You know, sometimes I just don't understand people. One of the other core Fink administrators got this e-mail a month or so back:
On Jun 15, 2005, at 6:06 PM, Dave Vasilevsky wrote: [misc fixes] Tell me how it goes! On Jun 15, 2005, at 2:47 AM, moldy icon wrote: it built and installed fine. Thank You very much. maybe this is some kind of general fink problem rather than anything to do with your package specifically because my fink is offering me that same 0.0.11-11 version of all the ggz packages... using FinkCommander. it can't be affecting just me, can it? I've done scanpackages, cleanup, and index several times. I wonder what else is not being shown? here's another unrelated question: why do the fink people burden us by forcing us to build almost every unstable (and stable) package from scratch? The i386 people using linux never have to build unstable packages. They never have to build anything... there's ALWAYS a binary of everything for them! Every fink package maintainer has created a binary... just to test their package. Why force us to build it? We don't have much diversity in the Mac OS X world... fink could easily separate G3 G4 G5 the same way it identifies darwin/x86. I think the fink bigwigs (especially "mister whacko elitist" ranger rick) are deliberately punishing us. Imagine the incredible CPU cycles they're costing the Mac OS X community making us build packages for which they could easily provide binary debs. What's your theory? For example, did they tell you NOT to give us a binary? Are you blocked from this somehow? I'm just wondering. I think I'll post this question to the group. Wanna bet they avoid the question and pretend this issue isn't an incredibly obvious deliberately-created annoyance we're not allowed to discuss? moldy icon
Any of you who know me know that I do my damnedest to help people and to not be "elitist" (although I can be a whacko sometimes <grin>).
Every fink package maintainer has created a binary... just to test their package. Why force us to build it?
Because while we have a huge user base, our maintainership is a tiny fraction of the size of other linux distros, we just plain don't have the resources to make it happen. We don't have a *secure* place to build binaries. Even if we did, if I built binaries on my development box, they'd link against stuff that's not even released yet, and so would be useless to you, the user.
It's not that we haven't thought of this stuff. We would love for there to be an unstable binary dist, but no one has the time or resources to make it happen (yet). I know there's been work to do so, but it's slow going.
The most frustrating thing in open-source is that while a ton of work goes into making things happen, there are a lot of people who only care about what they want out of it, without giving any regard to the people that made it happen. They are totally ungrateful for the time that has been donated towards making useful things for other people. It's seeing letters like this that make me completely sympathize with people who say "screw it!" and just give up.
I'm not going to give up just yet, though. There are plenty of people who do appreciate the work that goes into making any open-source project happen, and they're the ones who make it worthwhile.
So if you made it all the way to the end of this rant, please, think about the open-source software you use right at this moment, and send an e-mail to those responsible for it, telling them how much you appreciate it. I assure you, it'll do more than just make their day.
Share on Facebook
As a user-level example of Fink binaries that people can’t use: I have a build computer, and I grab binaries from there for my laptop. Do I publicize this? No. Why not?
1) I’m using X.org as my X11 environment and it produces binaries that are incompatible with earlier X11 flavors. Since the project’s build process is set up assuming the an early X11 version, which doesn’t result in problems for later ones, there hasn’t been a real need to create separate trees for the different X11s. Because of this the package manager doesn’t distinguish what X11 flavor was installed when a package was built, so people who are using Apple’s X11, say, can’t use my packages.
2) My server is _my_ server, and therefore I may inadvertently make changes that could create problems for anybody who uses it. This actually happened–someone had my .deb repository still in their list, but I had upgraded to Tiger without changing the URI structure and they were still on Panther.
Some measure of quality control is absolutely essential for distributing packages as binaries, particularly when, like Fink, the packages aren’t self-contained, but rely on other libraries being present.
Heh. Try to make a living on open source.
Actually, now that I think about it, it isn’t as bad as being a total volunteer. We have a small group of paying customers who obviously appreciate what we do. We can focus on them and ignore those that say “it does not work”.
As a mac user, I really appreciate what you folks at fink do. I take it for granted, actually. “fink install x” usually JustWorks(tm), and I think the OpenNMS install on fink is one of the easiest.
Anyway, stop when it stops being fun, but don’t let the lusers get to you.
If you need hardware and Internet connections to host build servers, I imagine there’s someone out there who could donate them. Perhaps try asking?
Servers aren’t our problem, there are enough people willing to donate servers. The problem is the infrastructure required to auto-build packages. msachs has done a lot of work in that direction, but we’re not ready to build a bindist from it yet.
I have been involved with Fink since nearly the beginning, first as an end user and now as a package maintainer and I can tell you that with very few exceptions, every fink developer I’ve interracted with online has been far from elist, rather they are willing to help and answer questions. They also take suggestions as well, if those suggestions are reasonable. Debian has a hundreds of developers and contributors, while fink has only a handful. Maintaining the old pile of kludge that is the fink core code can be a trial sometimes but they do it because they love fink and the community it serves.
I really don’t know how you people do it. I read the letter from the guy who gave up, and it described all of my questions about open source committers: how can you devote the time and effort that is needed to do what you do when there is no guarantee of success or appreciation? Yet, still, the open source community moves onward. I take my hat off to you.
I use fink and DarwinPorts for a few things, but what I really want is still unstable. Everytime I check to see if it has changed, I sigh when it has not. I am mainly frustrated at the fact that I want that software, but I don’t have the technical ability to contribute myself. I have been writing code since I was in the 7th grade, and people tell me that I am very good with it, but I don’t know anything about this sort of work. I would love to work on the Darwine project or in getting KOffice on Tiger, but I don’t understand enough about how the process works, or how UNIX works to be of much use. I know that I am not the only one who feels useless in this respect, there are many people, including some of my friends, who want to make this all work, but they don’t have the technical ability or they just don’t know what they can do to help.
Also, people who whine about open source incompleteness are the same people who sign up for experimental medical tests, then try to sue everyone involved when something unexpected happens (hence the idea of “experimental”), or they are the type who knowingly push something far beyond the capacity for which it was designed and then complain when the device fails. There will always be people (read idiots) who set unreasonably high standards, fuss about any flaw, real or imagined, and complain about the results.
I have been annoyed a few times when a fink or darwin ports build fails, but then I also understand that it is still in the unstable tree and this is something that has to be taken into consideration. Most people that I know understand this as well. I also believe that if you are not going to do anything to remedy the situation then you have no right to complain. So, the build fails, I don’t know how to fix it, I clean up after it and move on.
Anyway, I have a feeling that most of the people who utilize fink understand and appreciate what is being done and that there are probably many of us who truly would contribute if we knew how.
I for one greatly appreciate the fink project and all the good people who manage to devote so much time for the benefit of the rest of us. Thank you all.
I think that all of the maintainers and documentarians and all of the rest of the Fink team are incredible. I love open source and I love the entire concept. If I understood how porting and maintaining worked I would volunteer but I am a lowly user. Sure, some days I wonder why certain packages don’t just work or why I don’t have access to stable binaries for all of my needs. Sometimes when I update and a certain package just stops working and I have to work for days or even weeks to get it back up I get upset but I never am upset with the Fink team. The patience and tolerance that all of you (even the “elite” members) show for the incessant questions, many over and over the same ground, is an example of that we can all benefit from.
BTW Ranger Rick, how is the stable version of KDE3.4 coming? I appreciate all of the effort that you put forth. Thank you, thanks to all of you.
Hi
Where are the native binaries ??
You are not working on them anymore ! More than one year without updates ! You are working on x86 while the first Mactel are still not available !
Maybe someone will do it anyway …. Since we got a native Aqua port of scribus ….
Dave, please tell me you’re joking.
Two big reasons for user frustration that won’t go away in a hurry.
What people don’t get about OSS is that since it’s voluntary, programmers do as they seem fit. This can frustrate a user who’s waiting for X when programmer decides to do Y first. But as with the whole “free music” controversy, if you really want it all and for free, buy a guitar… OSS is great, but developers are also Free as in Free to make up their own mind about things.
And another thing the casual user doesn’t get is: the people communicating about the products are the programmers themselves. Not marketing or management. Which is refreshing, but needs its own brand of translation. If a developer runs buck-naked down the street screaming “Eureka, fetch me a towel” (or puts “it works, it works” on his blog) that doesn’t mean the product is finished. It only means that this developer has succeeded to make something run in a way that makes him confident the task of porting/developing a real stable product is FEASIBLE. He sees the light and the end of the tunnel. But isn’t there yet. Equally, if a programmer claims “the hard work is done” you must translate to-o-o…. yes, you got it: “getting it all working will take a long time indeed”.
So a casual user of OSS should know to translate “it works” to “one day maybe I’ll also be able to install and use this”.
And not be too put down by this. This is honest communication by dedicated people. Developers are not lying or grandstanding, they are just saying it as they see it.
It’s us users who have the choice of learning to “speaka da language” and using OSS or going back to the shrink-wrap experience of commercial software – where marketing and sales are the main communicators and where we are already trained to translate “insanely great” to “it works”.