Slashdot Mirror
Designing Good Linux Applications
An Anonymous Coward writes: "A guy from IBM's Linux Impact Team in Brazil has written a guest column on Linux and Main describing how applications should integrate with Linux. It's Red Hat-centric, but there is a lot of material about the FHS and LSB that most users probably don't know."
...'cos the common abbreviation for a Compaq Linux Impact Team would be interesting.
/. readers who have no Y chromosone and/or who don't appreciate South Park-style humour.)
(Apologies in advance to all
"Accept that some days you are the pigeon, and some days you are the statue." - David Brent, Wernham Hogg
before going to design NEW linux applications,
PLEASE take your time and DEBUG the current ones.
The collection of half-abandoned software that has tons of bugs that nobody uses (perhaps of those bugs) is absolutely huge.
While he doesn't mention Debian at all, it's clear that the article is strong on packaging. I actually prefer Debian's approach, having a list of sources from which you obtain software, and providing search tools for that list.
./configure command line to get a build that did what you wanted, you're likely to put off upgrading.
/etc/apt/sources.lists
;)
;)
The other important thing is that programs often don't work very nicely with each other, or need certain versions to work. This is where having a central system for controlling dependencies is rather important. I don't actually think Debian goes far enough at the moment (not really handling Recommends with apt), but it's getting there.
The other important part of packaging is handling upgrades automatically. Packages have security problems, they have new features added. If you have to work out (a couple of months later) which --gnu-long-opts-enable --with-features --without-bugs you had to put on the
# echo "http://debian.brong.net/personal personal main" >>
# apt-get update
# apt-get install bron-config
Whee
(note - that URL doesn't exist yet, but it's my plan for the future).
(note:2 - no ssh private keys in that
From the article:
I understand that this is directly from the FHS, and not some evil concoction from the mind of the author, but dammit, I think it's wrong. Perhaps /usr/local is obsolete with respect to package managers, and that makes some sense (because the package manager should handle proper management of placed files, though in practice that's not always the case), but as long as open source is around, there will always be software that is compiled rather than installed through a package manager. There will also always be applications that are not distributed in your package format of choice (as long as there is more than one package management system, this will always hold true). In these cases, it's still a good idea to keep around /usr/local and /opt. Personally, I'll have /usr/local on my systems for a long time to come, because I prefer to use the Encap management system.
Designing Good Linux Applications
The 'Linux' word is completely unnecessary - "Designing Good Applications" should suffice.
Application design couldn't care less of the OS that the application is planned to run on.
I do agree that some of what they talk about in this article would apply to most applications, but not everybody uses an OS the same way. Take this except, for example:
"Everybody loves graphical interfaces. Many times they make our lives easier, and in this way help to popularize software, because the learning curve becomes shallower. But for everyday use, a command at the console prompt, with many options and a good manual, becomes much more practical, making scripts easy, allowing for remote access, etc. So the suggestion is, whenever is possible, to provide both interfaces: graphical for the beginners, and the powerful command line for the expert."
This is wonderful advice in the Linux world. However, most Windows and Mac users, sadly, don't know what a command prompt is, let alone how to script it. This is a native concept to a Linux user.
I have no doubt that even in the Windows/Mac world a really powerful Command Line feature for any given app would be super useful, but it is only so for those who have climed that learning curve. In that case, it's better to focus on making the App do what it needs to do.
In any case, I'm sure I'll draw criticism for that comment. I'd prefer you didn't, though. The point I'm making is that slasho81's comment that all software should be the same despite the OS isn't quite so black and white.
"Derp de derp."
Documentation in
There will be things you don't like about the LSB and FHS. Personally, I reckon initscripts aren't config files and should live in
Anyone running Red Hat 7.2 or many other RPM based distributions can easily install apt (or a similar tool, like urpmi, tho I prefer apt) to do the same thing.
5 5-fr7.i386.rpm
The advantage there is that RPM is a standard - currently the older RPM (version 3) is included in the Linux Standards Base, but once Maximum RPM is updated for RPM 4, its extremely likely that RPM 4 will become the standard.
If you're using Red Hat I highly recommend installing it.
rpm -Uvh http://enigma.freshrpms.net/pub/apt/apt-0.3.19cnc
apt-get check
apt-get update
apt-get install
The author states that /opt is obsolete, and that everything should use RPM and install in /usr. Maybe this is the ideal in a system where everything is binaries-only, but I firmly believe it is poor administration practice.
The RPM database is binary and fragile. Once it is corrupted, the data describing what belongs to what goes out the window. RPM-packages have to be trusted not to clobber existing files or make changes to configuration files that one wants left alone. The alternative is per-application directories and symlinks (or a long PATH variable); there are tools which automate this, such as stow. The advantage is that the file system is - or at least should be - the most stable thing in the system. One can just examine a symbolic link to see what package it belongs to. This makes removing and updating applications very easy, and also makes it easy to see if there are any links left around from older installations. Removing an application is typically as simple as removing the corresponding application directory.
RPMs which install in the /usr tree will require root priviledges, whereas applications that can work from a self-contained directory can be installed by a non-priviledged user in their own directory,
Also, /usr in principle can be mounted read-only. This will certainly slow down any attempts at installing software in it!
I have had Redhat's installer corrupt the RPM database on multiple occasions; and I've had to override the dependancy checking innumerable times in attempts to update packages under both Redhat and SuSE, thus rendering useless the other purported benefit of RPM. New software typically comes in source form before RPMs; and the RPMs that do become available are almost always going to be third-party ones that don't necessarily play well with your system. By the time a vendor-created RPM becomes available, the distribution version you are using is no longer actively supported, and you'll need 300MB of updates to other packages just to satisfy dependencies. I've been there, it's horrid.
Seriously, a lot of Linux applications try to duplicate the Windows world and end up being just as bad. For example, for audio software, a monolithic executable with GUI is a Windows-style application--hard to reuse, hard to extend. A bunch of command line applications that can be piped together and come with a simple scripted GUI, that's a good Linux application because its bits and pieces can actually be reused.
Ditch the concept of spreading pieces of your app all around the FHS. This is organizationally similar to Microsoft's registry. It becomes a maintenance nightmare. Yes, RPM keeps track of some pesky details that let us get away with a messier install. Yes, the FHS does impose a common structure on what is an otherwise unstructured mess. But programmers are human beings, subject to the whims of ego, ignorance, and yes, even creativity and sheer brilliance. We're going to deviate from the suggested standards if given the opportunity, for one reason or another.
Give me one main point of access to everything the application does. If you need to use config files, give me the option of manipulating them through the application itself, preferably in the context of my current task. Give me one place to go looking for all the bits and pieces of the app. No, the FHS isn't simple enough. Give me context-sensitive documentation so I don't have to wander outside the app to get my job done. Don't make me wade through a spaghetti-code config file, with the documentation propped open on a separate screen to keep from getting lost.
Programmers are lazy. I should know, I am one. The last thing I want to do when I'm getting ready to release a program to non-techie users is tie up all the loose ends that seem ok to me, but not to the non-techie user. I'd rather document how to get a tricky task done than write the code that automates the tricky parts. I'd rather tell the user how to go tweak the flaky data in the database by hand than add another error-correcting routine. And it's more work to give the user one simple, full-featured point of entry to each piece of a complex application. But that additional work will make the application more usable, for the expert and the novice alike.
That is what I found in the fortune at the bottem of the this thread.
Ascii artist &
/opt is in FHS 2.2 at secton 3.12. It begins:
Doesn't look very depricated to me. I think the problem is your FHS link isn't really the FHS; it is the SAG (Systems Administrator Guide), which in section 4.1 clearly says it is loosely based on the FHS.
As for /usr/local, I do agree it should be off-limits to the distribution (besides setting it up if not already present). And packages in the package format of the distribution (e.g. RPM for Redhat, Mandrake, SuSE, etc ... DEB for Debian and any like it ... TGZ for Slackware ... and so on) really should stay out of /usr/local. What /usr/local should be is whatever is local policy (FHS doesn't say it this way). Packages that the administrator really wants to be separate from the package management system, stuff compiled from source, stuff locally developed, all is eligible to be in /usr/local. My guess is the author of the article has no experience doing system administration combined with a decision making role where he might have to choose to do something slightly different than what everone else does.
now we need to go OSS in diesel cars
After using many versions of Slackware, I finally tried Redhat at version 5.1. Actually I had tried it at a way earlier version and it never successfully installed. But 5.1 worked OK. The reason I tried it was I bought a Sun Sparc 5 and wanted to try Linux on it. Redhat seemed to be OK, so I later tried it on a couple other i386 systems, and that was working OK ... for a while. As it turns out, I needed to make upgrades before RPMs became available (see next paragraph). I also needed to make some changes in how things were built. The RPM system started getting out of sync with what was actually installed. The system ran just fine, but soon it got to a point where some packages I was installing with RPM would not install because the RPM database thought things were not installed which actually were (but weren't installed from RPM, so I can understand why it didn't know this). So I ended up having to do forced installs. And that ended up making it more out of sync. By the time I had gotten to Redhat version 6.0, I was getting fed up with it. I switched back to Slackware (and Splack for Sun Sparc eventually came out and I use that, too) and am happy again, with well running systems. And I am now exploring LFS.
You say the system administrator should know how to package applications? Why the system administrator? I'd have thought you'd have expected the programmer to do that. If I get some package which is just a TGZ source file tree (because the developer was writing good portable code, but not using Linux to develop on), why should I, in the system administrator role, have to be one to make a package out of it? I'll agree it doesn't take more brains than needed to properly install the majority of source code, but I won't agree that it is easy (in terms of time spent) to do. At least I have the brains to actually check the requirements of what a given package I'm compiling needs, and make sure it is there by the time it is actually needed. The dependency may not be needed until it is run, so I have the flexibility of installing in whatever order I like. Also, some "dependencies" are option, and don't need to exist unless a feature is to be used that needs it. For example, if I'm not using LDAP for web site user logins, why would I need to make sure LDAP is installed if some module that would otherwise use it is smart enough to work right when I'm not using LDAP.
now we need to go OSS in diesel cars
My sparc runs fine at run level 5. What would you like to see the various run levels be for? I changed mine around, and it looks like:
now we need to go OSS in diesel cars