Parallels' Bad Idea

As a chronic fan of the Parallels VM product I'm reluctant to criticize them for much, but I think this latest round of beta releases is a bad idea. The introduction of new razzle dazzle features into a universe of struggling users trying to get 1970 working right also introduced a new round of "it doesn't work!" that needs to be addressed.

Instead of solidifying the released product and thereby pleasing a large group of users we have only more chaos. Instead of talking about networking issues and USB isochronous video solutions we're talking about coherence. We have eye candy instead of functionality. We have bootcamp horror stories instead of Linux integration.

Stop the madness! Fork the project. Disassociate future beta releases from the existing released product. Don't leave people thinking that a beta release is an upgrade. Clarify the role of the beta product - state expectations you have of beta testers. Qualify them! Not everyone should be a beta tester. Take beta discussions to another forum - manage this project better.

 

If you read the beta relase note first

You will see "not recommended" for a critical environment like an office

Hugh W

 

I don't know what you're seeing, but 1970 works perfectly fine for me. The problem with these forums is that only people with problems post. There's a lot more of us that have a stable setup with 1970.

 

I don't have any problems with Parallels and never have. That's why I'm a chronic fan of it. I'm commenting on the hundreds of posts each week here from people who are not fortunate as you and I are, many of whom are wholly unqualified to be beta testers - it is becoming a mess. I think the press likes to call it 'quagmire'. You will agree I think, that 1970 is an incomplete work. And 3106 is not an effort to correct that so much as to correct problems introduces in beta 2.

 

My setup is also stable and I couldn't be happier - same for my wife's system - but this is not helpful to the project. Because I've set up my systems with care I can offer little useful information to the support team other than accolades. But Parallels seems to have thrown out a lot of new complexities in the recent betas to an unprepared userbase, unnecessarily.

The drumbeat of complaints and dissatisfaction is growing louder and the response is eyecandy (coherence) and features that have dire consequences for unqualified testers (Bootcamp integration, migrating existing installations, Vista/Microsoft padlocking). I think they are growing concerned about VMWare eating their lunch and are going for headlines offering features (broken as they are) instead of correcting the existing rough spots. You and I seem unaffected by much of these but they cannot ignore the trend of the posts here.

In hindsight it was probably a good move to remove the link to the forums from the front page as these pages are not good advertising.

 

I do think that it would make sense to separate the beta discussions and the release product discussions into two separate forums---and to make the beta forums an extra click away. The more popular Parallels gets, the more important it is that people running critical apps not be tempted to run the beta version. Nothing can sink a support effort more swiftly than a bunch of non-hacker-geeky-types (no offense intended) being frustrated by the inevitable beta bugs. There are enough issues in 1970 as it is!

 

Bad Idea...

I am in the realm of a semi-satisfied user. I have to agree with dkp about Parallels and this latest release of betas.

As a software developer from many years back (haven't done it in 15 years, moved to marketing and corporate management), I would say that the Parallels team needs a strong dose or release management discipline. Let me explain:

1. 1970 I agree was an incomplete release. (USB was the main shortfall IMO)
2. The betas purported to fix USB issues which they eventually did if not completely
3. The betas also introduced a new features previously unseen i.e. coherence, bootcamp partitions etc

I can understand their need to improve on features now that the VMWare Fusion beta is out (and it looks pretty solid for a beta). By way of an example VMWare USB support was much better than 1970! Yes the VMWare product does not have the eye candy of Parallels but the basics are right (USB support better, multicore virtualisation etc). If I were to characterise the two products I would say VMWare is substance over form and Parallels is form over substance. Having said this I have been a Parallels user since the beginning and have a paid up l/c since the beginning.

A word of advice to Parallels development team. You are not big enough to move on a lot of functionality improvements at the same time as getting the basic product solid. So develop a road map that is realistic wrt the resources you have and stick to it. Finish a feature properly then move on. Develop a beta system that restricts betas to people who are interested and competent to be beta testers and have the discipline to stick to that. VMWare will be a formable opponent and they certainly have the runs on the board as far as virtualisation goes. You don't do yourself justice by bringing out half baked betas (a release that doesn't work with FAT32 partitions for example!) and releasing them to all and sundry.

The development team deserves better as you have come a long way from the first beta release. The product as it stands now is at about 80-85% finished make sure you nail the remainder!

 

you're not being realistic. this is a beta program. if you're using this on a mission critical machine then you've ignored the clear, concise warnings regarding backing up and NOT using beta releases on important computers. why not stick to the last stable release?

don

 

Thing is, for some of us, these betas have been working flawlessly. I've had nothing but the best experience with all three of the latest betas, and in fact am using them for productive work. I'm keeping backups, but I'm still using it for real.

 

i think you should take a step back and take a look at the big picture. the original beta program allowed us (prospective customers) to beat up on parallels while the dev team worked to fix exposed bugs. we bled and the dev team went beyond just tossing bandaids at us...they shaped parallels into a very solid and stable 1970 release. we have over 200 clients set up with 1970 and they're loving it.

now repeat the cycle. we bleed again while parallels works to fix exposed bugs. eventually enough bugs will be released for a new stable version to be released. will it be perfect? when is ANY software perfect? will it be usable? absolutely.

the parallels dev team are directly responsible for the incredible press parallels has received. one look at macworld magazine, one visit to the various mac sites will contradict your view that parallels isn't all it can be.

don

 

I cannot for the life of me fathom where you are going with this. I have used every version of Parallels with extreme success from the beginning. I have access to a lot of machines to test it on and two systems that are "mission critical" - they run only production code. I've kept the processes separate and working very well. I'm well aware of the beta program and how it works and what can and should come of it. I have no problems to complain about - Parallels exceeds my needs and expectations. But that has nothing to do with what I introduced in this thread.

Read it again.

 

that's great, but let's not allow good luck to overshadow the real chance of failure when dealing with beta releases. with regard to 1970, you can count on seeing more posts from people having problems than you'll see from people not having problems. that's common in most beta forums.

don

 

your post is clear, you feel the forums should be split. that's up to the parallels developers. i take issue with your statements that parallels is misrepresenting their beta. this is simply not accurate. the description and warnings are very clear.

Don't leave people thinking that a beta release is an upgrade. Clarify the role of the beta product - state expectations you have of beta testers. Qualify them! Not everyone should be a beta tester.​

don

 

Bad Idea ....

Don,

What dkp is saying is, I think, that by making the betas universally available the fine print "do not use for prime time applications" gets lost in the noise.

It is one thing for a development juggernaut such as Microsoft to make a beta universally available it is quite another with a small developer like Parallels. It is very much a matter of scale and expectation setting. BTW when Microsoft releases to a public beta the code is pretty tight by that stage.

Oh BTW the VMWare Fusion beta is more functional in its first beta than the current Parallels product:

1. multi core support
2. 32 and 64 bit support
3. USB 2.0 and isochronous support

What is needed is the discipline and focus to get the basics right then move on to the eye candy!

IMHO.

 

Don - you are completely ignorant of (or are trying to appear to be) the draw for so many frustrated users to the sticky announcement on page one of this forum announcing Beta 3 (having previously spun the Beta 2 web). They are working hard at Parallels to overcome some nagging problems and right in the middle of it, even before anything else is settled, they churn out a series of betas that fix some problems (the bait) but also include a lot of new features including the very complex integration of Boot Camp. You don't have to be an Einstein to see that a lot of people have managed to trash their systems to the point of being unbootable in pursuit of USB 2.0 support and the lure of one Windows installation. The majority of people in this forum are looking for solutions - they see a new beta and swallow the pill. As I said I am not having any problems with Parallels and I never have. It's worked perfectly for me in every version and platform combination I've worked with. But I've been doing this stuff since the 1960's and wouldn't expect to have some of these noob problems.

My contention is it was not a good idea (see the thread title) to introduce to all these struggling frustrated customers so much new unsound software and little understood processes. Seriously - is this really the time to be handing Transporter to people who can't set the resolution of their screens? And with the additional frustration of the Vista padlock (MSFT sees your VM hardware emulation as a new platform - pick up the phone!), you got trouble right here in VM city. I remain amazed that there are as many success stories as they are because VM's are not obvious even when you know what you're doing. And if it isn't obvious that the typical visitor here is lost in space you're not paying attention.

So this is my editorial. My 15 minutes - Parallels is the most fantastic software I've enjoyed on a consumer class system in decades, and solves a lot of issues I have with desktop OS's. But they need to split the herd and put the gearheads in one group and the hapless registry hacking twist it till it breaks, manual ignoring klutzes in another, and provide channels of communication for peers rather than this maelstrom we have now. I don't see how that can be a bad idea.

And then they should fix the released produce in one fork (hell, call them service paks, not betas) and introduce the next version betas in a separate clearly identified fork and forum, and make certain the testers and users are getting what they need while avoiding what they really don't need.

 

fair enough, but i do disagree with the original poster's contention that parallels dev team isn't adequately warning users about the risks in using beta software. the sticky cleary states:

PLEASE REMEMBER: this is a beta software, and like any other beta software you should treat it very carefully. Do not install this Beta 3 into the production environment yet. Do backup your virtual machines before trying them with Beta 3​

don

 

I'll say this again. What Parallels needs to do is to make their latest final release and their new betas to exist on the same hard drive. That way people can do their work on their reliable final release and play around with the betas.

Also that way can release some of the panic. A lot of people use betas as their work. Which of course is asking for trouble.

 

That would be a good idea - getting more out of a platform is what they do, after all. A problem that can crop up is any kernel modules they load as loading order is important and conflicts are not impossible. We bump into this all the time in the data center with volume managers and clustered systems.

BTW, did you get my email a week or so ago about your morse code interface?

 

That isn't my message - but it's true - they have not taken the care needed to protect people from themselves. They can write all the caveats their legal department can craft and a large part of the frustrated audience is going to see 'blah blah blah enhanced USB 2.0 support blah blah blah Coherence...'. You know it, I know it, and everyone in this forum looking for solutions is downloading the beta for the next version with all it's new problems. Maybe even a simple notice: Hey! Listen to me!! Beta not equal service pak!!! Not the same at all!!!! Even then a lot of folks are going to grab the ring.

I recommend they isolate beta testers from production product users with problems, and address each group's problems in the proper context. I've also suggested they put out service paks vs betas of the next release. It's the right thing to do. And it was a mistake to go down the path they did. Damn, my 15 minutes are up.

 

What on earth ever happened to individual responsibility? Parallels does NOT have an obligation to protect people from themselves.

Beta software is, by its nature unreliable, and they said so. If you want to try it, you do so at your own risk, and they said that too. If betas don't work for you, you are NOT a victim of anyone but yourself if you don't have a backup to restore.

Betas are released to get feedback from people who are willing to test the software and report problems to the developers. Betas are NOT intended as service packs. There is no need to separate beta users from production users and support both because beta users shouldn't expect support at all. Beta users are contacted when the developers need something. Any help on betas is provided as a favor when the developers have a spare moment. That's just the way it works, even if some folks don't understand the industry or don't like it.

There has, so far, been no reason for Parallels to release a service pack so I have no idea where you are coming from on that one.

The bottom line is that they don't even have to warn you that beta software is unstable. Everyone in the software world knows that, and knows betas shouldn't be relied on and will break things.

If you buy an airplane and try to fly it without knowing how, and crash, I suppose you will want to blame the manufacturer for not sitting you down and convincing you that you need training to fly and stay alive.

This is just silly.