Slashdot Mirror
The Trouble With Software Upgrades
Carl Bialik from WSJ writes "When software makers urge upgrades, it isn't always in users' best interest, the Wall Street Journal reports. Many upgrades bring advertising or other unwanted features; some iTunes users felt this way about a recent upgrade. But for many programs, downgrading can be a headache--Yahoo generally doesn't link to old versions of software, and Apple says iTunes can't be downgraded. Some websites can help with the problem. OldVersion.com, for instance, offers more than 600 versions of about 65 different programs. The site's 16-year-old administrator says, 'Companies make a lot of new versions. They're not always better for the consumer.'"
Here is a problem for most software companies, and one reason I would never invest in a publicly traded software company. It is also the prime reason that I sold off my portion of a private software company I had owned.
The problem? Obsolescence.
Software seems to be obsolete almost immediately after it is released. If a better product doesn't replace it, the product itself contains bugs that require a new release or at least a patch. The difficulty in pricing software is figuring out what percentage of the sale profit needs to be held back to cover long term support (updates and customer service).
One way developers are recouping the expense of upgrades is by offering yearly support subscriptions, but these are better suited for corporations who desire a fixed budget. For the home user, I'm betting most prefer to buy a program once and desire a lifetime of upgrades. Recently I complained (to myself) about needing to rebuy a program that had been updated -- until I realized I hadn't bought a version from the company for 4 years!
The end result is for the company to find others willing to pay for the upgrades. Users who desire something at a discount should be willing to at least admit that they're also part of the problem -- they tell the developers that they'll buy a product at a certain price, and they give the developers reason for finding ways to pay for that product in the long haul.
In all the software I use (a ton of it between my businesses, my home, my side projects, my church congregation tech junk, and my family needs), very rarely does an upgrade work against me. In fact, I'd say 95% of upgrades I've performed in the past 10 years made me more efficient, even if they incorporated certain things I didn't like.
If software wants to do something you don't want it to do, block it with your firewall. For me, that's the only necessary step.
The final part of the quote: "They're not always better for the consumer" needs to be looked at differently. Updates that allow the developer to continue updating and supporting the software ARE good for the consumer, just maybe not in the "now" but in the long run. The time preference of the developer might be different than the consumer, but they have to be similar or the developer won't last.
I seriously hate it when someone says, "Here's the new release, it's going in right away!" That's where the term "Bleeding Edge" comes in.
I typically upgrade when I feel I need to, i.e. there's some new feature which really is great or required for the work you do.
Lastly, this guy is 16? Props!
A feeling of having made the same mistake before: Deja Foobar
Upgragde does not mean better. WindowsME was supposed to be an upgrade to 98SE. Trust me, it wasn't.
The company I work for writes software. Trust me, I would not recommend anyone to buy the first release of any upgrade we offer. Wait until it gets about 30 or 40 builds and becomes stable.
I guess you could say that the reliability of software is like a wave: It goes up until a major release, then it drops down to the bottom and starts working it's way back up again. When the software becomes perfectly reliable and feature complete, it's time to release a new version, and down we go again.
There is no "I disagree" mod for a reason. Flamebait, Troll, and Overrated are not substitutes.
Unfortunately most non-technical users have "Upgradeitis", which means that all it takes to get them to upgrade a piece of software is to tell them in straight forward language "An upgrade to XYZ software is available" and make the upgrade process painless. The average user (in most cases I think) doesn't stop and ask themselves, do I really need this? Is the software I have now doing what I want it to do? What's in this upgrade that I really want/need? How will this upgrade affect my data and/or other applications?
Seems to me that they've become almost programmed to think of anything new as necessarily "better" and thus desirable without ever thinking of the old addage "If it ain't broke don't fix it", couple this with the propensity of many users to load up on drive by download software that they'll probably only ever use once and you end up with a pay per incident support providers wet dream.
There's a part of the movie when Elliot Carver is having a teleconference with his underlings, and one of the questions he asks is:
Elliot Carver: "How about our new software?"
Underling: "We are releasing on schedule. As requested, it is full of bugs, and users will be forced to keep upgrading for years."
I wonder how valid that statement actually is.
...when you call the outsourced, scripted tech-cupport center.
"Are you running the latest version?"
"No."
"Well, we need you to install it so we can diagnose your problem."
---
More often than not the upgrade is better, it's the 5% of times when it's a pain. Due to software development and support being a resource hog, vendors are quick to abandon older versions instead of developting patch trains for multiple releases.
What's that? You say that they're talking about Windows? Pfff. Who uses Windows? I mean, do users have any clue how hard it is to uninstall... Oh.
:-P
(Yes, my tongue is again located in the cheek area.)
Yes, between Steve Jobs' cheeks it would seem...
Code, Hardware, stuff like that.
Will agree with you wholeheartedly.
The company I work for uses QuickBooks Enterprise. We started on Version 5 (Quickbooks 2005). It worked pretty well. There were a few very small hiccups, but mostly it did what we needed it to do. We had purchased upgrade protection because we knew a new version would be coming out shortly. About the 5th of December or so we received the 2006 Update.
Now, being skeptical to begin with, I was NOT going to install this right out of the box. It's one thing to upgrade WinZip or WinAmp to the next version, another thing entirely to take the company's accounting server down for an upgrade that hasn't been proven in the wild for more than 5 days.
Skip forward to the beginning of February. Two months have passed and the support forums on the QB2006 site are relatively quiet. There is no patch released yet, and no notification that they are working on a patch.
We decide to do the install.
WORST DECISION EVER
The system is completely unstable. It crashes repeatedly. We lose transaction data. It's not possible to 'downgrade' without completely knocking the server offline for baseline rebuild from ghost. The amount of data on the server would take about 8 hours to rebuild, and the server is being accessed about 18-20 hours a day by different shifts. We finally orchestrate a weekend rebuild about 7 days later, and then spend about 30 hours taking the data out of the new version and putting it into the old version.
I might add, when we called Intuit to tell them about our issues, here was their response: "Well, there's nothing we can tell you. It's a known issue. You'll have to downgrade to v5. We know the uninstall funcationality is broken, you'll need to restore from a previous backup. No, there's no ETA for when the patch is coming out."
To make matters worse, the version 6 update was a crock anyways. We've since installed it with the latest patch and it 'works', but it's slow as molasses, buggy as hell, and still notoriously unstable. If the Accounting department didn't habitually use two of its new features, I'd push us back to QB2005 just to stop having the issues.
To the darkened skies once more, and ever onward.
And I'll say this:
Piracy is not always a bad thing. I used to pirate 3d software back in high school before I could afford it (or justify to my parents that I needed it) Now, some years later, I happily own Maya and Modo. If I had never pirated other 3d apps in the past, I probably never would have gotten into 3d and eventually shelled out the money to Alias and Luxology (and various book publishers, training DVD resources, texture collection CDs, etc) Does it make the fact that I originally did pirate software? No, but like I said, it worked out in the long run. (Keep in mind, this was before there were readily available personal learning editions, and when 3D was even more expensive than it is now)
Winamp 5 includes 2 modes, one for old computers with low ammounts of memory, and the other for people who want the nice stuff. During install choose the classic gui and it should be alot faster for you.
What does one cow say to the other? Moo.
Why do you accept that? That'd be my question!
If I got a car that failed to start 5 mornings of the year, I'd be pretty pissed off. If I got a TV that wasn't compatible with channel three or seven, I'd be rather annoyed. If my car's doors unlocked randomly on the third Tuesday of the month, I'd be frustrated. If I got a VCR that couldn't timer record at 58 minutes after the hour, I'd be pretty pissed off too!
So why do we let our computer programs have these problems? Why do programs need to be killed or the computer restarted at random (freezing during startup isn't as common since win2k). Why do we accept a computer program that doesn't seem to handle the formats established at the time with ease (think Windows XP destroying exif information on jpgs)? Why do we accept holes in our software that lets crooks in along with their bots, spyware, and adware? Why Why why?!?
I've always been a fan of a certain car maker (and still am), but when I got a 2004 sport sedan and a few weeks later had my dash light up brighter than a christmas tree, the dealer tells me that their software/flash upgrade to the car should fix the issue. Sure enough, it did, and by the forums, it wasn't an uncommon problem with early production of the model. But this is stemming into other areas. An audio system I got in 2000 couldn't read any CD-Rs- obviously it didn't spend enough time in testing, as this should have come up. My car had bugs! The EPROM on my new furnace needed to be replaced. This is getting silly!
Sure bugs are bound to get through, but it is the programmer's responsibility to properly test their program. I'd rather Windows 2000 be released today and have it stable as anything and a solid performer. This isn't how things work. Microsoft spends more time making Spider Solitaire for Vista then they do testing the OS itself.
Updates aren't always better. Sometimes they add functionality, like additional CD Recorder capability, updated roster information for a sports game, security fixes, etc. Other times they add bloat and problems.
Anyone remember ICQ 99b and the 98's? Memory footprint of about 1MB, fast as anything, fixed the data corruption issues of previous versions. Good upgrade. Then recall late 99 versions and 2000+, where the memory footprint was about 80MB, the thing took a good minute to start up. It was buggy, and an ad-city. Then they wonder why it died a horrible death to the favour of MSN? Pack hundreds of features in there and make it slow as anything and nobody will go back. Wait... That's a good description of Windows.
-M
when you see the word 'Linux', drink!