Another glitch with Kontact. I use wordpress for this blog. And I recently started using kontact journal to keep some private journal entries. Now, Kontact journal allows you to use XML-RPC, which means I can use kontact to write and read blog posts from this blog. I tried that, and I changed my last blog entry in Kontact. But instead of changing the post, Kontact deleted the post. This is ridiculous. Seriously, Kontact people need to straighten up their basic testing protocols. The first and foremost thing is that data *must* not be deleted. And someone told me that open source developers are the most intelligent developers. I sincerely and with sadness doubt that.
So, I use a pop3 client for all my email accounts. I prefer to keep all my emails on my computer rather than on web. And I organize my emails properly. I have lots of folders, and lots of subfolders. And lot of subsubfolders. I liked the folder system most of the time. I didn’t like the label system because I prefer not to see the read emails (read ‘read’ as ‘labelled’). In my email client, I move my read emails to the right folder, so in my inbox, there are only those emails which either I have not read or I haven’t responded. But, now I am finding folder system a little annoying. The problem is that if in an email thread there are 10 emails, and 5 of them are already in some folder, and the new 5 are in inbox. Its difficult to switch back and forth between different folders. In label system, this is easy. Also, it makes it easy to leave all emails in inbox. So, I do not have to plan move emails all the time. I wonder how difficult it will be to implement label system in Kontact or Thunderbird. I have been noticing that Google interface seems bad and ugly, but there are subtle things which make it much more usable. Google HCI people are way too intelligent.
Continuing my post about business models for open source, here are some other things which make me think if open source is right way.
0) Well, as usual the first thing that I am able to think of is Office Suit. I am thinking of getting a tablet pc. I have wanted a touch screen laptop for long time time. Now, with touch screen I want to write notes and I want it to convert to text by itself. I owned a windows mobile for sometime. I used to use MS Word in it. And believe it or not, it was awesome. Handwriting recognition worked very good (after a small amount of practice, of course). Is there such thing in open source world? I bet not. Why not? Because it require research and lot of work. And then open sourcing it of course doesn’t make sense.
1) As all of you agree Evolution is a dead end. Thunderbird is good. I use Kontact though. I like to use default applications. This makes things look decent, and interconnectivity better. I did not shift to KDE4 untill KDE4.1 came. At 4.1, I expected things to work a little stable. Let me tell you a funny story about Kontact. In 4.1.0, if you close kill Kontact, you account information used to get vanish. Thankfully not the emails. So, you have to enter all the account settings again. In 4.1.4 things got improved, it stopped happening. But 2 days back, this happened again. I do not know why. I haven’t tried regenerating that problem. Another problem is that if I don’t give the kwallet password for sometime, it starts asking for password for accounts. Now this is ridiculous. There is no point asking for account password unless I am giving wrong password for kwallet. And when it ask for account password, and you don’t give it, rather you you logoff, it will ask for this password even if you give kwallet password. What a ridiculous sytem. You can not even properly store data. The worst part is the removal of all account settings. What a crappy thing must be going on inside that your data written in files is gone. Now this is called lack of structure. How did they even think of making a program which can remove data from a file. I am not able to think of any excuse for this problem.
2) PDF is an open standard. I use pdf a lot. But it seems that there is no proper software in ubuntu repository that could render pdf properly. I am mainly talking about different standard symbols used. Like ‘belong to’ sign, ‘equivalent’ sign and so on. Now these are very-very popular symbols. Open source people are not even able to support open standards, how worse can it get. (I tried using Okular and Evince, latest versions in ubuntu intrepid repositories.)