Having written open source software myself, and being a subscriber to mailing lists, etc, there is a realization that the number one thing missing from smaller open source projects is feedback from users.Editorial Notice: All opinions are those of the author and not necessarily those of osnews.com
I start with some background: I developed a small open source package to fill a need I had a while back. Wanting to give back to the community, and having a few friends express interest in it, I released it under the GPL and placed it on Sourceforge and Freshmeat. Being familiar with open source since 1991, this was trivial. Not being familiar with the amount of help I would receive in terms of user feedback was something I was not ready for.
Example, someone writes application X, which is open source and used on platform Y. This developer puts their package out and people (more than a few) find it useful. Unless this developer is extremely talented, there comes a point in time that the developer will run our of ideas so to speak. New features and bugfixes are easy to come by in the initial stages of development, but as time goes on it gets more difficult (in some instances) to think up truly useful features. Is it that the software is “complete”? Is it that this is all it was meant to do? In some situations the answer could be yes (e.g. software to check the time. There could be a limited amount of things to do with that 🙂 If the developer knows in the back of their mind that it is not complete, but has a hard time coming up with new features, they would (I think instinctivly) rely on the community feedback to come up with new ideas or catch bugs. This is not how it works out in some cases.
Large corporations have the resources to perform studies and hire many programmers who can share ideas and come up with new features and add ons for software. In an open environment, it is the users and developers that must come up with features, and user feedback will aid in bug tracking. I am not saying that this is not the case in open source software, but it does not seem to happen a lot for smaller packges. For instance, the linux kernel, gnome, kde, X, etc, are extensive packages that get a lot of attention. When one looks at smaller packages, for instance a small spam filter, the number of users might be high, but the discussions in the mailing lists and chat rooms are low, and generally fall into the realm of “how do I get this missing item I need to complie?” or similar questions which can be covered in a FAQ. It is a little disturbing for me as an open source participant to see that in some cases it is difficult for developers to get input.
My example: Looking over my logs, I see that many people download my software (if they use it or not is another thing). Again looking over my email list I see that only three people have evey written in to inform me of a bug (bugs that I missed), and apart from those, only two have suggested improvements to the software. To be clear, I am not here to complain, I am pointing out that I have spoken to others (with more popular software than I have) who claim the same types of response. On asking them how they think about new features, releases, fixes, and add ons, mostly they claim they would like more user input. After all, the end user is the person who will be most familiar with the end workings of the software.
I suppose this rambling is more of a call to the end users of software. You may think that developers do not want to be botherd with emails, chats or messages about bugs or added features and improvements. Indeed, some of them do not. However, I think that a lot of developers would more than welcome creative end user input. After all, it is input from the actuall users of the software that make it better. No number of geeky developers sitting in a meeting room of some large corporation will ever know what the end user really wants, and the information they receive could be filterd down to them through research groups and opinion polls. I am not knocking them, just pointing out (again) that end user input is critical. In my mind, it is the thousands of small items that make open source software great, and not just the “top 20” software packages alone.
Next time you download and use open source software that you feel is useful, and you think of some way to improve it or find a bug, do the community a favour and send a (nice) email/note/IRC msg back to the developer(s). They may appreciate it more than you know.
About the Author:
B.S. in Computer Engineering, M.S. in Physics. Currently
working as a microsoft server babysitter (Systems Admin) and writing software in my spare time. Looking for more research oriented work.
Having been on both sides of the Open Source world–as manager of a large project and as a contributor–probably the most painful thing about Open Source software development is the negative feedback going in either direction. You only need to deal with a few salty comments from OSS project managers or a hoarde of nasty OSS project “contributors” to learn this. Go browse the Trackers on SourceForge to get a feeling for how good/bad this process can be. Some people just seem to enjoy making others feel bad. It’s the fear of having this happen to yourself, whether as a contributor or as a manager, which prevents the feedback loop from working.
That’s just from my own experience of having done OSS development both ways.
Kris
Personally, as someone who has released a couple of projects to the world under the GPL, I like getting comments from users about my apps. It lets me know that at least one person out there is finding what I made useful, and encourages me to release more apps as open source.
As the main developer on Tux Typing, a cross platform SDL based typing game, it is interesting to see the difference between linux users and windows users.
While most of the 130,000 downloads we have had are on the windows platform, most almost all the comments/feedback/… are from linux users. Maybe this is because windows users do not know that the developers of OSS are interested in their opinions and would gladly work on making it better for them!
I agree with the author that feedback is minimal from all users, but I think there is definitely a “cultural” difference between Linux & Windows end users expectations and perspectives.
we are working on simplifying/enhancing our webpages to make it easier for end users (of all platforms) to communicate.
In response to the “bad feedback”, sometimes it is a little frustrating, but I think it is good at the same time. Too much negativity can drag you down, but candy-coating all comments doesn’t help anything!
Jesse – Tux4Kids
http://tuxtype.sf.net
Why not build a feedback mechanism into the software itself?
What, you mean you don’t spend all day reading OSNews/Slashdot/dot.kde.org/www.gnomedesktop.org looking for people bitching about your program? You mean if we want something changed, we should actually *contact* you??? Whoa, news to me!
I think you might get more comments if you set up a mailing list rather than just waiting to receive mails direct to you.
On a list, the other users (once they join the list) see the comments, bug reports, your replies, etc. They are then more likely to add to the list. You build up a community.
Most of the Amiga shareware/free/open programs that I use have their own list, and some of these lists are quite busy, even though the number of users is more likely to be dozens than hundreds.
See there to understand there are not a lot of OSS missing :
http://linuxshop.ru/linuxbegin/win-lin-soft-en/
That’s a good suggestion. If it’s too much trouble to build it in the program, they could at least include an URL in the program. It’s not enough to include an URL in the README text, because noone reads it unless they have a problem 🙂
What project did you contribute that you refer to?
Just curious.
I have contacted authors of OSS software in the past. In every case, it was to thank them for the apps they produced. However, I would be VERY hesitant to suggest new features or improvements. The main reason being, it seems to me that it would be a slap in the face of the developer to say, well this program could use improvement in this area, or I would like to see such and such feature. I know theses developers have given freely of themselves to produce the app, so I would hesitate to intimate to them, good, but not good enough.
As a participating developer of one of the Open Bioinformatics Foundation software suites, a lot of feedback comes back thru the support mail list. Having an IRC channel also helps. I suppose people will make the time to communicate with you if they think it will also get their problem sorted promptly.
Free support for a free product can be a bit hard to provide but the level of support we have had to provide has not been ridiculous and the feedback more than compensates in our case. Your mileage may differ.
In my experience, programmers are very pleased to have suggestions, and even more pleased to get well researched bug reports.
If a suggestion is not practical, the programmer will say so, but there is never any harm in asking (politely).
I’m sure they all like to be thanked, too.
Why peoples absolutelly want to put the commercial pressure on a non-commercial entity ?
Why peoples absolutelly wants to change the OSS for something else of what it’s really all about : a free way for developpers to have fun and learn and spread their accomplishment when they wants to.
Just let it go. I, as a developper, always seen OSS as simple as “take-it-or-leave-it”.
Don’t try to “fix” it, because it’s not broken, it’s just different of what some people expect.
I think you might get more comments if you set up a mailing list rather than just waiting to receive mails direct to you.
It is my experience that the reverse is true. On my main project, I probably get 4 direct emails for every one email to the public mailing list (not counting spam or follow-up messages). This may vary from project to project though.
As an aside, I enjoy recieving email from users – even when negative. In practise, most critisms are highly constructive.
I have worked on a few GPL applications and I can say that feedback and suggestions are great. Even if your suggestion does not make it into the project, it still shapes the direction of the software.
The best places to talk about software or make suggestions are on the mailing lists, bug tracking systems. Eugenia is always in the GNOME forums and is constantly making suggestions. She doesn’t just post her opinions here – and people listen to her. People listen to eveyone.
And what if you don’t get any answers from the developers you want to help?
I’ve given input several times on development but very often tend to get an answer like “do it yourself”… but if I was a developer I’d do it myself.
This is ESPECIALLY the case for GPL apps primarily developed for Linux.
So the best way to get input back would probably just to write that you very much would like suggestions and bug fixes on the page where the download URL is.
I can’t really say I’ve found this on many more places but Mozilla.org which could obviously be the reason why there is so little feedback..
Wake up Guys…you are the “users”
🙂
Maybe, there will be some others later…
R&B
I agree with others who say that if you are an OSS programmer and desire feedback, build a mechanism into the program that would allow a user to do give it to you.
For example, in the Help menu, add an option that says “Provide Feedback” and when the user clicks that option, it would take them to a web form (or whatever).
If more programmers started doing that, I would probably offer more feedback
In Firebird, whenever it blocks a popup, it will place an icon in the status bar. I woudl love for this icon to display the actual popups it blocked when double clicked. To me, that is valid feedback I would like to give, but I have no idea where to send it. Sure, they probably got forums or a newsgroup or mailing list (or whatever), but remember .. users are lazy and usually won’t make that much effort to go and find the ‘feedback forum’, and that especially goes for me too
And do keep this in mind also – if you ask for feedback, be mentally prepared for the feedback you’re going to get, be it positive or negative. If your program is slow, buggy, and sucks shit, I’m going to let you know about it
hehe
PS – To Zeb: Thanks for the link to Linux alterantives for Windows apps. Even though I think they are a little overzealous with their recommendations (Ogle as a replacement for PowerDVD? No way in hell.), there’s still a lot of good info in there!
I agree that the support mailing lists are a good place for feedback/info. Unfortunately I see a lot of them die out…
PS:fewmet which project do you work on? I visti the bioinformatics pages quite a bit.
It’s common knowledge that you should check to see if a bug has already been reported before reporting it. If the developer doesn’t use a publicly searchable bug tracking system like SourceForge (or a web page with a list of known issues), this isn’t possible (unless everyone who uses the software subscribes to the mailing lists, which I never do).
So that’s today’s lesson, quiz tomorrow.
I agree that the support mailing lists are a good place for feedback/info. Unfortunately I see a lot of them die out…
I don’t agree. I mean, how many users do you know are going to take the trouble to subscribe to a mailing list just to say “Hey, this is a cool program!” No .. probably the only time people are going to subscribe is when they actually are having problems and need to ask a question.
Personally, I can’t stand mailing lists. I got on the nano mailing list about a year ago because I needed help compiling it in FreeBSD – since I haven’t actuall taken the time to unsubscribe, I’m stilling getting nano emails. Huge pain in the ass.
I submitted a bug to Mozilla and it was fixed in amonth or two – large inline pngs did not display right until 1.3. I also submitted a feature request for sodipodi – bevel. I would be interested in cocoa or carbon open source applications.
Here’s a suggestion
I mostly do writing, web deign and graphics. One obvious feature would be an open source word processor that didn’t litter the html code with font or style sheet declarations and would use line breaks instead of paragraph breaks. I can’t use anything besides Appleworks for poetry or code it myself, because other WYSIWYG editors double space my poetry.
Why I don’t use open source more
I don’t use X-windows applications on my Mac because all the dependencies are screwed up because I installed UNIX Utilities for OS X from BSD Mall and nothiing works except the X server. I can use Abiword (that I installed from the web), but it is so flaky on the import and export and I can’t cut and paste between X and Aqua. I told them and the Mac store I bought it from as well as another one in my sister’s town.
The #1 thing missing from open source projects is more contributors! Do not ask what Linux can do for you, ask what you can do for Linux!
It’s really hard to get the ball rolling with an opensource project. Even if you get decent exposure, people just download the program and forget about who made it.
Have you ever emailed the author of df? or dmesg?
It’s really hard to get a momentum going with open source since users don’t stick around. Even the ones that are interested seem to show up on the mailing lists every once in awhile and then once bored they move on to other things.
In my experience only the users who end up contributing are likely to stay to allow a community to form. It’s very hard to come across those. You can have all the hits you like on your website,but if there isn’t anybody who is willing to add your project to their list of hobbies your software isn’t gonna form a community around it and it’s gonna eventually get boring and forgotten..That is until the users start wondering why their program isn’t being maintained and actually inquire about it.
Strange world this open source thing.
I would gladly contribute to open source applications.
Seriously, I have so much spare time, (being a student whose just finished college and can’t be arsed getting into a better job than part-time at McShitHole), that I’d love to contribute code and ideas to an establised project.
But you know what… I don’t know a damn thing about coding. C? Ha, yeah, that stuff confuses the hell out of me. Perl? Reading someone elses Perl is like beating my head against a plank of wood.
I’m willing to learn, but I have no idea where to start. The most I know is some PHP(/SQL), and perhaps some BASIC from my younger days, (read: 9 years old).
So, teach me C/C++ or whatever language you’re coding in, and I’ll make you some stuff. Until then, the most I can do is submit the occasional bug report.
Most users just get a different app if the one they are using doesn’t work like they want it to. Linux users are just more trained to think that they are dealing with a person not a big corporation.
Take my experience with GAIM. I thought that it was slow and really hated the way that it put every chat session into the same window. I guess I could have submitted a request for a that feature, but instead I got rid of GAIM and installed Trillian. GAIM also installed a bunch of extra libraries onto my system, so I wasn’t exactly heartbroken to see it go.
What do I want from Linux distributions. The most important thing to me is, that if the computer crashes while I’m using it rebooting will generally let me back into the computer. By that I mean the operating system will still boot into one form or another. Actually I would like to boot back into a GUI however I know that is not possible most of the time because I have a knack of killing the file system. Typically when I crashed Linux I never get past repairing the hard drive errors. I hate that the most. I am noob to Linux; however I have installed Solaris 2.4 on a Sparc 20, with no outside help. I have always wanted to be cool and use Linux full-time. I just can’t get past not easily changing my screen resolution/refresh rate and the whole killing the file system beyond my repairing ability. No way in hell could I ever installed anything into my computer and then find a Linux driver for that device. So I’m trapped in Windows. I am 25 years old and in 1993 I was using a 400MHz Cray at Sandia National Laboratories. The end
I disagree. I think mailing lists suck. I have so much mail already, the last thing I want to do is subscribe to some other mailing list just to get one response for a question or just to submit feedback. Mailing lists are great for developers on a project – you know, the ones that are actually INTERESTED in reading all mail posted to the list. For stuff like feedback or Q&A (tech support), message boards are a LOT more managable for the end user.
If these users can’t be bothered with going to Tools>Preferences>Tabs>”Display all chats in one tabbed window”[uncheck], then good riddance. As for your “trained to think..” comment, I don’t understand what you mean.
Wayne: Hmmm…when was the last time you tried Linux? These things have gotten MUCH better, with journaling filesystems and XRandR. You oughta give it another go…
Nice attitude there Salv.
I must say that I think you are the one making the mistake of confusing open source / free software with comercial proprietary offerings.
Here in the land of free software, two philosophical principals rule.
[1] There is no distinction between developers and users.
[2] Release early and release often.
The logic flows like this…
Because a developer of free software is not typically burdened by the overhead of a commercial enterprise
(hr dept, finance dept, endless process meetings, etc.) he *may* develop his project using the methods of the academy. Specifically , he opens his work to peer based criticism and review. This action gains the very real benefit of reducing the total labor for the developer. How so you ask? Ninety percent of the result of a given project is the consequence of merely ten percent of the total effort.
So then, cobble together a basically functioning program.
Release to community of user/developers for testing/review.
Receive responses from community (criticism and or code).
Integrate feature requests/bug fixes into project.
Release and repeat until:
– program stabilizes
– interest wanes
– project forks
If you do it all yourself, you_are_working_too_hard.
If you disregard the useful input of others, you lose the benefits of peer review and you program suffers.
If you really want to work alone, and don’t want to be *bothered* by others after you are done writing your *perfect* code, then why release your project as free software in the first place?
[remove CAPS to email me]
Say what you like about the technological merits of open source software, but the people who make it usually are selfish, ignorant, fun & power centric, extreamly arrogant, have very poor people skills and poor neurological intuititive abilities from consistantly being on the autistic spectrum, and don’t really care a damn about society at large. When you’ve got people like this making the software, and we do, this is why software is so bad. Having open source only worsens this effect because programmers rather than management have final say on design decisions. If society was different, then open source software would be good, but as it is, open source software is strictly for technical users and uses only.
I’m using Mozilla now, but I know its poorly designed gui & flamer slashscreen would confuddle many in their 40s. Try to get any programmer to see why… and they’ll never understand. A programmer with the people skills & intuition of the average member of the female species is as rare as a 4 leaf clover.
Many times I find that developers do NOT want comments sent to them directly. Most of the sites plea for users to read and post messages to the corresponding newsgroup. Problem is, there is WAY to much garbage and the developers certainly cannot read through all the junk all the time. It’s hard to find someone who is willing to sift through the junk and send only the good stuff back to the developers. This is where commercial software has an edge.
The #1 thing missing from open source projects is more contributors! Do not ask what Linux can do for you, ask what you can do for Linux!
LoL, you make it sound like Linux is the only OSS software out there. I can’t do much for Linux,it’s pretty much heading down the drain itself, and I can’t see anything stopping that.
What I can do for OBOS and BSD or Mozilla is a much better question, they don’t have a philosophy built on bloat and they have a promising future.
The #1 thing missing is not user feedback. But developer response to user feedback.
I am still waiting for so much as an acknowledgment of the bug report I filed with Mozilla. That was probably about a year ago. As far as I know, no one has even looked at it.
And then you wonder why businesess don’t take open source seriously…
yeahyeah:
You had a few valid points, but I just couldn’t ignore this:
Having open source only worsens this effect because programmers rather than management have final say on design decisions.
So you’re saying management is the answer[1]?
GG
——————————————-
[1]BWAAAAAAAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHA!!!!!!!!!!!!!!
I see so many similar projects on sourceforge and freshmeat. Like Unreal mods that look like brand new games, apps should be modularized so you can add functionality to an existing app instead of creating your own. Like plugins on steroids.