<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"
	xmlns:content="http://purl.org/rss/1.0/modules/content/"
	xmlns:dc="http://purl.org/dc/elements/1.1/"
	xmlns:atom="http://www.w3.org/2005/Atom"
	xmlns:sy="http://purl.org/rss/1.0/modules/syndication/"
	xmlns:georss="http://www.georss.org/georss" xmlns:geo="http://www.w3.org/2003/01/geo/wgs84_pos#" xmlns:media="http://search.yahoo.com/mrss/"
		>
<channel>
	<title>Comments on: Equilibrium in free software testing</title>
	<atom:link href="http://mdzlog.alcor.net/2009/04/10/equilibrium-in-free-software-testing/feed/" rel="self" type="application/rss+xml" />
	<link>http://mdzlog.alcor.net/2009/04/10/equilibrium-in-free-software-testing/</link>
	<description>a potpourri of mirth and madness</description>
	<lastBuildDate>Mon, 23 Jan 2012 03:06:44 +0000</lastBuildDate>
	<sy:updatePeriod>hourly</sy:updatePeriod>
	<sy:updateFrequency>1</sy:updateFrequency>
	<generator>http://wordpress.com/</generator>
	<item>
		<title>By: Jason</title>
		<link>http://mdzlog.alcor.net/2009/04/10/equilibrium-in-free-software-testing/#comment-558</link>
		<dc:creator><![CDATA[Jason]]></dc:creator>
		<pubDate>Thu, 16 Apr 2009 13:16:09 +0000</pubDate>
		<guid isPermaLink="false">http://mdzlog.wordpress.com/?p=282#comment-558</guid>
		<description><![CDATA[In many cases there is a lack enough information in the report, and a request for more information also goes unanswered. The outcome - the effort of &lt;a href=&quot;http://www.qainfotech.com/&quot; rel=&quot;nofollow&quot;&gt;software testing&lt;/a&gt; goes in vain. Any one with a practical suggestion as how can we overcome this?]]></description>
		<content:encoded><![CDATA[<p>In many cases there is a lack enough information in the report, and a request for more information also goes unanswered. The outcome &#8211; the effort of <a href="http://www.qainfotech.com/" rel="nofollow">software testing</a> goes in vain. Any one with a practical suggestion as how can we overcome this?</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: mdz</title>
		<link>http://mdzlog.alcor.net/2009/04/10/equilibrium-in-free-software-testing/#comment-554</link>
		<dc:creator><![CDATA[mdz]]></dc:creator>
		<pubDate>Tue, 14 Apr 2009 16:42:02 +0000</pubDate>
		<guid isPermaLink="false">http://mdzlog.wordpress.com/?p=282#comment-554</guid>
		<description><![CDATA[Thanks for your feedback.

1) This is about the sum total, and in point of fact we are moving toward a more unified way of organizing the package repository.  However, note that (in contrast to some similar projects), Ubuntu main has *always* been open to the whole community, not just Canonical.  There is a higher bar to gain privileges to work in main, but anyone who meets the criteria is welcome to develop in main.

2) Metrics like this don&#039;t work very well for us because of the relationship we have with Debian.  Primary maintenance for most of the packages in Ubuntu is provided by Debian, so there isn&#039;t a meaningful package/ubuntu-maintainer ratio.  Bug reports per developer is a more useful metric, and (notably) the number of bug reports scales with the number of users, while the developer base grows more slowly.

3) We started off with nearly zero barrier for bug reporting, so I don&#039;t think we could possibly have dropped it lower. :-)  For us, it&#039;s a key part of being an open community that we accept bug reports from anyone.  I wouldn&#039;t want to sacrifice that for pragmatic reasons, and if we can filter and prioritize them appropriately, we shouldn&#039;t need to.

4) One of the problems is that we don&#039;t have this data for most of our bug reports.  We&#039;re currently promoting the use of tools which help us by collecting that data (hence http://mdzlog.wordpress.com/2009/03/31/please-dont-report-ubuntu-bugs-directly-to-launchpad/), and over the next few months, we should get enough data to do more analysis of this type.

4a) Yes.  The bzr work will make this process more efficient, but even more importantly, it lowers the barrier to entry dramatically for new contributors.  Patches attached to bugs require much more developer time to review and merge, and aren&#039;t tracked as effectively as branches are.  Bazaar branches have their own review mechanism which is more convenient than using bug comments.]]></description>
		<content:encoded><![CDATA[<p>Thanks for your feedback.</p>
<p>1) This is about the sum total, and in point of fact we are moving toward a more unified way of organizing the package repository.  However, note that (in contrast to some similar projects), Ubuntu main has *always* been open to the whole community, not just Canonical.  There is a higher bar to gain privileges to work in main, but anyone who meets the criteria is welcome to develop in main.</p>
<p>2) Metrics like this don&#8217;t work very well for us because of the relationship we have with Debian.  Primary maintenance for most of the packages in Ubuntu is provided by Debian, so there isn&#8217;t a meaningful package/ubuntu-maintainer ratio.  Bug reports per developer is a more useful metric, and (notably) the number of bug reports scales with the number of users, while the developer base grows more slowly.</p>
<p>3) We started off with nearly zero barrier for bug reporting, so I don&#8217;t think we could possibly have dropped it lower. :-)  For us, it&#8217;s a key part of being an open community that we accept bug reports from anyone.  I wouldn&#8217;t want to sacrifice that for pragmatic reasons, and if we can filter and prioritize them appropriately, we shouldn&#8217;t need to.</p>
<p>4) One of the problems is that we don&#8217;t have this data for most of our bug reports.  We&#8217;re currently promoting the use of tools which help us by collecting that data (hence <a href="http://mdzlog.wordpress.com/2009/03/31/please-dont-report-ubuntu-bugs-directly-to-launchpad/" rel="nofollow">http://mdzlog.wordpress.com/2009/03/31/please-dont-report-ubuntu-bugs-directly-to-launchpad/</a>), and over the next few months, we should get enough data to do more analysis of this type.</p>
<p>4a) Yes.  The bzr work will make this process more efficient, but even more importantly, it lowers the barrier to entry dramatically for new contributors.  Patches attached to bugs require much more developer time to review and merge, and aren&#8217;t tracked as effectively as branches are.  Bazaar branches have their own review mechanism which is more convenient than using bug comments.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: mdz</title>
		<link>http://mdzlog.alcor.net/2009/04/10/equilibrium-in-free-software-testing/#comment-553</link>
		<dc:creator><![CDATA[mdz]]></dc:creator>
		<pubDate>Tue, 14 Apr 2009 16:28:36 +0000</pubDate>
		<guid isPermaLink="false">http://mdzlog.wordpress.com/?p=282#comment-553</guid>
		<description><![CDATA[This is interesting, and we&#039;ve talked about systems like this before, but it&#039;s aimed at a slightly different problem, i.e. getting more testing.  At present, many parts of Ubuntu are tested very well by people who are using pre-release versions on a daily basis, but we receive so much feedback that it&#039;s difficult to process effectively.

I like the general idea of making the feedback loop more visible, though.]]></description>
		<content:encoded><![CDATA[<p>This is interesting, and we&#8217;ve talked about systems like this before, but it&#8217;s aimed at a slightly different problem, i.e. getting more testing.  At present, many parts of Ubuntu are tested very well by people who are using pre-release versions on a daily basis, but we receive so much feedback that it&#8217;s difficult to process effectively.</p>
<p>I like the general idea of making the feedback loop more visible, though.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: mdz</title>
		<link>http://mdzlog.alcor.net/2009/04/10/equilibrium-in-free-software-testing/#comment-552</link>
		<dc:creator><![CDATA[mdz]]></dc:creator>
		<pubDate>Tue, 14 Apr 2009 16:26:02 +0000</pubDate>
		<guid isPermaLink="false">http://mdzlog.wordpress.com/?p=282#comment-552</guid>
		<description><![CDATA[This is a good idea, and we&#039;ve done some experiments in this area, trying to pull out bugs which have been filed by people who have filed good-quality bugs in the past.  We need to be careful, though, to avoid problems like a chicken-and-egg situation (where it&#039;s too hard to build a reputation) and missing the long tail.

I&#039;d be very interested to hear if this has been done successfully in another large-scale project.]]></description>
		<content:encoded><![CDATA[<p>This is a good idea, and we&#8217;ve done some experiments in this area, trying to pull out bugs which have been filed by people who have filed good-quality bugs in the past.  We need to be careful, though, to avoid problems like a chicken-and-egg situation (where it&#8217;s too hard to build a reputation) and missing the long tail.</p>
<p>I&#8217;d be very interested to hear if this has been done successfully in another large-scale project.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Interstellar Medium: the Free Software carnival &#187; Free Software Carnival: 6 – 13 April 2009</title>
		<link>http://mdzlog.alcor.net/2009/04/10/equilibrium-in-free-software-testing/#comment-542</link>
		<dc:creator><![CDATA[Interstellar Medium: the Free Software carnival &#187; Free Software Carnival: 6 – 13 April 2009]]></dc:creator>
		<pubDate>Sun, 12 Apr 2009 23:43:00 +0000</pubDate>
		<guid isPermaLink="false">http://mdzlog.wordpress.com/?p=282#comment-542</guid>
		<description><![CDATA[[...] Zimmerman wants to close the gap between projects in need of developers and projects in need of [...]]]></description>
		<content:encoded><![CDATA[<p>[...] Zimmerman wants to close the gap between projects in need of developers and projects in need of [...]</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Roger</title>
		<link>http://mdzlog.alcor.net/2009/04/10/equilibrium-in-free-software-testing/#comment-541</link>
		<dc:creator><![CDATA[Roger]]></dc:creator>
		<pubDate>Sun, 12 Apr 2009 21:31:36 +0000</pubDate>
		<guid isPermaLink="false">http://mdzlog.wordpress.com/?p=282#comment-541</guid>
		<description><![CDATA[All the upstream linking stuff etc didn&#039;t exist in lp way back then and obviously none of the other commentors knew anything about the various things you just did.

Here is another suggestion to make things better then in lp.  Have some sort of button I can click labelled something like &quot;I want to fix this bug&quot; that then starts a wizard - ie something that offers a few options at a time and shows a new page based on choices and other information that is filled in.

For example for someone that isn&#039;t the original poster it could try to confirm the problem.  It can ask about looking in other bug trackers and link if necessary.  It can ask about patches and guide the process of building a debdiff, doing a PPA and whatever else it takes.

From any starting point there are numerous things that could happen next and just having documentation won&#039;t help since it will only get bigger over time and will turn into spaghetti code (&quot;if confirmed, but no tags then blah blah else upstream blah blah&quot;).  Having a guide/wizard the follows the process means it can be far more intelligent about it based on what has already happened.]]></description>
		<content:encoded><![CDATA[<p>All the upstream linking stuff etc didn&#8217;t exist in lp way back then and obviously none of the other commentors knew anything about the various things you just did.</p>
<p>Here is another suggestion to make things better then in lp.  Have some sort of button I can click labelled something like &#8220;I want to fix this bug&#8221; that then starts a wizard &#8211; ie something that offers a few options at a time and shows a new page based on choices and other information that is filled in.</p>
<p>For example for someone that isn&#8217;t the original poster it could try to confirm the problem.  It can ask about looking in other bug trackers and link if necessary.  It can ask about patches and guide the process of building a debdiff, doing a PPA and whatever else it takes.</p>
<p>From any starting point there are numerous things that could happen next and just having documentation won&#8217;t help since it will only get bigger over time and will turn into spaghetti code (&#8220;if confirmed, but no tags then blah blah else upstream blah blah&#8221;).  Having a guide/wizard the follows the process means it can be far more intelligent about it based on what has already happened.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: jcastro</title>
		<link>http://mdzlog.alcor.net/2009/04/10/equilibrium-in-free-software-testing/#comment-540</link>
		<dc:creator><![CDATA[jcastro]]></dc:creator>
		<pubDate>Sun, 12 Apr 2009 18:41:32 +0000</pubDate>
		<guid isPermaLink="false">http://mdzlog.wordpress.com/?p=282#comment-540</guid>
		<description><![CDATA[Hi Roger,

I went ahead and linked that bug in launchpad. What happens in bugs like this is that someone finds the corresponding fix in upstream (or another distro) and then leaves a comment. What should be happening is that these bugs get linked. (I clicked &quot;Also affects distro&quot;, picked Fedora, and then pasted in the bugzilla URL).

This allows lp to track the status. Now when bug people go looking for &quot;bugs fixed elsewhere but not yet in ubuntu&quot; this bug should show up on that radar. I&#039;ve also added some tags, &quot;patch&quot;, and &quot;bitesize&quot; which put it on other reports/lists that are regularly checked.

We are constantly checking documentation to ensure that all bug reporters know this, but unfortunately sometimes things slip through the cracks. We as a project are doing better at linking bug reports than we used to be but we have more places where we could improve things like this.

Things like adding patch tags go a long way to getting it on the right radar, and an actual link to an upstream bug tracker will most likely be handled quicker.

(I couldn&#039;t find a newt bug with an upstream link but I found one with a patch and tagged it as such)]]></description>
		<content:encoded><![CDATA[<p>Hi Roger,</p>
<p>I went ahead and linked that bug in launchpad. What happens in bugs like this is that someone finds the corresponding fix in upstream (or another distro) and then leaves a comment. What should be happening is that these bugs get linked. (I clicked &#8220;Also affects distro&#8221;, picked Fedora, and then pasted in the bugzilla URL).</p>
<p>This allows lp to track the status. Now when bug people go looking for &#8220;bugs fixed elsewhere but not yet in ubuntu&#8221; this bug should show up on that radar. I&#8217;ve also added some tags, &#8220;patch&#8221;, and &#8220;bitesize&#8221; which put it on other reports/lists that are regularly checked.</p>
<p>We are constantly checking documentation to ensure that all bug reporters know this, but unfortunately sometimes things slip through the cracks. We as a project are doing better at linking bug reports than we used to be but we have more places where we could improve things like this.</p>
<p>Things like adding patch tags go a long way to getting it on the right radar, and an actual link to an upstream bug tracker will most likely be handled quicker.</p>
<p>(I couldn&#8217;t find a newt bug with an upstream link but I found one with a patch and tagged it as such)</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: GoblinX Project &#187; GoblinX Newsletter, Issue 195 (04/12/2009)</title>
		<link>http://mdzlog.alcor.net/2009/04/10/equilibrium-in-free-software-testing/#comment-539</link>
		<dc:creator><![CDATA[GoblinX Project &#187; GoblinX Newsletter, Issue 195 (04/12/2009)]]></dc:creator>
		<pubDate>Sun, 12 Apr 2009 12:41:04 +0000</pubDate>
		<guid isPermaLink="false">http://mdzlog.wordpress.com/?p=282#comment-539</guid>
		<description><![CDATA[[...] http://mdzlog.alcor.net/2009/04/10/equilibrium-in-free-software-testing/ [...]]]></description>
		<content:encoded><![CDATA[<p>[...] <a href="http://mdzlog.alcor.net/2009/04/10/equilibrium-in-free-software-testing/" rel="nofollow">http://mdzlog.alcor.net/2009/04/10/equilibrium-in-free-software-testing/</a> [...]</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Roger</title>
		<link>http://mdzlog.alcor.net/2009/04/10/equilibrium-in-free-software-testing/#comment-538</link>
		<dc:creator><![CDATA[Roger]]></dc:creator>
		<pubDate>Sat, 11 Apr 2009 18:11:44 +0000</pubDate>
		<guid isPermaLink="false">http://mdzlog.wordpress.com/?p=282#comment-538</guid>
		<description><![CDATA[Here is one:  https://bugs.launchpad.net/ubuntu/+source/vnc4/+bug/119982/

Can&#039;t find the other one as Launchpad is so slow, but it was similar and in the newt package IIRC.]]></description>
		<content:encoded><![CDATA[<p>Here is one:  <a href="https://bugs.launchpad.net/ubuntu/+source/vnc4/+bug/119982/" rel="nofollow">https://bugs.launchpad.net/ubuntu/+source/vnc4/+bug/119982/</a></p>
<p>Can&#8217;t find the other one as Launchpad is so slow, but it was similar and in the newt package IIRC.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: jcastro</title>
		<link>http://mdzlog.alcor.net/2009/04/10/equilibrium-in-free-software-testing/#comment-537</link>
		<dc:creator><![CDATA[jcastro]]></dc:creator>
		<pubDate>Sat, 11 Apr 2009 17:51:21 +0000</pubDate>
		<guid isPermaLink="false">http://mdzlog.wordpress.com/?p=282#comment-537</guid>
		<description><![CDATA[Can you please link to these bugs that you&#039;ve linked to the RH bugzilla?]]></description>
		<content:encoded><![CDATA[<p>Can you please link to these bugs that you&#8217;ve linked to the RH bugzilla?</p>
]]></content:encoded>
	</item>
</channel>
</rss>

