Fog Creek Software
Discussion Board




Bloatware - one more time

I'm sorry to come up with this once again, but since Joel postet another comment on Bloatware, I have to comment once again.
The argument that all users may use only 20percent of an office suit (I highly doubt that it is that much. Think of all the users that want to do mainly word processing and maybe a little bit of Excel, but buy the all-in-one version of office that comes with applications they won't even start a single time) but that it's always a different set of 20 percent may be true. But it helps a lot if you have to pay for the license yourself instead of using it at your company. Anyone who has seen an Office Suite like Applixware will just laugh at Joels infamous article about bloatware. And true, back when I bought my copy of Applixware, it bothered me that it didn't have the neat underlining of misspelled words feature. But since it cost me about 10-20 percent of what word would have cost, I could live quite easily with that. Plus one became addicted to using it because it was so streamline-designed and fast that I had already started it and finished my first paragraph when Word was still busy showing you a splashscreen (not literally, but you get the idea).
One might argue that Applix has disappeared from the landscape of office suits by now and that the market has decided to go for the feature-rich MS Office once again. Well, the problem is that sales have unfortunately nothing to do with quality. And in Microsoft territory, there are some additional 'mechanisms' that influence market shares. Sure, you might be interested in having your business going - but I suspect most people who visit this site want to build better software. Besides, do you really believe that people will spend a few hundred dollars on a product just because it has a word count??? And a product like Star Office which is  *free* and people still buying MS Office just demonstrates that the whole market is not sane.


PS: One more thing: In the Rick Chapman interview, Joel stated that there was no such things like secret Windows APIs - well, take a look at Wine development and how many undocumented interfaces they found so far...I guess those APIs were listed right next to the Unicode chapter, if you know what I mean...

Stefan Raspl
Monday, April 08, 2002

Stefan, I guess, most private MS Office users did not pay the normal license fee for their software. They either got it together with the hardware as an OEM version (and if you spend 1000 $ + for a computer, 50 $ for the software does not seem so much) or they copy it illegally somewhere else. So the price is not really the issue there. I bought a cheap German office bundle once (forgot the name of it), that looked an felt very close to the MS Office and was very nice to use. Unfortunately, you could not add footnotes with the word processor. So I could not use it for my research papers and that was the end of it. My next PC did have MS Office again.

Personally I think that bloatware is fine as long as you have a possibility to easily hide all the features you do not want to use, I would like better (more, easier...) customization features with big software bundles. It is very difficult to find solutions for this, which are easy enough to use, though.

As for the secret Windows, I agree with you. There are quite a few "Undocumented Windows" books out there and there sure is a reason for this. I normally refrain from using them, though, because I figure whatever is undocumented might change from one version to the other without keeping compatibility.

Have fun,

Jutta Jordans
Monday, April 08, 2002

Lets face it the only reason you need word is because so many documents are now in word format.

If you are going for a new job "Send me your Resume" really means "Send me your resume in Word format because if you dont you wont get the job because I wont be able to read it".

If everybody else is communicating by using X then if you choose Y then all you achieve is to remove yourself from the conversation.

Tony
Monday, April 08, 2002

PS , Yes word is bloated but my PC has a 40G drive and half a gig of RAM (like many others), and it opens word in the blink of an eye, and as for all those features that I dont use, well maybe my wife uses them, or one of my three kids, I dont know either way I can achieve what I want to with Word.

Tony
Monday, April 08, 2002

Kinda wanted to note that although Word is commonly used, a person can create Word docs and read them with WordPad (comes with nearly all PC's from 98SE to ME, as far as I know) and OEM distributions of Word or Office Suites often eliminate functionality. I paid for my copy of Office 2000, and although all I have used at this point is Word and Excel, I feel like I paid a good price for a decent piece of software.

However, on the other side of the coin, when I use my Mac (still my favorite computer, even though it is old and clunky [try more than 8 years old without a single upgrade]) I use Clarisworks (which is now Appleworks) and I have never had a single problem. That was most definently not bloatware, and it had every single feature in Word (and many more), and every version of since my old copy has barely changed. While there is a growth in the size of the program, most of the growth is in a key piece of functionality, the ability to translate documents from foreign formats.

I guess what I am trying to say is that while Word and Office are larger than they need to be, it is not a bad thing or a good thing. The programs work which is the important thing. If people were smart and understood how to write programs and interfaces in a user intuitive manner, often alternatives would seem more palatable than using Word. In fact, as I grew up on the Mac and Clarisworks, when I switched to Word, I had some very annoying problems, like: how to change margins, how to format line spacing, etc. But once I found the method to do the task, Word woked for me. I would use an alternative to Word if I liked it better or it offered me some thing I desired. However, large as it is, Word works and I won't change without a reason besides my hatred of Microshaft.

A better example of software cancer, aka bloatware, is probably the system software used by a computer. I had one cd for my mac. one. That cd held everything except my games. Right now, in order to complete a basic installation on my windows me computer, it requires 7 cds. And half of the stuff that is installed, I remove. I remove it without ever using it. And, of that stuff, 90% of it is a group of worthless introductory "how to use your computer" crap. A good 50% of that is only deletable by manually deleting it, instead of using the windows add/remove programs setup. That is not intuitive. People bash the Mac for not being intuitive, but at least they provide a standard way to do everything and hold themselves to that standard...

I guess me too tired. Rant is winding down. Microshaft bad. Apple good. Simple good. Overly complex bad.

PS, should mention that linux is not a good choice for a laptop, least not mine. Try 72 hours straight to set it up, and then the damned video driver doesn't work right.

Masterlode
Monday, April 08, 2002

So why don't people really say why software gets bloated?  The reason is that companies that need revenue need to have a reason to have customers upgrade.  Most new features may be nice but how many are needed.  I still use Office 97 and have no reason to move up.  It has all the features I need and doesn't chew up my memory.  Big harddrives and lots of RAM is no excuse to think bloatware is OK.

The other thing that bugs me concerning "upgrades" these days is the amount of bug fixes from the last version they contain and expect customers to pay for.  Would you expect to pay for things that go wrong on a new car in the first year?  I don't but we all do with software.  I have seen a bad habit of less and less minor upgrades that are a free upgrade in exchange for more frequent major upgrades.  Because M$ has now been releasing a new OS it seems like each year they and the thirdparty software providers think they can force all of us to upgrade and pay full price to keep OS compatibility.

So I do believe that the 20% rule does exist for some but the reason behind the bloat has nothing to do with cusomer demands only shareholder value.

Chris Woodruff
Monday, April 08, 2002

Do any of you remember Microsoft Works? It came free with most PC's and it costed like $30 if you wanted to buy it. However, it didn't have much success.

People want to use the most feature rich and newest app. Ask yourself, why are there so many people using Adobe Photoshop for there basic graphic editing needs? And why are there people using MS Word when they could've used WordPad? Because they go with the best, they're afraid that sooner or later they'll need that important feature.

Roger
Monday, April 08, 2002

To quote from the original article:
>In 1993, given the cost of hard drives in those days, Microsoft Excel 5.0 took up about $36 worth of hard drive space.

>In 2000, given the cost of hard drives in 2000, Microsoft Excel 2000 takes up about $1.03 in hard drive space.

While that might be true, hard disks (at least the ones regularly used in desktop PCs) have NOT become ten times faster in that time period.

>Ohhh. It eats up all your memory. I see. Actually, well, no, it doesn't. Ever since Windows 1.0, in 1987, the operating system only loads pages as they are used.

So you're relying more and more on the hard disk, whose speed has not (by a wide margin) kept up with the speed increase of the processor.

>And I don't think anyone will deny that on today's overpowered, under-priced computers, loading a huge program is still faster than loading a small program was even 5 years ago.

I'll deny that. I even have specific examples. Eudora 3.1 loaded (and, once loaded, responded) faster on my Mac IIci than Eudora 5 does on my G3/233.

>In fact there are lots of great reasons for bloatware. For one, if programmers don't have to worry about how large their code is, they can ship it sooner.

Unfortunately, programmers not only don't worry about code size, they often do not seem to worry about speed, either. And I think 'speed' is closer to the core of "why people hate bloatware" than size figures alone.

Too much software seems to be tested only on leading-edge hardware, the developers forgetting entirely that many customers will run their product on not-so recent computers. Very few companies replace their computers when they're three years old - and even a three-year old computer may not be recent enough...
The cynical will suspect an evil pact between software developers and the hardware industry, aimed at keeping up computer sales volume. This might be ridiculous, but pact or no pact, that's the effect of the "who cares about bloat" software development practice. 

Moore's Law is not universal, and you can't use it to defend code bloat. You'll have to take the *slowest* component in the computer into account, not the processor speed.

Ask yourself (or better yet, test it), "is this product usable on a five year-old computer?"

Harro de Jong
Monday, April 08, 2002

Taking a broader perspective, picture yourself the buyer for a large corporation. Nevermind the "nobody ever got fired for buying an IBM" herd mentality, let's say you're comparing one office suite to another feature for feature.

Being a buyer you can't know what particular feature set any one person will use, so buying anything with fewer features will mean potentially disappointing a segment of the population, thus bringing their productivity down as they either do something by hand or hunt down and purchase software that does what they need.

It therefore makes sense to go with the most feature rich software you can find.

It seems a lot of this argument is from the home-consumer point of view, and at home I don't use MS Office. I recently installed MS Works only because I wanted to use a spreadsheet that was in Works format (that calculated standing waves and harmonic resonances in a rectangular room). Otherwise I have no need for an office suite, so I don't install one.

Other considerations are the stability and future growth of the suite (i.e. will it be around in 5 years? can I get customer service or information as needed? What about patches and updates?). How accepted is it and/or does it reliably read and write to formats that can be shared across the organization and other organizations. I.e. I'd hate to make my vendor go out and install Star Office because of minor inconsistancies in the Project format that causes all of their dates to say August and all of ours to say September.

Mark W
Monday, April 08, 2002

Simple is hard

To make an easy to use UI is difficult to program.  That may be a key point many can take away from this web site.

Now that so many people have gotten over the hump of learning MS Office (Word and Excel)  to do something different requires learning a new UI. 

This, and not bloatware or any of the other things we discussed here, is the real cost of non MS office products.

I am not claiming MS office is easy or hard to use, just non-rtrivial to learn to get proficient; The point where users don't have to ask for help for most things they do.

As far as why the non-bloated ones fail:  Can't open office version X documents.  This is one of the key sticking points on the MS anti trust case. 

But you are right, you shouldn't sell a Word Process based on what it doesn't have.  Gnumeric (the Open Source Spreadsheet designed by Gnome guys) claims to have more features than Excel.

Adam
Monday, April 08, 2002

Adam - agreed. A large corporation switching to, say, Star Office rather than MS Office means a new set of Want Ads that read "Star Office Skills" instead of "MS Office Skills." The way things are you learn one set of skills and you get a job using them. If every company had their own office suite, it'd be impossible for anyone to go from one secretarial job to another, productivity would go down because of the ramp up time it takes to learn a new software, etc.

I'm sleightly exaggerating the point, but there certainly is something to be said for consistency, and perhaps the fact that neither the big boys (microsoft) or the little boys (dozens of office suite producers) want to create a consistent UI means it's more difficult to switch. (Didn't Joel have something on this? Making it really easy to switch back to the old product if you don't like the new product.)

Mark W
Monday, April 08, 2002

In regards to the "send your resume in Word format" comment, I'd like to point out that the one format which seems to be far more ubiquitous than MS Word is Adobe's PDF.

When I was doing the resume passing a while ago, I had been in the middle of a LaTeX 'phase' (I'm a math student, and "MS Equation Editor" is tiring).  Everything I wrote was done in LaTeX and I would use pdflatex (distributed with MiKTeX for Windows) to format all my .tex files for 'normal humans' to read.

Ever since I've been using MS Word and Excel more often, I find myself sending things in PDF format as a habit.  It solves the "I don't have that font you used" problem quite well, and is readable by just about every platform I've come across out there.

Also, it's a fairly good form of 'copy protection' for your document.  If I spent a long time formatting my resume, brochure, etc for aesthetic reasons, I'd be real upset to find someone ripping it apart and putting their name, product, etc on it.  Do employers need to edit your resume for any reason??

Just my 2c CDN.  That's 1-1/3c USD, so maybe you should find a more valuable opinion elsewhere.
;-)

Cheers,
Chris

Chris L
Monday, April 08, 2002

Hmm. You might even be able to subvert evil placement agencies who doctor your resume to make you more marketable to a corporation. I had this happen to me as an interviewer. I was looking for someone with super-strong JavaScript skills. They fwd'd on a resume (in Word format) of someone who looked promising.

He brought along his own resume and gave it to us during the interview and it didn't match the one the agency sent us. He swears it was the same one that he sent to them, and I believe him. The agency deleted the "6 months" in front of JavaScript.

I had this happen to me the other way on a paper application. I have two cats and the application for an apartment asked if you had any animals. The real estate agent told us to lie about it and just sneak them in, but I insisted on putting it on the application. Well, when the manager of the building asked if we had any pets I said I did. I saw the application he was looking at and they whited-out that line and photocopied the sheet!

Perhaps the measure of bloatware should be the number of viruses that can be introduced on the system. Linux has fewer viruses than Windows, MS Office has more than Star Office. I hear Acrobat is getting it's first round of viruses with the new PDF scripting language.

Mark W
Monday, April 08, 2002

Uhmmm...how about ASCII? Our HR department would throw both PDF and Word resumes in the trash.

pb
Monday, April 08, 2002

It isn't the size of the application that is the problem, it's the way that these applications take over your computer in the name of "helpful features". If Word was just huge and had a lot of features I didn't use, that's one thing. But in reality it is huge AND, among other things, it designates itself as the text editor in my e-mail program, puts a cartoon dog on my screen, and prompts me to insert the Microsoft Office CD whenever I try to edit a page in CityDesk. Its tentacles reach everywhere. In short, the pompous programmers who wrote it assume that it must be the single most important piece of software on my computer.

I always feel that these features do cause problems by changing my computer's behavior in ways that I don't expect or want, simply because "somebody" finds it useful. It becomes even worse when you install several of these self-important programs. (Real Audio. <<Shudder>>)

Nate Silva
Monday, April 08, 2002

Masterlode wrote:
> Right now, in order to complete a basic installation on my windows me computer, it requires 7 cds.

lol.. Where'd you get that install from, AOL?  I have an MSDN subscription with all of the Microsoft operating systems on less than 7 CDs.

Marty Eichelman
Monday, April 08, 2002

To explain the 7 cds.

2 cds hold the system and associated crap.
2 cds for Office and Office tools.
1 cd for cd burning software.
1 cd for applications like Photoshop, and WinRar.
1 cd for printer drivers.
1 cd for data that must eventually be installed.

so 7 cds to get the sys in the basic config. And one more for backup, which is completely full. I have some 40 other cds of programs and software that I use periodically.

Masterlode
Monday, April 08, 2002

You know what's kinda funny,
all these articles use roughly the same formulae as
highlighted out by Joel.

"I know, I think I might just go find something to review, modify joels article and see if anyone notices..."

only kidding,
but still, they seem to get published.

Ray
Monday, April 08, 2002

I think it is reasonable to expect hardware to enable the implementation of bigger and bigger executables.

Some time from now, not too far away, we'll be remembering the days when we only had a gig of ram in the same way  that some of us remember to 48K upgrade kit we purchased for our Sinclair ZX spectrums in the 70's.

A big engine is expected to be able to pull some weight.
Boundaries are made to be broken, the word "bloat" means nothing.

Tony
Tuesday, April 09, 2002

pb - "Our HR department would throw both PDF and Word resumes in the trash."

Well good luck hiring anybody but anally retentive Lynx fans. Sheesh...

DB
Tuesday, April 09, 2002

Tony, the zx Spectrum was released in 1982, I am old enough,    I don't want to feel even older ;)

And yes, I bought the 48k ram upgrade because of all those bloated games that required 64k

Andres
Tuesday, April 09, 2002

I've also noticed that increased speed, RAM and hard drive capacity don't lead to programs loading any faster. Sure we have nicer looking video games, but my word processor loads in the same amount of time that it took in 1993. On my 4 year old machine - a Pentium 233 - modern programs take forever to load.

Bloat's Law states that an upgrade in hardware requires an equivelant bloat in software such that load time is no faster, and v.v. (I just made that up.)

Sure the software does a lot more than it used to (try playing Everquest in 1993), and many video game designers are designing for tomorrow's machine so their stuff won't look outdated by the time it hits the shelves. My co-workers and I have joked about a "Photoshop Machine" or a "Photoshop Hard Drive" because the thing's gotten so darn big.

It seems like the basic programs we use aren't changing, they're just getting more sophisticated and more capable as time goes on. Sure they could be written in assembly language and operate 50 times faster, but it would take forever to do and nobody would buy it because it couldn't compete with Photoshop.

(I'm using PS as a non MS example on purpose).

Mark W
Tuesday, April 09, 2002

>Sure they could be written in assembly language and operate 50 times faster

I would say that this is not true. Mostly speed is lost due to bad design (for example when dealing with hardware bottlenecks), not due to choice of language.

I work in the area of image processing and I assure you that a well written assembly function is not necessarily faster than a well written C++ function compiled with decent optimization. (I tried this out when I changed some of our code from assembly to C++ for maintainability, I even gained a lot of performance in some cases because I could detect and remedy design flaws much more easily)

Many modern applications are so slow on startup because they read a lot of profile information from different places (ini-Files, Registry etc.), check and adjust system settings etc. resulting from features like a file history list, user configurations and the like. I did not ask for those things, I did not miss them before they were there, but I have to admit, that I would not want to live without at least some of them any longer (and my some of them might not be your some of them again).

Still I agree, instatanious startup is a nice thing. Last time I visited my parents, I switched on my old C64, just to see if it is still working, and I was flabberghasted because I had forgotten that you flip it on and it is just there, not even a second to wait. I would not want to work with it in ernest, though :-)

Have fun,

Jutta Jordans
Tuesday, April 09, 2002

I want to go in a different direction. Maybe this should be a new thread.
A few comments on the other thread said that bloatware is OK. Computers keep getting faster, Moore’s Law and all that. The problem is Moore’s Law applies to silicon chips, not software (http://www.salon.com/tech/feature/2002/04/08/lehman/print.html ).
As code gets bigger, more feature rich, and harder to maintain, the quality is reduced. As Joel’s faithful, we could never throw out the good code and build a more maintainable code base. All we can do is refactor and test and hope that existing features work the same version to version. Everyone who loves doing maintenance of someone else’s bloatware raise your hand. This makes the market ripe for the small, smart startup.
In some of his grow a business articles, Joel has commented that to be successful a good strategy is start small, get shipping product ASAP and pay attention to feedback.
If my Word Competitor Corp. tries to implement every feature in Word, I will never ship, or technology will change and leave us behind. To compete with MS, people have to learn from the marketplace, they have to sell to a few early adopters. To me these reviews are helpful if I want to be that early adopter.
There is a contradiction in thinking here. I do not think that Joel is saying no one should even attempt to compete with MS. Is he saying no small company can get market share? Is he saying that to compete you must deliver a feature complete product? Only niche markets are viable? Selling speed and compactness is a good way to get people to change. Why criticize people for that?

Doug Withau
Tuesday, April 09, 2002

At least for mainline desktop software, speed and compactness don't seem to be major selling points.

At least, I can't think of anyplace where a fast and compact app has won in the marketplace over a big, full-featured app in the last decade.

Dave Rothgery
Tuesday, April 09, 2002

I think file format compatibility is a lot more important than feature-parity. Wihtout virtually perfect file format compatibility, you lose 100% of the audience. Without 20% of the features, you lose 10-30% of the audience. Also, suite developers shold try to mimic MS Office interface as much as possible. There would need to be a very strong argument to depart from it, imo, unless you want users to feel unconfortable.

pb
Tuesday, April 09, 2002

Does it matter if software is bloated?  Certainly not for
the vast majority of software out there.  The majority of
users simply want software with more features...  Whether
they use them or not.  Does it make sense?  No, not to
me.  But then again it doesn't really matter.  I have yet to
convince one person that notepad can handle all of their
word processing needs, even though it would.  They all
prefer Word. 

What does this mean to me?  Well, it means that I will still
insist on using notepad (& VI), but if I want to sell software,
then I had better pay attention to what everybody else
wants.  It means that I have to take into account the
life cycle of my software.  What are the stats of the
average user's machine that it will be used on over it's
lifetime?  That is the key.  I am not going to spend a lot of
time making something fast and small if I don't get any
return on my investment. 

mlr
Tuesday, April 09, 2002

I think most programmers don't understand that bloat is liberating for users.  We programmers are pretty jaded with the ability to unlock all of the potential of our computers, and we play a game where we try to conserve it.  But users need us to unlock that potential for them.  Their view of the machine is fundamentally different.

I think companies like MSFT and Apple still have that Prometheus feeling.

Greg Neumann
Tuesday, April 09, 2002

Getting back to Joel's rant, the funny thing is that in one of the links he provided a new Office alternative is discussed -- written in Java! To me, the idea that (any) Java program is going to be less bloated than Office is hilarious...

Frederik Slijkerman
Tuesday, April 09, 2002

Oh, c'mon! Joel's obvously not referring to foot-print. He's referring to features. You know, the part of the app that people actually *use*.

pb
Tuesday, April 09, 2002

Since few of us will write a Word competitor, there's a more useful question to ask. How should programmers occupy their limited time, after fixing bugs: 1) Adding new features, or 2) Improving code size, memory usage, and performance.

I lean towards 1), for a simple reason. Human CPU cycles are far more valuable than computer CPU cycles. New functionality that lets customers do things they couldn't do before, or makes tasks dramatically easier, signifcantly increases customer satisfaction.

If customers start complaining about performance, I'm happy to spend time working on it. Until then, saving a customer a week of effort helps them more, and is more profitable, than reducing code size.

Jared Levy
Wednesday, April 10, 2002

Replying to several messages at once...

>How should programmers occupy their limited time, after fixing bugs: 1) Adding new features, or 2) Improving code size, memory usage, and performance.

>I lean towards 1), for a simple reason. Human CPU cycles are far more valuable than computer CPU cycles.

But that ("Human CPU cycles are far more valuable than computer CPU cycles") means you also need to pay attention to performance. Avoiding bloat per se is not the issue, but it *IS* important not to make the user wait for things that should be trivial.

>If customers start complaining about performance,

"Lack of performance" is not the sort of complaint that usually makes it back to the company that created the software (unless they really dropped the ball). Users know that complaining about performance isn't worth it: there's nothing the software company can do, short of writing a new version of the program.

>Oh, c'mon! Joel's obvously not referring to foot-print. He's referring to features. You know, the part of the app that people actually *use*.

The two cannot be separated. If the footprint is huge, it usually will influence user experience, too. The app will take forever to load, and (example: Outlook) simple tasks like moving 10 messages to the trash will take 25 seconds - instead of being instantaneous, as it should be.

>I think it is reasonable to expect hardware to enable the implementation of bigger and bigger executables.

As I've shown, you'll have to take the slowest component into consideration, not the 'average speed increase'.

>A big engine is expected to be able to pull some weight.
Boundaries are made to be broken, the word "bloat" means nothing.

The hardware is meant to break boundaries. The faster, the better. But developing software to run only on the fastest machines available is, in most cases, just plain dumb. For games, I can understand this drive; in this case, something is actually added to the user experience. But all too often, the line "but it runs fine on the latest and greatest" is just an excuse for sloppy coding.

Harro de Jong
Wednesday, April 10, 2002

I'm suprised no one brought up the real danger of bloat, which is security holes.  And customers are starting to shout at these problems.

Another hole in Joel's argument is that bloatware is tolerable while hardware prices fall so quickly -- but now we see that Dell and others are starting to RAISE prices.  This assumption of Joel's is starting to be blown away.

Now, I am actually a fan of bloat.  Annoying as it is, bloat is a sign of easy times, when poorly skilled programmers with good ideas can cobble code together.  Bloat is a common symptom of innovation, and Emacs was once a good example of that.  But looking at the situation coldly, the good times may be waning.

Sammy
Wednesday, April 10, 2002

Many people have commented that its most important to optimize software performance for older computers.

I disagree.

For many home users, the only time that they actually get any new software versions is when they buy their new machine. So, when grandma bought her Pentium II three years ago, she got MS office bundled with it. If she needed a program to organize her digital photos, she probably bought that within a few months of buying her computer.

Statistically (based on statistics I'm making up, of course), the rate of new software purchase decreased as time passes. So, when I write software, that software will probably be installed on a fairly new (last year to year and a half) system.

Benji Smith
Wednesday, April 10, 2002

>Many people have commented that its most important to optimize software performance for older computers.
I disagree.
For many home users,...

But for most software, home users are NOT the target market. It's businesses you'll mostly be targeting, and there the 'new software on three-year old computers' model is very common.

I think it's very important to make your app feel fast/responsive etc. Having a user wait for one second may not seem like much, but it adds up. As I said before, the user won't complain to you, but his impression of the app will be "that dog-slow POS" and he'll try and find an alternative next time. Example: I've had to use Outlook for a few months now. Even on a 1 GHz PIII it's so slow I'd never recommend it to anyone.

Harro de Jong
Thursday, April 11, 2002

"I've had to use Outlook for a few months now. Even on a 1 GHz PIII it's so slow I'd never recommend it to anyone. "

I've used outlook for 6 years now, and even on the old PI/133, it was so quick I'd recomend it to anybody.

Tony
Thursday, April 11, 2002

Harro said:

But for most software, home users are NOT the
target market. It's businesses you'll mostly
be targeting, and there the 'new software on
three-year old computers' model is very
common.

And then he said:

As I said before, the user won't complain to
you, but his impression of the app will be
"that dog-slow POS" and he'll try and find
an alternative next time.

----

Don't these two statements conflict with each other? If the target audience is a corporation where a single person or group of people make the decisions on what everyone should use (both hardware and software), then how does the individual sitting at his 3 year old computer waiting for Outlook to load and not complaining affect the people who make the purchasing decisions? He's not going to be in a position to find an alternative next time, he's going to take the corporate standard computer and like it.

I've tried to load PegasusMail on my computer here at work because I like it, but I can't use it because the mail servers use some weird HP version of IMAP with special drivers you can only find for Outlook. I couldn't find faster software even if I wanted to.

Mark W
Thursday, April 11, 2002

When I'm at home, I often use Word 2000, Excel 2000, Illustrator 9, and Photoshop 6 all on my 3-year old laptop. What's more, I sometimes use them all at the same time. It's an AMD K6-2 400 mHz computer. And I haven't died from boredom or gone crazy waiting for the applications to load.

Word takes about two seconds to launch. Excel takes about one second. Photoshop and Illustrator take ten or fifteen seconds each to launch. But you know what? I don't care. Ten seconds is nothing.

Once I'm actually inside the applications, they perform very well. In Word, when I type an 'A' on my keyboard, an 'A' shows up on the screen. Instantly. I have never had an occasion when I had to wait a few seconds after typing very quickly to wait for a paragraph to finish rendering on the screen.

Performance is a **non-issue**.

If I'm doing complex things in Photoshop using huge files with a dozen layers each and performing processor intensive things like gaussian blurring a 1000x1000 pixel layer, then, yes I'll wait a few seconds for that operation to complete.

Buy I still don't care.

If performance were a huge issue, I would take the files to work, where I have a faster (1 gHz) computer. Or I would buy a new home computer. But those circumstances are so rare, that it doesn't really matter to me.

What does matter to me is that Photoshop 6 lets me organize my layers into folders. This feature is so handy I couldn't go back to a previous version. Does Photoshop 6 take up more space on my hard drive than Photoshop 5? I haven't the foggiest idea. Does it use more memory? Beats me? Is it slower? Not that I've noticed.

What I do notice is that the new version has handy new features.

New features do not necessarily make me more productive, but as a user I _feel_ like they make me more productive. Reducing latency for a function from a .5 seconds to .25 seconds doesn't usually make me feel more productive, even though the software has undergone a 100% performance improvement. It just doesn't matter anymore (most of the time).

Benji Smith
Thursday, April 11, 2002

>Don't these two statements conflict with each other? If the target audience is a corporation where a single person or group of people make the decisions on what everyone should use (both hardware and software), then how does the individual sitting at his 3 year old computer waiting for Outlook to load and not complaining affect the people who make the purchasing decisions? He's not going to be in a position to find an alternative next time, he's going to take the corporate standard computer and like it.

That may be what often happens (except for the "and like it" part), but that's not the way it should be.
If the decision-maker decides to foist crapware on the thousands of users he's deciding for, he's the wrong man for the job. There may not be formal complaints, but issues like these are easily sniffed out by asking a few users whether they like the software.

>New features do not necessarily make me more productive, but as a user I _feel_ like they make me more productive. Reducing latency for a function from a .5 seconds to .25 seconds doesn't usually make me feel more productive, even though the software has undergone a 100% performance improvement.

In the examples you mention, you're right. Performance isn't an issue once you get past the point where actions are perceived as either 'instantaneous' or, for large/complicated operations, 'reasonable'.
But this is not always the case. Taking Outlook again, when I select a message, it takes a noticeable amount of time (up to five seconds) for the message to be displayed in the preview pane. This again on a 1 GHz PIII. Those are annoying delays. Waiting 30 seconds for Visual SourceSafe to delete a file is another example of unacceptable performance. Or try paging through a Word document with a lot of images. Press page down, and it's "one banana, two bananas" before Word has drawn the new page with its images.

Harro de Jong
Friday, April 12, 2002

>If the decision-maker decides to foist crapware on the thousands of users he's deciding for, he's the wrong man for the job.

And if this so-called "crapware" is the right tool for the job? One person's crapware makes another person productivity tool.

Here's an idea. Why doesn't an "anti-bloatware" proponent describe their ultimate corporate desktop/office suite, and we can analyze for ourselves.

This will give us a common basis for discussion. We may not agree on metrics (features v. speed or whatever) but at least we can get a list of pros and cons out there that we can pick and choose for our own needs.

Mark W
Friday, April 12, 2002

hmmm.. when I found this site, I was looking up the definition, believe it or not.. "Bloatware" From what I can tell from my research on the term, Bloatware means small softwares that run together to complete a task quicker than what one software was intended to do. Basically in theory, if you have 1 secretary doing all the file working for a 1000 employee firm... it'll take forever, but if you have more secretaries doing it, ta-da ;)  layman's term, but that's what I figure it means.  Just wondering if anyone can deter or correct me if I'm wrong :D

Wolfric V. Elkrose
Wednesday, April 02, 2003

erm.. I got it reversed >.<  Bloatware = the one secretary in my defination, one person doing all the tasks.  What one wants in a program.. is anti-bloatware. :D

My apology for the confusion.

Wolfric V. Elkrose
Wednesday, April 02, 2003

*  Recent Topics

*  Fog Creek Home