Hello - I'm afraid I've joined this forum just to complain

I've used Linux & KDE for about a decade now - ever since I escaped from windoze land. I've loved (and occasionally hated) both. One of the best features of KDE for me has been kmail - a simple, light & fast mail client that had all the features I wanted (well, since it was made possible to delete attachments), and none of the bloat & **** that typically comes with software products. I could receive, read, file, compose & send mail with no trouble at all; it was completely solid, and it even had a useable address book. I wanted nothing more.

Then a weekly Fedora update killed my Kmail. It installed a new version - KDE 4.4, I think - and Kmail worked no more. It's been made dependent on something called akonadi (I've no idea what that does) and akonadi in turn depends on something called Nepomuk. I've little idea what Nepomuk does, either, although I visit the town regularly. Something to do with social networking, apparently. Nepomuk in turn works with mysql . . . . . and mysql refuses to talk to my mail account. So - no more email.

I've spent half a lifetime with SQL servers, so I should be able to sort it out. But I can't - I don't want to. I don't need SQL on my machine, and I don't want it. All these servers may do wonderful things for millions of people, I don't know. For me, they just kill my email and I'm **** off about it. So, with great regret I've left Kmail. I've chosen another client, Thunderbird, which is big & not beautiful, and it uses ugly old mbox file formats. Converting my decade++ of emails, carefully structured into dozens of directories, has taken me half a week. I'm by no means sure I've preserved them all intact, and I've got a mail client I never wanted: but I won't be going back.

I wish software developers would learn - after all, it's a mature profession now - that because something may be feasible, that doesn't mean it's desirable. I can't count the number of times I've heard someone say 'I can wrap that in another layer', or words to that effect. Removing **** can be far more beneficial (it takes more skill, too). There may be benefits to some users, some of the time, but you have to count the costs as well. Tying Kmail up in all this clobber has transformed a competent, reliable tool into a useless lump. It's probably done a lot of damage to KDE as well - it begins to look like windoze itself!

This is a bit of a rant, I'm afraid, but I'm angry & I'm sad. I've loved working with KDE, but it seems to be going somewhere I don't want to go - and from comments I've seen on the web, it seems I'm not the only one not wanting to go there. I'm sorry to complain after so much good work has been done, and I hope that the sun shines on KDE. I'll be taking less of an interest, though.

I also am a long time user of KDE. One who has welcomed and embraced the changes KDE4 brought us, clumsy introduction notwithstanding. I was just preparing to update my wife's laptop from KDE4.3 to KDE4.4 to receive all of the latest and greatest KDE enhancements/advancements - until I read this post and started looking a little deeper into the kmail changes and usability. Unlike you alanjpraha, I don't have much of a problem with the use of Akonadi -> Nepomuk -> MySQL (or whatever) --- as long as they're largely transparent, don't hinder performance or create confusion for the end user.

Incoherent is the word that now comes to mind regarding the kmail/kaddressbook pair. I can't make sense of it and I KNOW my wife won't be able to. So, I'm faced with the prospect of upgrading and then having to explain something I don't understand to my wife or not upgrade at all. Of course, there's a third option - the one you chose.

This is most discouraging. The list of once Superior, now diminishing, apps is growing ever larger. Four apps come immediately to mind - konqueror, konsole, amarok and now kmail/kaddressbook. Each of these apps - once towering symbols of quality and usability - now seem to somewhat lack direction and seem to be driven largely, not by community need, but by technical ideology on behalf of the key developers. This is evidenced in many threads whereby the users' voices are routinely ignored or (more disturbingly and with greater frequency) are criticized and mocked. Couple quick examples. Konsole took from KDE4.0 until KDE4.4 to re-introduce a VERY popular feature lost from KDE3.5. Countless hours of debate over it's merits between the users and the devs. In the end, the devs relented and the feature is back. Or, more accurately, a new developer decided to get it done. But why did it take so long? Why the heavy resistance? In the Amarok case, I requested a simple feature restoration, also lost from Amorok 1.x series, only to be rather bluntly told that there was no chance in hell to have the rather simple and highly useful feature restored. That if I wanted it - write it myself. Konqueror now languishes in limbo regarding it's rendering engine as the users endlessly debate the devs. And now I can't comprehend kaddressbook/kmail.

Of course, some technically proud, well intentioned developer or user will quickly come to my aid in explaining the merits and usage of the overly complicated new kmail/kaddressbood design, but why should they need to. I've been working with computers longer than many of them have been alive. Intuition is key. Users need to be able to intuitively master the tool or THEY WON'T USE IT! I don't know where this idea is getting lost.

Although I'm not likely to abandon kmail, I DO HEAR YOUR FRUSTRATION. Truth be told, our little rants here will sadly fall entirely on deaf ears. Trust me on that one.

I would really like to know about the incoherency that you have found, because I have not found any.

When I click on Addressbook from kmail tools (for updating, editing, etc.), I'm presented with a totally different A/B choice than from kmail send To/CC: selection. In fact, the To/CC: selection doesn't even list that (std.vcf) A/B. Thus, any updates I make via tools -> A/B are apparently unavailable to me for sending. Kaddressbook itself is fairly coherent - until you get into kaddressbook properties dialog vis-a-vis Akonadi. You're options there are Filename, Access Rights, Monitoring with no real overall context. The average user has NO idea what Akonadi is and how it's different than std.vcf, etc. Clumsy at best. I'm sure there are really good explanations for my fumbling. Just do this... THen do that... And I can figure it out (as I will) eventually. But my point about intuition still stands.

When I click on Addressbook from kmail tools (for updating, editing, etc.), I'm presented with a totally different A/B choice than from kmail send To/CC: selection.

That has always been so. The selection you see in the To/CC section is taken from Recent Addresses, not from the main addressbook. IIRC that selection is stored in kmailrc.

I have to agree with the point about intuitiveness. I'm a power user and will do a bit of work to learn how to use a good tool, but I am the exception. If you want a wide user-base it MUST be accepted that most end-users expect to be able to use a piece of software they've never seen before with a reasonable degree of competence. If they can't they'll simply never use it again. My wife is an example.

annew wrote:

When I click on Addressbook from kmail tools (for updating, editing, etc.), I'm presented with a totally different A/B choice than from kmail send To/CC: selection.

That has always been so. The selection you see in the To/CC section is taken from Recent Addresses, not from the main addressbook. IIRC that selection is stored in kmailrc.

Actually that hasn't always been so. In prior versions, the To/CC selections gave the user a choice of a/b (or filters on the default a/b?) from which to choose. Every main a/b entry, plus the Recent Addresses were all available to the sender (as they should obviously be). There was a definite (albeit messy) relationship between the two. The new version appears to provide ONLY recent addresses. I certainly can't find any of the new entries I just added. Bottom line, there seems to be no coherent or intuitive link between addresses in KAddressbook and those available to kmail.

Look, we can debate all day long about our preferences and our kmail likes or dislikes (I personally like the tool), but when it clearly starts to let users down at these very basic levels, somebody needs to speak out. Also, I don't think this discussion is really so much about this particular bug as it is about the bigger question - How does something this blatant happen in the first place? I have my theories on that, but that's for another thread.

Then a weekly Fedora update killed my Kmail. It installed a new version - KDE 4.4, I think - and Kmail worked no more.

so an upgrade broke your application due to missing dependencies, and in stead of solving those dependencies, you chose to move to another mail client in stead?

The new version appears to provide ONLY recent addresses

new in which version?

i checked this with kde 4.4, and it lets me choose from a variety of resources: several available address books, recent addresses and several categories..

Actually, I mis-stated. Autocompletion always used the Recent Addresses list. The selection that you mention does indeed offer me my Personal Contacts addressbook, from which I can select any entry. The problem, therefore, is somewhere in your setup, so it's time to try to identify what's missing or mis-configured.

<<< so an upgrade broke your application due to missing dependencies, and in stead of solving those dependencies, you chose to move to another mail client in stead? >>>

You bet - gone and left it! I'm sure it hasn't escaped your notice that <all> non-trivial software is buggy, defective and potentially dangerous. We tolerate the inconvenience because software can do useful and magical things. But, take away the utility & magic, and you're left with a packet of bugs & vulnerabilities. I simply don't want stuff running on my system that I don't need, and know nothing about. Akonadi & Nepomuk may be the saviours of mankind for all I know - may solve the resource problems, bring peace to this earth . . . I just don't know that. What I do know is . . . I don't need them to do my email.

This shouldn't be a big thing: there's a simple approach, well-known when I was a BAL programmer nearly half a century ago. If you've discovered the greatest piece of software ever, put it forward by all means BUT make it a) modular and b) optional. That gives your miracle a fair chance without **** off half your users. After all, you can run an address book without SQL; you can even do email without an address book, if you have to.

Where I live we know a thing or two about being told what's best for us

Incidentally, I didn't last long with Thunderbird. It's ugly. Opera on the other hand seems to have the most intelligent attitude to email that I've seen so far, and its GUI would challenge even the once-simple & -elegant Kmail. Time will tell.

I too am really, really disappointed with the new KAddressBook.

Like the other posters, I am long term user of KDE and have suffered from the rocky upgrade path of KDE 3.5 -> KDE 4.3. And I thought that everyone had learned their lesson - KDE really does have users who rely on it for their day-to-day use.

All I can say, staying within the bounds if what is acceptable in this forum, is that I am very, very disappointed. This is a major upgrade, not a 4.3 - 4.4 upgrade. We - your userbase - should have the choice of whether we accept this upgrade. Like the other poster, I upgrade my KDE to the latest version whenever I can, just to winkle out the bugs and inconsistencies in KDE4. But now without warning my distribution lists and contact groups have gone missing, and I can't see any of my contact information in KAddressBook.

And I am sure that there some reason for it, and I am equally sure that it hasn't gone missing. But the point is that I will have to work it out and that will distract from what I actually want to be doing.

Yes, I know that you all do it for free, and yes I know that I am coming across as a whinging Aussie/Pom, but I am really, really disappointed off.

We - your userbase - should have the choice of whether we accept this upgrade

AFAIK you have that choice: upgrading from kde3 to kde4 isn't something your computer silently does.

But now without warning my distribution lists and contact groups have gone missing, and I can't see any of my contact information in KAddressBook.

you can use kde3 and kde4 side by side. As a consequence of that, kde4 addressbook doesn't use the resources of kde3 addressbook by default.

This is actually something that you asked for in your reply: you should have the choice which application you want to use

What's missing is a decent import tool dat allows you to move the kde3 resources to kde4.
But you can do it manually, by adding the resources to your kde4 version of kaddressbook.

i suggest you start a new topic about that if you don't know how it is done.

Seems to me that the defensiveness about complaints misses the point of the complaints--is there a reasonable issue? Where the complaint can be addressed satisfactorily, that's the end of it.

Otherwise, along with the supportive posts, perhaps the people who decided to change things could explain why. There's an old maxim--if it ain't broke, don't fix it.

Not everyone got KDE 4 by explicit choice--mine is on Karmic (Kubuntu 9.10) and one doesn't have an immediate choice. I can just imagine what a mess it would create with Plasma and all the other bling that's been added to try to retrogress installations like that.

I can't use HTML to send a message, apparently because someone decided I shouldn't be allowed to. Or have I missed the switch somewhere in the arcane panoply of widgets? This means that my last best route to send links to friends that don't wrap has been trashed. Or other irritating things one can't turn off, like all the pretty day-headers in the mailboxes that require you to click on them, then click on the message for that day. If you want the 'feature', fine, but the whole point of LINUX for me is that it is configurable, and this is rapidly destroying that, unless I drop KDE as my desktop, not an attractive approach, either.

If anyone has any suggestions about the two minor issues that I've raised, I'd be most happy to hear them!

Thanks,
Joe

Joe, I'm going to argue that you DID have a choice... if a distro is dumb enough to put something in the repos that isn't ready for prime-time it's clearly time to go shopping for a new distro. The current stable version of Debian is still using KDE 3.5; I only started using KDE 4.x a couple of months ago.