Scalix 11.3 is on the way ...

[TCWardrobe]

Hi Florian,

Michael,

sorry, can't really agree here.... couple points:

I would be rather disappointed if you'd agree so easily

- There are major releases, minor releases and patch releases. The first couple releases this year have been patch release (11.0.1 through 11.0.4) and we absolutely made sure that everyone who was on 11.0.0 could use these as we really wanted to get rid of the most common issues that appeared with our customers.

Nice to know... I never used anything between 10 and 11.2 but it sounds like it was a very buggy era... but from my point, there are very nasty issues left, not many but enough to get my head aching, I'll name them later.

Maybe you also should be aware that we ourselves are probably one of our most demanding customers, as we use our product - one of the advantages of being an email company - we really and truely eat our own dogfood.

And there I don't get how those issues could slip through... but it may just be that Murphy does like me more than you

I wasn't happy with the quality of the initial Scali x 11 releases myself, however, since 11.0.4, I've personally had very little serious trouble and I jump between all possible kinds of clients all the time. For me, 11.2 and 11.3, and now 11.3 have truely been what they've been named for, releases that add functionality and capability or change platform support; some bugs were fixed as well, but would I be ok to use 11.0.4 today - surely I would. We originally wanted to make the license change at the 11.1 point, then postponed it for two further minor releases - I think that's a fair game.

The only "fair" thing would be to announce such critical license changes way before they happen. Don't you agree that this will have a major impact on how scalix is used right now? Good or bad for your company does not matter here, I just say it will have a big impact. And speaking of "have truely been what they've been named for", I would have hoped you'd rather called it stable and usuable here and not what you named it, but as I see the issues with 11.2... again, we'll come to that later

There are some annoying issues left, not lack of features but bugs, I would not call scalix to be as stable as theyself say it is but that's a marketing issue, no real problem for me there.

Can you get more specific here? Again, as I said, I'm pretty happy with what 11.2 provided to me. There are some issues left, in particular in specific markets such as Japan, but by in large there are very few big points left, according to my tracking.

NOW we'll come to the point we were all waiting for!! *flourish*
As said, it is not much. In general scalix performs well but some things I don't get why they are still there, even if at least one was called to be fixed in 11.2 (more or less official statement "will be fixed in the next release" dated may 07). The thing with "unread mail count". In general if userA does something, moments later userB notices that... GREAT! But it seems this just works if userB has this place of action (a folder, calendar, whatever) in the focus? The same thing just local for a users mailbox. As the forums indicate this has to do with the rules which move mails out of inbox (too fast?) and the message count lacks way behind this action. This is the most annoying thing I can't imagine how this could happen if you folks at scalix used your own stuff internally... in other words, I would NEVER call anything stable with such an issue, this IS critical! Network wide or just even local work in Outlook must be reliable when it comes to this. If changes in the network would take a few seconds till they reflect at ones mailbox, fine, things need time. Same on the clients inbox itself. If it would take a few moments till the unread message count refreshes, fine, things need time. But the wrong count is there for ages, just restarting outlook solves that issue for the moment. I give a damn about features, just basic operations are needed here.
The other thing is the connector crash... how long is that known? Not critical at all, technically, but how does it look like? Sorry, I just have installed the connector a several times and got almost ever the crash. How often did you folks install this thing? Again, it is not critical, but I can't understand how THAT could slip through too and of course I would have released a fix as soon as possible.
The last thing I can remember right now (users are sometimes just... argh!) google Desktop. Ok, it works for some, maybe most people but most definitely you (scalix) are aware of folks where it does not work. If installed Outlook crashes fairly often, mail indexing does not work at all, tried different versions and of course the exact same versions where it worked for others. Here I have the impression it depends on at what time google desktop was installed, before or after scalix deployment. I did not have the opportunity to look into this any further yet, so this is just a assumption right now. But again, very official statement from scalix "certified to work with scalix", how often did you folks install this in what different environments and what versions? What means "certified" there? Normally I'd say something is certified to work exactly in a well defined environment, like exact version, exact service pack, exact patch level, exact anything, there it is certified... but I did not find ANY hints on what "certified" means in this case for you folks. Seeing mailboxes around several gigabytes in size, users just LOVED google desktop search, they got used to it, it made life a lot easier... but I can't get it to run reliable and properly and I can't promise the users they will be able to use it some time soon. But said this, I am VERY happy nevertheless that at least this thing works at the most important mailbox here. Not that the mailbox is that important, but the user behind it is somehow... oh, let's net get too much in detail here
But again, maybe not as mission critical as some users say it is to them, it is fairly nasty and I've seen several folks having the same issue and are lost too... is it so hard for you folks to reproduce this issue?
Another thing, fairly critical, the activesync thing. It may work more or less, there are issues left... but despite that officially it is... oh, let me paste it:

Code: Select all

Certified ecosystem components.
[...]
Wireless Email: Cradle Synch
Blackberry Desktop Manager
ActiveSync (for Windows Mobile)
Chapura PocketMirror for Palm OS

Again, certified... but deep in the release notes more or less "nahh... works somehow, some issues left", ok, a marketing thing maybe, I changed my mind, I hate marketing!

But changing the license model which WILL have a real big impact for the users between minor releases without further notice or time to react. I know, you may say no one is forced to jump to release 11.3 but otherwise they will stick with some very annoying issues forever without jumping to newer releases and for sites using more than 10 users without paying money (and most annoying annual fees).

- Again, please name the issues that you believe we should fix for free for all users.

Done

- There is no such thing as "annual fees" for any Scalix version, ever - all our licenses are perpetual, the only thing you pay for is software subscription, i.e. continued access to all new minor and major releases going forward, i.e. instead of buying future versions of the software you pay 20% per year of upgrade insurace. As we've released at least 3 versions per year since Scalix started, this will give you another 15 new versions of the product until you've paid the original license price again; fair deal, IMHO.

You like to be fair, aren't you? The only fair thing is basic operations, working as stable as you folks say it is. At least I would suspect things that are "certified" DO work and not with issues or unreliably. And of course "You get fixes if you pay for it" is everything but fair. Don't get me wrong, I understand that a commercial company has its priorities at other areas we'd love to see their priorities. I would not mind if the free edition would get the patches a way later... but I don't want to argue about the words we both use, in general it means, I get patches and new versions and support for money. It is like it is, a subscription fee... this may make sense with virus scanners and other things which needs to be updated at a regular basis but not with normal software regardless if it is DTP, messaging, office suite, whatever. True, other software has issues and bugs too, but imho not issues and bugs that obvious like those mentioned above. It is ok that the solution for those issues take time and may not be so easily fixed in any way but to force users to pay money to get those solutions (hard-spoken) is everything but ok. And of course I get patches and fixes years after I did purchase the software without paying anything. It is as simple: bugfix for free, feature enhancement > new release if it's minor it may be free, if it's a major feature or lot's of, new version, one need to pay for an update or a new version. You may say Scalix CE does not cost anything, but with just 10 users it is almost useless for the current user base. You should know how real OpenSource software works... am I limited with the use of apache or bind or anything else? I know scalix needs to make money one way or the other and that is absolutely ok, but why not like all other companies? I am not in the position to argue with you or anyone else at scalix about the way they make business, but I am in the position to say I don't like it.
Oh, almost forgotten... I don't mind how many and how often you release new versions a year, I really don't mind! New versions introduce new bugs, I have no problem with using old software, I still have some Debian woody boxes around, they work as a charm. Also the almost 5 year old WIndowsXP or 8 year old Windows 2000, they still work great! Sure they have issues and bugs, but nothing critical, nothing that could be a showstopper in any way or at least nothing anyone realised yet to be critical. Or formerly I used to work with Exchange 5.5 or 2000... all basic operations did work without a hassle, there were issues but also workarounds. And that is the one thing I can't understand. Software is complex and so spoken it is almost impossible to code perfect bug-free software, but knowing this the extinction of bugs should be the top priority, then lots of QA and testing and after all that new features. I know QA and testing does not make money... BUT, that's the most obvious benefit of something free and open scalix COULD be. The community as a very wide scattered testing and QA staff! I don't want to see them as guinea pigs, but for sure 1 million users find more bugs as 1 thousand. It may be hard to find a balance between a commercial and an opensource company there but the step you are about to do is not a wise one. See MySQL AB for example. I am not deeply involved there, but this seems to me the almost perfect balance between both worlds. Again, I am in no position, it is just my opinion. I know Scalix and MySQL AB may not be comparable when it comes to the user count... but don't you want to be there one day? How should that work without showing the world how great your software is (or can be, see the issues )? If you drop the 25 to 10 free premium users, I bet scalix will lose a great deal of interested "testers".... uhm... where was I again?

I know a commercial company like Scalix needs to pay attention to their shareholder value but changing the license model to rise the income (sorry, but it is nothing more than this) at this time is very disappointing!

Sorry, but that's simply not true; we don't believe we'll be raising our income based on the change, at least not in a very relevant way; it would take a lot of commercial 20-user customers to keep a company our size afloat - it's certainly our overall customer base, including some larger enterprise accounts that make the difference here, as is the case for most software companies.

Ok, provide me with the financial reports of scalix and some numbers and give me deep insights in all financial related stuff at scalix to prove you're right I am not a market-researcher but I think my assumption there was correct, the future will tell you, the books will tell you, the market researching you did before you announced this step told you already. To be honest... I have no Idea if you will raise the business volume or if it will drop, but I think the overall user count will drop most definitely. And an extrapolation theory:
We have 13 users, would be happy with the CE version up to now, short of money sort of... with the new step we'd be forced to a.) spend money b.) live with issues c.) drop scalix sooner or later. "b" is not a long-term option, "a" and "c" may be, but I don't like one of them anyway.
Either option... the one will raise your income the other will drop your overall user count. I think there are some more companies like them and these options may be very common to those too.

We've simply seen a good number of cases where the balance between free and commercial, especially when it comes to additional services like support, etc., doesn't play out as well; actually, especially in small business space this is more true for our reseller partners than ourselves - and all of them we've talked to have blessed the change so far. Again, the theory is that a company or other kind of commercial organisation running an email system for 20 users or so requiring groupware functionality is a serious operation with serious availability and support requirements and should be treated as such - this is much easier with a commercial license in place.

Can't agree there. If one needs SLAs, guaranteed service and response times, sell it to him. What has that to do with how many free premium users are there? YOU have the possibility to sell anything you want. 5-10-15-20-25-30... premium users packages WITH (or without) support. Maybe 1-2-3-4-5-6 would not make that much sense but... if one demands it, sell it. Plenty of room for market opportunities and resellers. I'd say it would make sense to limit the general "free" options somewhere. I don't know if 25 free premium users is a good limit, but 10 is definitely not, too many medium sized business that won't fit there anymore! I'd also say a limit of features (not basic ones! But the iCal thing maybe and other comparable things) for the free edition would make sense in the one or the other way, this is common practice. It is easy to understand to pay for more features but not so easy to understand to pay for bugfixes, very uncommon.

I firmly believe that, with the new offering, we simply provide the world's best and strongest free email and groupware solution; given the new features, we'll be expanding that role and there are many new things to come for commercial and free users - always keep in mind that we provide pretty rich functionality even for standard users, and they are gaining capabilities with every version we publish.

How can you speak of free groupware if that is exactly what you are about to trim a great deal? Sorry, I don't get you there, not even a bit. This is marketing speaking! Folks beware!!! Serious, how can you believe to be the leader of something free if it is trimmed that way so it is almost unusable for most situations?
You do have an eye-candy webmail client, agreed, but even old-fashioned SquirrelMail does not have this performance issue with laaaaarge mailbox folders. SWA may have real nice features and may be the beauty of all webmailers but the most obvious thing is utterly broken or at least more than just suboptimal designed. And again, features are nice, but stable base functionality is mandatory!

In short words, is it really necessary to change this in between minor releases? Doesn't this sound like a change one should apply in a major release?

In short words, yes - this also has to do with our overall release strategy; I do agree that Scalix 11 was too disruptive, partially because we put too much change in various areas in a single release. Snce then, our strategy has changed and minor releases are our main vehicle to bring careful, incremental change to our userbase, in a system of continous improvement, so the next major one will most likely be out no earlier than a year from now.

Agreed, many mucho grande new features is almost ever a bad idea without major testing and QA taking place. But things need to go on, so basically it does not matter what releases you release and what version numbers they have, But don't you think you should be at least very careful what you call stable? I do remember the late scalix 10, early scalix 11 times... so many had used early 11.something releases and where shocked about the issues they had... those days I had the impression (compared to the old linux kernel development) you released a development software as stable (using kernel 2.3 or 2.5 in production is really fun!). In my opinion version numbers are nothing. See, version 11... but still basic issues with for example the unread mail count? Or is this a new feature introduced in the 11.something release? I am used to use real stable software, so bare with me if I sound a bit unfair but talking about version numbers is crap. If you intend to release less often but more stable... GREATLY appreciated but I must say at least "stable" is imho mandatory all the time with such "mission critical" software.

Given the change in 11.3 and subsequently 11.4, we could have made it a major release, by the way, but decided against it, for many reasons - as explained above.

I didn't get the explanations... but that may be because I see the things a bit different... as described above but again, history told, version numbers are in general irrelevant and overrated.

Again, very ready to discuss the case you're making in more detail - but by in large I think we're very much on the right track and again, I can assure you that the reasons for moving in that direction are quite different from what you suspect and name......

IMHO this is not the right track... not when it comes to raising the userbase and after all to take some market share... but you said above "incremental change", let's see in what direction it goes. As Scalix in general is a wonderful product, both directions are possible even if I have my personal thoughts about the direction it will go.

Cheers, looking forward,
Florian.

Sure you will? As I think your speech is full of marketing patters which I hate... could get quite annoying methinks

As a side note, I almost ever used "you" but I think you will know when I mean you as Florian, a man his work and community presence I appreciate a lot, or Scalix as the whole company, the release team or whatever department fits.

greetings
Michael

[jaime.pinto]

Please let's put a stop to all of this!
Scalix position regarding the new price structure is final, at least for now. Florian and other scalixters, there is no need to repeat yourselves. Everybody else, there is no point on bitching around.
Bottom line: the only thing that really matters is the 11.3 release *OF THE PRODUCT* and when it's out, hopefully bug free. What it doesn't have yet or feature enhancements is a subject for 11.4, in another thread, and for the moment I don't really care. What it does have and weather or not it works properly is a subject for the day it's released onwards. Until them let's everybody please just be quiet and return to our 11.2 routine!
Thanks

[da]

Based on the announcements in this forum, it looks like community edition users with 11-25 users will have two options if we remain with Scalix:

(1) remain with 11.2, with no fixes for yet-to-be-discovered bugs or even known bugs (e.g. smartcache problems) and without support for Outlook 2007 (which is now the only version of Outlook available at retail outlets or installed on new PCs)

(2) move to a commercial version of Scalix version 11.3.

Please correct me if this is wrong. It is a significant change in policy.

To enable us to make an informed decision, and include appropriate expenses in company budgets, I would appreciate it Scalix would provide details of the new pricing structure, or at least some indicative costs as soon as possible.

[TCWardrobe]

Please let's put a stop to all of this!
Scalix position regarding the new price structure is final, at least for now. Florian and other scalixters, there is no need to repeat yourselves. Everybody else, there is no point on bitching around.
Bottom line: the only thing that really matters is the 11.3 release *OF THE PRODUCT* and when it's out, hopefully bug free. What it doesn't have yet or feature enhancements is a subject for 11.4, in another thread, and for the moment I don't really care. What it does have and weather or not it works properly is a subject for the day it's released onwards. Until them let's everybody please just be quiet and return to our 11.2 routine!
Thanks

I do not agree, not at all! Nothing is final until it is released and I read the release notes. And even after the release, nothing is final, everything can change. And after all Florian was prepared to get some "noise" around this. But I'd not say my words are noise, they are comments, thoughts, suggestions and of course a bit of a worry here and there. Haven't you seen the sign on the front door? "Yes, we're open!" but closed for discussing a really serious topic for at least some of us? You must be kidding. And most obviously, you know what "feedback" means? No one is repeating anything (in general), we're discussing and for heaven's sake, giving feedback!

greetings
Michael

[TCWardrobe]

Oh, just realised, this "unread mail count" issue, it is rather a "overall object count". I just moved several thousand mails from one folder to another, besides that Outlook went a bit crazy (I don't complain there, for me just cosmetics) it still showed thousand something mails in there, not only unread mails but all. As before, just restarting Outlook helped. But there I saw for the glimpse of a moment the count still at thousand something but it dropped completely just a second later so it seems to be really just a refresh issue... isn't that all push technology? Or does the connector pull in intervals those informations? Or after all, does it just count them for itself?

But... wrong topic I assume, but as I did raise this issue here and Florian asked for it and of course as the edit button does not work... hum... you get it.

greets
Michael

[florian]

Please let's put a stop to all of this!
Scalix position regarding the new price structure is final, at least for now. Florian and other scalixters, there is no need to repeat yourselves. Everybody else, there is no point on bitching around. ....

I do not agree, not at all! Nothing is final until it is released and I read the release notes. And even after the release, nothing is final, everything can change. And after all Florian was prepared to get some "noise" around this. ...

Jaime, I'll agree with Michael in this case. We have taken a decision, we believe we have had good reasons doing so and up until this point (and presumably beyond!), we stand up behind it. It's not one that we've taken lightheartedly, and it's absolutely one that raised a number of concerns, also internally and also for me myself - therefore, I appreciate all feedback and we're allocating enough of my and other's time to read, understand, comment and discuss things - that's why I've opted to post this to the feedback forum instead of the announcements, which does not allow for comments.

So thanks for the input and I'll be back with more,
Florian.

[florian]

Based on the announcements in this forum, it looks like community edition users with 11-25 users will have two options if we remain with Scalix:
(1) remain with 11.2, with no fixes for yet-to-be-discovered bugs or even known bugs (e.g. smartcache problems) and without support for Outlook 2007 (which is now the only version of Outlook available at retail outlets or installed on new PCs)
(2) move to a commercial version of Scalix version 11.3.

As a fact, this is correct.

Please correct me if this is wrong. It is a significant change in policy.

Now, this statement let's me think a bit. Technically and legally this is correct as well, any change in licensing is also a change in policy.

However, for me, policy is about more than just parameters, prices and implementation, it's a lot about intention.

And the intention of Community Edition hasn't changed, not at all. When it was created, as I explained above, Community Edition would have Community Users (equal to todays Standard Users) in unlimited amount, plus 5 Enterprise Users (as a trial pack). So the intention of Community Edition was to be a free integrated email and calendaring system with a state-of-the-art web user interface, integrated administration and a scaleable and mature store, directory and routing engine behind it. The thinking about Community Edition was all about what Community Users (i.e. home users, students, people running a mail server for their private initiatives, friends, clubs, non-profits, etc.) would need to have something that is a bit more than imapd, postfix and SquirrelMail.

Actually, Outlook Support and some of the groupware functionality enabled for Premium users on SWA were not on the radar at all, because the organisations and people we thought about were typically thought to be the ones that wouldn't want to spend any money on Outlook and Office licenses, Windows Operating Systems and similar things - after all, all this stuff is a lot more expensive than the actual mail server mailbox license of any commercial offering. We thought of people using new, open stuff to do their email - and personal calendaring - the fact that open standards for the latter - namely CalDAV - took way longer to become available than we originally anticipated was an unfortunate fact of the matter.

Now, with this intention in mind, we didn't go down the route of other companies that allow you to use limited versions of their software for "non-commercial" use - we placed no restrictions on the use of the software at all - actually mainly because we didn't want to go into any discussion if hobbyists clubs or similar things would constitute a commercial operation, therefore not qualify - it was just simpler that way and we assumed that if a commercial organisation was happy with the Community-level functionality, they would certainly not bother to consider Exchange-type solutions and alternatives.

The 5 Enterprise users were originally put in to "demo" the capabilities of the commercial product, so if any of what we ever did was driven by marketing, this was it. With the number limited to 5, the intention was quite clear and I hope this makes sense. Now, with this definition of "Community" (and I think it is a very balanced and fair one), a 15 person company (i.e: commercial organisation making money from their business) would not really be considered a community user. As I said, however, we did not have any commercial product in place to cover those and there was some interest - and again, they were willing to pay for it. The product just didn't fit.

So we decided to give it away to them as well at the time, and in hindsight, the way we did it was a little bit of a mistake. We simply increased the number of Enterprise/Premium users in Community Edition. What we should have done is that we should have given away a free 25 user license key for the commercial product to those people, to make it clear. The reason why we didn't do it that way was that these were the Scalix 9 days, and there were no license keys back then, just different downloads - and adding a third one to the mix would have been pretty cumbersome, hard to maintain and difficult to explain. So we actually expanded the coverage of Community Edition by adding users to it, however, watering down it's purpose, definition and intentions.

With the introduction of SBE we assumed that things would clear up automatically, given that this 15 user company described above would possibly want some of the add-on features, but certainly want support coverage and services not available in the same way on top of CE. This, however, didn't happen - instead there was a lot of confusion on when to go commercial, what to use, etc., although the best path for commercial organisations of that size seemed very obvious and clear to me from the beginning - if this was my organisation, I would always pay $995 once and $249 every year (price of SBE-50), to get all the goodness of having a commercially supported product, no troubles and insurance. In fact, I've run a couple small companies in the past, and that's what I've always done - as long as running email servers is not my core business, I don't want to bother with the details, just get going. And the price, with all respect, is pretty cheap for the functionality provided, and very, very competitive with every other offer in the market.

The ones that suffered from this the most actually were our partners and resellers, because not seldom would such 15-user companies actually connect with one of them, let them talk through the product, even help and support with the setup, until then, in the final stage basically telling that partner that they would go with the free version, because they could now maintain it themselves - this put a number of our partners in difficult situations where they suddenly had to consider charging for presales activities, etc., to compensate - not really a good scenario, I believe that if you're a commercial IT reseller, you should be able to establish a trust relationship with your clients, help them through the initial discussion and then at some point, make some well-deserved services and also product licensing revenue, based on their margins. So these folks have asked us to change the CE definition (some suggested really adding the "non-commercial" use to the terms) for a long time. A working reseller channel is crucial for us, because it ensures maximum customer satisfaction, local service and support, good localization (important for a global product, we have customers in 50+ countries), etc.

So I think if there is any change in policy for CE, it is to refocus on the original intention of the product, which is a good one, and which we'll drive forward by also adding stuff to it. Everything that's based on open standards should and will be added to it in the future, everything that enhances the web user experience. Everything that provides for better access to the systems data. And certainly all and everything that can be used as an API for own devellopment - and that again is another reason why CalDAV was very clearly to be included - so will future additional WebDAV based access methods.

As soon as 11.3 is out, I'll hopefully be able to start talking about 11.4 (coming next spring) and I think this will more clearly show up.

So I hope this explains what is - and was - our policy and why I don't believe it has really changed.

To enable us to make an informed decision, and include appropriate expenses in company budgets, I would appreciate it Scalix would provide details of the new pricing structure, or at least some indicative costs as soon as possible.

The amended price list is already available through our resellers and by contacting Scalix sales. I don't want to go into pricing details on this forum. By in large, pricing for SBE-50 and EE is unchanged. SBE-20 is a new product designed for an even lower entry point (around $/€ 600 I believe) for those small commercial customers described above. Additional SBE users now also come in packs of 10, not only 25. We've expanded our range of support offerings. We are now offering solutions for Hosting Providers. And we continue to offer commercial per-incident support (sold in packs of 5) for users of all editions.

We've started community/evaluation support for CentOS4/5 for 11.3, so that those Fedora users don't have to complain about having to upgrade their operating system with every Scalix upgrade. We'll fully support this for commercial use in the next version. We're looking into server support for debian Etch and Ubuntu. And we'll continue to provide good feedback here on this forum and through other channels, to commercial customers and community users alike. And we'll continue to provide new features and functionality and bug fixes for all editions as well as to help fueling the drive to remove the need to use expensive commercial clients such as Outlook as fundamentally it is my belief that of all things, eMail client software should be free, as should be browsers and other such components, even possibly operating systems, unless they come bundled with some specific hardware, such as my much liked trusted and highly reliable MacBookPro.

The last paragraph also summarizes what really matters to me about this product, and because our management fully supports this strategy, that is the very reason why I'm still with this company. And for those who now call me a marketing buff, be warned - my major is physics!

Hope this helps,
Florian.

[florian]

And with this being said, now to my friend Michael (and don't get me wrong on this, please) !

Nice to know... I never used anything between 10 and 11.2 but it sounds like it was a very buggy era... but from my point, there are very nasty issues left, not many but enough to get my head aching, I'll name them later.

Indeed it was. And I'm not proud of it, although I didn't have the same job as I have now w/rgds to the product, so I could point fingers to others, yet I don't. I wasn't what it should have been and I'm working very, very, very, very hard to never let that happen again.

And there I don't get how those issues could slip through... but it may just be that Murphy does like me more than you

Well, I can tell you why - email is one of the most personal applications of all - everyone uses it in a different way, everyone just uses 10% of the functionality, and given that we're a pretty technical company our dogfooding is valuable, but incomplete. We have to do hard QA as well and we've ramped up on this quite a bit over the last year. And it shows and it helped.

The only "fair" thing would be to announce such critical license changes way before they happen. Don't you agree that this will have a major impact on how scalix is used right now?

For some it will, for some it won't. I think for the ones that I like to protect the most, the ones that I call community, it most likely won't and that's the feedback I've received so far, as well in personal communication. It's pretty clear that the ones that will be impacted the most will be commercial companies (who else really needs >10 Outlook clients or deep groupware functionality for meeting planning, etc. - for all the non-profits that are impacted, please contact me personally, make your case and I'll see if I can do something for you) with anywhere between 11 and 25 users. Now for those, we today start asking them to pay $/€ 650 for SBE-20 (incl 5st year support, if you order this year or early next, you'll get 5 extra users in addition) and then $/€ 200/250 (10 per user) per year for upgrade protection. that's between $/€ 13 and $/€ 27 per user per year over a 5 year period for a system that is as capable as Exchange. I think that while it's more than nothing, it is a very reasonable amount to ask for of a corporate customer.

As said, it is not much. In general scalix performs well but some things I don't get why they are still there, even if at least one was called to be fixed in 11.2 (more or less official statement "will be fixed in the next release" dated may 07).

Yup, some things slipped over the sommer, we did lose some time during the acquisition phase.

The thing with "unread mail count". In general if userA does something, moments later userB notices that... GREAT! But it seems this just works if userB has this place of action (a folder, calendar, whatever) in the focus? ...

Yes, I'm aware of some remaining issues there. They don't affect me personally because I don't use rules, I keep everything in my Inbox basically, archiving stuff manually once/twice per year. I just don't believe in too many folders, they confuse me. And search works so well now that I find everything that I need very quickly. I understand that this is a bad spot. However, the underlying technical reasons are deeply architectural and we've decided to move them out a couple times as they are also considered somewhat complex and risky. I'll make a promise now, understanding that it's a relatively weak one: You won't see any improvements in 11.3 on this, but you will see them next year. And I mean it because I agree that this is one of the types of user experience that really sucks.

The other thing is the connector crash... how long is that known? Not critical at all, technically, but how does it look like? Sorry, I just have installed the connector a several times and got almost ever the crash. How often did you folks install this thing? Again, it is not critical, but I can't understand how THAT could slip through too and of course I would have released a fix as soon as possible.

Are you talking about the one at the end of install? We don't understand it. I know that (a) it is guaranteed to happen if you start setup.exe from within a ZIP folder. I know that (b) is is guaranteed to not happen if you use the .msi-based installer over the .exe based one. For the things in between I don't know anything about the cause. It does not happen on our QA systems and I haven't gotten a hand on a test system that reliably reproduces it for me. If you get me a VM image that I can hand to engineering so that they can reproduce - and then fix it - I'll get you a Scalix T-Shirt.

The last thing I can remember right now (users are sometimes just... argh!) google Desktop. Ok, it works for some, maybe most people but most definitely you (scalix) are aware of folks where it does not work. If installed Outlook crashes fairly often, mail indexing does not work at all, tried different versions and of course the exact same versions where it worked for others.

Another bad one. Outlook plugins. It's extremely hard. You have some 3rd-party closed-source code, calling another bit of closed-source-code (Outook), calling into our code. Next to impossible to debug. We've made some progress with Plugins, but we are not there yet We've made progress on this with both 11.1 and 11.2, so it should work better than before, but so.... this will continue for as long as we have the Outlook connector, it's next to impossible to resolve all combinations of weird (and partially buggy, even when from google) 3rd party components. The only way to get rid of that issue completely for us is to get rid of the connector, and if you've read the Xandros-Microsoft announcement, that's one of the things we're planning to look into for late next year.

But again, maybe not as mission critical as some users say it is to them, it is fairly nasty and I've seen several folks having the same issue and are lost too... is it so hard for you folks to reproduce this issue?

Yes it is, because it depends on the data you put through it. Again, very bad one.

Another thing, fairly critical, the activesync thing. It may work more or less, there are issues left... but despite that officially it is... oh, let me paste it:

Certified means that we'll work on resolving known issues and we test it regularly. We've improved active sync support considerably for 11.2, and if you're using OL2K3/XP/WM5or6/AS4.5, you should be ok - what's left that doesn't work there with this combination? There are 2 open bugs in the area I believe, but they are not totally major.

Again, certified... but deep in the release notes more or less "nahh... works somehow, some issues left", ok, a marketing thing maybe, I changed my mind, I hate marketing!

Certified means - we have tested, it generally works, there might be issues, if issues exist, they should not be major or critical, and we know about them and work to remove them.

The only fair thing is basic operations, working as stable as you folks say it is. At least I would suspect things that are "certified" DO work and not with issues or unreliably. And of course "You get fixes if you pay for it" is everything but fair. Don't get me wrong, I understand that a commercial company has its priorities at other areas we'd love to see their priorities. I would not mind if the free edition would get the patches a way later... but I don't want to argue about the words we both use, in general it means, I get patches and new versions and support for money.

That's nonsense, sorry! You don't get patches because you pay, you pay for the use of the software and updated versions of the software because of who you are. If you are a commercial shop with more than a certain number of employees and pay Microsoft a ton of money to get all those Outlook licenses and beyond, so you already spend quite a bit on IT stuff, you should pay us for the use and care of our software. That is really fair.

Small guys, home users, community folks typically don't need the Outlook part and some of the other stuff. They should use the software for free. And they should certainly get the same quality of the same software at the same time. Especially with email where some bugfixes have security impact, I am fundamentally opposed to hold back software from those who do not pay; this has been proposed internally in the past and I've always opposed it and will continue to do so - I really want the free version to become available same day as any commercial product. I want money from commercial users, but I don't want the folks that I really consider community to use old stuff, with more issues. That wouldn't be good.

It is like it is, a subscription fee... this may make sense with virus scanners and other things which needs to be updated at a regular basis but not with normal software regardless if it is DTP, messaging, office suite, whatever.

No, it isn't. Subscription, by definition, is nothing perpetual - if you end your subscription, you lose a service. This is not the case here, because you can continue to use your same version. Some other companies call this software assurance, others upgrade protection. Some don't offer this at all, requiring you to pay fresh for each new version. We've decided for this model, because we want to keep people up to date, we believe that email systems because of common threads should be kept up to date and we make minor upgrades really easy - a Scalix 11.1 to Scalix 11.2 upgrade takes an average of 15 minutes downtime, even on a large server. Unlike Microsoft, where every upgrade is a major migration, Scalix is - and has always been - very upgrade friendly. The one from 10 to 11 was a bit more intensive, but given what we've done under the hood, works still fairly well, also in larger organisations. Doing the Exchange 2003 to 2007 upgrade is a much bigger pain, promised. You need 2 servers to start with.

True, other software has issues and bugs too, but imho not issues and bugs that obvious like those mentioned above.

Well, Microsoft released a service pack once that made Lotus Notes stop and they've broken their own clients with some of their server upgrades. I don't think Scalix has many issues that really bring you system down.

I'm taking the above serious, but as I said before, it all depends on perspective and it's my job to get priorities right - sometimes I hit, sometimes I miss, that's life. If there is a particular bug that you want to escalate, please ping me with the reference and I'll see if it makes sense, given what's on the table. That's, btw., another advantage of commercial customers - quite naturally, we prioritize bugs reported by them a slight bit higher, and I think again, that's fair!

It is ok that the solution for those issues take time and may not be so easily fixed in any way but to force users to pay money to get those solutions (hard-spoken) is everything but ok. And of course I get patches and fixes years after I did purchase the software without paying anything.

Who do you compare us with? the fact that we opted to combine patches with new features and as a compensation make upgrades really easy is something that can be held against us, of course. However, you'll confirm that it's pretty common with any software company that you get bugfixes basically for the most current version of the software, unless they are so critical that old versions must be backpatched. That usually happens for security issues, less so for functionality that doesn't make the system unusable. The things that you mentioned above would, IMHO, fall into a major category, but not in the critical one that I would consider for backpatching.

It is as simple: bugfix for free, feature enhancement > new release if it's minor it may be free, if it's a major feature or lot's of, new version, one need to pay for an update or a new version.

Oracle does not give you patches for free; Cisco doesn't without a yearly support contract. Microsoft requires you to buy there software at a high price (think Office), then they give you updates for the lifetime of the software for free. I don't know any commercial software company that gives you free patches for free software, all the time. For the users that fit our community edition definition (see the above), we actually do it.

You may say Scalix CE does not cost anything, but with just 10 users it is almost useless for the current user base.

According to a survey within the CE user base (that is the ones that left us with their email address on download, which are presumably the larger ones) last year with about 600 responses, 90% of them are well convered with 10 users or less. So this statement is simply false and not based on any data, just speculation - or am I missing something.

You should know how real OpenSource software works... am I limited with the use of apache or bind or anything else? I know scalix needs to make money one way or the other and that is absolutely ok, but why not like all other companies? I am not in the position to argue with you or anyone else at scalix about the way they make business, but I am in the position to say I don't like it.

And I certainly accept that statement. However, there are various kinds of software and the stuff that we're doing doesn't exist yet as full open source. All the projects to create an open source Outlook connector have failed so far. The reason is simple. Ours has about 150 man years of development time into it and while I don't say that this is the amount you really need if your budget is tight, it's way beyond the scope of a small open source project. And large open source projects such as Linux itself or Apache get created around stuff that is open, interesting, fun and appealing to a large number of people (on the developer side, not so much on the user side). And providing a connector for the worlds's most proprietary (and probably most hated) eMail client is simply not such a thing.

Commercial software companies, on the other hand, do what their customers want and what they pay for. And they pay their (poor MAPI) developers to stand the pain!

Oh, almost forgotten... I don't mind how many and how often you release new versions a year, I really don't mind! New versions introduce new bugs, I have no problem with using old software, I still have some Debian woody boxes around, they work as a charm. Also the almost 5 year old WIndowsXP or 8 year old Windows 2000, they still work great! Sure they have issues and bugs, but nothing critical, nothing that could be a showstopper in any way or at least nothing anyone realised yet to be critical. Or formerly I used to work with Exchange 5.5 or 2000... all basic operations did work without a hassle, there were issues but also workarounds.

Well, one of the fundamental issues with software is that it tends to get more feature-rich and larger and more complex over time, people want and take that functionality for granted, and with size come bugs. Scalix - as it was still OpenMail - was pretty much without any such issues in the days when Exchange 5.5 was current, however both fail to satisfy modern users needs, otherwise there would be no business case for upgrades. In email systems, which are connected to the outside world, keeping the system up to date is almost mandatory because of additional threats and challenges coming in the form of Spam, Malware and other things. Next to the need to support new client platforms. The organisations I know that are still using Exchange 5.5 today are not held back because they don't want/need new functionality - the bigger issue is that they are afraid to move because of Microsofts link of this upgrades to other upgrades (think: Active Directory, etc.) at the same time - which is bad. It's keeping old, polluting technology in place because there is no gas station that serves the new and improved stuff that you would need to get it going. We try to be different that way. Conservative and mature is good, IMHO, old is not.

And that is the one thing I can't understand. Software is complex and so spoken it is almost impossible to code perfect bug-free software, but knowing this the extinction of bugs should be the top priority, then lots of QA and testing and after all that new features. I know QA and testing does not make money... BUT, that's the most obvious benefit of something free and open scalix COULD be. The community as a very wide scattered testing and QA staff! I don't want to see them as guinea pigs, but for sure 1 million users find more bugs as 1 thousand. It may be hard to find a balance between a commercial and an opensource company there but the step you are about to do is not a wise one.

I disagree and do that based on experience. We did think the same way and that's why we started giving out pre-releases of Scalix 11 well before the actual GA release. We had 5-digit download numbers, but only about 2 hands full of bugs were reported, and none of them we wouldn't have found ourselves. The problem is that email is mission critical and you can't test an email system without putting it in production. And noone would put a alpha or beta release in production in a large, challenging environment. Therefore, the best feedback we got from our internal QA - though it needed improvement as well, from our internal dogfooding - we used it in production very early - and from the early adopters, right after release, although the pain there was too large. For that reason, we've decided to release production versions more frequently, with less change per release, but then more focused QA (on the area of change) and tight feedback loops with customers and support. This has worked out really well as most people here would be willing to confirm I hope.

See MySQL AB for example. I am not deeply involved there, but this seems to me the almost perfect balance between both worlds. Again, I am in no position, it is just my opinion. I know Scalix and MySQL AB may not be comparable when it comes to the user count... but don't you want to be there one day? How should that work without showing the world how great your software is (or can be, see the issues )? If you drop the 25 to 10 free premium users, I bet scalix will lose a great deal of interested "testers".... uhm... where was I again?

Different business model. Which works for development products, but not for an end user application product. MySQL makes money from Support and Services, which is very expensive. The reason why it still works is that when they have a support issue, it is mostly from one of their OEM customers that embed MySQL software into their own solutions, so the person affected by the support issue or requiring training is usually a highly-paid developer and the person's employer is willing to pay any price to unblock him. eMail users are the you's and me's in every organisation, "ordinary" users, and very few organisations are willing to pay similar levels of support.

Also, MySQL has an OEM model, so their direct customers don't pay them out of their own pockets, but those customers customers pay MySQL indirectly through the runtime license fee. Again, MySQL selects the ones that pay not based on what they get, but on who they are (commercial companies producing commercial software and their customers/users), so it works for them, however, it wouldn't work in the same way for us and the likes.
[
quote]I know a commercial company like Scalix needs to pay attention to their shareholder value but changing the license model to rise the income (sorry, but it is nothing more than this) at this time is very disappointing![/quote]

First of all, Xandros/Scalix is a private venture at this point, so there is no large pool of shareholders. There are investors interested in a good ROI but that's based on building a great software company, not really on license revenue with Small customers - we'll all be happy if we IPO one nice day and for that, the change doesn't make a difference. The main driver, to say that again, is to recreate balance between commercial and free users, our resellers and partners and make sure that everyone gets their appropriate share. And I still firmly believe this is contributing to that very goal.

Ok, provide me with the financial reports of scalix and some numbers and give me deep insights in all financial related stuff at scalix to prove you're right I am not a market-researcher but I think my assumption there was correct, the future will tell you, the books will tell you, the market researching you did before you announced this step told you already. To be honest... I have no Idea if you will raise the business volume or if it will drop, but I think the overall user count will drop most definitely.

Given our overall growth in user count, I don't think so. It might make the user headcount grow slower, but even that I don't believe and I've seen proof to the contrary. As a privately held company, Scalix/Xandros does not provide financial reports to the public, so that's one thing I can't really help you with.

We have 13 users, would be happy with the CE version up to now, short of money sort of... with the new step we'd be forced to a.) spend money b.) live with issues c.) drop scalix sooner or later. "b" is not a long-term option, "a" and "c" may be, but I don't like one of them anyway.

As calculated before - 13 users is $/€22,50 per user, per year over 5 years, all updates included. That's less than your office license costs you for the same timeframe, so I hope (a) is a viable option for a 13 usre commercial organisation.

I don't know if 25 free premium users is a good limit, but 10 is definitely not, too many medium sized business that won't fit there anymore!

Average business sizes vary with market. In Austria, e.g., about 90% of all companies are 10 employees or less, so our Austrian partners actually asked to lower this further. I have received feedback from all over the place that the number chosen is a good compromise.

How can you speak of free groupware if that is exactly what you are about to trim a great deal? Sorry, I don't get you there, not even a bit. This is marketing speaking! Folks beware!!! Serious, how can you believe to be the leader of something free if it is trimmed that way so it is almost unusable for most situations?

Again, I think we fundamentally disagree on what most situations are, what situations constitute typical community use and what is commercial use for which I think money can be paid. I find it simply difficult to accept that you're much easier paying Microsoft money for an office suite for which there is a perfect free replacement in the form of open office, while you say that for the same organisation, a mission critical server application must come for free.....

You do have an eye-candy webmail client, agreed, but even old-fashioned SquirrelMail does not have this performance issue with laaaaarge mailbox folders. SWA may have real nice features and may be the beauty of all webmailers but the most obvious thing is utterly broken or at least more than just suboptimal designed. And again, features are nice, but stable base functionality is mandatory!

Yes, and the one thing I currenty push hardest for the SWA team is performance and stability. You will see some improvements in 11.3, and there will be more in the next one, plus some nice functionality enhancements as well, That's the second promise here and today. It's one of my super-hot focus areas. And noone who has ever used SWA will want to go back to SquirrelMail, ever!

... those days I had the impression (compared to the old linux kernel development) you released a development software as stable (using kernel 2.3 or 2.5 in production is really fun!). In my opinion version numbers are nothing. See, version 11... but still basic issues with for example the unread mail count? Or is this a new feature introduced in the 11.something release? I am used to use real stable software, so bare with me if I sound a bit unfair but talking about version numbers is crap. If you intend to release less often but more stable... GREATLY appreciated but I must say at least "stable" is imho mandatory all the time with such "mission critical" software.

Either you're contradicting yourself (as you said above you didn't use any version between 10 and 11.2) or you're repeating yourself. All appreciated. Most of our major customers are on 11.2 now and they are happy, by in large. I wasn't happy with what 11.0 turned out to be, since then I'm primarily responsible for this in Program/Product management, and since I've taken over, we've made good progress. My further promise is that I plan to go on like this. I don't want to go into all the details why 11 wasn't as good as we expected, and the story would be longer than all that i've typed so far today, but we've analyzed and understood some of the mistakes that were made in great detail and we've adjusted our processes to improve. Being German.... , Mercedes cars had a longer period of really bad quality in the late nineties - such phases occur and happen, but they have recovered. Fortunately, in software, everything happens but also recovers a bit faster than when building cars. I think we're on a good trajectory when it comes to quality now.

For the release number, I don't like marketing numbers, for me major releases indicate major change. Scalix 11 has a much extended architecture, converting a mail/calendar server into a collaboration platform through foundation services like search, header caching, web service APIs and others, this was enough to flip the switch. Add to that the full internationalized architecture, opening markets in asia, etc., and you have your major. With the fixes and additions we provide inside the 11.x series (lately CalDAV), we still deliver on that promise and it's an ongoing story.

12 will be, again, a different product. It will get rid of the X.400 legacy, possibly of the Outlook connector, will have a different way with integrating with directories and also a different way of storing data. This is a new promise and step, and another good, non-marketing reason to flip the numbers. However, until then, there is still work to do to fulfill the larger and ongoing Scalix 11 promise and that's what we're working on, every single day. Changing a single parameter in our licensing structure does not change the product in such a fundamental way that it should be linked to a major release number change. That would really give this topic a lot more attention than it deserves and distract from the true and fundamental changes that should occur at such time.

I didn't get the explanations... but that may be because I see the things a bit different... as described above but again, history told, version numbers are in general irrelevant and overrated.

Well, at this point I assign them for this product and I'm trying to give them good meaning.

As Scalix in general is a wonderful product, both directions are possible even if I have my personal thoughts about the direction it will go.

Without knowing - or wanting to know - the details, I'll do my best to prove you wrong!

Cheers from Frankfurt,
Florian.

As a side note, I almost ever used "you" but I think you will know when I mean you as Florian, a man his work and community presence I appreciate a lot, or Scalix as the whole company, the release team or whatever department fits.

I might have used I and We interchangeably in my writing above and I'm not going to read through it again to catch any mismatch to reality; I have a life outside Scalix, yet I do identify with the company I work for, it's more than a job for money. In that sense, some of the opinions stated above do not necessarily represent official positions of my employer, but by in large I think it represents what the Scalix team thinks and feels about our product.

[Valerion]

Yes, and the one thing I currenty push hardest for the SWA team is performance and stability. You will see some improvements in 11.3, and there will be more in the next one, plus some nice functionality enhancements as well, That's the second promise here and today. It's one of my super-hot focus areas. And noone who has ever used SWA will want to go back to SquirrelMail, ever!

Actually, the fastest way for me to find out if my email address or phone works is to kill the Squirrelmail config file on my server

Yes, SWA is a lot better, but currently in some situations unusably slow, if you have slow links and serve it out over HTTPS. My own users do prefer SWA, but 5-10 mins of loading it every time will drive you away from it quickly.

Hopefully the new release will be a step in that direction. We will have to wait and see, but from my last phone conversation with Florian I am confident. Mm ... maybe if I can use some kind of trick to get the SSL connector to tunnel through SSH we will be able to run Outlook for those users as well now. We will see when it ships

[florian]

Hopefully the new release will be a step in that direction. We will have to wait and see, but from my last phone conversation with Florian I am confident. Mm ... maybe if I can use some kind of trick to get the SSL connector to tunnel through SSH we will be able to run Outlook for those users as well now. We will see when it ships

Why would you need the SSH tunnel? Can't you open additional secure ports? the one that you'll be needing will be 5767/TCP, which is the offical UAL-SECURE port number once registered by HP, but then never used.

On Webmail and slow loading, are you already on 11.2.0.58 (patch available for registered customers and partners through support), or are you using 11.2.0.52 - ping me through eMail if you need help.

Florian.

[a.schild]

The other thing is the connector crash... how long is that known? Not critical at all, technically, but how does it look like? Sorry, I just have installed the connector a several times and got almost ever the crash. How often did you folks install this thing? Again, it is not critical, but I can't understand how THAT could slip through too and of course I would have released a fix as soon as possible.

Are you talking about the one at the end of install? We don't understand it. I know that (a) it is guaranteed to happen if you start setup.exe from within a ZIP folder. I know that (b) is is guaranteed to not happen if you use the .msi-based installer over the .exe based one. For the things in between I don't know anything about the cause. It does not happen on our QA systems and I haven't gotten a hand on a test system that reliably reproduces it for me. If you get me a VM image that I can hand to engineering so that they can reproduce - and then fix it - I'll get you a Scalix T-Shirt.

I will have to tripple-check, but I think I have a VMWare image where the installer "reliably" crashes.

[florian]

Well, I'll extend the T-Shirt offer to you, then!

Florian

[grahamk]

Wow, what a great discussion. Only a handful of contributers not associated with Scalix, so I thought I'd lend another opinion.

I have been using scalix since 11.0.0. Prior to using Scalix, I was using SME Server. SME sucked, but it was better than paying for anything from MS imho. The company I work for has 30 staff in the office, and all servers are running linux services. (Samba, Freeradius, MySQL, Apache, etc).

I was so excited to find Scalix, 25 free premium users (a complete groupware solution, that was essentially free). And while testing 5 users on it... it worked!! Sure, it was buggy, but lets keep our expectations in check, what idiot would install .0.0 version and expect anything different!

So I would say that a lot of people on the forums, are people who use a lot of Open Source software, and love free stuff. From an ICT Manager point of view, free is easy to get past the board, but with a lack of understanding (TCO) people dont see that sometimes free is not "free".

So, Scalix moves from 25 users to 10 users, and as an argument against the move, we cry "but 10 users is just not enough!!!". Want to use Outlook for 11+? IMAP, POP, SMTP (its FREE). Want to put in a business solution that gives you the functionality of Exchange, then pay $20 - $40 USD per user for the privilege. I am in Australia, and to put Windows + Office on a workstation, costs around $600AU, and that is perspective in a nutshell.

Rather than focusing on a single implications of the term Open Source, Free, I tend to focus on what I believe Open Source is really about, Open Standards. True business value, from where I sit, and (mostly) delivered by Scalix.

I feel like I am starting to rant, so I will finish up. Open Source software, developed by people without a commercial focus, has a hefty TCO, often long install and configuration time. I'm no expert, but as I mentioned all of our services are running on linux, and I have tasted a lot of Open Source software.

The worst thing from my point of view, is that 25 Premium Users to 10 Premium Users does change the sell i've been working on with my boss. Luckily for me, i've been talking about buying scalix for long enough, and when I found out the news that Scalix was changing to 10 Premium users, I just told my boss that we have run out of Premium users (ie, over 25), and now is the time to buy. He agreed. Done.

If you are onselling Scalix, to your customers, or your boss, Scalix has moved the goalpost, and it may be painful. Lets see how good your customers are, and how much they trust "YOU". It is very obvious to me that scalix is a commercial company, and I think thats great, because chances are, they are not driven by their own interest level in the project, instead they need the money to eat, etc. We might get really lucky and Scalix might still be around in 6 months

To sum up, thanks for providing a linux based solution to groupware. There is a huge lacking for solutions like it in this space.

[Valerion]

Why would you need the SSH tunnel? Can't you open additional secure ports? the one that you'll be needing will be 5767/TCP, which is the offical UAL-SECURE port number once registered by HP, but then never used.

Well, opening another port on the firewall is easy, however we are in a situation where most of our employees are permanently based at a customer's site, or travels a lot. Due to network security issues where they work, connections are not allowed to non-approved ports, they have to be specifically requested, approved, etc., which takes time and effort. The only ports we can guarantee will be open are 80/443 and 22. In that case doing a port translation or port mirror makes things doable without all the politics involved.

On Webmail and slow loading, are you already on 11.2.0.58 (patch available for registered customers and partners through support), or are you using 11.2.0.52 - ping me through eMail if you need help.

Will do, Florian, thanks.

[konman]

Florian - you asked for opinions - so here is mine.

First - Friggin fantastic product - set it up with limited Linux experience - more than happy to pay for it now (got about 14 users).

Things that would make it absolutely rock are -:

Integrated Anti-Spam
Integrated Anti-Virus
Backup

The support and guides in the forums are first rate to set the above up - but it is nerve racking for newbies.

Cheers,
KonMan.