Why use downloads as an indicator of OSS success?

re: Why use downloads as an indicator of OSS success?

wrowe — Thu, 21 Feb 2008 08:35:54 GMT

Yes, downloads are an entirely useless metric.

Open source is ... open. Others are free to mirror it, and frequently do.

So the total number of downloads is usually vastly understated ;-)

re: Why use downloads as an indicator of OSS success?

Fluke — Wed, 20 Feb 2008 21:44:45 GMT

SourceForge (not to be confused with Microsoft “Source Fource”) seems to address several of your comments. If knowing the total number of downloads is of limited usefulness then project page views is even less helpful. Yet, all that Codeplex seems to provide is page views for the last 7, 30 and “all” days. SourceForge provides graphs to check if it is “steady/growing/declining.” If it is conversations that matter then the mature interface for mailing lists and forums that SourceForge provides is critical. While Codeplex does provide for “comments/discussions,” it's interface not as polished.

-----

To Drive: Henaway appears to be at a Microsoft “Open Source” site to provide feedback. The front page of Port 25 claims that “[Port 25] represents an open conversation dedicated Linux, Windows and open source interoperability.” If instead MS would prefer the community to “chill” such that those which discontinue to provide feedback won't be missed then the mission statement of Port 25 is a lie.

At the heart of Henaway's feedback is that there exist people that are upset with the misconceptions Microsoft claims about OSS. The misconceptions he complains about is the type provided by “Get the Fact.” But another popular Microsoft promoted misconception is that “MS OSS is OSI approved.” Not everything that MS claims to be OSS honors the Open Source Defination provided by the OSI. Instead, the OSI only approved *two* licenses. That means so-called “MS OSS” covered by non-OSI approved licenses are *NOT* OSI approved even if MS chooses to call it “MS OSS.” Bill Hilf has explained that microsoft.com/opensource “clearly outlines Microsoft's position on OSS.” Part of that clear outline is a document stating that MSCompBio covered by the MSR-LA license is part of “MS Open Source.” The OSI has not approved MSR-LA as open source and the MSR-LA violates requirements of the Open Source Definition. Back in December Jamie Cannon of Port 25 fame had acknowledged the problem needed to be reviewed. The document remains today to illustrate that the *clear outline of Microsoft position* regarding OSS is to *provide misconceptions and lies*. Hence, rather than give Microsoft a pat on the bad for it's empty rhetoric, we continue to use the “open conversation” forum to bring up that Microsoft has fallen far short of being a F/OSS friendly company.

re: Why use downloads as an indicator of OSS success?

drive — Tue, 19 Feb 2008 05:48:04 GMT

Henaway, chill dude - learn Linux and ruby or Mac and ObjectiveC - What in the heck are you doing at a Microsoft Open Source site? And where in the world did the OP even mention Linux (you ranted this: "I realize that working for Microsoft it's your job to tell us how Windows is the best, and all this open source/linux talk is all talk with no walk") I'm sure MS won't miss you (and did you even read the blog?). MS OSS is OSI approved (see www.opensource.org/.../207). I've got the code for more than a dozen or so great projects that are public open sourced from them. As far as downloads being an indicator of success, sounds like a reasonable way to do it. I don't want to phone home to someone's basement when I install their software. It may be tolerable for Adobe, Apple, Microsoft, and other established vendors to "activate" or "update", though I'm not happy about it. So short of having every piece of OSS phoning home, downloads seems to be a reasonable measure.

re: Why use downloads as an indicator of OSS success?

mtz — Tue, 19 Feb 2008 03:02:10 GMT

..how many different ways can an open source project track its popularity? most companies will quote sales figures to show how popular their products are. how often are their products used after they are purchase? who cares as long as they keep on buying ..its the same thing with the open source ..the only way tp track popularity is from downloads and bug reports .. do you have any better ideas?

re: Why use downloads as an indicator of OSS success?

Joe — Mon, 18 Feb 2008 03:56:20 GMT

You've defined the problem in a way that yields only one answer. But it is a misleading answer.

You say "by success I mean sales". By that measure, downloads as well as use of the free-as-in-beer software are worthless to the distributor. But there are other sources of revenue, like advertising and services. The Mozilla Corporation earns substantial fees from Google for driving search queries to Google. In any situation where revenue is proportional to the number of users (as it is for Firefox) the number of downloads <i>does</i> correlate with dollars. Granted, there aren't that many good examples of this, but it does occur.

re: Why use downloads as an indicator of OSS success?

Henaway — Fri, 15 Feb 2008 23:10:32 GMT

Can't take the heat of letting through any comments that don't paint Microsoft in a glowing, Godly light, eh? Why am I not shocked?

re: Why use downloads as an indicator of OSS success?

Henaway — Fri, 15 Feb 2008 03:51:31 GMT

Sorry, but I have several computers that Microsoft will tout as "sales" and tell the world how because I was forced into a license for Windows that it MUST be the world's greatest everything... but I take ONE DOWNLOAD, burn it to a CD, and install it on several of MY OWN computers, as well as some family and friends' machines ... so no, the download metic isn't necessarily good because 3 sales may equal ZERO used products, but ONE download could equal 7 or 8 installations and used products.

I realize that working for Microsoft it's your job to tell us how Windows is the best, and all this open source/linux talk is all talk with no walk, but there are many out there like me who are SICK of the way Microsoft abuses us and have found a BETTER alternative that lets me do the things my computer is capable of doing without limits.