Perfect Software?

[Ralf Maximus]

I've noticed a kind of zeitgeigstial interest in "old" software, or previous versions of current software.  There appears to be a resurgence of web sites devoted to past releases of MP3 players, IM clients, and productivity suites.

This leads me to a thought and a question: does anyone know of a software company that said at some point, "this thing we make is as good as it needs to be, so we're not going to improve it anymore?"

I'm not thinking of projects where the developers ran out of steam, lost interest or capital -- I mean cases where somebody made an executive decision that anything more would be needless and bloatful.

Think how different the world would be if instead of Vista we had XP 2007, or Outlook went feature-complete in 2000.  Or if WinAmp stopped at 2.x.  I'm beginning to wonder if the "always add new stuff" mentality is starting to wane, if companies will notice the trend and question their direction.  Are we outgrowing the "newer is always better" mentality of the 90's?  Or are we forever doomed to next year's model?

So... any stories of arrested development?

[Grorgy]

Thats a coincidence, or maybe not, appaholic, relequestual from the forum here, featured metapad, a notebook replacement just the other day which hasnt been updated for years, tho the help at the website has been updated to to include instructions for vista.

[icekin]

I can think of CATIA, a 3D CAD tool that has not been updated much since 2001 when it hit v.5. Since then, it has had several revisions, currently on R17, but no major rewrite or feature changes have occurred. The advantage is that even the newer revisions run on the same hardware that the older ones did.

In a shop, the average consumer is simply going to buy the biggest box with the most features written on the cover. Few would stop think about how the extra bloat would slow their machine or whether they even need all those features in the first place. Companies end up bloating a software trying to capture this market. In some rarer cases, the developers start bloating the software because they start to believe that those added features will greatly help users.

I would personally love a world where companies focused on improving stability and portability of their programs rather than features. I especially dislike companies that release the same program with the higher version number and claim that new stuff has been added (e.g. AIM, Nero). Its best to have a modular structure and allow independent coders to write their own plugins. Firefox has this system and it has worked out quite well.

[Darwin]

This is an excellent thread, raising and really interesting question. Sadly, I can think of many wonderful apps that were abandoned (but released as freeware) after a major OS upgrade (Win 3.x to 9x or 9x to XP etc.) but nothing that has been maintained and updated for years without a significant upgrade.

Unfortunately, as much as I recognize that this is desireable, I'm like those people in the box stores looking at packaging. I find it almost impossible to resist the urge to upgrade to the latest version of an application, even if my needs are more than met by the previous version. This is a complusion, and an often expensive one! I'm pretty sure that I'm not in the minority. I'm sure that software developers are aware of this mindset and keep piling on the features/tweaking the GUI/changing the code to accomodate it. Why? Two conflicting reasons: 1. This mentality will drive many users to DIFFERENT solutions, even if there is NOTHING wrong with the current one. There is the perception, familiar to anthropologists and sociologists, that stasis in culture or evolution is backward despite the fact that it is seen by practitioners as indicative of success and 2. It keeps revenue coming in because people like to buy actively developed software so either you've got a lifetime licence model and need to keep evovling your product to attract new users or you've got a paid upgrade model and need to keep your existing users coming back to fill your coffers (even if only a little)...

Both developers and end-users are caught in a vicious cycle - one that is ultimately counter-productive.

[CWuestefeld]

A couple of examples come to mind. I'm not sure if it was a conscious decision on their part, but the fact was that these applications languished, and are now really historical footnotes.

• PKZip - While this did everything it aimed to very well, a "bloated" Windows app (WinZip) eventually cleaned its clock despite using the same underlying algorithms.
• Lotus 1-2-3 - Languished in the version 2 realm for ages, waiting for others (particularly Microsoft Excel) to catch up. Eventually a version 3 was released, but it was too little, too late. At one time "Lotus" was synonymous with spreadsheet, but now that name seems to carry the connotation of "Notes".

In both of these cases, while the app itself may have been well-suited to the platform it targeted, the evolution of platforms (particularly the triumph of Windows over DOS) sealed its fate.

A similar story is in progress for the super-organizer Zoot, but I hope that the sequel comes out before the story ends. This is still a 16-bit app, and has been passed by by the need for rich text and volumes of data beyond what its 16-bit space can handle. It's been tweaked continually, but hasn't had a major revision in several years. Luckily, there is now a public beta for a 32-bit port, so hopefully it will fight its way back (it deserves it).

This all seems to imply that even when an application is "just right", the developer can't just rest on his laurels. The evolution of the platform underneath him will eventually kill his product just as surely as poor quality might.

(edit: fix formatting)

[app103]

I only know of one such application...a silly little clock. (there are probably loads of small single purpose apps that are the same story)

Back before I learned any programming, a friend of mine made a small desktop clock for me. When I requested it, I was unaware just how much of a stubborn minimalist he really is.

I gave him an image to use as a guide.



This is what I got from him:

• When I asked to be able to change the colors and the font, he said no...that would be bloat.
• When I asked for a border on the clock, same answer.
• I asked him to remove the X and put an exit option on a popup menu...nope, he wouldn't do that either.
• How about remember position? nope.

Everything I asked for, I got that same answer....no, that would add bloat.

He says the clock is perfect the way it is and there will not be any adding of features to it or any 'beautifying' ever. (If one discovers a bug, he will fix that, but it does what it does and does it well, so there is no bug fixing needed.) He says it could only be more perfect only if he can find a way to remove more 'bloat' from it and make it even smaller.

This was the application that gave me the push to learn programming. My frustrations from dealing with him lead me to learn how to create my own version with all the bloat features I wanted.

We still argue about this clock, even today, with him still insisting that it is perfect.

For anybody that wants his clock, it's attached to this post. He gave me permission to distribute it. (earlier 'bloated' versions and source are available, bundled, on his site's files page)

[Ralf Maximus]

We still argue about this clock, even today, with him still insisting that it is perfect.

This has me so conflicted I'm almost shaking.  ("Norman, please coordinate.")

On one hand, I want to buy that guy a beer.  Talk about sticking to your guns!  He should be held aloft on the shoulders of every developer forced into a feature-creeping deathmarch by the Suits above.

On the other hand, people like that cause me to look for places to hide the body.  How frustrating.  Such simple changes to make a satisfied user your pal for life!  I can't fathom ignoring such harmless requests for refinement.

On the third hand, he *did* do it for free, so you've gotten what you paid for.  Unless you've been paying annual support.  :-)

In any case, it's a perfect (extreme?) example of a developer putting his foot down and saying no to bloat... so I guess.. I ... *reboot*

[Beth UK]

This is another good thread!

I have yet to see an adequate description of the term 'bloat' and wonder if, in reality, it is one of those semi-abusive terms we throw at software developers when something doesn't work? Nero is often attacked with the word 'bloated' yet some people swear  it is the best burning software ever.

The concept of 'perfect' is an interesting one too because you could argue that 'perfection' is a destination never reached. We can only approximate 'perfection'.

I guess we could say that the concept of a triangle is perfect - especially if we consider the tautological statement (true in itself) that a triangle has three sides. Well, yes - it must absolutely have three sides to be a triangle, and in having three sides can only 'be' a triangle. It can be nothing else, and therefore might be said to be 'perfect'? But in reality (outside of the concept) no true triangles can exist 'perfectly' because of the imperfection in construction and can possibly only meet the requirements of 'triangular'.   - I have had too much coffee today!!

app103 - your description of the development of the clock software is interesting because IF the goal was to meet YOUR requirements then the resulting clock failed to meet the 'perfection' of your design brief. Also, as the measurement of time itself is fundamentally flawed you could say the clock (as a mechanism to measure time) is fundamentally flawed and could never be perfect anyway. So your friend would need to say how exactly it is perfect. Without knowing his 'criteria' of perfection we cannot accept it as perfect in any way? Also, if the clock was perfect there would be no possibility of bugs ever existing - so in saying he would fix a bug isn't he accepting the possibility of the clock being less than perfect?

I don't think the visual design of software can ever be described as 'perfect' - only popular or unpopular. Lines of code can be 'perfect' if they subscribe to all the rules and are free of errors. But then again, code may be flawed within it own logic and therefore can never be perfect.

I think we can maybe talk about software that is 'good enough' for the task in hand in that it is useful for most people, most of the time. For someone it will NOT be good enough (that is almost certain), and can therefore not be described as perfect.

I think if we began to look for a 'perfect' definition of 'bloated' (the opposite of perfect software) we would find it impossible to agree. Likewise, maybe, the idea of 'perfect' software is purely conceptual, and not obtainable in real life.

[jgpaiva]

I think i'd go for allsnap (see review here). It hasn't been updated since 2005, and i also don't think there's any need to update it. I am so used to it, that i don't even notice it there anymore, only when it's missing

I don't think that having microsoft/apple/linux people staying in one version of the OS and not updating it would be good though, the OSs need to keep up with the new technologies, since our life is in a constant change. Regarding winamp, for example, i think it's a good thing it has been updated, i like to be able to connect my zen micro to it, and i love the new interface in the most recent version!

[Dirhael]

Nero is often attacked with the word 'bloated' yet some people swear  it is the best burning software ever.

The burning application itself still is one of the best ones you'll find on any platform, but the problem is that you no longer can purchase *just* Nero Burning ROM. It's bundled with just about every single type of application you can think of, and you can't just choose not to install them (well, you can deselect most but it will still install a whole bunch of stuff completely unrelated to burning CD/DVD's).

Just to experiment for this thread, I installed the latest trial of Nero 8 and deselected everything I could see, and yet this is how many directshow filters it forced down my throat (see attached image)...and then there's the fact that it adds several startup processes that just isn't needed for burning at all. Really, if I wanted a complete media center I'd have gone out and purchased software dedicated to such a task.

When the English-only version of the Nero 8 trial is 640MB large (the 178MB download is compressed so it will extract the installation files to a temp folder, which is that large) you know something is very very wrong.

Thankfully there are certain enthusiasts that create custom installers without the bloat, despite the fact that the Nero guys keep trying to shut them down. The fact that these guys managed to keep the installer down to just 13MB and still keep all the burning functionality is quite interesting...

EDIT: I also just wanted to add that if an application (like Nero) has a add/remove feature list during installation that is longer than a very bad year and every single feature selected by default, it should be punishable by death not to have a check/uncheck all button in said list!

[Ralf Maximus]

Thankfully there are certain enthusiasts that create custom installers without the bloat, despite the fact that the Nero guys keep trying to shut them down. The fact that these guys managed to keep the installer down to just 13MB and still keep all the burning functionality is quite interesting...

I've got Nero 6, with zero desire to upgrade beyond that.  Even Nero 6 has loads of crap I never use and a bazillion codecs I don't want.

Know of any stripped-down installers for v6?

And if I uninstall what I have now and reinstall using one of the enthusiast packages, will I get rid of the silly unwanted codecs, or do they stay forever lodged in registry pergatory?

And can I have a pony?

[Dirhael]

Thankfully there are certain enthusiasts that create custom installers without the bloat, despite the fact that the Nero guys keep trying to shut them down. The fact that these guys managed to keep the installer down to just 13MB and still keep all the burning functionality is quite interesting...

I've got Nero 6, with zero desire to upgrade beyond that.  Even Nero 6 has loads of crap I never use and a bazillion codecs I don't want.

Know of any stripped-down installers for v6?

And if I uninstall what I have now and reinstall using one of the enthusiast packages, will I get rid of the silly unwanted codecs, or do they stay forever lodged in registry pergatory?

And can I have a pony?

Unfortunately, I haven't seen a lite/micro version of any version below 7

The codec situation is easier to solve though, as I would recommend you to just download and run the lovely freeware DirectShow Filter Manager 0.5. As you can see from the screenshot, it lists all directshow codecs installed on your system. Just select the ones starting with Nero and hit that tempting "Remove Selected Filter" button

Alternatively, if you're unsure about if you might need the codec in the future then hit the "Show filter properties" button instead and set the merit to unlikely or lower. This should ensure that it's only used if nothing else on your system can play a video/audio file in a DirectShow player like WMP. My personal experience is that most of the codecs doesn't do that much harm as long as you remove all the Nero splitters and quicktime codecs. The splitters in particular have caused all sorts of strange problems in the past for me, so that thing isn't getting anywhere near my system again except for testing purposes like in this thread.

EDIT: Oops, wrong image attached...and I also forgot your pony

[Ralf Maximus]

PONY!!!  :-)

The codec situation is easier to solve though, as I would recommend you to just download and run the lovely freeware DirectShow Filter Manager 0.5.

What a nifty little tool, and perfect in its simplicity.  Not to look a gift-pony in the mouth, but any idea what the green/yellow/red color coding means?  All my Nero filters are red.

[Dirhael]

PONY!!!  :-)

The codec situation is easier to solve though, as I would recommend you to just download and run the lovely freeware DirectShow Filter Manager 0.5.

What a nifty little tool, and perfect in its simplicity.  Not to look a gift-pony in the mouth, but any idea what the green/yellow/red color coding means?  All my Nero filters are red.

The colour coding comes from the priority of the codec. The ones "preferred" is purple, "normal" is green, "unlikely" is purple and "do not use" is red. It's important to remember though that "do not use" doesn't really mean what it says (and this is not a problem with the filter manager, but AFAIK it's rather with DirectShow in general). They will still be used if nothing else can play a file, and in certain cases they might even end up being the first choice for some reason despite your settings (I don't undertand the system well enough to explain why exactly this happens though) so it is always a good idea to remove the ones you don't absolutely need.

[Ralf Maximus]

Awesome!  Thank you.

[justice]

Perfect is boring and that clock is a good example of it.

[Armando]

Perfect is boring

MS understood that a long time ago. 

[Beth UK]

Perfect is boring

MS understood that a long time ago. 

Yes! The thrill of a system crash adds variety and flair to the computing experience!

[Ralf Maximus]

Perfect is boring

MS understood that a long time ago. 

God, yes.  And we have companies like MS to thank for the current state of things.

The sad part is, I believe they could still make their recurring revenue from improving their existing stuff rather than the wholesale replacement strategy they pursue now.  Vista is a total departure from XP; Office 2007 is a complete rewrite.  And when the dust settles, after all the millions have been spent, the end result is only marginally superior to what was replaced, if at all.  Sometimes it's just *different*.

The world would embrace something like XP 2008, where the core product remains the same but additional drivers are included, the Aero interface is an option, and loads of cute new themes (shudder) are thrown in for fun.  Charge $99 for the upgrade and people will pay.  MS could even make more money with such a product, since support calls will be fewer and (one would assume) the development team much smaller than what Vista required.

This breaks away from the original "perfect software" question I posed, but compliments the underlying "when is enough is enough" theme quite nicely.

[CWuestefeld]

The sad part is, I believe they could still make their recurring revenue from improving their existing stuff ...

The world would embrace something like XP 2008, where the core product remains the same but additional drivers are included, the Aero interface is an option, and loads of cute new themes (shudder) are thrown in for fun.  Charge $99 for the upgrade and people will pay.

I dunno about that. People have come to believe (probably correctly) that the additional drivers are theirs by right, that they shouldn't be charged for it. Likewise, you can find cute new themes for free out on the Intertubes.

It's funny, though, that when referring to software versions we still generally use the nomenclature "version 3.0" -- but how often is there anything other than a "0" to the right of the decimal -- particularly in major products. Think, for example, of recent discussions around these parts concerning Paint Shop Pro versions X1 and X2 -- there's no thought of revision numbers in there.

The exception to that rule, of course, is open source software, and this is a pet peeve of mine. One way that you can tell something is open source is that it perpetually has a "0" in front: "0.8.7.1". The authors never have the confidence to say "I think this is a good baseline; let's bless that and work from there." Years ago I mentioned this to the guy responsible for POPFile (one of the best spam filters out there), and he assured me that he had a clear roadmap to get to v 1.0. Years later, they're still stuck in the 0.9's. And of course, this leads to either the software languishing, or feature creep as they asymptotically approach their v1.

[Darwin]

The world would embrace something like XP 2008, where the core product remains the same but additional drivers are included, the Aero interface is an option, and loads of cute new themes (shudder) are thrown in for fun.  Charge $99 for the upgrade and people will pay.

Agreed. This is essentially what Mac does with their OS - these new versions that roll out every 18 months or so are paid upgrades, but the price is very small. AFAIK, you simply go buy the latest version and install it. I *belieive* the price is in the $100 to $149 range. Compare this to the pricing scheme for XP and Vista! Nontroppo would know better than I (my OS-X machine is simply too old to run anything newer than OS-10.2 and is a glorified toy WRT how I use it, so I don't pay attention to the cost of OS-X!).

[nosh]

Compupic stopped being developed years back and it's still my first choice of a gallery-based graphics viewer. Photodex even had a mad-fast DOS graphics viewer called Sea, wonder how many people here tried that one...

I have an OCD like a lot of others to go with the newer,bigger,faster software but whenever I've combated it and decided to stick with an older version, the decision has invariably paid off. I'm even going to try and boycott the FooBar2000 upgrades. Lets hope it doesn't go the Winamp (sucks the llama's ass) way.

I'm sure I'm not the only software fan who experiences these info-overloads either, we've all seen enough good programs go down the crapper.

[Darwin]

Lets hope it doesn't go the Winamp (sucks the llama's ass) way.

I have an OCD like a lot of others to go with the newer, bigger,faster software

You and me both!

but whenever I've combated it and decided to stick with an older version, the decision has invariably paid off.

I've been trying to do this for years. Have only succeeded a couple of times. Other times I've succeeded in "holding out" for a full release or two and STILL break down and upgade. I need help...

[Ralf Maximus]

I've been trying to do this for years. Have only succeeded a couple of times. Other times I've succeeded in "holding out" for a full release or two and STILL break down and upgade. I need help...

The cure for this: download the trial version of TantalizingNewUpgrade v33.5 and install it in a virtual machine.  Play with it.

I did this with the latest WinZip, saw nothing spectacular (still doesn't play MP3 files while unzipping) and was instantly released from my compulsive urge to get the damned thing. 

Curiosity can be a seductive mistress; be strong.

[app103]

I have an OCD like a lot of others to go with the newer, bigger,faster software

You and me both!

but whenever I've combated it and decided to stick with an older version, the decision has invariably paid off.

I've been trying to do this for years. Have only succeeded a couple of times. Other times I've succeeded in "holding out" for a full release or two and STILL break down and upgade. I need help...

The cure: Use an old 233mhz machine with 64mb ram as your only pc for the next few years.

You will start to seek out and love the small single purpose tools that have no bloat. You will lose your desire to upgrade and begin to look for things in which you can actually downgrade to the most efficient version. You will begin to really look at features offered and how you actually use the software to decide if the added features are of any real benefit to you.

Then return to your latest & greatest PC, bringing the best tools with you. You will love them and appreciate them enough to stick with them and resist the temptation to upgrade them or throw them away, unless there is a security issue that dictates that you should.