Thursday, August 14, 2008

</gsoc2008>

<humour>Dipesh, I think I'm gonna sincerely hate you... You know I don't like blogging, I already told you, and despite of this you send blog posts that I have to answer if I don't want to look impolite...
That's a dishonest way of doing things, a treachery.</humour>

So, I won't make the whole history of the page master thing, it has already been described by dipesh, in commits, on koffice mailing list and so on... Doesn't really matter.
The changes we made in the branch have been reviewed by Thomas Zander on Monday, I spent a lot of time to fix every problem reported (mainly unit tests needing upgrades), and also improving the changes, Sebastian did the UI changes. The branch has been merged into trunk today, so it's done : KWord does support page styles.

Else, I started looking at the other KWord problems (not ODF specific). There are a lot of cool things to fix, like for instance cursor blinking (too late, it's done), table support (too late, you're doomed if you try this), working support for variables in headers/footers (if possible without requiring hacks, it's gonna be hard)...

This year, the summer of code ends quite early : on monday, it'll be over. So I'd like to thank my mentor Sebastian (despite his treacherous blog posts), the whole KOffice team, the KDE community and the folks I met at the Akademy (I won't give any name, I'm too bad at remembering names and it would hurt people if I forget them I suppose)...
I do not thank nokia for giving away free N810 to developers and thus killing the wireless router at the Zandpoort Youth Hostel (and I didn't manage to get any N810 :/ too bad, I wanted to port parts of KWord to it).

Well, I think this blog post isn't messy enough to be a good one, but I have no idea to increase the mess.
Ho sure I have things to add.
For instance, I promise I'll do my best to continue working on KWord during the next year. There are so many bugs to fix, features to add...
Too bad, I had other things to add but I forgot them.

8 comments:

Sebastian Sauer said...

> That's a dishonest way of doing things, a treachery.

But it worked :)

> I promise I'll do my best to continue working on KWord during the next year

yay, that *is* fantastic!

Rasi said...

"If you try that you are doomed"

This means its there, but we better dont use it or it is still not there at all?

Pinaraf said...

It means we don't have any table support in KWord so far. It's not really worse than KWord 1 which had a terribly poor and buggy table support...

Anonymous said...

Than you so much for proper page styles. Their lack was really annoying in Kword and made writing really large documents impossible for me... and I love the app!

So, again: thanks!

elvis said...

Love the mess Pierre! Keep it coming! *blink* :)

Anonymous said...

You can hate people all the way for urging you to blog. Truth is: You're likely to fail SoC if you fail to communicate to the community what you're doing. It's not only bad practice in teamwork, it also makes KDE look bad in the eyes of Google, our and your sponsor. Only one blog at the end of the project is on the 'too meagre' end, IMO.

-- sebas

Pinaraf said...

Come on, sebas, it was just humour, relax.
If you think I didn't make it clear enough, ok, I'm gonna write a new post to fix that. But I hate the idea of writing a piece of text that thousands of people will see just after I click the send button, that's all.
I still don't understand how other people do to blog that much. But that's definitely impossible to me.

Sebastian Sauer said...

y, probably a bit strange kind of humor that deserves <Monty></Python> tags around them (similar to some of your commit-messages though I enjoyed the poetry, heh).

Re more blogs; valid point, sebas. Guess that's something I could have done better by forcing Pierre a bit more to do so. I guess what Lydia wrote here makes quit lot of sense.

My personal preferences there was to use IRC and to don't only "monitor" but to participate by reviewing patches, by looking at the code, by landing own changes, by having tons of design-discussions and by providing input and expecting output in real-time. In the second half, pagestyles, this was in my eyes the best way to progress at all cause the amount of work was just to much. I tried to get that done by myself before and did break after reaching to much frustration and that's something where teamwork (as in pair-programming) in the form of realtime IRC-communication can help a lot to get over that mountain.

But there is for sure also the PR-aspect we owe google and for that things like regular blogs and/or ongoing visible progress beside commits, devel-mailinglists, IRC and such internal communication is a rather important part of the deal.

(This time public) conclusion? Well, great work on the code was done. Visible progress was not so good. We need to work on the second part to improve it (beside gsoc since we learn for life here :)

Anyway, thanks Pierre for the great work and thanks sebas for the feedback!