Re: Computer time -> Developer time -> User time?

From: Beth (BethStone21_at_hotmail.NOSPICEDHAM.com)
Date: 01/22/05 

Date: Sat, 22 Jan 2005 17:34:32 GMT

Frankie say:
> Randy wrote:
> > Beth wrote:
> > > ...hence, tactics used to maximise programmer "productivity" came to
> > > the fore: "portable source code", "black box libraries", etc....this
has
> > > lead to "bloatware"...
> >
> > No, the thing that has lead to bloatware is the insistence on the
user's
> > part that their software have more features than the "Jones' software".

But isn't that an industry "self-fulfilling prophecy" rather than a true
"user demand"?

You know: Swamp the stores with "version 2"...brand new users can't buy
anything else because they don't stock anything else...these move to
"version 2"...the stores threaten others with "it will be obselete!
Obselete, I tell you! Change now!" (in a grim-reaper-dressed "Ghost of
Christmas Future" kind of way: "Buy version 2 for tiny Tim, Scrooge, or
this grave you see before you shall have your name carved into it!"
;)...people "panic" and move to "version 2"...well, lo and behold! Now that
everyone's been "herding" and "panicked" onto "version 2" (and the store
isn't even stocking "version 1", so, even if you _wanted_ it, how can you
get it?), there's few sales on "version 1"...they discontinue it and make
it "obselete"...then a year or two later, repeat the cycle...

A demonstration of this is Microsoft's "5 year rule"...basically, Microsoft
retract "support" for each of their OSes, 5 years after release...

Now, first, if this is a "reaction" to user "demands" then how on Earth can
they declare when it'll be "obselete" _BEFORE_ they've even released it!
The way the "5 year rule" works is simple...if Longhorn is released this
year - 2005 - then it'll be officially "obselete" (and Microsoft then
refuse to "support" it in any way) in the year 2010...just add 5...as
simple as that...

It's a "prediction"? Then it's not a terribly good one because, for Windows
98, they've had to _break_ the "5 year rule" and extended the
"support"...why? Because, completely _contrary_ to this "users are in a mad
rush to upgrade, upgrade, upgrade!" theory, far too many people _WEREN'T_
upgrading from 98...a sufficiently large number weren't upgrading to be
significant to Microsoft to "extend" the "support"...the retraction of
"support", in truth, is a "threat" to force "constant forward motion"...if
everyone refused to "upgrade", then they'd be forced - as in the Windows 98
example - to stay "supporting" it, otherwise they'd be cutting off their
own nose despite their face...while 90% of users are still on Win98, they'd
"retract" all "support"? No, they wouldn't...they'd be risking their
"monopoly"...

Sir Bill has used this "poker bluff" repeatedly to make himself a
billionaire...developing software for Apple (for their original Mac) ready
for "launch date", Bill used another "threat": We will not provide you any
of the software...Apple panicked at a "launch party" where they couldn't
actually show the machine doing much of anything because there were no
software applications and they didn't call his "bluff"...worse, part of the
"extra terms and conditions" Sir Bill demanded? "Some" rights to Apple GUI
technologies...

He "threatens" retracting his software and then people panic that it might
disappear and leave them in a terrible situation so "surrender" to his
demands...but, really, people should start calling that "bluff"...a
software company that refuses to sell any of its software? You can't keep
up that "spoilt brat" attitude for long...unfortunately, Bill will keep
getting away with it because most people are not "wise" to his "poker
bluff" tactics...and there's no way to get this message across and for
everyone to believe it, in order to get the whole world to "call that
bluff"...so, he'll get away with it...

Users _DO_ want _better_ features...but they don't want "random, completely
arbitrary features, so long as there are numerically more features than
before"...

The "feature bloat" is caused by one thing...and it's not users "demanding"
it...the industry wants to make profits...if it did this in the more
"honest" way, then they would release a good application and then the next
upgrade may not happen for many years (and the "new version" isn't simply,
unlike Windows, a case of "better graphics, USB support and that's it" -
after all, why exactly do you need to buy a boxed copy of that? Microsoft
can upload the "new bitmaps" onto "Windows Update" or a few DLLs for the
"USB support"...it _could_ all be done by "Windows Update" but that's free
and they want to make money...so, they draw a "line in the sand" and it's
at those points they change the graphics to make it look "brand new" and
sell it under a new title with a new logo - but an actual fundamentally
significant improvement of functionality)...

But that doesn't make too much money...so, they turn what should be like
buying a house into what's more like buying your weekly
groceries..."threaten" things going "obselete"...bring out "new versions"
that aren't appreciable that different from what you've already bought
(let's look at it...Windows 95...+ USB support + IE built-in...= Windows
98...umm, did ME have _ANYTHING_ new to it at all from 98? All superficial,
anyway...ME -> XP? An illusion, really...XP is simply 2K "in disguise"
(though not disguised too well when 2K binaries are totally compatible and
it can even be found identifying itself as the same as 2K)...hence, really,
they are just "herding" the "home" users over to the "business" NT kernel
and have just added "backwards compatibility" support...of course, this
"support" isn't a "new feature" at all...it's a "must make sure there's no
barriers to forcing people to upgrade to XP" feature _ONLY_ to support
their "forward constant motion" plans...if you didn't "upgrade" then you
would need no "backwards compatibility feature"...in short, just like the
ten years before 95 (where they remained "16-bit" far too long - on the
"home" front (the fact OS/2 and NT weren't like that shows they _could_
have it...because, in fact, _THEY DID_! But only sold it to "business" ;) -
despite the hardware having been perfectly capable for a decade...this also
sheds light on "commercial software develops faster"...not by Windows
standards, it doesn't! It takes them 10 years to write a new OS...that
easily sounds like _less_ than "hobbyist" rates of production (accounted
for by the "extra work" needed to create the in-between "false progress"
versions ready for release)...the "versions" in between have all been
superficial modifications to make it _look_ "different" for the purposes of
duping users into believing they aren't doing what they actually are doing:
Paying for something they already own)...

Users don't care for "technology for technology's sake" (well, "geeks" do
but they are a minor fraction of the "audience" here)...users just want to
get their work done...in fact, the less "nonsense" from the OS in doing so,
the better (features that "enable" work to be done, great...features that
are never used and "bloat" and slow it down and get in the way of doing
work because it's all far, far too confusing: Nope, no-one wants
that...absolutely no-one...even Microsoft would wish that upon
themselves...they put it in because it "justifies" the next version so they
can sell it to make profits...as simple as that)...

You're usually the one with the statistics, Randy...what are the statistics
about _USER PRODUCTIVITY_ since all this "feature bloat" has been
increasingly introduced? Well, lo and behold, it's getting _WORSE_, despite
the bold advertising claims, used to push people onto the new
version...there's "memetic reality" - a constructed self-delusion of "how
we would like things to be" - and then there's the _real thing_...a
"connection" between the two, apparently, is not always necessary when
"profits" are involved ;)...

It is an industry "self-fulfilling prophecy" to keep pushing this "constant
forward motion" hard because you can double, triple your profits by selling
some "technology" in "baby steps" over a decade with minor "version
enhancements" rather than just sell a new version when there is a
_significant_, fundamental change and improvement to the product...this
isn't merely a personal opinion, it has been investigated as charges in law
courts against Microsoft (typically, the ones that ended up as "out of
court settlement" that we never heard the court's verdict on this
practice)...

Users "demand" it? Well, they "demand" getting something worth the money
they hand over...they are _EXPECTING_ actually better software...what they
are typically supplied is an "illusion" of this...and all the "extra
features" are a "blind them with science" tactic (because most users don't
really know what 90% of these claims on the side of the box actually refer
to), to "convince" them that they really are buying a "significantly
improved" product...once bought, though, actually listen to them _voice
their opinions_ on what they bought...if you don't outrightly "hate"
Microsoft, then you're certainly "disappointed" in what they do...I have
met _ONE_ person who defended them and was enthusiastic (who suspiciously
has an incredibly detailed knowledge of NT kernels, that one wonders if
he's an ex-employee or being "hired" by Microsoft to "be on the look-out"
for the "next Linux" developing on the OS development groups ;)...and
challenging that enthusiasm demonstrates no actual solid logical base for
it, anyway...indeed, comes across more like some kind of S&M "fetish" or
something...enjoying being pissed upon by contemptuous Microsoft
policies)...

One has to be careful of "industry theories" from the perspective that, for
instance, the industry is grossly unlikely to suddenly say: "you know what,
we should make ourselves less money because, really, it isn't
justified"...or a developer declare: "There's too many developers here...so
you should fire me in order to do things properly"...money is a poweful
motivator and it motivates as many (if not more) _LIES_ and "self-delusion"
than it does "competitive progress"...it has to work this way, of
course...but, as always, the presumption of something that's 100%
"advantages" and 0% "disadvantages"? Not anything that exists in this
universe, sorry...money isn't the root of all evil - that _IS_ an
exaggeration - but it's certainly not the root of "only good" either...

The industry as a whole might not have quite the same levels of "contempt"
that Microsoft specifically do...but we're all "motivated" in the same
direction, so to speak...indeed, Microsoft are "king of the pile" almost
certainly because they were the ones _without_ "conscience" or "guilt" to
push it all the way "to the max"...right up to 11...an unfortunate "strange
attractor" in the system we build for ourselves, it does actually favour
"bastards", so long as they are smart and devious enough to "get away with
it" (because so long as they remain _legal_ "bastards" - and even then,
money can "oil the wheels" that various things are "overlooked" - they are
allowed to be that way as much as they like)...

> I don't doubt that this is true, but I don't think it tells
> the whole story.

It doesn't...now that Randy has brought this "jigsaw piece" into the
picture, it brings the topic around for me to introduce another "jigsaw
piece" above that not all of this "progress" is actually "demanded" by
users but is rather "persuaded" for "constant forward motion" because,
simply, you only make money when you _SELL_ the software, so you want to be
_selling_ as much software as possible...a simple "trick" is, therefore, to
basically sell the same thing over and over with "minor modifications" and
"more features" - pretending in the glossy adverts that it's some kind of
"revolution", not "minor evolution", that you're selling - to keep
generating that money...the "dealers" also make their money that way, so
they support you on the other end of things in ceasing to stock the older
software and to "push" the newer software for you in the stores...

Not that this is any "conspiracy" (well, looking at Microsoft's "Halloween"
documents, perhaps it _IS_ a "conspiracy" in their case ;)...it's just the
natural "strange attractor" to this "chaotic system"...that is, dealers
make money selling software, developers make money selling
software...marketing agencies make money _pushing_ the sale of
software...it's all _pushing hard_ to "sell software"...the user is
"swamped" by various "constant forward motion" tricks (sometimes, of
course, they are willingly going along...I mean, sometimes the software
developers actually _DO_ have something worth paying money to "upgrade"
to...but, often, this is a "forced" thing...one needs to not solely "look
at the figures" but also "sample" the user opinions voiced...you know,
sure, they bought it but did they actually _LIKE_ what they bought? It's
not solely programmers and "technical" users who voice anger (and sometimes
hatred) at Microsoft or joke / complain about the "constant push forward"
tactics and that they _DON'T_ feel they really are getting much of anything
that useful out of it :)...

You know, if I set up a "system" where you get $50 every time you jump in
the air, it's not unusual to find big crowds jumping up and down...not an
actual indication that jumping up and down is a "good" or "right" thing,
it's just a way to make a living under this kind of "system"...it's a
"strange attractor", so to speak, to borrow terminology from "chaos
theory"...systems inevitably have their "natural attractors"...things tend
to "settle down" in certain "patterns" (oh, and the word "strange" only
refers to the fact that in "chaotic" systems, these attractors aren't
necessarily "simple" or "obvious"...put them into a graph and the shapes
can look rather "strange"...hence the name...with a "simple" non-chaotic
system, the "attractor" might be obvious...some "all the 'virtual ants'
following these simple rules, all end up gathered together in the top-left
corner of the screen" conclusion...you can even work that out pretty easily
from looking at the "rules"...but with "chaotic systems", the same thing
inevitably happens, it's just not "simple" and the "shape" of this
"attractor" not "obvious" at all...but it's the same thing...indeed, the
underlying point in "chaos theory" is that these "systems" _LOOK_ "random"
and "unpredictable" but they aren't really...they just become "far too
complex" to defy attempts at "prediction" and come out _looking
"random"...they actually ain't...there is no inherent difference between
"order" and "chaos" (this is the "chaos theory revelation", so to
speak...don't go by "appearances" because they _are_ the same thing, just
poor measurement, complexity, interaction, etc. all conspire to make it "go
wild" sometimes ;))...

Hence, these things aren't "conspiracies" and you _DON'T_ need any
"conspiracy" (not to say these don't actually happen from time to
time...the Holocaust is a clear "conspiracy"...the neo-con agenda of
"control through fear to 'unite' the American people against the 'chaos' of
the liberal society" is also another "conspiracy", though being "on-going",
some people refuse to see it)...but when a number of "personal agendas" all
match up, they create a "thrust" in a certain direction...everyone is
making their money through "selling software", so the "thrust" becomes "get
these users to buy things! Buy! Buy! I need to feed my kids, darn it! BUY
SOMETHING!!"...and, on the "making a living", then who can blame them? BUT,
this doesn't mean that it's a "reality" that we actually really do "need"
to buy these things...very often, we don't need to do any such thing at
all...but the media and advertising agencies create a climate of
"consumerism" to constantly convince of this false "need" to "buy" all the
time...

They all make their money that way...you can't really "blame" them...but
this does not translate that users "demand" this or "need" it...they get
it, like it or not...

> I don't think that the reasons Beth gives
> provide a complete explanation, either.

Well, it wasn't meant to be some "Hitchhiker's Guide to the Entire
Universe" or anything in one post...even if *ahem* the size of my posts
sometimes make it look that way ;)...

> Sure, "portability"
> is going to increase the code size, but to the extent that
> it provides "functionallity", and isn't just a "mantra",
> that isn't what I'd call "bloat".

Neither is it what I'd call "bloat" either...the distinction hinges on that
"isn't just a mantra" clause...

After all, with LuxAsm, I'm proposing "modular architecture" (that's NOT
going to be "flat", "specific" and "simple" but, for the "integration" of
the project (and as "foundations"), I do believe it's _NEEDED_ to make sure
the project can reach its objectives :)...we also have the "Unified
Model" - which really annoys the "master of specific", Rene...so it's not
meeting his "specific" approval - which is inherently placing a highly
generic abstraction tool (capable of being used - especially with macros -
to create very "HLL" HLLisms ;) _NATIVELY_ into the syntax of the tool
itself...again, I consider this _NEEDED_ (that's why I put it in) in order
to reach the objective of simultaneously pleasing "real assembly" coders as
much as HLA users (that is HLA as in "high-level assembler" as a
_category_, not solely Randy's tool...hence, yes, HLA but also MASM, TASM,
etc. :)...without - this is what I tried to "delicately balance" - this
"compromising" one for the other...make it all 100% "user-defined" then the
_programmer_ writes their code in the "style" they like and it should
Hopefully result in both being able to use it in their preferred "style"
without this ever "getting in the way" of other people using the "other
style"...not just that you aren't being "forced" to do any of it...but the
"extension" I've also added is that you're also not "forced" to do it in
any particular way, either...it's all "user-defined"...

Hence, in practice, we can see (as well as the fact that I have said this
many times too...nearly every time I criticise "portability for the sake of
portability", I pop in the "condition" that I'm not criticising the
practice itself but that I'm criticising the _ABUSE_ of that
practice)...that I don't have a problem with "abstraction" or other such
things...indeed, that it requires "abstract thinking" to even notice these
things...

> "Black box libraries"
> *can* be bloated, but not necessarily so. There *is* a
> difference between libraries, and a "concise" library would
> be just as effective in reducing developer time as a
> "bloated" one.

Ah, but the point is not so much the "library" part, as the "black box"
part...and it's not that it's "inherently bad" (whatever Rene says ;)...

It's all to do with what Randy says in his radio interview: The difference
comes when the programmer _CARES_...

And the problem can exist that the "black box library" is employed exactly
because the programmer _DOESN'T CARE_...but, true, it is a _symptom_, not a
"cause"...when you see the smoke, it's usually coming from a fire...

Plus, there is something to note about "black box"...in order to
"modularise" and "encapsulate", this typically demands
"indirection"...every "indirection" is working through a "proxy"...going
"via", not direct...and, simply, you know the equation, yes? "speed =
distance / time"...well, works for "logic" distances, as well as physical
ones...the computer runs at a set "speed", it has a certain "distance"
(measured in "instructions to complete" rather than miles or kilometers but
it can still be thought of as a "distance to travel")...thus, it's going to
take such-and-such a time...if you "indirect", then you increase the
"distance" but the "speed" does not change (that is, the 3GHz machine keeps
ticking away at 3 billion (American) ticks a second...indeed, the only
speed changes are _downward_ - waiting for I/O devices - never
upward)...hence, the "time" takes longer...

It's a minor thing in a single instance (usually) but it's an
_accumulative_ thing...indirection to indirection to indirection to
indirection, then loop back round and do it all again...and again...and
again...you need the "indirection" to keep the "black box" "decoupled" from
the other code...you know, you can't jump straight into the other
module...you call via a "jump table" to an entry-point that makes a
"transition" to a "layer" which communicates to a "library" (with another
call via a jump table to an entry-point that makes its own "transition"
which returns) which returns to call another "library" (which calls via a
jump table to an entry-point which makes a "transition" and returns) and
then returns...

Of course, there is nothing "inherent" about "black box" or "libraries"
that demands things work this way...BUT, inevitably, that's actually how
it's working...

Basically, let's "anthropomorphise" this...you're working in a "team" of
people...each "team member" does some work...BUT the other team members
actually _DON'T_ know what they are doing...indeed, it's a "secret"
("information hiding")...perhaps you need a hammer ("printf") to do your
work and so do they...but you don't talk to each other, so you both might
go and buy two hammers when only one is really needed..."team members" are
only permitted to know what the others are doing at certain "key points" of
interaction between them...

In such a situation, how do make sure they are working to maximum
efficiency? _CAN_ you ensure this when you have this "information hiding"
secret that no-one knows what the other is doing?

Well, it is at least "slightly possible" from the perspective of one person
having an "overview" of the whole process and co-ordinating the "team" to
work efficiently (they don't necessarily need to know everything the other
people are doing, so long as _someone_ has sat down and worked this out
:)...indeed, you could think of this as one of those "top secret" military
projects where they get commercial companies to build the various "parts"
but as each has no idea what the others are doing and are under "secret
orders" then only the military themselves know what all the bits are
destined to be used for (some new "fighter plane" design or
something...like the radar-invisible Stealth bomber / fighter...or, if you
believe those "flying saucer" fanatics (the Earth-bound ones, not "space
aliens" ones...though, the "Roswell" folks would insist that "space aliens"
told them how to build these UFOs ;), the "flying saucer" craft they are
building at Area 51...oh, and boiling an egg? 3 minutes, of course ;)...

But if the programmer doesn't care - doesn't take that "overview" - then
these things can't be expected to necessarily work well together...

It's not the "cause", true enough...but it's often a "symptom"...and, of
course, there's "theory" (where it's "not so bad") and then there's
_practice_ (where it can often get bad, as libraries are used to access
libraries which use libraries to call libraries and the whole thing
"bloats"...not because any particular "library" itself is necessarily bad
but because the programmer has stopped taking that "overview" and caring
about what's going on...

> It seems to me that to account for the size of "modern
> apps", I have to assume that needed/desired features are
> being implemented in a "bloated" fashion, and thoughtlessly
> pasted onto already "bloated" code.

Yes; Because you have to think about how modern instruction sets can do
more (more "addressing modes", atomic single instruction "exchange", etc.)
in a single instruction than those older 8-bit instruction sets could
do...the machine has more things "offloaded" from the CPU too...and _code_
is actually a typically _very compact_ thing...

It just doesn't seem right that _code_ (as in _instructions_) for some not
terribly complicated "feature" is bloated the size by 1MB or
something...that's a _lot_ in the physical code terms...but these
applications seem to be throwing away MBs and MBs on "extra features"...

It's not, of course, possible to "generalise" because of all the different
applications with different "features"...other than to say that it
"intuitively" doesn't seem right...doesn't seem right at all...

> This *may* be
> intentional - the "average software consumer" may think that
> a program that takes 3 CDs to install is a "better value"
> than one that fits on a floppy, even if the functionallity
> and "feature set" is identical.

Mind you, though they may want to make their software seem "better", this
can only extend so far...what I mean is, if it's 1 CD rather than 3 CD then
the _price_ can be lower...and I think a user will find a lower price
"better value"...

After all, even if it fits on a floppy, they could put it on a CD
_anyway_...just waste most of the space on it...why not? The CD typically
doesn't have anything else on it...and, to be honest, it's probably cheaper
to buy CDs than floppies these days, anyway...less materials, easier
construction...not that this is any great cost, anyway...as think how cheap
blank floppies and CDs are to consumers and we ain't buying "in bulk" of a
few thousand they will be buying in (which reduces it further: "economy of
scale" :)...plus, they are getting it "wholesale" rather than "retail" too,
I guess...no, this _ISN'T_ a big cost at all...

So, sure, it fits on a floppy but stick it on a CD, anyway...you can put a
better "picture" on the CD...they, to be honest, are more "convenient"
too...you don't have to fill up the CD, in order to give the user the
"impression" that it's "better value" than a floppy disk...

And, you can "change strategies" too, anyway...if it only takes up a
floppy, then make a "compilation" of small programs and put them all on one
CD...now, if the users sees "DOOM, DOOM 2, DOOM 3 plus 27 mission packs
(and bonus Space Invaders, Asteroids and 100 card games)" all on _ONE_ CD?
Well, I think no-one's going to then think 3 CDs for just the one game is
"better value", eh?

Plus there are other simple tricks...put on the cover: "New
Supersqueeze(tm) compression technology! Exclusive to this CD! Squeezes 3
CDs' worth of data on 1 CD!!"...wow! It's got 3 CDs' worth of data on just
the 1 CD because of "better technology"? That means I'm getting the "better
value" of 3 CDs _AND_ it's all on just 1 CD (much more convenient :) _AND_
they must be "geniuses" to be able to come up with this really clever new
"technology" that can do all this! (It's just programmed properly - no
"bloat" - and ZIPped up with ZLib or something...that's all...but your
users will think you're the greatest genisues of "compression technology"
there ever was! ;)...

[ Did you ever download that "the product" demo thing? Basically, it's a
64KB file but it uncompacts itself, then it "dynamically creates" all the
graphics and sound...and manages to compress what's effectively 9GB of data
into *ahem* a 64KB file...30,000:1 compression! 15 minutes of high quality
music on just one of the demos alone...now, if anyone asks me what is
"better value" then these guys win hands down without any hesitation...and
it's not just being "technically" impressed by it...I let some
"non-technical" friends know about the demos and could stick _tons of them_
on a single floppy disk (about 23 of them if all just 64KB big :)...yet, it
provides _HOURS_ of demo "entertainment" of a very good quality...they were
jaw-droppingly impressed at all of this...and, ever since, I've never had
to explain why I keep complaining about "bloody Microsoft!"
thereafter...they now get the point perfectly well...they've seen what
_could_ be possible...that the problem is not their hardware but the
fracking bloated software they are running on top of it...remember, we all
know this but this is "revelation" to many an "average user", who always
thinks it's that they didn't buy the more expensive processor - the
hardware - not that the software ain't using it properly...well, it makes
sense from everything else they encounter...if the picture is crap on your
TV? Well, that's "bad reception"...a "hardware" problem...if the car won't
start on a cold morning? Well, that's an ignition problem...a "hardware"
problem...when they see that the window is just sitting there hanging,
unpainted, unresponsive while the disk is whirring? Ah, must be a "hardware
problem"...the disk can't be fast enough or something...need to buy a
better disk...and, of course, if they phone the "support line", what do
they say: "hmm, yes, you need to buy more RAM and a better disk
drive"...they "confirm" it all...it's a _cruel_ illusion because the user
is basically being _deceived_ into thinking that their "hardware is no
good" when it's perfectly good...but just isn't being used
correctly...indeed, perhaps Rene is at least correct to start bringing up
words like "moral" and "ethical"...I mean, if we willingly indulge
ourselves in supporting and continuing this _deception_, do we deserve to
sleep snuggly at night? Think of those parents working hard in order to
_only just_ be able to pay for a reasonable PC for their kid's
"education"...then they are _scammed_ to buy this thing (which isn't
needed) and then convinced that their "hardware is not good enough"...and
they both work _overtime_ and stuff like that, in order to
"upgrade"...because they can see their kid likes and has an "aptitude" for
computers (the computers teacher at school says so too :), so they want to
"encourage" their kid to get on in life...to do what they like
doing...starting to feel like *** yet? To screw them out of every last
penny _unnecessarily_, just to afford what for yourself? A bigger "plasma
TV" or something you don't really need and you know you don't really
need....but the truth is that this probably _IS_ a "true story" many times
out there...but, on the other hand, if you do the best job possible and are
selling software on the fact that it _factually is_ "good" then you've
nothing to be guilty for because you've not let anyone down...certainly not
yourself either (yeah, go on, feel smug...you've _earnt_ it, so why
not?)...then, in that case, get richer than Captain Kirk...honest day's
work...you've _raised_ yourself and others, not let anyone down at
all...ever see those Spiderman movies? This is the "hero" in everyone:
"with great power comes great responsibility"...people who take that on -
without fear, without "cheating", without "avoidance" - and triumph are
"little heroes" :) ]

And all this doesn't quite add up with "reviews", anyway...what I'm
thinking is, if Half-life 2 came on a CD rather than a DVD (they squeeze it
into a CD)...then I don't think anyone would think "this is not good
value"...but they are going to be judging it by magazine reviews, "preview
movies" on the website, friend's "word of mouth" and so forth...I can't see
any user picking up a CD of, say, DOOM 3 and then throwing it away: "Huh!
Must be total crap because it's not on a 27 DVD boxset format!"...

On the contrary, people with only CD drives and not a DVD player (okay,
very few these days probably :), which will be saying "Great! I can still
play this without upgrading to a DVD player!"...

I mean, at the end of the day, it's quite simple...all these "illusions"
might be "clever" but they can't beat something that actually _IS_ "better
value", "superior work", "high performance", "small footprint"...rather
than just "pretending" to be so...in the words of Vic Reeves, "you can't
compete with concrete!"...

> I recall reading a post -
> long ago - from the author of a "password recovery" program,
> intended (nominally) to recover encrypted files to which the
> owner had forgotten the password (for some specific word
> processor, IIRC). He stated that his algorithm was so fast
> that he'd added a delay, so that his customers would think
> the program was "doing something".

Missing a trick, really...people ought to learn a touch of
"marketing"...simply market it as "the FASTEST recovery program on Earth!
No program does it faster!"...then leave the "delay" out...

Study "P.T. Barnum"...when his show was crap, he actually _SOLD IT ON THAT
FACT_: "Come see this terrible show! See how fake the illusions are!
Witness how hilariously bad the magician is!"...and so forth...you know, if
it's bad, then sell it as a "parody"...if it's short, then sell it as
"minimalist art"...if it's incomprehensible, sell it as a "mystery"! A
"riddle to be solved" ;)

The customers will see that it's "doing something" from the fact that the
recovered files have re-appeared, for Pete's sake...sell it as being "super
fast!" and then they won't be in the slightest way
"disappointed"...basically, the Barnum "trick" is simply to "prepare"
people for what to expect...when your "show" meets that expectation (or
beats it), they are impressed by it...

Anyway, the most obvious "trick" of them all is to put up a
_prompt_...something like "file XYZ has been found, shall we recover
(Y/N)?"...then the "delay" is the customer's own reaction...

Nope, this is just stupid to make your program seem _WORSE_ by adding a
"delay" into it...I mean, if the files weren't there and then you run the
program and the files are recovered (and presumes that you're "recovering"
the files because you, you know, want to _USE_ them or something ;)...then
that's all you need to tell the customer it's "doing something"...

Indeed, in a sense, this is just how _BAD_ things have gotten...when
someone does a _GOOD JOB_ then they are "frightened" that because it's so
much better than everyone else's program, they should _RUIN_ it and _MAKE_
it _WORSE_, of all things, so that it doesn't "stand out" as being, you
know, good and fast...?!? People have become so "brain-washed" to expect
slow programs that they are "disorientated" by programs that don't take
forever? Scared that customers will reject programs that _WORK PROPERLY_?!?

This is a demonstration of why this industry problem has gotten so, so
serious that it can't really be "overlooked" anymore...that's just severely
_twisted_...

Now, I understand something like a "dealer" in "luxury items" like sports
cars or perfumes or whatever, delibrately keeping the price high (so it
doesn't come across as "cheap" = "tacky"...you know, it's the standard
"Lottery win" dream: Go out and buy a Ferrari or Porsche with your
millions...so, these things should always actually _delibrately_ remain
"out of reach" so that they stay as "dreams" for people :)...but it's just
_TWISTED_ to ruin the _quality_ of a product to make it "the same as all
the other crap out there"...you know, the customer wants a plasma
TV...quick! Crack the screen and cover it in dirt and pull the power chord
until there's a "loose connection"! "Could I take a look at your hifi
players?"...quick! Pour coffee all over it! Smash up the radio component
with a hammer so it doesn't work!! We can't sell the customer a good
product because it might "look out of place" against all the other crap
around here...

Totally and utterly _TWISTED_...if this is what the industry is coming to,
then it has a _mental disease_...it needs _treatment_ fast...that's not
normal...that's not natural thinking at all...if this is what the industry
is coming to - delibrately creating _crap_ products, just because they are
"frightened" of "looking out of place" amongst the "rogue's gallery" of
utter crap - then these points aren't minor things to make...the industry
must be diseased and rotting to think in these terms...to be creating
people who think this is "normal" thinking...something that they "should"
do...it needs a "cure" and it needs one _FAST_...and, no, this time, I
think we _shouldn't_ be adding any "delay" for that "cure" to come ;)...

> > Money talks.
>
> Yes, and what the "money spenders" like (apparently) might
> not match what someone who actually *knows* something about
> software would like.

No, sorry...I just don't buy this crap at all...and "crap" it is...

It's all "self-fulfilling prophecy"...developers, as noted, are delibrately
putting in a "delay" to make a program _WORSE_ and then offering it to
customers...I bet if you _ASKED_ the customer: "Should I unnecessarily
delay the program?", then they'd say "huh?!? Are you mad? You were
seriously going to delay it for no good reason? Why on Earth did you think
I wanted to be slowed down? Look, I can tell it's working from the fact
that my recovered files have returned...that's what it's supposed to do,
the fact that it's clearly happened tells me that it worked...the 'delay'
tells me nothing"...

The industry wants to sell software (that's where money is made)...it's NOT
about what the "customer wants"...indeed, if it was, then we'd simply be
_ASKING THEM_ what they want...I mean, they aren't going to object to being
listened to for making better software...

But, no, instead, as the industry _wants_ to hear "add more features! add
more features! Doesn't matter what they are! Just add more features so
there's a new version I can spend more money on!"...so it convinces itself
that this is what users want...

But they don't...

And there's a simple way to prove this to your own satisfaction...find
"average users" then _ASK THEM_...of course, do it properly with "unbiased"
questions...

I _did_ try this out before and the suggestion of "faster operation" / "no
'please wait' messages" / "faster loading of programs" was _UNIVERSALLY
APPROVED_ (not a single objector to it at all...logical enough, though, why
would anyone _want_ a program to run more slowly than it needs to? Only
"twisted" people who get "frightened" that their programs are doing a good
job (?!?) ;)...contravertially for this discussion, there was _NOT_ any
significant call for "more features" at all in any of those I asked...or,
at least, the point was that "more features" _ONLY_ rated when it was
suggested that they were "useful features"...

On the contrary, the industry has been trying to sell people the "universal
device" which has all the MP3 player, TV, video phone, text messaging,
personal organiser, calculator, tea-maker and so forth...only "geeky"
people ever buy them...because people prefer the _SIMPLE_ concept of "one
device, one purpose"...

The "3G" phones are having a _HARD TIME_ being shifted...and many were
trying to blame "technical problems put people off buying it"...ummm, NO!
Because the people who haven't bought "3G phones" would _HAVE NO IDEA_ that
the "network went down" last Tuesday...you only find out such things _once
you've already bought_...

It's a _MYTH_...it's an "urban legend" that the industry "self-deludes"
itself with...

It works like this: One makes money buy selling software...someone gets the
clever idea of: "hey! If I just add on two extra features, change the
bitmaps then I can sell it all over again and market it as 'new and
improved'...most users will fall for this"...so they do it and they,
indeed, "hype" well and sell more copies of something that isn't really
substantially improved at all from the last version...this person _KNOWS_
that it's an "illusion" and a "scam"...they've _delibrately_ chosen to do
it...but when they are successful, everyone asks "what is the secret of
your success?"...well, the developer ain't going to reply: "By
contemptuously deceiving my users with an illusion that things are better
and new when, really, they ain't"...not least, the developer _can't_ say
that _publicly_...their future customer might be listening in on the
interview, after all...they want to be able to "get away" with the trick
once more...that won't work if they admit openly that it's a
"trick"...others thus _COPY_ what they see...why, look, if Gill Bates is
successful with this strategy then it must be "what customers want"...we
should do this too...

And then it becomes this grand "myth": Customers want random, arbitrarily
"more features" so long as it is numerically more than the last version and
the bitmaps are different...

Now - STOP! - apply logic..._WHY_ on Earth would a customer want that? You
know, they would want it for a _REASON_, yes? What is the logical _REASON_
that they want "more features" for no useful purpose? Why do they want to
be confused with too many buttons and options?

Ever helped out a "clueless newbie" with their computer? Do you know what
causes all of their troubles? _TOO MANY_ buttons..._TOO MANY_
"options"..._TOO MANY_ "features"...it confuses the hell out of them...I
don't blame them...it often confuses the hell out of me too and I'm
supposed to "know" about this stuff...

And, as I say, the matter _CAN_ easily be settled...construct a proper
_unbiased_ set of simple questions to ask users about "what they want" and
then see what the results are...one thing that might surprise is that the
"wants" _DON'T_ differ that vastly from what "technical" users want...the
only really difference is technical people say "I don't want automatic"
while the "newbies" say "automatic would be nice" (but, even then, they
don't want things so "automatic" that they have no idea what's going on)...

And this "abstraction helps them"? That's Sandra Bullocks too...I often go
and help "newbies" with their computer problems...and it's amazing how the
things that supposedly "help" do the _complete opposite_...for instance,
they put in the CD and suddenly everything is running automatically and
they are like "no! Stop! Why are doing that? I didn't want you to do that!
How do I stop it?"...

Despite what Microsoft think is "useful" and "helpful", I found that if I
made a series of "shortcuts" to each individual drive (labelling them with
"hard drive", "floppy disk", "DVD drive" and then one "icon" to open up
their internet connection and so forth :) then turned off "autoplay", it
was _MUCH MORE USEFUL_ to the "newbies"...exactly like it's much more
useful to me - it's what I have too - because it's much more logical...

If you want to play a DVD then put the DVD into the DVD player...then
double-click on the DVD icon I've created on the desktop...there, that's
it...

And they are perfectly happy with this...users are NOT stupid...they are
simply _CONFUSED_ by all this "autoplay abstraction" nonsense...you know,
they put a DVD in and it starts playing...no, no...I don't want you playing
now...they hit the "close" button...right, okay, _now_ I want to play
you...ummm, there's no "icon" for the DVD player...how do I start this? All
this has done is play things when they didn't want it to play...and then,
when they do, the means to be able to do that has been "abstracted" away...

Look, it's on the front of the box...there, see? A DVD player...it has
"DVD" written on it...in order to play a DVD, the user has _GOT TO KNOW_
which slot is the DVD player and how to put a DVD into the DVD player...

So, as the user _HAS TO KNOW_ what slot is the DVD player in order to use
it...and it's a _PHYSICAL_, _VISIBLE_ thing right there in front of their
nose...how on Earth is it "useful" to "hide" this in an "abstraction"? It
isn't...this is not only "nannying" and highly patronising...it is so much
so that it defies all logic...the user does not find it "helpful", they
find it vastly confusing...they fracking _KNOW_ what a DVD is and which
slot is the DVD player! If they didn't know this, then there's no fracking
way they could ever actually play their DVD movies, right? You can't wave
the disc around in front of the monitor and pray it'll work...you've _GOT
TO KNOW_ which is the DVD player and you've _GOT TO KNOW_ how to put it in
because that's a _PHYSICAL_ thing...once they know which _PHYSICAL_ thing -
that they can plain see right there in front of them on the front of the
box - is the DVD player, what is the point of "hiding" this from them on
the desktop? That just confuses them because they can see the _PHYSICAL_
player right there but they can't see anything on the screen that
corresponds to the player to click on...

_ASK_ the users...go _HELP_ the users with their problems...the truth is,
what they want does NOT that vastly differ from what you'd logically expect
(_IF_ you are in possession of any logic at all ;)...

> Money talks in another way, too:
> advertising. That has to be a factor in what people buy...
> or use. If Windows and Linux were compared on the basis of
> "users per advertising dollar spent", I'll bet Linux would
> look pretty good!

Agreed; I did, a while back, suggest that perhaps something like Linux
might actually try selling itself at a "modest" price (perhaps as
"shareware"...those who really can't afford it could *ahem* just kind of
"overlook"...those that can...well, come on, pay up! ;)...and this isn't
for "profit" or "wages" (although, if any developer _were_ to "fork out" on
something resources for the project, then they could be "compensated" back
:)...but is put into a "pot"...if you like, it's not the developer's
money...it's the "project's" money...this could be used to take out adverts
in magazines or TV or whatever...to fund "support"...that kind of thing...

Now, bringing "money" into "open source" would alter the "dynamic"...you
know, money has a strange way of changing people's behaviour (just check
out how families suddenly start ripping each other to shreds over an
"inheritance" from a dead relative...and, bizarrely, they do this even when
the amount is next-to-nothing...all that "I was closer to them than you
were!" / "You call yourself a relative! Where were you?" / "I was with my
own family...we _need_ the money, unlike you greedy lot, because I have a
family to support...that's why I didn't get in contact all those years, I
was working to make enough money for my family" / "A likely story! You're
just trying to steal it all when everyone knows I was the 'favourite'" /
"So? None of you came to visit her in her final days at the home! I was
there every day" / "No, you weren't! You only turned up on weekends! I did
visit and the nurse told me so!"...blah-blah-blah...they fight and
fight...then, in the end, the "inheritance" of only $1000 is shared 10 ways
between them...the family is destroyed forever...well worth $100, eh? ;)...

But it's a "thought", anyway...you know, kind of like how Red Hat
commercially sell Linux...but, instead, the Linux developers themselves did
that kind of thing...then the money is used for "support" and
"advertising"...probably is, what happens if the money generated becomes
more than needed for such things? Perhaps an agreement that anything deemed
"excess" is given to, ooh, the "Tsunami Relief fund" or an AIDS charity or
something?

Ah, it's a possibility...but getting something like that to work in
_practice_, however "nice" it sounds in "theory"...well, I'm not deluded
that it would be "easy" to get this to work, especially when the developers
are on 4 continents or whatever ;)...

> ...
> > "Sure Word is big and bloated. But that's what people
> > want. They won't pay for apps without a lot of features..."
>
> I don't know how big Word actually is,

[ Massive; Especially when you consider it's only a word processor...heck,
it's not even a DTP program, where you'd kind of expect it might be
bigger...there's a _reason_ why it is the "proverbial bloatware" in
everyone's examples ;) ]

> or exactly what
> features it has (I understand that, unless told otherwise,
> it'll execute any macro it encounters, possibly executing
> malicious code - not a feature I'd put on my "must have"
> list, but thats a different argument...)

You know the "proverbial" toolbar button that "makes tea"? If any
application will eventually include the "make tea" toolbar button then
it'll be Word that gets there first...indeed, I haven't seen the latest
version, perhaps it's already got there...put it this way, Word had so many
"features" (90% of which are of "dubious" value...or make you think: "Oh,
for frack sake, if you want to be a DTP package then become a DTP package
and stop trying to this weird "highly extended word-processor" thing ;)
that if someone did say "it now has a 'make tea' button" then this wouldn't
raise an eyebrow or anything...not at all surprising...

Indeed, I'd even say: Microsoft are getting _DESPARATE_ in thinking up "new
features" to constantly "justify" why they need "new versions"...because
they don't, of course...it's just that "constant forward motion"
thing...indeed, they change the file format every version...do they do this
because of "new features"? Nope, they do it just to introduce
"incompatibility" to force people to "upgrade" all the time...

Oh, remember, "Clippy" was one of the "features" of "Office" and
"Word"...have you ever met little old "Clippy" the paperclip? I think this
annoying little pest makes the case all alone about the dubious quality of
how Word features "help" the user, eh? Enough said ;)

> So perhaps Word
> isn't a good one for me to use as an example... but I'll
> take the chance. I don't think anyone is going to write a
> "feature for feature clone" of Word in "carefully
> hand-crafted assembly".

Perhaps not in assembly language but Star Office and Open Office - both
free...while "Office XP" was some £400 on release here...£400? Microsoft
software is the only software that _DEFIES_ the "economy of scale"...being
the most popular software in the entire world (other than Windows in order
to run the darn thing), they should be generating this thing in _massive_
"bulk" to get "bulk discount", "amortised costs" and so forth...on the
contrary, though, Microsoft's "off the shelf" stuff with the largest sales
of any software on the entire planet is almost costing more than "custom
software" (okay, not quite...but it's close when it should, by almost
rights, be "no contest" and a massive division by all "mass-production" /
"economies of scale" arguments) - might not manage "feature for feature"
(but, to be honest, with the amount of useless bloated "features" on Word,
that's a _GOOD THING_) but if had these then you wouldn't in any way miss
"Office"...it can read all their files, do the job just as well...

BUT - here's the point - if there are no "developer time costs", then,
sure, it's going to take longer to develop at a "lower-level" to get the
better quality program...but the whole point is: Still costs
next-to-nothing...this is all "hobby time"...

It's a _change of paradigm_ here...that's kind of the point I was
making...instead of "bloatware" every year or two, there's one "kick arse"
version every 4 or 5 years that has _taken_ that "extra time" to do a far,
far better job...

Yes, it takes more time...but, in "open source", time isn't an expensive
cost anymore...hence, that's the "paradigm shift" itself...release
"versions" less frequently because you are putting _MORE EFFORT_ into each
version...remember, the "commercial" needs the semi-frequent release to
_MAKE MONEY_ to pay all those "developer wages"...again, not applicable to
"open source"...

Okay, not the best example maybe (though, it's more "design" we disagree
with than "implementation") but Rene has put together RosAsm _IN ASSEMBLY
LANGUAGE_ in 5 years (and was doing it single-handedly for a while too...he
has "contributors" now, though :)...this isn't a good enough "rate" for
"competiting" but it's a "hint"...and, as noted, I'm not talking "assembly
language" but just more "low-level programming", "careful programming",
"optimising code" and so forth...you can do it in C...it's more the
"attitude", not the programming language...

If you like, the comparison is: The commercial releases every two years and
adds on "some new features" each time...bloated and tends to get more
"bloated" all the time...they _need_ to release these frequently to make
the money to afford the "developer time" and make a profit...

"Open source" has different costs and priorities...so _should_ "open
source" be otherwise following the exact same commercial assumptions and
practices? Why "release frequently"? Why care about "rushing" because of
"time"? Releasing frequently is needed in commercial development to make
money to afford it all...of course, with "open source", it's kind of
"constantly released" because anyone can dive in to download the source as
it is being developed...but you know what I mean?

The "paradigm shift", in a sense, is to realise that Rene could work on
RosAsm for the next 20 years, if he liked...and they'd still be no
"developer time costs" for him to really worry about...so, how about -
because of the changed "priorities" in "open source" - traded some of that
"rush" for "quality" instead? The idea is instead of "bloatware with some
extra features" every year or two, we have "leanware with fundamentally
improved performance and features" every 4 to 5 years...you see what I
mean? As there is no "developer time costs" to consider then there's the
possibility of thinking more _LONG-TERM_...taking that "extra time" to the
better job...

> Wouldn't have to be all that
> "hand-crafted" - such a thing might be done using the
> high-level features of HLA, for example - not the smallest
> possible code, but not that bad (it's "not so humble" of me
> to have *any* opinion of Randy's code, perhaps... but I do)

Ah, yes...the point is that things have got so "bloated" everywhere else
that even with every possible "bloat" you could try to incur using HLA's
most "HLL" features, it'll still kick the "bloatware" size and
performance...I mean, getting that confession out of Rene might be
impossible but, as "bad" as HLA is in his eyes, it's still not as bad as
using Java or VisualBASIC or whatever instead...doing our _WORST_ with
proper "low-level programming" is still typically going to weigh in better
than the majority of HLLs (C / C++ is an "exception" here because it may be
a "HLL" but it's one with a distinctly "low-level attitude"...indeed, some
call it "medium-level language" to point out this little "exceptional"
place it holds...it is a HLL but it's almost as if it's trying its best NOT
to be as far as it can - except for that "portability" clause it needs to
keep to)...

Indeed, I sometimes tempted to "re-classify" to consider C as being more on
"our side" than the _TRUE_ HLL side...the interpreted Java running
"scripts" on top of "virtual machines" on top of "scripts" type of
nonsense...

> Even using C - in the manner that Herbert suggests: *learn*
> assembly so you understand what you're asking the processor
> to do, *know* your compiler well enough to understand what
> code it will emit (Herbert sometimes neglects to mention
> this part), and then write "thoughtful" C based on that
> knowledge - would do the trick.

I completely agree; I did note that my points were more about "low-level
programming"...in which I would include "thoughtful C"...indeed, C without
the "standard libraries" is just a kind of "machine-independent ASM" in
some regards..."what the compiler does" is somewhat trivial from C to
ASM...indeed, when MASM and HLA introduce their "IF", "WHILE", "PROC" stuff
then they, in fact, almost make the "jump" up to what C does in reverse
(but without the "portability")...

> "Just suppose" that someone wrote a "feature for feature
> clone" of Word - not necessarily squeezing out every byte,
> as the "smallest code" contestants do, but with *some*
> thought to writing "reasonably concise" code - *some*
> thought to avoiding obvious bloat... Any guess - just a
> rough estimate - of how the size would compare? I'd bet on
> "half" - possibly much less - without even knowing how big
> Word is now. Do you think I'm wrong?

No; You're basically "getting the point" generally...

Redo Word with "careful programming" instead...no, as you say, not
necessarily "all in ASM"...but you're not just grabbing any old "bloatware
libraries"...you're taking your time to select things carefully to do a
_good job_...you optimise, you work on "squeezing" it down (but, again, no
need to necessarily be "every last byte"...indeed, Microsoft's Word is so
bloated, you don't need to go anywhere near "every last byte" to kick their
arses over the Moon ;)...this might squeeze it down to half size, to take
your "estimate", and it runs quickly (the only "delays" are the ones
waiting for the user to type things in...word-processors aren't naturally
"high performance" things - as it's all "user input" driven - but, anyway,
it's better performance where that applies :)...

And here's the comparison...Microsoft release a new version two years later
(call it "Version 10" for discussion...version "9" was that which was
released when you started your project :)...it's much more "bloated" and
slow once more again...then they release another version ("version 11") -
even more bloated and slow yet again - another two years later...

In this time, you've worked on just _one_ equivalent "version" but taking
the extra time to _DO A GOOD JOB_...right, Microsoft's "version 11" has
been "bloated" to the order of _two versions worth of continually
increasing bloat_...this ain't just "bloated", it's "double-bloated" since
version 9...the performance is not fairing any better...

And, against this, you are offering something that's _half-bloated_ of
version 9 and _double-speed_ of version 9...

Hence, by the time you release (and Microsoft release "version 11") you've
got an _EIGHTFOLD_ decrease in size (version 11 is "double-bloat" of
version 9...your "clone" is "half-bloat" of version 9)...and, though Word
isn't a "performance" kind of application, if it was and the ratios were
similar (that is, your "clone" is "double-speed" of V9...while Microsoft's
V11 is "quarter-speed" of V9) then you could have _EIGHTFOLD_ on the
"performance" too...

Do you also see the little "trick"? Because it's a _relative_
measure...Microsoft are doing nothing and simply getting worse and
worse...while you're taking steps to _improve_ and make software
_better_...yes, you take twice as long but you're getting _EIGHTFOLD_ the
"spoils" for doing so...and the fact you take twice as long to do this
better job has no "cost" implications like with the commercial
stuff...Microsoft _NEEDS_ to release that fast and, in a sense, "rush" the
job because it has mouths to feed...this isn't applicable in the same way
to "open source", so, perhaps it's time "open source" _TOOK ADVANTAGE_ of
its "differences" to do what these "bloatware" coders can't do?

Of course, these numbers are all made up...so purely "illustrative"...

Also, there is the "features" argument...the "V11" would have "more
features"...then again, who said your "clone" couldn't also "add more
features" while it's being developed...BUT, again, I'd suggest a "paradigm
shift" here too: Not "more features" but _BETTER FEATURES_...for example,
taking the time to add "incremental" operation and "modular architecture"
and such to LuxAsm for _better_ features...rather than just add on lots of
toolbar buttons...or, to take some "bias" out of it, to use Rene's
"right-click feature" for "easy navigation" in the text editor - take the
time to put that in - rather than just add in lots of "search this",
"search that", "advanced search" option in the "search" menu...indeed, add
_smart_ "features" or don't add any at all...

There _IS_ such thing as "feature overload" (some believe it's a "myth" but
it's NOT)...a point where you've got so many features, it becomes "top
heavy" and starts working _in reverse_: The user is now _CONFUSED_ by too
many buttons and options and dialogue boxes...what does that button do
again? Oh, crap, where was that dialogue box option again? And their
productivity vanishes as they are trying to "re-learn" or are
systematically going through each "dialogue box" and every "tab" to find
the option they are looking for (to be honest, I often get this problem
with Windows itself...someone says "can you switch this off?" and I say
"sure" but then it's like "oh, crap...I know it's in a dialogue box...ummm,
is it 'display properties'? Nope, then is it in 'control panel'? Nope, what
about 'administrative tools'?"...going through it all, trying to find out
where the option is (I remember there _is_ an option for it...just can't
remember where Microsoft have "hidden" it anymore ;)...

And, indeed, if it's getting "monolithic" then you should be considering
"divide and conquer", not bolting on "more features" yet again...similarly,
"open source" has different considerations...not all of that which apply to
"commercial" continue to make sense with "open source"...

This, in a sense, is the underlying point...it _works_ differently, it has
different "expenses", different "objectives", etc., etc....therefore, we
probably _SHOULDN'T_ - in "open source" mode - be thinking of projects
necessarily in the same "commercial" terms that we usually do...for
instance, a more "specific" version? A more "long-term vision"? A different
"attitude" to using "bloatware features" (basically, umm, just "don't" ;)?
And so on...

It's a thought to ponder - a "meme" to "inherit" - anyway ;)

Beth :)