I won't dispute what you said. You laid out what the right thing is to do. It doesn't always get done unfortunately. It's one of those things I mean to change soon. I include myself in that because often I too am a little low on time. It takes a few seconds to update a package's version entity, validate it, commit it, and done. At least the package is updated after the test says it compiles properly and it can be thoroughly tested over the course of time.
Rechecking the descriptions takes a lot more time. Existing descriptions might not be valid anymore as functionality is added or removed. Programs might be added and removed entirely. New programs requires additional research to figure out what they do and how to best summary it. During a development cycle like we currently are in, the package descriptions aren't always considered high priority. Often we find ourselves in a position where we can do a quick test compile & install and if there were no failures update said entity, validate, and commit. It might take a day or two before there might be time again. I'd rather get the package in without updating sizes, SBUs, and descriptions, than wait longer to get the update. Then there is the part where every now and then this kind of work is a waste of time, really. Let's take the lfs-bootscripts for example. Yes, the descriptions should be updated to properly explain the new scripts: hotplug and syslog-ng. Except syslog-ng will probably be going away sooner than later. Why explain a bootscript that will most likely be gone soon too. Not to mention the package itself too along with libol. I'm not actually talking here about what is right and should be done. It's more about what is practical. Also note that I'm talking about the quick upgrades here that don't get done by rebuilding an entire LFS system. When an entire LFS system is built to test a few things, then the 10-20 minutes it takes to build Glibc (twice even so we're talking 30-60 minutes here depending on the speed of a system), one could take the time to fill in a few already known missing descriptions. Sort of like "last time I upgraded lfs-bootscripts there were new scripts. Let me document them while I wait for Glibc to finish." The latter is the more practical approach. Unfortunately it's also the part that rarely if ever gets done. Anyways, I don't mean to make excuses, just explaining from a different point of view, even though I agree it's partially a wrong thing that I'd like to see changed. I've thought about making it a general policy too that if you upgrade a package, do it all at once including sizes, SBUs, descriptions, the whole nine yard. Honestly, if it takes an extra day or two to get a package updates it's not the end of the world. Exceptions do apply of course, but, as the saying goes, those make the rule. Thanks for your input, though, Randy. It's one of those things I hadn't gotten around working out since my "return" a little bit ago. There's lots to do yet. But at least it's out in the open now and we can think about the issue in general. -- Gerard Beekmans /* If Linux doesn't have the solution, you have the wrong problem */ -- http://linuxfromscratch.org/mailman/listinfo/lfs-dev FAQ: http://www.linuxfromscratch.org/faq/ Unsubscribe: See the above information page
