Getting Software Added to Unix Distributions?

suso asks: "I've been working on a set of programs called num-utils that I would eventually like to be considered for inclusion in some of the many free Un*x distributions (on the install CDs, etc). So my question is, how does one put their applications on the track to be included in the main distribution of Red Hat, Debian, SuSE, *BSD, and so on? Is this just something that is up to the maintainers or are there submission forms of some kind?"

Linux is not Unix

unless, of course, SCO wins their lawsuit.

Re:Linux is not Unix

Did I just notice a recursive acronym here?
Linux Is Not UniX

Prepare to be modded down.

Re:Linux is not Unix

Un*x is traditional. AT&T can't sue you for trademark infringement if you don't say the whole word.

*nix is much newer (circa mid-90s, rather than the late 80s Un*x started in) and is actually much less accurate. Un*x refers to Unix(tm), I've heard *nix refer to just about anything POSIX. Why don't more people refer to things by the standards they're actually judging Un*x-a-like systems?

Usefulness and Popularity

If your programs are genuinely useful and well-written, they will build up a user base over time. Eventually they will become viewed as worth putting in a distribution.

Re:Don't use the GPL

Drop the GPL. Your software is much more useful and far more likely to be included in non-Linux systems if the license isn't tainted

Which systems are you referring to? I know the BSDs avoids GPLed code for drivers and core programs, but user packages is a different story. And OSX includes GPLed code. *BSD, All GNU/Linux, and OS X, I'd say that the GPL can get you into plenty of OSes.

-t

Write a text editor

Write another text editor app, then it will be sure to be included in the distro. Distros dont have enough text editors in them.

Simple

1. port it to as many systems as possible, even non-targert systems. possibly AIX, old Digital Unix.. you name it.

2. get the werd out. If people know about your package, it could solve a problem somewhere that would get it installed.

3. support it. if you support it, people will keep using it. even if it is initially crappy, you'll get bug fixes and advice.

4. package it. no one more than me.. 'cept for those that hate it more than me, hate doing custom compiles on a system that doesn't have /usr/ports or emerge.

Then you live on w/ your life. If your software is good and fulfills a need, you'll see it get put in.

Then you can go onto 5. Profit. or ????. YMMV

Re:Simple

2. get the werd out. If people know about your package, it could solve a problem somewhere that would get it installed.

Somehow I get a feeling that he has that one covered. :)

Make noiseb about it

Not to sound flamebait, but you're quite right in doing it : giving it the maximum visibility (for example by posting a link to it on a popular news discussion site) will make a few people notice it exists.

Now, the main question is does it do ogg ?

Re:Make noiseb about it

Now, the main question is does it do ogg ?

It doesn't matter about ogg. Once it reads mail, then it is feature complete.

Feedback forms

For starters, you could try looking for feedback forms on the distributions' web sites, such as these forms for SuSE [www.suse.de]. Forms like these are often intended to bring suggestions to the attention of the distribution developers.

Step-wise procedure...

1. Write software.
2. Take out a coyright in your name.
3. Apply GPL notices to code.
4. Publish code via ftp.
5. Send code to Source Forge and Freshmeat.

Very difficult?

-

Publish first to website.

Two ways:

How we did it (fpc, a pascal compiler)
- First the app was published on our site only, and gained momentum and peer review. This stage took several years.
- for the distributions where ordinary users can submit packages (*BSD ports and Debian) somebody
will do a port in time. You could do that yourself of course and speed up the process.
- After a time the commercial ones pick it up if it is really good. You can lobby for that too, but maintainers might also contact you if you have critical mass.
I found SUSE always the most responsive. RedHat is the only major that doesn't include it, and has been promising it for the next major version since 6.x times.

About SUSE there is a nice anecdote. I mailed our contact that a new version was out, and got a reply back that the final ISO had already been made. Two days later I got a mail back that they had to update a critical bug, and also updated our package to the newer version (which was a fixes only release btw)

The second way is to try to submit your packages to the FSF, so not just GPL it, but really get in bed with the FSF
FSF stuff more readily gets into distro's than third party projects. Of course again, they will only be really interested if your work is phenomenal.

Re:Publish first to website.

The second way is to try to submit your packages to the FSF, so not just GPL it, but really get in bed with the FSF. FSF stuff more readily gets into distro's than third party projects. Of course again, they will only be really interested if your work is phenomenal.

That is not true. The GNU project (the FSF doesn't do this directly, the FSF is the foundation that sponsors GNU) will take on any software, as long as it is free, conforms to the GNU coding standards, and is not yet covered by other GNU packages.

Savannah (the GNU equivalent of Sourceforge) also carries a lot of near-dead projects.

Re:Publish first to website.

Savannah (the GNU equivalent of Sourceforge) also carries a lot of near-dead projects.

Savannah, where GNUs go to die...?

Reputation + Packaging

Half of it is about reputation and having a good following. The other thing that helps is to supply packages for each distro. i.e rpm for redhat, deb for debian. For others like FreeBSD which have the ports tree see if you can get your files commited

HTh

Rus

GNU?

I think Gnu should help. Try submitting on their site. If you code according to their guidelines, and if your software is useful enough they will include it in some package (something like what binutils is).
Your program will then automatically get into *all* distros :)

Nandz.

Is your app a virus by any chance?

If so it will add itself eventually.

Get eyeballs..

The distro maintainers make these decisions based on popularity and dependency. Why include software nobody ever uses?

The larger distributions will not carry your tool until it's become widely adopted by the Linux community - be thankful, otherwise RedHat 9 would require a DVD or two, instead of (just!) 3 CDs...

These utilities you have here, while useful, will probably not see much user adoption. However they would be very useful in shell scripting. If a more mainstream user application requires your utilities to function, the distro will be forced into including your stuff - as a dependency.

Simply send it to SCO

They will take care of it and will find evidences that your code is already illegally included in all major distributions, the kernel and the rest of the world. And they will offer a license for using it.

Money talks but here's some free advice...

$100,000 buys you 10 minutes of face-to-face time with Dubya and I bet a similar "donation" would get you some time with the guys at Red Hat, SuSe, etc.

But, of course, that's not what you wanted to hear. I'd check out their FAQs, ask questions in their relevant forums, usenet groups, etc. I'd imagine that each distribution has its own criteria for inclusion so your approach to one vendor might have to be completely different to your approach with another.

Whatever you do, bear in mind two (slightly paradoxical) things:

1. They probably get asked to include lots of software, some good, some bad and some downright ugly.

I know of one major magazine that was sent an application for inclusion on its cover disk that calculated sales taxes for you - by taking the figure you gave it and multiplying by the relevant amount. That's the chaff. You need to be the wheat. So make sure that your software is truly worth inclusion (Does anybody already have a similar offering in their distribution? How does yours differentiate itself? How is it superior?) before you start investing serious time and effort into promoting it.

Also, remember that there will be great pressure, both internal and external, for vendors to keep their distributions free of bloat. Even if your software is unique, does it really offer something that a significant proportion of the target audience will want and use? You could develop the best doll's house design software for Linux ever but if nobody wants to design doll's houses on their Linux machines then you're screwed.

2. If you really do have a product worthy of inclusion then persevere.

Once you find the distributions' relevant contacts, harrass and hassle them about it until you get some sort of feedback. If they say 'yes' that's great, but if they say 'no' ask why it's a not a go.

But remember, although it might be their jobs to deal with new submissions, it isn't their jobs to deal with crap. Don't be offensive, advesarial or overly aggressive and don't become a stalker (leaving two voicemails a day is a no-no) or the only answer you'll ever get is 'no'.

Good luck.

On Perl and command-line utilities

Your mathematical utilities would be more useful if you had programmed them in C. Your choice of language will limit their adoption. Basically because using Perl scripts is not as fast as calling compiled C programs. This fact alone will make people reductant of using your utils in their code.

Because FreeBSD doesn't ship Perl as standard part of their distribution anymore, it'll be likely that your utils will not get included in any BSD software because it would pull in Perl. It may be a reason for Linux distributions too for not using your num-utils. Debian may be the only distribution which relies on Perl.

Re:On Perl and command-line utilities

Your mathematical utilities would be more useful if you had programmed them in C. Your choice of language will limit their adoption. Basically because using Perl scripts is not as fast as calling compiled C programs.

I have actually written a lot of this kind of code, working with huge datasets, and you are simply wrong. The performance of these kinds of utilities is dominated by I/O, not CPU. Furthermore, even the CPU-intensive parts of these utilities (conversions, etc.) are implemented in highly optimized C code inside Perl.

Because FreeBSD doesn't ship Perl as standard part of their distribution anymore, it'll be likely that your utils will not get included in any BSD software because it would pull in Perl.

I think Perl is a pretty awful language, but not including it in FreeBSD strikes me as a stupid decision. Perl is useful and works well across many systems. It's also not particularly big. If FreeBSD doesn't include Perl in the standard install, then FreeBSD has a problem, not people who write scripts in Perl.

What is BSD doing instead? Implementing all utilities in C? Gee, that's bright: let's create lots of unnecessary work for ourselves, increase maintenance hassles, make it really difficult to use good algorithms, make sure that there are plenty of opportunities for pointer bugs. That's negative progress: operating systems need less C programming, not more.

Re:On Perl and command-line utilities

FreeBSDs decision to move Perl from the base system to the ports tree was because it was easier to handle that way, both for developers and users. It was not about not supporting it anymore. In fact, it is supported quite well, including automatic integration of CPAN packages with the package management utils etc.

This seems to be easily misunderstood, probably especially by Linux users where no distinction between base system and third-party apps exists (or in a less visible way, at least).

It did indeed mean that some tools in the base had to be reimplemented, either in C or as shell scripts. Obviously the majority of developers decided that this was less pain than having to keep one version of perl around even when users want a newer one, because you could break their systems otherwise, to have to check various important parts of the systems when you integrate an updated perl etc. The result of all this is that having an up-to-date perl is just one "portinstall perl" away, the system is more stable and modular, and, indeed, that trivial perl utilities like those in the article are unlikely to become base components. Big deal.

Debian

If you want your program to be included in Debian, you may package it yourself. Debian New Maintainer's Guide [debian.org] is a good place to start at.

Or, you could file an RFP (Request For Package). See instructions [debian.org].

I don't think it's really all that hard...

The people that tend to do packaging are not likely to be influenced by you pestering their Inboxes, or filling out forms, or posting to forums, etc. Instead, ensure that your program meets the following requirements, and you should have no problems.

- It should fulfill a genuine need. If you're aiming for wide distribution you can't expect to achieve it with a something that's only relevent to a few people or in a few circumstances. You should also have some sort of document that shows how someone would save time or accomplish new things with this tool.

- It should be small yet robust, minimalistic yet powerful. I don't think anyone would consider adding a tool to a default install that is either too large for the features it offers, or two pedestrian in the type of features that it offers.

- It should be packaged well. Ideally it should compile and install in the proper locations out-of-the-box on a variety of systems. Make sure that it uses well-known methods, such as autotools (i.e. "./configure --prefix=/usr/local") or some other well-know "make; make install" type of setup.

- It should be well documented. At the very least you should have full manpages that your install script puts in the right place. Also consider man2html output on a web site, an possibly texinfo for the purists. You can't expect to get away with "just run --help and figure it out" or "look in the README."

- It should be licensed sanely, and should have reasonable dependencies. No one like a bizarro license, and no one likes a tool that takes sixteen different libraries of particular versions to compile.

- It looks like you're trying to get these tools standardized so that they could be relied upon for scripting... this will always be very hard to accomplish, but you might look into getting them merged with some popular packages, i.e. 'fileutils'. If there's a particular program that they are well-suited to being used with (like awk or something) then see about getting them added, perhaps in a "contrib" dir, to a project like that.

Frankly, though, your post was a little worrysome... in the sense that it almost seems like you're trying to get everyone to use these tools because they're there, not for some intrinsic reason. That just won't work, they have to do something really well or make it much easier to do some other task, etc.... You can get the word out and announce to various interested parties, but you will never be able to force anyone to do anything. In other words, view the situation as one of wanting to make the best programs you can, and if they receive universal support that's icing on the cake.

e-mail

Make sure it can check your e-mail. No software is complete until it can check your e-mail.

HE's from SCO!

He's trying to sour the batch! Don't let him near it! :) j/k

Meanwhile for Windows developers...

Thankfully in the Windows world I don't have to concern myself about getting included in "official" distributions... I prefer to distribute my software via self-propagating emails. ;)

Build Mindshare

Some distribution, I'm thinking of gentoo, have a user submission feature for new packages.

Just submit a new ebuild as a bug report. (No, that IS actually the proper way.) After a few weeks in the mill, your package will be out and about and happily rsynced in with every gentoo user. Gentoo are working on porting portage, their source distribution mechanism, to MacOSx and Window (running CygWin).

Of course, instant gratification is not a hallmark of the portage system.

You are competing with everybody else's widget in portage. So just make sure you get in cahoots with the folks who write the install docs, and have your software be made the subject of a few ZDnet articles. Writing a HOWTO based around your product is also a good idea.

How it worked for me ..

Once upon a time I wanted an MP3 streaming server, none of the ones I looked at did what I wanted. So I did the standard thing and designed my own.

After releasing my first version to freshmeat [freshmeat.net] I had about five subscribers to the project.

These subscribers gave me patches, feedback, and encouragement.

Doing a websearch [google.com] for the project name I discovered by accident that the the package made it into Gentoo [gentoo.org], and similarly Netbsd without any feedback or involvement from myself!

The next step was my becoming a Debian Developer [debian.org] so that I could upload it there - and not worry about other people doing a bad job without me. (Not a real concern; I had wanted to join Debian for some time anyway).

Now life is good - I've no idea if it's in RedHat because I've not touched it for years, but SuSE include it the *BSD's and Gentoo cover it, and Debian gets the latest versions all the time.

Freshmeat lists 120+ subscribers to the project, and it's probably on the verge of becoming an official GNU package sometime soon.

If you use it and like it buy something nice? [amazon.co.uk] </ObPlug>

Version?

Dude, you're only at version 0.3. Don't you agree that it's a little early to be talking about adoption by the distributions?

unique?

Your commands are already installed on my system, or aren't needed.

average
awk '{sum+=$1}END{print sum/NR}'
bound
awk 'NR==1{min=$1}{max=$1>max?$1:max;min=$1 &lt; min?$1:min}END{print min,max}'
interval
awk '{print $1-last; last=$1}'
numgrep ( 500 < x < 1000 or x is a multiple of 3)
awk '$1 < 1000 && $1>500 || $1%3 == 0 {print $1}'
numproc
awk '{print $1 [[your math expression expressed in standard infix notation]]}'
numsum
awk '{sum += $1}END{print sum}'
random
awk supports rand(), bash has $RANDOM, Linux has /proc/random for a stream of random data. Any range can be chosen using 'numproc', for instance /1..10/ is 'rand() * 10 + 1'
range
Why are you ever instantiatin ranges? It wastes space. Ranges should be abstractly manipulated.
round(floor)
awk '{print int($1)}'
round(nearest n)
awk '{print int($1/n)*n}'

These commands are easy to use and have a transparency that makes it very clear what the bugs would be, where as num-utils has warnings like

round will drop off the decimal places in decimal numbers. This may cause some calculations to be in error, depending on how you are using the data.

that make me wonder what round does if it has problems with decimal numbers.

Joe

Transparent Debian

I don't know about the rest of the world, but the process for getting something into Debian is fairly transparent.

Either add your package to the wanted [debian.org] list, or become a Debian Developer [debian.org] yourself.

I'm not saying becoming a Debian Developer is quick or easy (though few would describe it as really hard), but the process is very transparent, and available to anyone. In part, I suspect this transparency, in combination with its maturity, is why Debian has so many more packages than any other software distribution.

Never underestimate the power of transparency. :)

It's in Gentoo...

alexscomp alex # emerge -s num-utils
Searching...
[ Results for search key : num-utils ]
[ Applications found : 1 ]

* sys-apps/num-utils
Latest version available: 0.3
Latest version installed: 0.3
Size of downloaded files: 28 kB
Homepage: http://suso.suso.org/programs/num-utils/
Description: Set of programs for dealing with numbers from the command line

HAHA don't get so excited...just pulling your leg.

the FreeBSD perspective

First off, we dislike having gpl code in the base system, but we do have gpl stuff (but always a bsd fall back) for things like tar, gcc, dialog, rcs, sort, gzip, and just a few others. We keep our /bin and /sbin clean of the gpl. We typically think that awk or perl is capable of this sort of thing, and to get a program like this into the base system requires at the very least a commit bit in the cvs tree, or approval form the core team, or membership to the core team (yeah right). Since a text-processing, or rather a number processing, program is considered to be strictly useless to a base system's functionality, it would most certainly be religated to the ports tree. Don't kid yourself, you might think your software is the best thing under the sun since sliced bread, but to be so bold as to think it needs to be included in the base of any Unix-like-system is a pipe dream.

easy

Just put a blatant plug for it in the form of a question on Slashdot.

Oh, wait. Nevermind.

For real though, just post it on freshmeat and if people have an interest it'll get popular quick.

To be included in Solaris...

For Solaris, once your util becomes an essential application used everyday by 99.9% of Solaris users, Sun and your good self can follow this procedure:

- Use the Solaris package tools to create a package for your program, make the default install directory somethig sensible such as /opt/SUNWats/sun4u/bin/thiswillneverbeinyourPATH/p kg1921u9238/

- make sure the package requires a few libraries that will take a least a day of pain to install on to any Solaris box.

- Ensure to include a man page, avoid using words with less 5 syllables, refer to everything as n.

- now do nothing for roughly six years (more if the program is required for other popular applications).

- Once that is done, send the package to sunfreeware (because downloading endless packages from the designed-by-satan website is by far the quickest way of installing essential programs via a text based console).

- It can sometimes only take a few years from this point for Sun to include it on the Solaris CDs!

- Of course, they will first need to put it through the flag-randomiser to ensure no command line switch is the same as what it is for every other OS in known universe. It will also remove --help and -h, to avoid you having to do this yourself.

- Just think, by Solaris 27 (aka SunOS 2.9.1) you can see your package installed by default from a Solaris CD!

cjk
PS remember, if your program involves text editing, ensure it implicitly uses ed, lord knows what confusion it will course otherwise.

For Mandrake Linux ...

... you can follow the guidelines on their site [mandrakelinux.com], although it may also be helpful to post a link to an SRPM in your mail (which you should also cc to the cooker list [mandrakelinux.com], since the employee handling contribs is very busy, but there are a lot of non-employee contributors who will be able to add your package.

More and more development is being done by the community, so you may want to stop by the cooker wiki [mandrakesoft.com] which may have more up-to-date documents than those on the Mandrakesoft website.

Oh My God.

I'm sorry, but I think we've all just been trolled. I don't believe there's really an attempt to ask a valid question here. This individual has written a couple of perl scripts, and truly believes they will change the world. He hasn't done any research (no mention of CPAN, thinks that FreeBSD does Perl, etc.), and truly believes that a few Perl math routines will change the world. Can you spell "ego trip"?

But just in case I'm wrong, here's what you do: Point your browser to CPAN [cpan.org]. Carefully read the instructions. Submit your scripts. If they're good, they'll get used, you'll make a name for yourself, and will be on the way to The Big Time.

I really can't believe this made /.

Re:learn awk

so let's see:

you wrote 8 apps, in perl, that should take any perl programmer 2 minutes to do (and are so simple, one should use awk for them).

the body of your apps has more text that is dedicated to license and documentation than actual code.

your implementation is crap. i quote from your 'random' man page:
random is slow when dealing with large ranges to randomly find a number from. This is because it creates a list of all potential numbers before picking one. So it can be memory intensive for large ranges.
wtf?? if you were my employee, you'd be fired for writing code like this.

your apps are of very little use to most people. and the ones that do need them likely have just the right variations on what they're doing that they might as well write their own.

i would say your chance of getting included in any big distribution is approximately zero.