A New Imaginary Parallel

I like Cheerios: or "How I learned to despise net marketdroids"

HOW many lines of code? or What do these 80,000 lines do?
 
 

Rant 1:

Time for new imaginary parallels.  The most common one we've all heard is that if Detroit had advanced as quickly as the computer industry has, we'd all be driving incredible distances at ludicrous speeds on miniscule amounts of fuel.  The new one that comes most to my mind derives from MSN Messenger.  My oldest email account is my Hotmail account, from over a year before they sold out.  I consider Messenger's popups, telling me how many new messages I have and when I get a new message, to be incredibly convenient, but I have no use whatsoever for any other feature of it.

I especially have no use for one feature in particular: the updater.  Messenger works, does what I need it to, and that makes me happy, but I scream every Friday because the updater reminds me that there's a new version 'available'.  I don't want it.  I know it can only be larger, and packed with more crap I won't use, but I only have three choices.

The quicker of mind will immediately note that there is no "Piss off and leave me alone" option.  So, if I want to use it, I have to put up with pressure to get the newer version that I neither need nor want.  This seems to be the typical tactic nowadays... not running and throwing a big bucket of water that soaks you instantly, but dancing around with a plant mister, invincibly agile, soaking you gradually.  I'm almost ready to give in, even thought I know that fairly soon they'll make some change 'for the better' that will render previous versions unusable, so I'll have to get the 8 Mb version, if I want to stay appraised of my Hotmail account without checking it all the time.

The new parallel, you ask?  Just this.  Stipulating the previous comparison between the Big Three and Big Silicon, the more appropriate one is the matter of planned obsolesence.  Sure, Detroit invented the concept, subtly changing each year's new model to distinguish it from the last, but at least Detroit couldn't instantly render the Buick you've got in the garage an inert hunk of metal, totally unusable.  And even if they could, I don't think Detroit would be crass enough to do it just so they could sell the new model cars, with no brake pedal or windshield, but with extra wipers, an operational extra steering wheel in the back seat, a paisley paint job, and the front axle perpendicular to the ground... in the name of 'enhancing the user experience'.

I'll take that back... they would be crass enough to do so but they couldn't get away with it for a second; Big Silicon itself, however, hails these actions as triumphs, and we dumbly nod along and grin.  Progress is fine and dandy, but there's something to be said for the limits that Detroit labors under.  They'd sell the idiotic car I described above if people would buy it, but aside from the fact that there exists actual competition in the auto market, no car for the general consumer is ever sold without some significant options.  You'd still be able to buy the base model (a fairly normal car), the intermediate (perhaps with just the extra steering wheel, in a more subdued paisley), or the full-on idiot-model.

I like my old Buick, even if it is loaded with options that I never use (who really needs a phone booth sticking out of the trunk?), and I'm not looking forward to the letter from Detroit that says it will be so much scrap metal in a week.

I remember reading somewhere that the simplest machines are the best, even if elegant and efficient design is more difficult, and so the complexity of the engines turned out by automakers today are generally the result of lazy thinking.  Who would use a ten pound, spring-loaded, gear-driven, remote-controlled... paperclip?  Such is the bloated code on which most programs today are based, but we use them because no one can be bothered with making regular paperclips anymore, and the general consumer lacks the wit to even want a little, bent, piece of wire to hold papers together.  Consumers demand the Ultimate Paper-Smasher-Together because it's sexier, even if it doesn't hold papers together as securely as a little, bent, piece of wire.
 
 

Rant 2:

I'm reminded of x10.com's faq about the advertising practices of theirs that I find so very obnoxious.  They used the phrase 'enhancing the user's web experience', I presume not tongue-in-cheek, to explain that their ads pop-under the browser window because it enhances the user's experience of the web for them not to pop-over.  It is true that this is preferable to popping-over, and it just takes a click to close the beast when you see it, but it is most emphatically not enhancing to the user experience to be less rude than possible.  I wrote to their marketdroids and said:

"It may be that urine is easy to separate from milk, but please do not whiz in my Cheerios and smile and tell me that you are enhancing my Cheerios-eating experience because all I have to do is separate the urine from my cereal."

They never replied.

How odd.
 
 

Rant 3:

Perhaps MS addressed the issue of new window sizes in a Service Pack, but my system is secure enough, I don't run anything even vaguely mission-critical, and I flatly refuse to download 5-50 Mb service packs or updates on the basis that software should work without needing massive downloads to fix everything that didn't work or didn't work correctly while the product was initially rushed out the door, despite known problems.

The interesting thing is that I've seen these service packs, updates, etc. also available from MS on free CDs, but shipping and handling is about $5.  By that measure, AOL has spent about $80 trying to get me to sign up (I have many AOL coasters), so... figure they've direct-mailed, what? 50 million CDs total?  That's probably conservative, so they've pissed away at least $4 billion dollars just mailing them?  I think not.

Obviously, I also think every user of Windows should have a stack of free CDs, saved from each mail-out of the latest set of patches and upgrades.  Of course this is an unreasonable position to take, from a business standpoint, but is it any more reasonable a position for consumers to take, that it makes perfect sense to plunk down hundreds of dollars for software we'll have to go to unreasonable trouble or expense to maintain?  Not because software is of such nature that it always 'goes bad', like the deli you got last week, or because it acquires 'wear and tear' from normal and responsible use, but because Windows is phenomenally complex and impossible for a single person to truly understand.

This is not a good situation from a design perspective.

Engineers walk the streets who completely understand the workings of engines, from the metallurgical aspects to matters of dynamic balance to petrochemical reactions, but no single person can sit down and  fully describe Windows or reproduce it given sufficient raw materials.  For every regular developer and programmer who asks, "Why on Earth does Windows do that!?", there is probably a very small percentage, if any, of the Windows team who can give an answer to each particular 'that'.

It's true that no software on a large scale can be bug-proof, and Win2000 is incredibly stable and usable, but why is it a triumph worthy of praise that they've produced a stable OS?  Do you applaud the ex-bank robber for walking past a bank and not robbing it?  I'd have a much higher opinion of MS if Win2000 had been offered to the public with apologies for past sins replacing the hysterical self-congratulation that was generally aped by everyone else with a demented and sheep-like attitude toward major pieces of software.

This is not to say that Windows should be so simple that a single person can encompass it all, but that anything as complex as Windows currently is must be reducible to simpler, more elegant, and more efficient principles.  Anything this complex has gotten out of hand and can only become more incomprehensible, and this can only make it more difficult for the Windows sub-teams to make its many parts work together correctly.  The designers of an engine know where all the parts go and exactly how they work together, but can the same be said of the designers of an OS like this?

One cannot expect wond'rous great things from a team that is literally developing they know not what.
 

Return Home

Bryan Flynn

IE 5.0, IE 5.5, IE 6.0, Internet Explorer, and Windows 98se, Me, and 2k/2000 are all trademarks of Microsoft Corp.

All rights to page contents reserved by author, copyright 2001.