that is all anything else in it, is bloat.

    If you just want pictures and text, then why not run Mosaic 2.0 or Netscape 1.1? That's pretty much all they they do. And they are small and fast. Nevermind that they won't display 90% of all web pages properly, since most web pages contain a lot more than just pictures and text now.

    No-one forces you to upgrade. If you want to use old software with fewer features, go for it.

  • by Wholeflaffer (64423) on Wednesday July 07, 1999 @06:52AM (#1815155) Journal
    muLinux is probably the most versatile and interesting one-floppy-disk distribution of Linux out there. (Mr.) Michele Andreoli has written some incredible apps, such as an http server that's less than 1500 bytes in size. You can get X-windows on a second floppy, too. If you want it all (gcc compiler, ssh...), you'll have three floppies to deal with. I've been having lots of fun with this package...and very little bloat.

    Check it out at []
  • "Look, Navigator is bloated on every platform, not just Linux. What an advertisment for
    cross-platform-icity: "Navigator! Bloated on over forty platforms!" "

    Navigator is bloated because it doesn't know what it is. "I'm a browser" "I'm a newsreader" "I'm an editor" "I'm a mailer" "I'm a JVM" "I'm a javascript

    I really don't understand why Netscape don't break it into components and market it as a "suite" of nice, small, streamlined programs.
  • by Bastard Child (66982) on Wednesday July 07, 1999 @07:09AM (#1815174)
    "Create [software] that even and idiot can use, and only an idiot will want to use it."

    I don't know who came up with this quote, but it seems to fit. Seems that bloated software is mostly for people who don't "get" computers, but who *do* use them every day. This accounts for over 90% of the computer-using populous.

    The downside is that the other 10% (or less) are nearly forced to use the bloatware also because the first 90% say "Here's that file I wanted you to look at, it's saved using Office 97." These are the same people who use MS Excel (or Word) to store record-and-field (read database) type info because they don't know what a database is or does.

    I spend approximately 5-10% of my time writing scripts that do data conversion because someone decided to use "Bloatware" to do a particular job rather than the "correct" software. Why? Because "It's so easy to use!"

    On the flip side, I *can't* use some of these products, because they're "too easy" to use. I still create HTML using a text editor. I'm the only one at my company (a multimedia company that produces web sites) who does this. Everyone else uses Frontpage, then wonders why the pages don't work right in Netscape. I will not use Word because it *insists* on correcting my "mistakes," and tries to anticipate what I want to do. If I put the letter 'c' in parentheses, it automatically converts it into a copyright symbol.
    If I wanted a freekin copyright symbol, I would have used charmap.exe.

    Enough soapboxing. =)
  • by Hard_Code (49548) on Wednesday July 07, 1999 @07:10AM (#1815180)
    I don't buy that bloat is caused solely (or even *predominantly*) by user-requested (user-wanted as per market study) features.

    For instance, who the heck asked for that damn paper clip and his moronic friends!!?? Who thought to themselves: "Damn it! I can't do a thing with this Office97...hey! I know! A dancing paper clip would really help me a lot!"

    Also...Users don't buy upgrades JUST for new features. In fact, I'd argue users buy upgrades JUST so that they can stay current enough to work with everyone (and everything) else. Maybe if Microsoft didn't break its old file formats and introduce new ones every rev then people wouldn't upgrade as much. I got Word 6.1 for free with a new computer and have to date not installed a new one. I can't use documents from work though (and NO, it was not my choice to use MS Office...).

    In the same vein, I think embedding HTML functionality into everything is stupid. I hate getting those bloat laden HTML messages from people in Pine, which I have to decipher for myself. Sometimes these messages include two attachments: one HTML attachment for bloated email readers, and one plain text attachment for normal ISO8859_1 readers. Over double the bandwandth of the simple plain text message. Most people don't even know this feature is on when they sprinkle bold, italic, and font tags all through there 1 line message and 6 line sig.

    Bloat is also caused by laziness. I guess nobody at Microsoft ever thought "Hey, if we actually took the 3 extra months to get this really tight we might have a product consumers would buy and not feel like it was being stuffed down their throat." The lazy approach to complexity is to just scale up the same techniques and practices you were doing before. It's easy to copy your O(2^n) parser from Bloatv1.0 to Bloatv2.0 to use with files 10 times bigger. It's harder to redesign the whole architecture.

    In any case, it seems to me the bloat coming out of MS is more due to two things: 1) Introducing useless technologies and additions to 2) Position themselves strategically for even more dominance,
    than attempting to add things users *want*. Heck, I've *wanted* a smaller, tighter office suite for years...they haven't added *that* feature.
  • Ah, Emacs. It's not just a text editor, it's a calculator, web browser, mail reader, lisp interpreter, integrated development environment, psychologist, word processor, dictionary, and King James Edition all in one! gack. I'm a big fan of the theory that a mail reader is for reading mail, and a text editor is for editing text; never shall the twain meet. Give me vi anyday.

    As for fast lean web browsers, chimera, arena, and amaya all come to mind, except they've all been in beta-state with no development since 1996. Arena would be great if it didn't segfault trying to load most web pages. Amaya would be even better if it didn't try to be something I don't need (an html editor) and crash whenever I try to follow a hyperlink. Chimera looks nice but I have yet to see it do anything Mosaic couldn't do. And the newest version won't even compile for me. Of course, the newest version was released over 2 years ago. I'd be willing to work on some of those older browsers, trying to get them to a functional state, if there was any interest. Anyone else? mail me [mailto] if you're interested in something like that. I don't want the newest and shiniest with all the features, like the Mozilla team, just something that works right and doesn't take up more than 4 megs of ram to run.

    Netscape is bloated because of the mail, news, composer, instant messenger, and everything else even vaguely internet-related built in. I remember that 3.0 was a lot better for not using up as much ram but I had to dump it because it was hideously unstable. Heh. Now I can't even surf without filling up my 32 megs of ram and watching netscape fandango on core.

    I'm really looking forward to Opera's linux release. Unfortunately, it's payware, but if the linux version is as good as the windows version, I'll shell out the $20. It's definately worth the money. Until then, I'm finding Mozilla to my liking. Everything except the hideous "chrome" bits. When I use mozilla, I only use the "viewer" part, with the bare-minimum user interface and the "my, that's alpha" feel. And it only consumes 10 megs of ram running. (heh. Only. I seem to recall running netscape 2.0 in 4 megs of ram sometime long, long ago.)

    Maybe I need to try Mosaic again. If I remember my specs right, it didn't support any of the things I dislike about the "modern web", things like animated gifs, java(script), CSS, dhtml, and frames. Maybe I should just get off my duff and start coding something better. Mozilla tries too hard to be like netscape. I want something for just plain old browsing the web. Is that too much to ask? Oh, and it has to have pictures. lynx is great, but I need to get my pr0n somehow.

    Enough of my ranting. Please feel free to point me in the direction of any other projects like this, or if there isn't any, I can damn well do it myself. Or die trying.

    Leapfrog, (

  • I am absolutely sick of hearing about how C/C++ causes bloat because of the size of "Hello, world" programs. Use the right tool for the job. If you want to write "Hello, world" use a shell script or Perl.
  • Best article I have seen on the causes of bloat in MS products is R.A. Downs's analysis of bloat in RegClean Version 4.1a Build 7364.1 []. In a program of 818KB, he finds 350KB (that's over 40% of "bloat," including unused cursors, dialogs, string entries, tool bar, menus, icons, etc. You might quibble with some of what he counts, but the basic point is powerful.

    A. Michael Froomkin [mailto]
    U. Miami School of Law,POB 248087
    Coral Gables, FL 33124,USA
  • Unfortunatly you are right.

    I come from a sales background dealing with the
    average "joe user." (I worked sales while getting
    my comp-sci degree). The only way to successfully
    make a sale and compete against you competitors is

    1) Find out the persons "needs" (read *wants*)
    2) Tie down the customer to those "needs" that
    map to a feature set in your product. Then
    hammer on the features that are unique to your
    product over the competition.
    3) Explain each feature, the benefit of it to the
    customer, and then tie down (i.e. get him to
    agree with you) and then close.

    It goes like this.

    Feature -> Advantage -> Benefit -> TieDown
    Feature -> Advantage -> Benefit -> TieDown
    Feature -> Advantage -> Benefit -> TieDown

    (i.e. close the sale)

    In the mainstream world of computing (joe user),
    it's features that sell. The feature of stability
    is hard to sell to the average user b/c they
    a) expect it to be stable (and if you try to
    push this too much they'll lose confidence in
    you and your product and go somewhere else
    where your competitor will tell them all
    their products are stable and get the sale).
    b) It's not glamorous.

    The "feature" of not being bloated doesn't sell
    because it has much less "features" compared to
    the "bloated software" (which normal people
    really care very little about until they have
    to upgrade). The software industry read *MS* has
    done a good job of ingraining people with the
    mentality that they NEED the latest bloat.

    High tech, enterprise companies are different
    b/c they have IT staff. But small businesses are
    the same as joe user b/c usually the suits are
    dumber than stumps.
    ***************************************** ***
    Superstition is a word the ignorant use to describe their ignorance. -Sifu
  • "if we software developers were really doing our jobs instead of resting and vesting our stock options, word processors would have already bloated into 99.999 percent reliable voice-recognition software. Your computer would have fused the functions of your telephone, television, and fax machine into one seamless whole. Your computer would have become the instantly searchable repository of all your correspondence, financial transactions, data searches, and phone conversations. Plus, it would be making smart connections between your data and actions."

    Hmmm, I thought your point was that
    satisfying consuder demand was paramount. But I'm not convinced that consumers *want* all their information tools merged into 'one seamless whole'. What's so great about seamlessness?
    I like modularity. Let's not forget, the entity
    doing the "seaming" has vested interest in seaming together useful tools with gargabe we
    don't want. Remember push-technology seamed-in
    with the browsers... wasn't that just grande; advertising seamlessly packaged with the
    browser (and OS?), pushed right into our faces.
    Is this what we really want?

    Do we really want our software to anticipate our needs. Or is MS and other corps. telling US this is what we need? When you type a search term into a search-engine, do you prefer the engine to respond to your direct request for pages containing keywords, or do you prefer the engine to figure out what you 'REALLY' are asking for?

    Modularity over seamlessness, Responsiveness over anticipation, any day.

  • Why doesn't anybody put any effort into making their software elegant, internally coordinated, optimized, and expandible for the future?

    I can give you a very serious answer to that. I write bloatware for a living. I am directly responsible for a handful of bloated DOS apps.

    And the reason is this: Nobody pays me to spend the time to rewrite a system from scratch. They just want features added. And they always want the lowest short-term cost they can get.

    It's just like evolution. Code is bloated for the same reason that your optic nerves are wired backwards, or that giraffes have an esophagal nerve that goes down, around their heart, and back up to the brain. Greedy optimization algorithms.

    For example (this is a real life example): I just added Yet Another Report to a clinic's billing system. The report in question is very similar to one that I did last year. So what I did was this: I copied the old code and modified it. Now the program has two modules that are very similar with a lot of (nearly) duplicated functionality.

    I could have rewritten the old code to be more general-purpose and called it twice: once from the "old" version of the report, and once from the "new" one. But this would have taken me a little bit longer. Likewise, you're probably wondering why I didn't just write it in a general way to begin with. Well, I try. When I know that I'm going to end up reusing some code, I'll do that. But if I don't know, then it's sometimes hard to justify the additional time taken to do the Right Thing.

    That's the problem with doing the Right Thing: sometimes it just takes a little bit longer. Someone has to pay for that time, and non-programmers usually don't get what I'm talking about when I bring up issues related to the "cleanliness" of code. Quick and dirty almost always wins over long-term maintainability and elegance.

    Oh, and there's another reason for bloat too: Once you have an installed base, you can never remove anything, no matter how braindead you think it is. Why? Because I never know how many (if any at all) of the end users are using some feature. Making the program better isn't worth the risk of getting complaints like, "I loaded the update and now my old inventory system doesn't work."

    FWIW, in amateur projects where I don't have to be accountable to anyone, I do try to do the Right Thing. Heck, I used to be a VIC-20 programmer who would actually spend hours on a program to make it 12 bytes smaller. The part of my brain that used to do that, is still with me, it's just not the part that gets the bills paid.

  • The reasons for the phenomenal bloat in many software packages (particularly MS's) my include user requests for features, but that's a minimal part of it.

    In order to sell a product that's competing against another, you have two options. You can advertise that it simply does its job better, that it costs less, or that it does MORE. "More" is generally the easiest to sell. "It works better" is kinda nebulous, and doesn't hold up against a product that has "100 additional features."

    Users don't ask for these additional "features"... software developers come up with them in order to better compete. This is especially true in the case of Microsoft, which is often primarily competing against itself. Need to convince users of MS Squeegee87 to upgrade to Squeegee2010 EX Plus Beta Turbo Edition Gold? You pump up the feature list. Saying "we stripped it down so it'll run faster than Squeegee87" or "Yeah, so we made it $30 cheaper" doesn't so much work when they already have a version that "does more".

    It's not the coders, mind you... I'm sure the coders would love to strip Office down to a clone of Notepad and a calculator. And this brilliant professional would have you believe that the coders do the design... generally not true.

    This article is a perfect example of the arrogant attitude that seems to pervade Microsoft and its ilk. "Why don't you just get a bigger machine like everyone else, you idiot?" seems to be their mantra. My reply: I will. And since it's running a real operating system and decent apps, it'll actually run FASTER than the 233MHz box I'm running now. I don't upgrade to maintain poor performance...I upgrade to better it.
  • by Croaker (10633)
    I don't know who came up with this quote, but it seems to fit. Seems that bloated software is mostly for people who don't "get" computers, but who *do* use them every day. This accounts for over 90% of the computer-using populous.

    Trust me, I work with people who don't know computers. This is not the reason for bloatware. People who aren't nerds don't give a damn about bells and whistles, as far as I can tell. They really care if the can use the software. If they can open it up and start using with the least amount of fuss, then they are happy. Things like wizards are what they crave, not the latest bell and whistle. Wizards aren't all that bloaty... they are basically scripts that just sit on the application, taking the user through steps he or she wouldn't be able to figure out by themselves.

    In addition, the vast majority of people in this situation do not have a choice as to what they use. They have to follow along what their IS department or computer manufacturer installed on their system. Blaming them is like blaming drivers for poor highway design.

    The major reason for bloatware is, of course, to continue the cashflow of the manufacturer. Force people to ugrade by selling them bug fixes (which they should get free) or make the file formats incompatible.

    Of course, there is also the programmer's seeming inability to declare a project "done." Take EMACS for example. It used to be a text editor. Now... it's... well... more. A lot more. Creeping featuritis is a disease that can be caught by OSS, as well as the commercial sector.

    Personally, I think that every application gets to an ideal point of features/bloat. Wordprocessors, for the most part, reached that level years ago. To hammer on them more is counterproductive. Other applications, such as some 3D programs I use, have plenty of room to grow... there is always more things you can do that will add realism, for example. These I don;t think of as bloat, as long as they are easily accessable from the main application. When you start to create obscure little corners of your application that take the user minutes of hunting to find... that's when you've gone into pur bloat mode.

  • by Kaa (21510) on Wednesday July 07, 1999 @07:45AM (#1815239) Homepage
    Well, not always -- he does talk about the elegance of Windows [shudder] -- but his basic point is valid: people like to have features not necessarily to use them. IMHO bloat is caused by:

    (1) One-program-does-it-all philosophy, which by, the way, is a valid design viewpoint. Emacs belongs to this school of thought, while Unix takes the opposite extreme (plenty of small interacting programs).

    (2) Monolithic design, which is NOT a feature. MS Word has features targeted at lawyers (and useless for everybody else), at accountants, at writers, etc., etc. You don't need most of them, but get all of them anyway. Pluggable modules would have been a much cleaner solution (you are a lawyer? plug in the "Lawyer" module...)

    (3) Feature competition between programs, which is driven by users: "What, your program cannot do a mail-merge to an index which includes animated GIFs and print out each third line?? It sucks, mine can do it!".

    (4) The need for backward compatibilitly. This is less visible in application programs and more visible in system tools which often must be bug-for-bug compatible with everything going back ten years or more.

    (5) The need to support all hardware under the sun. And the number of cool devices that you can plug into a computer grows and grows and grows and ...

    (6) In the trade-off between a clean/tight code and speed of development, speed almost always wins. In the current business environment projects that are 50% over budget and on time are much much better than projects that are on budget but 50% late. Basically, the slogan is: "who cares whether it is optimized, if it works, ship it!" (in case of MS or games it is often "who cares if it works properly, ship it anyway!")

    So I don't believe it is the malice of Microsoft or the incompetence of programmers that gives us bloated programs. Basically the definition of a bloat is "this program demands more resources than I expected it to". Having more resources available is a (not necessarily the) solution. Yes, Office 2000 needs ~200Mb of disk space to install. So what? I recently bought myself another hard drive -- it cost under $200 and is 10Gb in size. Do I care that much about allocating 2% of it to MS Office? Guess.

    Bloat is bad in that it adds complexity which is the enemy. Insofar it consumes computer resources it is tolerable.

  • by cje (33931) on Wednesday July 07, 1999 @08:06AM (#1815245) Homepage
    The author of this article has apparently fallen into the trap that so many other people have; namely, the mindset that the larger a piece of software is, the better it must be. I base this conclusion off of the fact that he seems to correlate bloat with "nifty features" such as voice recognition and "HTML mail."

    Is this the bloat you know? Because it isn't the bloat that I know.

    We're seeing a trend in software, and it's not a Microsoft-only trend (although one might argue that Microsoft is perhaps the best example.) When you see version 2.0 of a product that has double the system requirements (disk space, memory, processor speed, etc.) of version 1.0, it's pretty hard to explain away the bloat as being "nifty new features" if version 2.0 only provides a few new pieces of functionality. This author would have us believe that if the system requirements of a particular package double from one version to another, then that means that the functionality/usefulness of the new version is double that of its predecessor?

    Does anybody really believe that?

    For a period of about eight years, I used AppleWorks to do most of my home productivity work. For those of you that may be unfamiliar with it, it was basically an integrated applications suite that contained a word processor, a spreadsheet, and a rudimentary database. The integration was pretty nice; you could insert spreadsheets into your WP documents and the database had features like the always-useful mail merge. And it fit on one side of a 5-1/4" floppy disk, which was 144K.

    Now, nobody in their right mind is going to argue that AppleWorks is more powerful than Office 2000. But let's use the author's estimate that a full install of Office requires 200M of disk space (this seems a bit low to me, but what the hell.) By the author's assumption, this means that Office 2000 has 1,422 times the functionality of AppleWorks. Whee! Viva la Office!

    It doesn't matter, really; he's preaching to the pew. Microsoft has a large base of loyal customers who have demonstrated time and again that they're really not interested in issues like this. And quite frankly, maybe it's because most of their customers simply don't know enough to complain. (It's hard to suggest this without sounding like an elitist bastard, I know.) In reality, maybe it's pretty tough to blame Microsoft for consistently cranking out some of the most shameless bloatware around. After all, if their customers don't care, then why should they?

    The bottom line is this: if you have to "up" the processor requirements and grow your executable size by 20% to accomodate a new, sophisticated rendering engine, then that's one thing. But if you throw in a flight simulator, 600K of unused bitmaps, and a mishmash of other junk and then start complaining about "deadlines", that's something completely different. If it makes me a "whiner" or a "malcontent" to complain about the latter, then I guess I'm proud to stand among the rank-and-file of whiners and malcontents.

    Sorry, Redmond; one of the reasons I don't use your stuff is because I don't want to buy a new machine every year. And guess what! Thanks to Linux, I don't have to.
  • Even the WindowsCE devices are prone to feature bloat (that was the original point of this thread). The classic example: PalmPilots are strictly single-tasking environments, while WindowsCE still multitasks. Unfortunately, you can't see all the applications running in the background, so eventually if you keep switching apps as you would on a PalmPilot, the device crashes.

    Microsoft not only insists that it knows what the consumer wants, it insists that its way is right for all platforms. I don't want my VCR, my organizer, and my remote control to crash because they're running WindowsCE... but hey, if "the consumer" wants it, "the consumer" can have it!
  • Windows 3.1 apps that need WINSOCK.DLL to function don't work right under Windows 98. Windows 98 doesn't have enough conventional memory for most good DOS games, and there's no way to control a lot of the stuff that gets loaded. Backwards compatibility is a joke in M$-world.
  • >Why does Excel need a 3D flight simulator?

    Because Microsoft doesn't credit individual programmers unless they sneak it in. Most game programs have a list of contributors in the manual, down to the guy who beta-tested it for a couple of hours one day. Microsoft software? Rarely is there credit. So Microsoft programmers find a sneaky way to sign their names to the work.

    If you play the flight sim, you'll notice a scrolling list of the people involved in excel on one of the terrain features.
  • by Anonymous Coward
    The point of the paperclip was to make Office easier and more friendly towards people with little or no computer experience. Natural language help systems and attractive graphics go a long way to making people feel comfortable with these systems.

    YOU may not like them, but there are people that do. I'd wager that there are more people that DO like them than don't or Microsoft would not be continuing to improve them. (The agent in Office 2000 has had quite a bit of work done to it).

    Have you been to a trade show (such as Comdex) and watched MS pitch these features to the crowd? They *LOVE* this stuff.

    This is the reason so many purely technical based solutions fail in the marketplace. They fail to take into account ergonomics. They fail to take into account cosmetic appeal. Given the choice of two similar products, the vast majority of users *WILL* choose the "prettier" product over the "simple" product, even if the prettier product is inferior in most other ways.

    THIS is why Microsoft keeps winning. Sure, their monopolistic practices help, but they would STILL be winning without them.

You can not get anything worthwhile done without raising a sweat. -- The First Law Of Thermodynamics