What does Zimbra have against OpenSUSE

[langs]

I think there are some very long bows being drawn here, model t fords and windows 98 are not good examples to be using with regards to support and EoL.

Yes everything has a use by date and gets EoL, it's the timing of it however that makes it critical. RHEL4 is still very very much a production platform in use by a great number of people and guess what it still will be next year and the year after that.

While I accept that Zimbra only have the resources to build, test, and support ZCS on a finite number of platforms, the selection of those platforms should be carefully considered and linked to who is paying your bills. If your resourses are so thin you shouldn't be axing support of dominant in use platforms for new ones and should mirror the life cycle's of your distro partners, ie 7 years for RHEL4.

Just because there is a new version of the distro doesn't automatically mean you go out and upgrade to it for the hell of it, not broke don't fix it springs to mind.

You'd do well to remember in the enterprise world RHEL has at about 50% market share and most of that is RHEL4. Their biggest worry btw comes from Sun and Suse, not a Ubuntu as it's not even in the same market space, most people using it are NOT in the production business. Seems to me your business model is focusing too much on FOSS and not enough on the paying side of the ledger.

Should we be worried that the ZiHoo model is now FOSS with adverts plastered though out like normal Yahoo media tart crap?

If we don't build Zimbra for this version of OS, is that keeping people away from our product?

Absolutely it will, and it will give you invalid stats to a degree. Example I use Solaris or Debian for every other box I have, but because 2 years ago when I wanted ZCO I had to choose a platform you supported so I've been using CentOS every since for it and our testing machines. People that don't have the ability to leverage other distro's may well just walk to a competitor instead over looking ZCO.

To say the community can just build from source so therefore it's not the deathbed of say OpenSUSE thats being either extremely naive or insulting, without Zimbra to drive it nothing will really get done and it will be ever more a niche userbase, ie look at the solaris builds there aren't alot of us using it or compiling.

Face facts part of the appeal and major selling points for Zimbra is it's ease of installation and upgrade (hello MacOSX users), your targetted market is not the hordes of unix boffins that feel comfortable with g++ etc. If I was Tgx I'd be dirty too as OpenSuse is every bit as popular with the masses as Ubuntu, and the OSX version can't be that stellar on take up either in comparision, unless you've hit that market hard.

I'd like to see Zimbra as a company engage far better with it's paying customer base about what they want. Ie out of the box AD support, extending Zimbra into document management and support tracking via 3rd party relationships at a company level, make life easier for your clients to achieve things that help them reduce costs thereby driving the ZCO ROI up. Everyone I talk to locally agree's these things are a no brainer.

[uxbod]

Just to throw one thing into the mix ... how many platforms does Exchange work on ? Does Exchange include document management ?

[langs]

uxbod what the hell does exchange have to do with it.. but to answer you yes it does have interoperability with document management and it has never tried to be an open platform system, but they haven't EoL'd it for 2000 server either and won't till 2010. MS also support exchange on 16 non EOL versions of windows (platforms) to ZCO's 10? versions of linux (platforms).

Can I also point out Microsoft don't need market share like ZCO does to survive and as such have to do less to entice customers. It's moot anyway as it does everything we're talking about.

Using the argument that your opposition doesn't do it so you shouldn't is just bloody stupid and ANY business man would agree. With that kind of attitude we'd never see innovation.

[uxbod]

Actually comparing to what a competitor does is useful IMHO ... With the number of *NIX variants out there, and the number increasing, I cannot see how Zimbra would be able to develop, QA and support all of them. Perhaps this is the reason why other companies adopt the same approach.

I'd like to see Zimbra as a company engage far better with it's paying customer base about what they want. Ie out of the box AD support, extending Zimbra into document management and support tracking via 3rd party relationships at a company level, make life easier for your clients to achieve things that help them reduce costs thereby driving the ZCO ROI up. Everyone I talk to locally agree's these things are a no brainer.

Perhaps if RFEs and support tickets were created (unless they already are and I have missed them) and people voted for them we would see a lot of the changes you have expressed you would like.

Perhaps this early post shows that Zimbra does listen to its customers :- Potential new Zimbra user.

[langs]

Perhaps this early post shows that Zimbra does listen to its customers

pity there are by far more examples of RFE's going unresolved to say otherwise, they may listen but we all know it's about the follow through that matters.
I'm not saying ZImbra are deliberately trying to upset customers or not listening, but I think they just don't ask. I doubt 80% of their NE customers use the RFE's and that is not their customers fault.

They have accounts executives and reps "apparenlty" but they never engage us to talk to us about concerns etc, the only time you hear from them is when you get;

I hope you are doing well and had a nice long weekend. I wanted to touch base with you to see if you need anything from me in regards to renewing your Zimbra license.

The number of *Nix variants hasn't really increased, unless you mean linux distro's and really who cares, production enviroment market share is all we care about. They need to support RHEL and SUSE or they'll go bust, having Ubuntu support gives them a Debian based distro through I still say it's not a production platform, Mac OSX is probably a smart play as it's a niche and they don't have to contend with exchange.

Anyway this thread is about EOL, and I am adamant support for RHEL4 should not be pulled, but mirror the life cycle of Redhat, I mean with LTS are they going to dump support for it before it's 5 years is up? End of the Day ZCO is a great great product, but I think the company itself could do more to engage it's customers about issues that could have drastic effects for them.

[Jbrabander]

uxbod, I think some of the grumbles with RFEs is that if you browse through them, there are a number of them that show the ticket was made back in the 3.x version and yet is still unresolved. I do agree with langs in that having tickets that are 2 to 3 years old does seem pretty bad...votes or not.

langs, from what I've read in these posts thus far, yes it sounds like 6.x will be the last version for RHEL4. And if I read pmweb correctly, giving about 1 year between major versions, Zimbra 7.0 will be due around mid 2010 or so. So RHEL4 should be supported until then if I understood the post from mmorse earlier. How does that affect fixes and such for the 6.x version afterwards? I don't know, I haven't been on Zimbra long enough to know how it was handled in the past. But yes, support for RHEL4 will probably end earlier than the "official" Red Hat life cycle for version 4. But I'd guess that by the late 2010-2011 range, you'll might be looking at getting off RHEL4 since the Red Hat support is coming to a close.

I do have to disagree that not doing something because your competitor doesn't kills innovation. I think you've got to ask yourself why aren't they (my competition) doing something. Is it a good idea that I could add and make my product distinctive? Or is it more trouble than it's worth and I'll spend a ton of resources producing it for little gain? Like so many things, it's a balancing act.

[Jbrabander]

Oh, and since Solaris was mentioned earlier in this thread:

http://www.zimbra.com/forums/general-questions/24459-solaris-version-available-download.html

[jholder]

Hi All,
Thought I'd jump in here to address a few issues.

1) There has been much discussion on this thread about how "Zimbra doesn't care about RFE's" or votes, or how we simply don't care period. I've even gotten a few PMs accusing me of being silent because "You know it's true".

This is a "Glass half full" argument. While pointing out how many rfes are unconfirmed, or targeting, people are forgetting the number of those that have been completed or targeted.

Once an enhancement is completed, it is ignored by the user-base at large because their need has been fulfilled. Remember, just because it hasn't been targeted doesn't mean we're ignoring it. It means we haven't decided how to include/implement it yet.

2) Zimbra has limited resources, and we will do what our paying customers want first, then votes second. This is why it's important to let support/or your account exec know what you want. We try to balance this with the need of our open source users...and sometimes people feel as if we don't pay enough attention..or pay too much attention to a group/enhancement. The fact that paying customers get priority with enhancements should be considered when looking at how we prioritize it. If an enh has 2 votes, but a large customer like comcast requests it, then that enh will make it sooner than a bug with 0 customers and 30 votes. Once again, this is why it's important to let us know that you're paying, and want a feature.

3) Release cycles are not only judged by when the OS vendor releases their release, but by what our components operate and user demand. Mac OS X Leopard is a great example of this. We simply cannot release NE as GA because of a bug with LDAP/Postfix. Sometimes there are OS-specific bugs that just won't work well with Zimbra. In other cases, we need to phase out a OS because we're going to introduce a feature that will only be available in the few version of the OS. Let's take Ubuntu for example. For a LONG time, they didn't have iSCSI support. This was an OS issue that impacted Zimbra. On the Mac side, it isn't the OS specifically that's causing issues, but the way we build or a component within Zimbra. These are mitigating factors when choosing OSes.

4_ Finally, user input is important. This thread title is a great example: "What does Zimbra have against OpenSUSE". Here's another question: "What does Zimbra have against Knoppix" or Mandriva, or Arch Linux, or BSD, or Sun? The answer? Nothing. Absolutely nothing. If download and install notify numbers aren't up to par then we just can't support it. We can't support every OS in the binary form, which is why we provide source. To say that we're ignoring user demand by saying "You can build it yourself", isn't quite fair. Many software vendors take that approach because it's impossible to support and build for every platform.


With that all said, I have specifially asked our PM team to reconsider OpenSUSE. This doesn't mean that they will, but it does mean that they will look at this thread, download/install numbers, and employee resources, and make the decision.

Whether they do change their mind or not, this thread is an example that we do care and do listen.

Yours faithfully,

john

[Jbrabander]

1)While pointing out how many rfes are unconfirmed, or targeting, people are forgetting the number of those that have been completed or targeted.

You're right. That is easy to forget. Personally, I was just trying to say that there are some old ones hanging on out there.

2) Zimbra has limited resources, and we will do what our paying customers want first, then votes second. This is why it's important to let support/or your account exec know what you want.

True, but to some degree I don't want to pester my account exec every time I put in a vote on bugzilla. (yes, we're a paying customer)

3) Release cycles are not only judged by when the OS vendor releases their release, but by what our components operate and user demand.

Well put. (although I still wish you supported Outlook 2K with a ZCO ) And it can be hard to remember that new software sometimes comes with new components that work great with newer OSs, but have to be "broken" to work with older OSs. At some point you've got to cut off the older OS to get the best function from newer software. Heck, didn't Microsoft just finally cut off Win98 licensing completely?

4) Finally, user input is important. This thread title is a great example: "What does Zimbra have against OpenSUSE". Here's another question: "What does Zimbra have against Knoppix" or Mandriva, or Arch Linux, or BSD, or Sun? The answer? Nothing. Absolutely nothing.

Again, well put. And I agree, if you try to support everything, then nothing truly gets done. The explanations are much appreciated though. It's nice to know why some things are done the way they are.

[langs]

Originally Posted by jholder

2) Zimbra has limited resources, and we will do what our paying customers want first, then votes second. This is why it's important to let support/or your account exec know what you want.

John, what happens to an RFE when you have alot of customers asking and 150+ votes? Why does that sit "Not Committed"?

I am yet to see a single RFE I have voted for in 2+ years implemented, I have submitted them as cases and alway get referred back to the RFE / Bug listing for it.

In that time I have had 3 accout reps, Jovi, Ty, and now Janie and not one of them have ever engaged us about what we want or followed up with issues. Not even a yearly customer survey. Reps get changed with no notice as well.

It seems unless your a comcast etc you don't count.

1)While pointing out how many rfes are unconfirmed, or targeting, people are forgetting the number of those that have been completed or targeted.

Couldn't agree more, but you know what when I front up to operations meetings every Tuesday with our exec board do they care that we've closed out 300 jobs? No they care about the 10 we didn't get closed out and the longer they sit unresolved the louder they get and the more heat I get.

Communication is key and the bulk of it from my point of view shouldn't be happening here on the forums but via the account execs you put in place, as a paying customer I shouldn't have to chase them or Zimbra. Grief follows the money, I pay you so the grief goes from me to you not the other way around.

Cutting out RHEL4 support without consultation is causing me grief, while I will migrate to CentOS 5 notice via the right channels would have been nice. You know what NE customers are using RHEL4 why didn't you get the execs to talk to them and warn them, give them a heads up so plans could be put in place a head of time.

It comes down to attitude for me, does Zimbra want to be my partner or just another vendor whom I have no loyality to?

Oh and just to harp on for one my second about CentOS why do we see posts like this from Zimbra employees just days ago;

Bleeding edge distributions are not the best choice for a stable mail server, why don't you migrate to something stable such as CentOS, RHEL or Ununtu LTS?

Why are they setting people up by recommending CentOS as their first choice for a stable distro, knowing full well it's not offically supported and that Zimbra painfully make that clear all the time. John you have posted very harshly about it before to make everyone aware of this fact.