User: Password:
Subscribe / Log in / New account

el trace narcotico

el trace narcotico

Posted Jan 28, 2010 19:18 UTC (Thu) by bkoz (guest, #4027)
Parent article: Back to the drawing board for utrace?

Thanks for providing some coverage of the languid soap opera featuring a charming but troubled kernel and the ardent suitors from the userspace development community who care about debug, trace, and analysis.

Pointing at a still-unmerged tree for realtime and saying "that's the way to do it, see?" seems disingenuous at best. Saying the maintainers of ptrace have no comment is also questionable, when in a related article Oleg indicates long-term solutions outside of ptrace are mandatory.

From outside the kernel community, tuning in at regular intervals to observe the serialized show, it looks to me that these two groups are just passing in the same development space without any real progress.


THE KERNEL. We watched him grow up, gain power and prestige while trying to remain sane in a world gone mad. There's a constantly changing field of opportunity and adversary that he attempts to navigate deftly. Detractors, even gentle ones, whisper of a touch of dementia, a rather narrow and introspective view of the world. Me! Me! It's not what I say or do or even what I say I plan to do (API), no it's what I've done (ABI).

SUITOR U. Has loved and maintained key parts of THE KERNELs business for years. Fix this. Fix that. Has proposed a grand utrace building for the main kernel grounds, and when the blueprints were shown, and told to adjust and update the main house kitchen in addition (ptrace), did so. While doing this, the groundskeepers stopped by and asks the suitor about using the utrace building basement for some new scheme. Last week's episode was when SUITOR U took the latest in a long series of plans back to THE KERNEL, and due to the general inclement weather and dour moods, was told something else entirely. Oh dear!

SUITOR L. Remember me? We danced way back when. I had a real pretty dress, a special bias cut, and caught your attention for a few months. You still let me come to your balls but now I only dance with an advisor or two, and console myself by saying I have the pick of the entourage. Call me. I follow you on twitter.

SUITOR F. It's my cotillion!!! OMG. He danced with me. Was told he's a rake but in this light, doesn't care. The night is young, whooopeeee!

(Log in to post comments)

el trace narcotico

Posted Jan 28, 2010 19:43 UTC (Thu) by dlang (subscriber, #313) [Link]

the realtime tree is a success because huge portions of it _have_ been merged over the last couple of years. In addition, both sides now consider this process to have improved not only the kernel, but also the realtime support compared to what would have happened if the realtime tree had been accepted as-is when it was started.

this is even allowing for bugfixes over time. They (the realtime people) have explicitly stated that breaking out portions from their tree and submitting them as individual features (and dealing with the demands to justify and clean things up before they are accepted) has significantly improved the realtime tree itself.

Copyright © 2018, Eklektix, Inc.
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds