Hacker News new | past | comments | ask | show | jobs | submit login
Top ten reasons why I won't use your open source project (thechangelog.com)
31 points by netherland on Jan 31, 2011 | hide | past | web | favorite | 35 comments

My, that comes across as a bit preachy. You get something for free. So what if it doesn't have a twitter account associated with it.

The important things are 1) does it do what I need 2) is there a good community and 3) is the licensing clear, and compatible with my needs.

It's great feedback for people who are interested in attracting more people to their project, which is clearly (to me, at least) the manner in which it's intended.

It's pretty good marketing advice for many things. "How to market your open source project" might be a nice, less linkbait title. I guess it worked, though:-/

Perhaps but he's preaching against the very attitude you have which is "you're getting this for free so you'll take what you get and like it"

If you're someone writing an open source project you're wasting your time if it goes completely unused. So even as an open source developer you're still trying to sell your project to people. You're just giving it to them at a VERY reasonable price.

So what he's saying is all the same "selling a product" rules apply even if it is open source. You can't just throw it out there and expect people to use it

> you're getting this for free so you'll take what you get and like it

That's putting words into my keyboard. I think most of it is reasonably good marketing advice, and some of the items listed are horrible reasons to not use something.

The title is the problem, really. If it were something like "how to help people find your open source project" it would all fit into place much better.

You are not wasting your time if you learn something.

Not intended that way, in fact, that's the purpose of the article - how to build that community. Even if the code is good, it takes time and there are practical steps you can take.

#1,2, and 7: Yes.

The others: not so much. When it matures (jQuery) perhaps worrying about branding is in order. But if you are spending your hours making a tool for me to freely use, I'm ok if your design isn't so pretty. And if you don't want to spend much time on Twitter. And ...

I don't think the issue is as much about design as it is about seriousness. Having a homepage for the project, spending some time or money on design and reaching out to the community shows you're committed to the project which will get others to commit. That's the best insurance a user can get against the project being completely abandoned one day.

As someone whose tried to use open source whenever possible I can tell you there's no feeling worse than using an open source component that's been left for dead

I can tell you there's no feeling worse than using an open source component that's been left for dead

I'd say using a commercial component that's been left for dead is much worse.

It's a good list for "What you should be doing to call attention to your open-source project (when it's ready for prime-time)."

The current title reads like: "this is what you owe me if you want me to look at you." Well, ok, but I don't feel that way. I'll use smaller, humbler projects. But I do agree that in doing so I must be prepared for the day that it is no longer updated and the bugs aren't being fixed. I'd certainly feel more comfortable with projects that met most of the lists' criteria, and would be willing to bet more on them for the long haul. But I would still use smaller projects.

Some things (couchdb) seem riskier than others (php), and some very small projects (php-on-couch) seem quite risky indeed. However, I could pick up and go if php-on-couch stalled, I could probably manage if couchdb stalled, but I would be in trouble if php did.

> I don't think the issue is as much about design as it is about seriousness.

You get what you pay for. And your best insurance against the project being completely abandoned is to have the source code.

6. You don't have a Twitter Account

Seriously? This would never in a million years have crossed my mind as a reason to avoid using something.

Depends on the product - this pitch is aimed at 'cool new web framework' type opensource projects.

If I'm looking for a heavy duty technical library, say 3D mesh generation or image processing I want someone serious behind - preferably with a full time academic job in the field and some papers, more than a twitter account and a cool logo.

Not a reason not to use. A reason I might have found you in the first place.

Perhaps, but the title is specifically "won't use".

If you don't use an open source project because they don't have a Twitter page or they don't have a domain name or gasp they don't reach out to you then you are going to miss out on really solid pieces of work. I don't think marketing prowess is something that should impede you from using these projects. On a lighter note I am imagining Stallman reading this.

Although the headline kind of loads the answer, and sets the whole tone at a preachy level, his feedback is good - if you want to drive attention/usage of your open source project, you have to put in a minimal amount of marketing.

If that's not something you're worrying about, ignore the article.

Ready for linkbait, I read the article and...it's good!

If open source figures into you company's strategy, as it does more often these days, you need to do it professionally. Yes, even if it is free.

Great article, dodgy headline. Altogether though, some solid advice. My project(s) meet a number of those criteria, but fall short on a couple of them. So you've given me some food for thought, regarding things I can do to better market my own stuff.

I disagree that GitHub's project pages "suck for SEO". GitHub's pages rank quite highly on Google. A search for "cancan" for example returns my project as the first result. A search for "homebrew" returns GitHub's page as the 3rd result and the custom project page as the 4th result.

Also, creating a separate page for a project means there will no longer be one page for people to link to. Links will be split across the two pages, and links are important for SEO.

I agree a separate project page is great for branding and makes it look like a dedicated, serious project; but I wouldn't do it for SEO purposes.

Otherwise thanks for the great list of ideas.

Thanks for the comment, Ryan. I think you make my point. If I know the project is called 'cancan,' you're right GitHub results are favored. But when I search for 'ruby authorization,' the only CanCan hit is a blog post from someone else giving an overview of the project.

Love CanCan btw.

Thanks for the clarification. I misunderstood your post as saying GitHub's project pages don't rank well in Google in general. I agree that a custom project page allows you to better target other keywords (such as "authorization") for SEO.

Add: your source has so many accidental dependencies that its agony to extract the bit I need.

Actually, I do use open source, even so. Its just lots harder than it needs to be.

To elaborate: if your IP is exhibited with an app, please, please put that IP into a lib, or a separate API. Do NOT blend the app with the lib e.g. do not mix in dialogs, command line stuff or even platform calls if possible.

I do embedded work a lot, and it doesn't have any of that, or its so different as to be unrecognizable.

IP examples: encryption, protocols, file format interpreters, translators. I (and very probably others) need your magic code 'naked'.

I have been elucubrating on the topic for a while; I think some of your points are not so important (a pretty homepage will be needed once the community is quite large, not for a just-started project).

By the way, my thoughts on community projects are here:


it's a multipart post, not yet finished.

We think Wynn's points are good points, we (the Nuxeo project) have been using similar ideas to guide our community effort.

See: http://blogs.nuxeo.com/fermigier/2011/02/top-10-reasons-for-... for more details.

…you used HAML

Good article. Does having a full documentation substitute the readme file?




I am glad, my open source project doesn't have any of those issues...

Wow, that post is just wrong.

Open source is not a product. Open source is a community effort. If you don't want to use it - well, it's your problem. No one will shed a tear about it.

If you actually would care about open source you wouldn't be making shitty top10 lists. You would get your ass up and contribute to the project of your choice. If you think twitter is important ... fine! Start a twitter account for that project. Draw fancy icons. Do "SEO". Do something to help. And do something to get rid of that false sense of entitlement.

I know Wynn and his sense of humor, and I can see how easy it might be to totally miss his comedic tone with the loaded title and seemingly "preechy" top 10 list. When/if you meet him, you will know what I mean after a few minutes into a good chat with him.

I will say though, that Wynn does contribute to Open Source - we all do at The Changelog. He also did apologize in the opening sentence of the first paragraph for the "loaded headline" as a hat tip to Mashable for an article they posted two years ago titled, "FOLLOW FAIL: The Top 10 Reasons I Will Not Follow You in Return on Twitter". It's Wynn's character to reference and relate, but I can see how that could be missed. It's hard to "hear" someone's tone in plain text, especially when you are quickly scanning a top 10 list.

This post was more about concrete ways to build community around your code and things you can do to promote your open source projects.

Wynn is one of the nicest guys around and is very passionate about OSS. He's the polar opposite of entitled or preachy - he writes, codes, interviews, and speaks regularly. Perhaps his linkbait parody missed....

Yeah, it did miss. He might have good points when you understand he's just using that style as a gag, but after a few points, I just gave up.

"Open source is not a product. Open source is a community effort. If you don't want to use it - well, it's your problem. No one will shed a tear about it."

Bullshit. Open source or not, a product is a product. If nobody uses it, you've waisted your time (and the time of the other contributors) building it.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact