Thoughts for the day

* Each day, life will send you little windows of oppurtunity. Your destiny will utimately be defined by how you respond to these windows of oppurtunity.

* As you live your days, so you craft your life.

* Screw-ups are the mark of excellence.

* Real risk lies in riskless living.

* Reading a book by someone you respect allows some of their brilliance to rub off on you.


An interesting conversation :-)

This happened sometime back. I had it in my disk.

21:34:46) mqu: kjetilho: comments on #203206 sound quite shaking. »no one understands etable« and such.
(21:35:34) mqu: would look like being arrogant if someone actually started working on this, as it seems quite some people didn’t work it out
(21:35:49) mqu: ?
(21:39:30) kjetilho: mqu: no, of course not, patches are always welcome 🙂
(21:41:03) mqu: yes, but if »no one (of the evo team!) understands etable« …
(21:41:04) fejj: heh, everyone who has ever understood that code has left the project/company :p
(21:42:12) fejj: it’s like it takes people a few months of reading that code to figure it out, and then a few months later they leave
(21:42:53) mqu: fejj: does this make sense?
(21:45:00) mqu: fejj: u make me not want to look into the code because i don’t want to find out »few months later« i wasted 0.5% of my life 😦
(21:45:00) chen_afk: mqu:, may be this could help
(21:48:13) You are now known as chen_dinner
(21:48:33) mqu: chen_dinner: wow, someone understood it 🙂 quite a new post, thanks

Srini – the only person now who can keep ETable alive 🙂

re-entrant calendar cache

Yesterday I submitted the patch which makes ECalBackendCache thread safe. Initially i just thought of making it thread safe using GMutex, but this will block mutliple threads reading the Cache. Looking through devhelp, found GStaticRWLock which sufficed my need. With this multiple threads can read the cache at the same time, but if a thread wants to write something into the cache, all the other threads would be blocked. After testing the patch with various operations, attached it to bugzilla 🙂
And while testing i found the a major bug, where the groupwise calendars got deleted when a new calendar is added. I debugged it for a while and informed the issue to harish. Filed a bug for the same.

go crashers go !!!

I was working on some of the crashers last week. Have fixed around three major ones. Have attached the patches for 343396, 351330 (this made my head roll :-)). The third one is on the way for making ECalBackendCache thread safe, am testing the patch right now, it would fix 346673. Thanks to dave for providing the valgrind traces which was very useful in sorting out 346673 and 343396. There are still some invalid reads in weather calendar backend and during the EDS shutdown when the last client goes down. We should manage the operations in a better way before closing EDS. This holds good for evolution too when EDS goes down due to some crash. And for 2.8.1 me and harish are planning to work on all the major crashers in bugzilla. Lets make things solid 😉