Coding Horror: Why Does Software Spoil?

[tinjaw]

The latest posting by Jeff Atwood has sparked a very interesting, and long, comments thread.

Coding Horror: Why Does Software Spoil?
http://www.codinghor...archives/000973.html

Sorry for the drive-by posting. I'm on a short break at work.

[Ralf Maximus]

Wow, that was a good read.  The article was thoughtful and the comments spot-on.  Nice find.

The only thing I can think of that nobody mentioned was how much *nicer* it is to run old software on new hardware.  Stuff that was slow and resource-hungry 10 years ago runs like a dream on a modern dual-core PC with a gigabyte of RAM.

As a developer (and business owner) the best way I've found to keep customer-driven feature creep down is to make them pay for requested modifcations.    If it's really important to them, they'll open their wallet.  I charge them a token amount, but that's all it takes to filter out the crap.

However, I cannot fathom being a freeware author in this position.  The stuff's FREE and you want more??  I'd go postal in a week.   

[tinjaw]

Yes, the money barrier is an excellent filtering mechanism to find out who is serious and who is just whinnying because it is easy to whip off an email or comment on a forum. I'm not sure if I will ever sell conventional software applications, but if I did, I would think seriously about having a basic core provided as freeware, a shareware version with added functionality that most users that would use the software more than just occasionally would find worth the small price tag, and then as many plugins as were called for. I would build the software so that the basic functionality of the shareware version would be all that loaded initially, and optimize the hell out of it to load quickly. Any additional functionality would load in the background as the application idles. I am sure other applications do this as well.

And bloat is definitely a subjective thing. However, when there are significant hits to performance and usability, for any reason, that is a bad thing. If the cause of that is new features you never use, it can only lead to anger. I find it is always a good policy to wait six months or more to find out what the user base reports back on the new version. The exception being, of course, if you are being held back in a present version when the new version has what you need to move forward. Then, in that case, I say jump to the new version and try the 30-day trial almost all programs have these days, and if there is no significant tradeoff that you aren't willing to endure, then pay the upgrade price.

[app103]

I bought Snagit 5 when it first came out. I have been using it ever since, never bothering to pay for an upgrade. It worked, did what I wanted, did it well (even on my old slow pc)....never had a reason to ditch it in favor of anything else or pay for an upgrade.

I recently was given a free upgrade to v7. They were hoping that I will like it enough to want to buy v8.

I tried it out...did about 3 test screenshots and uninstalled it, going back to v5.

This is a fast machine, 3.2 ghz with a gig of ram. Snagit 7 was s-l-o-w. I'd hit the hotkey and have to wait for it to allow me to select the area I want captured.

I don't want to wait. I shouldn't have to wait. I never had to wait before. I am generally a patient person, but not that patient.

It was slower running on this machine than v5 was on a 233mhz 64mb ram machine, if you can believe that. 

Whatever they did, they ruined their product, in my eyes. No amount of extra goodies they could add could make up for the slowness that came along with it.

Now I suppose that if I had upgraded from v5 to v6...and then to v7, maybe I wouldn't have noticed the speed difference as much. But that isn't the case here, and I noticed it enough to hate their product.

It's a real problem for developers that make the decision to make users pay for upgrades, rather than giving them lifetime free upgrades. It puts them into a certain way of thinking where they have to keep banging out newer versions with more & more features to encourage people to pay again and upgrade...enough features to make it worth it to the users. So the applications become bigger & bigger and more bloated and including tons of silly things that hardly anybody would want or need.

It's sad really. Because the best version of their product is pulled and you can no longer buy it, if you are a new customer. As a user, the solution is either find something else (company loses your business) or do something slightly less than honest, like paying for latest version and then never using it, opting to pirate an older better version, instead.

I think if the trend continues, we may start seeing software that includes lifetime free downgrades, or free upgrades but pay to downgrade. 

[zridling]

Great find, tinjaw. Here's the line that is true to my experience this past year:

...eventually you begin to loathe and fear the upgrade process.

Look at Apple's recent Leopard release. It was immediately hailed as a Vista-killer, and then people starting using it and quickly found a dozen problems, among them very serious ones. Look at Vista. No more need be said. CorelDRAW was one of my favs and I stuck with it for 10 versions! until they simply made it impossible to install without the damn thing taking over your system. And it's slower than Photoshop now.

I also hate to see a great, fast program be feature-bloated by bitchy forum users. I've seen it happen over and over and over. Some guy writes in and makes it his mission in life that his FTP program has a spellchecker — huh?!! Donald Lessau of XYplorer does a great job of cutting these type of requests down fast. If such requests fit within his roadmap for the program, he schedules it for an upgrade. Otherwise, he explains why he doesn't want to do that.

As Falstaff taught us, discretion is the better part of valor!