JP Vossen on 10 Jun 2009 23:14:10 -0700 |
> Date: Wed, 10 Jun 2009 17:10:27 -0400 > From: "David A. Harding" <dave@dtrt.org> > > On Wed, Jun 10, 2009 at 03:48:20PM -0400, JP Vossen wrote: >> http://www.google.com/search?q=csh+considered+harmful > 9,200 Results > > http://www.google.com/search?q=bash+considered+harmful > 44,100 Results > > :-) Nice! Got me on that one. But wait... I was sloppy. I'll see your results and raise you (add quotes): http://www.google.com/search?q=%22csh+considered+harmful%22 467 http://www.google.com/search?q=%22bash+considered+harmful%22 4 (8) > (Although I agree that scripting csh/tcsh is unwise.) Yeah, I'd never bet against Tom Christiansen... >> the major difference between Debian and Red Hat is the packager, APT >> (Advanced Package Tool) for Debian and derivatives and RPM (Red Hat >> Package Manager) for RH and children. > > To be fair, APT is to YUM what DPKG is RPM. Comparing APT to RPM is as > unfair as comparing YUM to DPKG. (My goodness, that's a lot of > abbreviations.) I agree that dpkg == rpm, but APT == yum is harder because "APT" is a suite of tools whereas yum is *a* tool. More on that below. > Also, how come Debian gets derivatives and Red Hat gets children? I was just trying to avoid being repetitive. Consider the terms interchangeable in this usage. >> IMO APT is superior, > > YUM was written after APT became popular, so it includes what it's > developers thought were interface improvements. For example, it includes Agree. That yum behavior where it updated its cache every time you did *anything* drove me nuts, but it seems like that's fixed in CentOS-5 at least. (Yes, I know about yum -C... :) > in a single program, yum, functions that are separated into two APT > programs, apt-get and apt-cache. I think these interface improvements > make YUM easier to use for newbies but make no significant difference to > advanced users. I agree that yum is a cleaner and easier interface. I'd even argue that APT is much worse than you've said. It's not just 2 programs, I have more than 8 listed in the top table in http://www.jpsdomain.org/linux/apt.html, and too many to count in the bottom one and I'm probably missing some. It's arguable how many are "APT" (whatever that is) and how many are addons. I also just noticed that I should change my page from 'apt-get' to 'aptitude' which is actually newer and better. (OK, done.) yum, sometimes rpm and rarely rpm-build do everything you need, indeed can easily do, with RPM. (I'm ignoring edge cases using rpm2cpio & cpio.) With APT there are a lot more tools to keep track of, to the point that I opted out of that and wrote a shell wrapper (http://www.jpsdomain.org/public/source/deb) to remember all that crap for me. There is also a *much* more extensive wrapper in Python called Wajig (http://linux.togaware.com/survivor/Wajig_Overview.html). You can argue if all of the APT stuff is a bug or a feature. I'm undecided. It's a bug for newbies, though apt-cache and aptitude are probably enough. (apt-cache for search because aptitude's default search method is totally non-intuitive.) I should have been more clear and said that what I prefer about APT actually has more to do with process and philosophy than technology, implementation and user interface. The killer feature for me is that APT-based systems are infinitely upgradeable, whereas RPM-based systems **strongly** recommend clean installs for major version upgrades. The joke goes that the Debian installer sucked so bad for so long because you only ever used it once. And that is true in my experience for both Debian and Ubuntu. Also, I'll say that APT-based packages seem to Just Work out of the box more often and to a greater degree than do RPM-based packages, but that may be only my perception. Later, JP ----------------------------|:::======|------------------------------- JP Vossen, CISSP |:::======| http://bashcookbook.com/ My Account, My Opinions |=========| http://www.jpsdomain.org/ ----------------------------|=========|------------------------------- "Microsoft Tax" = the additional hardware & yearly fees for the add-on software required to protect Windows from its own poorly designed and implemented self, while the overhead incidentally flattens Moore's Law. ___________________________________________________________________________ Philadelphia Linux Users Group -- http://www.phillylinux.org Announcements - http://lists.phillylinux.org/mailman/listinfo/plug-announce General Discussion -- http://lists.phillylinux.org/mailman/listinfo/plug
|
|