**** BEGIN LOGGING AT Sun Jan 07 02:59:58 2007 Jan 07 11:23:29 <[cc]smart> I just noticed angstrom is in the MasterMakefile Is there reason that this is not the recommended way of baking Angstrom ? Jan 07 11:38:19 lots of reason Jan 07 11:38:49 the most important one is that people learn better by doing things themselves Jan 07 11:40:00 [cc]smart: MasterMakefile is an nslu2-linux thing - nothing to do with Angstrom, other than a canned way for all nslu2-linux developers to build it the same way. Jan 07 11:42:16 <[cc]smart> Does "it's an nslu2-linux thing" make it a bad idea to use it for angstrom in itself ? Jan 07 11:42:42 <[cc]smart> i would think "a canned way to build it" would be a good thing (TM). Jan 07 11:43:07 <[cc]smart> so. i don't get what you want to tell me. Jan 07 11:43:33 the makefile is a good thing Jan 07 11:43:46 but too many times people hid behind it Jan 07 11:43:58 it is not a substitute for common sense Jan 07 11:47:18 <[cc]smart> er. no, it ceratinly isn't. but it is still a darn good way to get going i'd think. the startup wall is high and i'd think it's easier to chase reasons why one thing doesn't work once you're confident the other ones are ok. however, this is philosophy, not science. Jan 07 11:48:36 <[cc]smart> MasterMakefile should typically give a working example the least. Jan 07 11:50:28 <[cc]smart> in the essence though, there should be no reason for me to keep two separate setups. that is, you do not sense any issues if i go and use the same environment for ngstrom i use for openslug... or are you ? Jan 07 11:51:30 [cc]smart: the master makefile builds OE in the way that the OE doc recommends, so as long as you use common sense (I agree with koen on that one), then there should be no problem. Jan 07 11:51:58 if you use the master makefile to set up your environment, you can still cd into the subdirectory and use bitbake as you would have done without the master makefile. Jan 07 11:53:17 <[cc]smart> right, this is what i would intend to do. what i don't see right now is where angstrom would differ from what i think is usual OE way based on openslug fiddling... Jan 07 11:59:48 [cc]smart: SlugOS has a lot of slugos-specific stuff in it which shouldn't be slugos-specific, due to early slugos development being done by peole who were not as familiar with OE as the core OE developers (and therefore felt they needed to do things in a SlugOS sandbox instead of pushing ideas into the wider OE arena). The intention is to change that in the future. Jan 07 12:00:04 So from that aspect, SlugOS is not a good example of what to do with OE. Jan 07 12:01:05 From a getting a build system up and running point of view, I agree with you. As long as people actually look and see what the makefile is doing when something goes wrong, instead of complaining to #oe from a position of no understanding of how OE or bitbake works. Jan 07 12:05:47 <[cc]smart> ah i see, you mean because they are using org.nslu2-linux.dev Jan 07 12:05:53 <[cc]smart> right Jan 07 12:07:58 org.nslu2-linux.dev is just the Master Makefile and a couple of other setup files. it is not an alternative to org.openembedded.dev - nslu2-linux uses org.openembedded.dev directly, we do not fork the metadata. Jan 07 17:09:01 <[cc]smart> een chacko Jan 07 17:09:05 <[cc]smart> ~seen chacko **** ENDING LOGGING AT Mon Jan 08 02:59:57 2007