On 11/07/2011, at 6:26 AM, Rui Maciel wrote: > On 07/10/2011 05:35 PM, Tom Albers wrote: >> You can talk all you want, you called akonadi and friends 'cruft' and I find >> that disrespectful and unneeded, you could have conveyed exactly the same >> message without the addition of the word 'cruft'. > > Listen, I was very clear on what I've said. So you can at least try to > understand what has been said, which wasn't the insult you perceived, or > you can keep on overreacting because someone used a word you didn't > approved. Yet, the latter doesn't help anyone in any way, including > getting kmail to build on a platform which doesn't provide Akonadi, and > only adds noise to this mailing list.
Heh, heh, with friends like Rui who needs enemies? ... :-) Seriously, though, I was greatly surprised to wake up this morning (in Australia) and find that this storm in a tea-cup had taken place. Are we not professionals here and can we not discuss the issue without focussing on one pejorative term? Well, I have made my choice and given up KMail. Three out of four of my immediate family have between them two Macbooks, two iPads and two iPhones, so I have acquired a Macbook to stay in sync with them. I already had an iPhone which holds all the contacts, etc. that I need and it syncs automatically with my wife's iPhone and iPad contacts and appointments over our house's wireless network. So I have no need of Akonadi and friends, hence my original query. However I daresay I will install a Linux-KDE development setup on the Macbook some time ... ;-) My concern now is the tightly bound design of PIM software and also, for that matter, Nepomuk, Strigi, Soprano and friends, which forced this choice upon me. Is this the way KDE is going to be for evermore? Surely, PIM could be designed around a shared data source (a relational database if you must) in such a way that the various applications can exist independently of each other, in a loosely bound form. Wasn't that why some of us adopted and championed the separation of applications and data in the 1970s and 80s? KMail could even survive without the data source. I have used it without an address book for years and just relied on its memory for recently used email addresses. And if the user wants Strigi and friends to index his/her emails, cannot KMail simply tell them where to find the emails by updating a shared file or database? Re Strigi and friends, why do we have to build all of them and their dependencies (including Redlands, etc.) as a pre-requisite for building KDE libs and other parts of KDE? In the past few years I have spent numerous hours chasing after Strigi and friends when updating and building a development setup, but never have I had a second's worth of benefit from them. I used to think that KDE 4 is so highly complex that it needed the resources of a semantic textual database to build itself, but that does not seem to be the case. I do not know what other KDE developers think, but for me Strigi and friends are an intolerable overhead on development work in KDE. Cannot they be de-coupled? All the best, Ian W. >> Visit http://mail.kde.org/mailman/listinfo/kde-devel#unsub to unsubscribe <<
