plans vis-a-vis inline functions and templates

[Dirk Mueller]

On Tuesday 16 December 2003 17:46, Darin Adler wrote:

> Dirk, your comments would have been even more welcome if they had a
> less condescending tone. "Feel free to waste your time on it" is a
> really insulting thing to say.

Darin,

I'm sorry, it wasn't intended to be insulting. I was just trying to express 
that there are imho more productive things to spend time on, which includes 
but is not limited to actually clean up previously applied patches, or to 
work on areas that improve something visible to the end user of 
Safari/Konqueror. It is even more productive trying to get our trees more in 
sync (doing so would have prevented your recent v100.1 security update), 
although I meanwhile realize that our goals are too different, so I'm not 
trying to push this too much anymore. 

Its not as if I never "wasted time", and I don't see it as a negative thing - 
quite the opposite. Its a fun thing to do in spare time - much more than 
trying to debug a horribly complicated dynamic webpage to find an obscure bug 
that you have no idea where it actually comes from. I've done that almost 
daily the last 4 years, and I do understand quite well that there are so many 
other things that are a lot more satisfying.

For example, two years ago I started a little "time waster" when I was annoyed 
by the amount of compile dependencies when touching a header file somewhere 
in kjs/khtml. Look here:

http://webcvs.kde.org/cgi-bin/cvsweb.cgi/kdesdk/scripts/fixkdeincludes

Looking at it two years later, I realize the time I spent on this little 
script would've been better spent on anything else, since it turned out that 
the effect the script has on the build time after it finished the 
optimisation is in no relation to the amount of work that we put into the 
script to make it work. Not trying to destroy your fun, I think it will be 
the same if you decide to waste^h^h^h^h^hspend time on your 
deinline-one-liner-virtual-functions patch. 

> It's hard to take you seriously when you 
> write things like that.

Look, I'm not in the lucky situation to have enough time that I'd even feel 
like ever considering to do such a compile-time improvement which you 
were/are thinking of. I'm not paid for anything I do here, I have a day job 
which I need to be able to pay my rent and to buy food once in a while and a 
lot of other "obligations" inside the KDE project, besides being regularly 
buried under several hundreds of time consuming emails each day. In total I 
might have at most 8-15 minutes per day available for doing KHTML related 
work. I haven't been able to look at or even think about integrating the 
patches that were posted by the safari developers within the last 5 weeks. 
I've several unfinished patches in my local tree which I haven't been able to 
work on for months, so I'm trying to save time where I can, including not 
trying to spend more time than necessary on the daily load of email. 

I agree, and I admit that saving time by not spending an extra minute to 
carefully crafting and reviewing the way I write the emails to this 
particular list, might cause that some of the emails too sound harsh or 
snobby. It is not my intention, though I use a casual and informal style as 
it is the easiest (read: quickest) way to write my answer down. 

In the concrete case, there were two options: a) not answering your mail at 
all or b) quickly listing the possible areas for improvement that I would 
think would make more sense than trying to move code around that is both 
error prone and would eat up even more of my extremly spare time trying to 
merge something useful from the safari tree in an hopeless attempt to get a 
somehow barely working tree finished for the imminent KDE 3.2 release. 


Dirk