-60

As many others I am looking forward to the release of KDE SC 4.5 which hopefully will contain the Akonadi-based version of Akregator, Kmail, KOrganizer. But at the same time I am worried that these release will have regressions which may be only discovered after release when running in non-testing environements or are considered as minor by developers. The experience with KDE4 as such and also Kaddressbook shows that regressions of any kind are a delicate matter.

Therefor I suggest that KDE SC 4.5 will be released with the new and the old KDE SC4.4 versions of this apps, which would allow users a smoother transition and/or profiting from KDE SC 4.5 features without being forced to switch the KDE-PIM apps immediately at the same time.

I do not know how much trouble (extra work) this would make, I hope not to much and would really appreciate if this would be possible. I believe other users will appreciate this too, that's why I post it here, so that other users can show there support and make suggestions.

While this could be done by distros, the whole idea is still valid. Let's see what other users think...

This should be done by distributions, KDE-PIM devs should focus on making it work with akondi so that it is not needed to ship multipple versions of each app in the first place.

i voted no
it is time to finally go for the full akonadi. Also there are date-converting mechanisms to make sure your data doesnt get lost.
Also this has been tested and used by alot of devs for quite some time so i trust the kde devs that the change to akonadi will be smooth and hasslefree

I know that in theory this would be a distro job, but in practive many distros released KDE 4.0, when it was not ready for general use. It would need at least a release not recommending this to distros. But there may be some things like naming or the data-transition not automatically kick in.

I also generally trust the devs in wanting a to do a good job, but this also a very complex thing and data transition e.g. from the old addressbook to the new one did not go as smoothly and there are still apps like Kopete having trouble with it. Also Nepomuk integration did not go smoothly. Even if no data gets lost, if there are problems, there may be problems to access them temporally. Then downgrading may be a hassle or even impossible.

As it is now, KDE is not a repository where you pick and choose among a lot of options - there is only one PIM.

So when distros use KDE, they use the KDE that is available. After all, you would expect that a software in version 4.4 is featurecomplete compared to 4.3 or 4.2? WRONG!!

The devs treat KDE like a playground with total disregard for people that rely on continuity.

It should be the responsibility of the devs to NOT let any application be included in any of the versions that did not as a minimum cover the functionality of the former version. Yes, develop it, announce it as an alpha - and keep it there until it is ready. Does that mean there would be a version of KDE with no upgrades of e.g. the PIM? No problem!

As it is now, if you upgrade to 4.4, you loose access to a lot of your data.

4.4 Addressbook does not have multiple IM addresses.
You can access this data by using the Akonadi console. So the data is not lost.

BUT - try to find the notes field! There are no personla notes anymore. And if you have done as me - stored essential data in that notes field, this data is nowhere to be seen. I already tried looking for it using Akonadi Console. It is not there.

But who in their wisdom decided that people using KDE are perfectly ok without access to their data for 6-12 months? Because this is the effect if you are not VERY into digging for lost data. Try telling a normal user that they have to use Akonadi Console to access their data... This is an excellent way to push people back to Windows and Mac!

I am really disappointed at this.

We do not need two versions of the PIM in 4.5. We need it in 4.4! In 4.5, we need one single PIM that really works. I don't care if it is on Akonadi or not - it only has to work. If it does not work on Akonadi yet, let us have the old one until the Akonadi version is finished. TOTALLY finished.

Oceanwatcher wrote:
The devs treat KDE like a playground with total disregard for people that rely on continuity.

Hey Oceanwatcher, I think you miss the tone a little. After all KDE is a community and a big one. There are many people with different needs. And allthough the community is growing, I guess there will never be enough devs for all the wishes and needs. So things like this happen. So the community has to find ways to deal with it constructively. That is what I tried to do with this suggestion. And it seems like that it needs some convincing and although you describe the problems you have, the tone is not convincing.

Markum

Sorry, that was not the intention.

I was preparing a long response here and decided to drop it. I think it is better to sum it up with this:

We can not wait for 4.5. This means we will not have access to our data until very much later this year. The dual solution is the solution we should have from 4.4. I think the best that could be done is to release this as part of 4.4.1. Then we would have it in 1-2 months.

I don't know if I understand connection between Akonadi and KDE PIM,
but there should definitely be option to use KDE PIM without akonadi, especially on low end devices like netbooks.

I really hope that the transition will be smooth and no data will be lost... anyway before upgrading to a full akonadi system I'll do a full backup...

Okay... I'm not much in favor of having multiple versions of the same applications in the repos. What I AM in favor of is thorough development and testing before something is released. Remember KDE 4.0? KDE can post all the warnings they want about platform releases or whatever else; most of the the major distros are stupid and WILL put the newest versions in the repos. The answer is to not release anything that isn't completely ready. KDE would do well to adopt the Debian philosophy: it's not ready until it's ready.

In theory no-one would argue with you. In practice it's not possible to check for every possible variation of users' hardware, software choices and personal usage habits. And of course distros such as Debian do benefit from waiting to see what other users experience and how they fix it - IOW, to some extent the rest of us are beta-testers for the more conservative distros.

Devs: Implement nice Akonadi-Stuff and port all the old stuff.
Users: Install the 4.4 version from repositories. I'm currently using openSUSE's 4.4 KDEPim-packages and openSUSE's 4.5 snapshots for all the other applications, it works very well.