Fog Creek Software
Discussion Board




running project on sourceforge.net

Please excuse me if you think that this question should not be posted here.

I have a part of research project which I decided to make open source. Two monthes ago I published it on http://sourceforge.net and announced on freshmeat. This is a simple but nice tool (at least on my opinion) for c++ developers.

So far, statistics show more then 100 downloads. The problem is that I get no feedback from people. Nothing. No bug reports, no problem reports... Just nothing.

I would like to ask if this is normal? Does anyone have the same experience?

dk
Friday, November 29, 2002

I'm not running on SF myself, but so far my little pet project has received 2 emails from people. One was a diff, one was a bug report. The two worked out well.

I found this surprising at first also, but then I thought about how much I send to authors. I maymake it a personal habit to send a ``thank you'' to OSS authors, as I know I appreciate it.

I think it's normal not to generate a big buzz, because not many products are new: mine is a command-line cd/file catalogger: there are graphical ones out there that most users will prefer, and I expect little.

Most projects, I think, are like that: one or two fill the niche, and the others mostly remain obscure. That's OK though: I wrote my software for me, because I needed something that did XYZ. I'd like others to use it, but if they don't, it's no big deal because I never meant for it to be some new trend.

Just curious: what have you written?

Mike Swieton
Friday, November 29, 2002

This is shameless self promotion, but I can't resist: dgd.sf.net

dk
Friday, November 29, 2002

I am running a free project since 2000. I hosted it in 2001 on SF.
In almost 3 years, I received about 30 emails. Most of them were saying "Hi, great piece of code! Do you plan to support feature X or Y?"
I received only one bug report and had a total of 3000 downloads. So your statistics are not alarming, they are just normal.

I had a look on the web site of your project. It seems to be a "highly specialized" software. So the audience is very limited. You cannot have as much downloads as other star open source applications.

My advice is to continue its development and to ignore statistics.

XRD
Friday, November 29, 2002

First, you of course have all read Cathedral and the Bazaar. I had 3 copes of that book , and due to lending it out, I now only have one copy left (and that is lent out also!!).

Anyway…read the book…I am sure you all have. The book explains EXACLITY WHAT motivates people to contribute to project (a combination of interest, and also ones ego).

There are several important things that will help a open source project.

#1, the existing functionally of the product must be of good use. If the product has too limited in terms of functionally, then no uses will care to use the product.

#2, to get any kind of real feedback, the product must attract developers. So, a program that manages your cooking recipes might attract some downloads, but few developers are really into cooking (on the other hand, I did cook a fabulous plate of pasta tonight!). Hence, if the product is not used by a developer type of person, then you will not get much back in the way of developers will you?

For example, a lot of web developers use Apache, and thus it is somewhat easier to get them to give you feedback. If every single user of Apache was JUST an end user, then no code updates will occur. However, many users of Apache are also web masters with a lot of coding experience. The more the product is useful to developers, the more likely you can get developers on board.

#3, The product is a obvious way to reducing licensing costs, or replace an existing commercial product. This type of product also carries some prestige to the developers, as success with this type of product can make you famous. Samba is real cool in this way, since it reversed engineered windows networking, and lets you now use a Linux box as a file server to a windows network. Samba has to got to be the #1 thing that MS hates right now!, and as result seems to be the #1 thing that open source people love!. People will contribute to products like this just for the principal of the matter. (ie: getting rid of a very strategic commercial product, and replacing it with a free one!).  Of course windows networking is free, but Samba just means I don’t have to purchase a windows server. I see this product cropping up every where these days, and can even smell the joy those programmers have for this product.

#4, There must be clear path, or idea that much more can be done with the product. As each step, or large goal falls, then the developers will loose interest.  Now that Samba works, and  a great leap/benefit was created for the general community, , many developers will loose interest and move on. Hence, what are the possibilities, and results if I as a developer contribute?

This list could get a lot longer, but open source is not so simple.

In addition, it is very important to realize the good developers are hard to come by, and they may already be involved in other projects, or sitting here writing this post in place of working on your project!

In other words, there is competition for developers in open source just like the commercial market, and they are a limited resource.

I am always amazed how people think that open source is a un-limited resource. It is nothing close, and developer time is hard to come by.

Albert D. Kallal
Edmonton, Alberta Canada
Kallal@msn.com

Albert D. Kallal
Saturday, November 30, 2002

I agree with everything said thus far, especially the point about realizing how rarely any of us submit feedback on the numerous bits of software we all try out. There just isn't enough time in the day.

I've been on both sides of the fence on this issue. I've setup two projects on SF, and while one of is somewhat popular (13k d/l in 19 months + 220 forum posts), the other died a miserable death. And, from a technical perspective, I always felt the latter project was more impressive. Go figure.

One recommendation: make a sample of some kind available and *quick/easy* to get up and running. Too often I find something on SF that, while it sounds interesting, takes far too long to get working on my machine just to see what it really does. Most people aren't going to spend the time. And if any level of frustration arrises from getting the code to do what it's suppose to, I'll likely just shut-down the environment, delete the downloaded files and move on with life. I've wasted enough time as it is, so spending more to send an email isn't likely to happen. If a sample isn't applicable or possible (rare, I think), then at least some screenshots of some kind.

Documentation is almost always lacking or non-existant in OSS projects, so the author has to do *something* to make it easier for people to check out and use their work.

Anyway, just my two bits thrown in. The most important thing to keep in mind, I think, is to just write the software for yourself - for your own needs. If you want to make it available to the rest of the world, that's great. If you expect to be crowned the new "King of OSS" overnight, well...that may be a bit unlikely.

Ryan LaNeve
Saturday, November 30, 2002

dk,
I wouldn't worry, they'll let themselves be known if something does not work :D

I also host a project on sf.net and after 4500 downloads I have seen about 20 direct emails either requesting features, bug fixes, or wanting to help. My project is also very niche oriented and does not pull the attention that large projects get. But as it and the technology (C#) matures it does seem to be gaining momentum.

I once complained about this very subject on /. and got a response from the creator of a specific linux build with over one million downloads and his response was (in so many words) it's normal and get used to it, lol.

Ian Stallings
Saturday, November 30, 2002

Just be thankful that you aren't getting any hate-mail.

I used to be active on the Xvoice project (voice recognition on Linux, aimed primarily at people who absolutely couldn't type, and were willing to put in some effort up front to get voice-reco in a Unix environment, so they could work & earn a living), there was a very active core of users submitting regular bug reports, all great, but because the software sat on top of ViaVoice for Linux, and had a few other dependencies, it could be tricky to get going.

Its hard enough staying motivated to work on OSS when you're trying to fit it in between work/study, and that's made a lot harder when every evening some script kiddies who expected god knows what from the software flamed me and whoever else was associated with the project because they couldn't install ViaVoice, or because when they did get everything going it wasn't as "cool" as they thought it would be.

Of course, for every irate passer-by there was someone who stopped, got the point, smiled and took a brochure, or occasionally became part of the core user group.

I guess the point is that you shouldn't be worried about lack of response. I kept working on Xvoice for as long as I did because I thought it was going to solve a problem - giving access to Unix systems to people who otherwise would not have had it. The _bad_ part was that the project had this "cool" factor, attributable to voice recognition, that attracted a lot of people to try it out, people who for the most part were not the users that the app was really for.

For the time being, be thankful that its all quiet. Having one real user supplying real feedback on usability - identifying problems and desirable features because they are actually going to continue to use the app and really want them, is much better than a having hundreds of the type of people who trawl the net all day looking for "cool" software to stem the tide of boredom, who download, run, flame, and move on. Beware, because they exist, they're out there, and they WILL piss you off.

:)

Dooris
Thursday, December 19, 2002

*  Recent Topics

*  Fog Creek Home