CINXE.COM

<?xml version="1.0" encoding="UTF-8" standalone="yes"?> <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"><html xmlns="http://www.w3.org/1999/xhtml"><head><meta http-equiv="Content-Type" content="text/html; charset=UTF-8" /><title>LKML: Adrian Bunk: RFC: Starting a stable kernel series off the 2.6 kernel</title><link href="/css/message.css" rel="stylesheet" type="text/css" /><link href="/css/wrap.css" rel="alternate stylesheet" type="text/css" title="wrap" /><link href="/css/nowrap.css" rel="stylesheet" type="text/css" title="nowrap" /><link href="/favicon.ico" rel="shortcut icon" /><script src="/js/simple-calendar.js" type="text/javascript"></script><script src="/js/styleswitcher.js" type="text/javascript"></script><link rel="alternate" type="application/rss+xml" title="lkml.org : last 100 messages" href="/rss.php" /><link rel="alternate" type="application/rss+xml" title="lkml.org : last messages by Adrian Bunk" href="/groupie.php?aid=1112" /><script> var _paq = window._paq = window._paq || []; /* tracker methods like "setCustomDimension" should be called before "trackPageView" */ _paq.push(["setDoNotTrack", true]); _paq.push(["disableCookies"]); _paq.push(['trackPageView']); _paq.push(['enableLinkTracking']); (function() { var u="//m.lkml.org/"; _paq.push(['setTrackerUrl', u+'matomo.php']); _paq.push(['setSiteId', '1']); var d=document, g=d.createElement('script'), s=d.getElementsByTagName('script')[0]; g.async=true; g.src=u+'matomo.js'; s.parentNode.insertBefore(g,s); })(); </script></head><body onload="es.jasper.simpleCalendar.init();" itemscope="itemscope" itemtype="http://schema.org/BlogPosting"><table border="0" cellpadding="0" cellspacing="0"><tr><td width="180" align="center"><a href="/"><img style="border:0;width:135px;height:32px" src="/images/toprowlk.gif" alt="lkml.org" /></a></td><td width="32">聽</td><td class="nb"><div><a class="nb" href="/lkml"> [lkml]</a> 聽 <a class="nb" href="/lkml/2005"> [2005]</a> 聽 <a class="nb" href="/lkml/2005/12"> [Dec]</a> 聽 <a class="nb" href="/lkml/2005/12/3"> [3]</a> 聽 <a class="nb" href="/lkml/last100"> [last100]</a> 聽 <a href="/rss.php"><img src="/images/rss-or.gif" border="0" alt="RSS Feed" /></a></div><div>Views: <a href="#" class="nowrap" onclick="setActiveStyleSheet('wrap');return false;">[wrap]</a><a href="#" class="wrap" onclick="setActiveStyleSheet('nowrap');return false;">[no wrap]</a> 聽 <a class="nb" href="/lkml/mheaders/2005/12/3/55" onclick="this.href='/lkml/headers'+'/2005/12/3/55';">[headers]</a>聽 <a href="/lkml/bounce/2005/12/3/55">[forward]</a>聽 </div></td><td width="32">聽</td></tr><tr><td valign="top"><div class="es-jasper-simpleCalendar" baseurl="/lkml/"></div><div class="threadlist">Messages in this thread</div><ul class="threadlist"><li class="root"><a href="/lkml/2005/12/3/55">First message in thread</a></li><li class="origin"><a href="/lkml/2005/12/3/57">Adrian Bunk</a><ul><li><a href="/lkml/2005/12/3/57">Jesper Juhl</a><ul><li><a href="/lkml/2005/12/3/126">Greg KH</a><ul><li><a href="/lkml/2005/12/3/132">"M."</a><ul><li><a href="/lkml/2005/12/3/141">James Courtier-Dutton</a></li></ul></li><li><a href="/lkml/2005/12/3/136">Greg KH</a><ul><li><a href="/lkml/2005/12/3/140">"M."</a></li><li><a href="/lkml/2005/12/5/344">Florian Weimer</a></li></ul></li><li><a href="/lkml/2005/12/3/164">Adrian Bunk</a><ul><li><a href="/lkml/2005/12/3/171">Greg KH</a></li><li><a href="/lkml/2005/12/3/172">Chris Wright</a></li><li><a href="/lkml/2005/12/4/7">Arjan van de Ven</a></li><li><a href="/lkml/2005/12/5/220">Florian Weimer</a></li></ul></li><li><a href="/lkml/2005/12/3/207">Jesper Juhl</a><ul><li><a href="/lkml/2005/12/4/19">Matthias Andree</a></li></ul></li><li><a href="/lkml/2005/12/4/80">Jakob Oestergaard</a><ul><li><a href="/lkml/2005/12/4/124">Greg KH</a></li></ul></li><li><a href="/lkml/2005/12/5/90">Florian Weimer</a><ul><li><a href="/lkml/2005/12/6/235">Greg KH</a></li></ul></li></ul></li></ul></li><li><a href="/lkml/2005/12/3/59">Arjan van de Ven</a><ul><li><a href="/lkml/2005/12/3/63">Adrian Bunk</a><ul><li><a href="/lkml/2005/12/3/65">Arjan van de Ven</a><ul><li><a href="/lkml/2005/12/4/37">Denis Vlasenko</a></li><li><a href="/lkml/2005/12/5/84">Michael Frank</a></li></ul></li><li><a href="/lkml/2005/12/3/80">Matthias Andree</a><ul><li><a href="/lkml/2005/12/3/81">Otavio Salvador</a></li><li><a href="/lkml/2005/12/3/84">David Ranson</a></li><li><a href="/lkml/2005/12/3/91">Arjan van de Ven</a></li></ul></li><li><a href="/lkml/2005/12/3/111">Jeff Garzik</a><ul><li><a href="/lkml/2005/12/3/127">Greg KH</a></li><li><a href="/lkml/2005/12/3/157">Matthias Andree</a></li><li><a href="/lkml/2005/12/4/71">Michael Frank</a></li></ul></li><li><a href="/lkml/2005/12/5/108">Rob Landley</a></li><li><a href="/lkml/2005/12/10/95">Ryan Anderson</a></li></ul></li></ul></li><li><a href="/lkml/2005/12/3/105">Ben Collins</a><ul><li><a href="/lkml/2005/12/3/120">Adrian Bunk</a><ul><li><a href="/lkml/2005/12/3/123">Lee Revell</a><ul><li><a href="/lkml/2005/12/3/131">"M."</a></li><li><a href="/lkml/2005/12/3/165">Matthias Andree</a></li></ul></li></ul></li><li><a href="/lkml/2005/12/5/268">Bill Davidsen</a><ul><li><a href="/lkml/2005/12/5/353">Jeff Garzik</a><ul><li><a href="/lkml/2005/12/6/72">Matthias Andree</a></li><li><a href="/lkml/2005/12/9/129">Bill Davidsen</a></li></ul></li><li><a href="/lkml/2005/12/5/366">Horst von Brand</a></li></ul></li></ul></li><li><a href="/lkml/2005/12/3/110">"Dr. David Alan Gilbert"</a></li><li><a href="/lkml/2005/12/3/130">Lars Marowsky-Bree</a><ul><li><a href="/lkml/2005/12/3/137">Dave Jones</a><ul><li><a href="/lkml/2005/12/3/139">Lars Marowsky-Bree</a><ul><li><a href="/lkml/2005/12/3/144">Michael Buesch</a></li><li><a href="/lkml/2005/12/3/159">Adrian Bunk</a></li></ul></li><li><a href="/lkml/2005/12/3/166">Matthias Andree</a><ul><li><a href="/lkml/2005/12/3/168">Dave Jones</a></li></ul></li></ul></li><li><a href="/lkml/2005/12/5/305">Florian Weimer</a><ul><li><a href="/lkml/2005/12/6/114">Lars Marowsky-Bree</a><ul><li><a href="/lkml/2005/12/6/134">Florian Weimer</a></li></ul></li></ul></li></ul></li><li><a href="/lkml/2005/12/4/28">Indrek Kruusa</a><ul><li><a href="/lkml/2005/12/4/31">Arjan van de Ven</a><ul><li><a href="/lkml/2005/12/4/59">Indrek Kruusa</a></li></ul></li><li><a href="/lkml/2005/12/5/388">Rob Landley</a></li></ul></li><li><a href="/lkml/2005/12/5/275">Bill Davidsen</a><ul><li><a href="/lkml/2005/12/5/280">Adrian Bunk</a></li><li><a href="/lkml/2005/12/6/74">Matthias Andree</a></li><li><a href="/lkml/2005/12/6/118">Lars Marowsky-Bree</a><ul><li><a href="/lkml/2005/12/13/297">Bill Davidsen</a></li></ul></li></ul></li><li><a href="/lkml/2005/12/12/99">Felix Oxley</a><ul><li><a href="/lkml/2005/12/12/130">Horst von Brand</a><ul><li><a href="/lkml/2005/12/12/161">Felix Oxley</a><ul><li><a href="/lkml/2005/12/13/180">Horst von Brand</a></li></ul></li></ul></li></ul></li></ul></li></ul></td><td width="32" rowspan="2" class="c" valign="top"><img src="/images/icornerl.gif" width="32" height="32" alt="/" /></td><td class="c" rowspan="2" valign="top" style="padding-top: 1em"><table><tr><td><table><tr><td class="lp">Date</td><td class="rp" itemprop="datePublished">Sat, 3 Dec 2005 14:56:08 +0100</td></tr><tr><td class="lp">From</td><td class="rp" itemprop="author">Adrian Bunk <></td></tr><tr><td class="lp">Subject</td><td class="rp" itemprop="name">RFC: Starting a stable kernel series off the 2.6 kernel</td></tr></table></td><td></td></tr></table><pre itemprop="articleBody">The current kernel development model is pretty good for people who <br />always want to use or offer their costumers the maximum amount of the <br />latest bugs^Wfeatures without having to resort on additional patches for <br />them.<br /><br />Problems of the current development model from a user's point of view <br />are:<br />- many regressions in every new release<br />- kernel updates often require updates for the kernel-related userspace <br /> (e.g. for udev or the pcmcia tools switch)<br /><br />One problem following from this is that people continue to use older <br />kernels with known security holes because the amount of work for kernel <br />upgrades is too high.<br /><br />These problems follow from the development model.<br /><br />The latest stable kernel series without these problems is 2.4, but 2.4 <br />is becoming more and more obsolete and might e.g. lack driver support <br />for some recent hardware you want to use.<br /><br />Since Andrew and Linus do AFAIK not plan to change the development <br />model, what about the following for getting a stable kernel series <br />without leaving the current development model:<br /><br /><br />Kernel 2.6.16 will be the base for a stable series.<br /><br />After 2.6.16, there will be a 2.6.16.y series with the usual stable <br />rules.<br /><br />After the release of 2.6.17, this 2.6.16.y series will be continued with <br />more relaxed rules similar to the rules in kernel 2.4 since the release <br />of kernel 2.6.0 (e.g. driver updates will be allowed).<br /><br /><br />Q:<br />What is the target audience for this 2.6.16 series?<br /><br />A:<br />The target audience are users still using 2.4 (or who'd still use kernel <br />2.4 if they weren't forced to upgrade to 2.6 for some reason) who want a <br />stable kernel series including security fixes but excluding many <br />regressions.<br />It might also be interesting for distributions that prefer stability <br />over always using the latest stuff.<br /><br /><br />Q:<br />Does this proposal imply anything for the development between 2.6.15 and <br />2.6.16?<br /><br />A:<br />In theory not.<br />In practice, it would be a big advantage if some of the bigger <br />changes that might go into 2.6.16 would be postponed to 2.6.17.<br /><br /><br />Q:<br />Why not start with the more relaxed rules before the release of 2.6.17?<br /><br />A:<br />After 2.6.16.y following the usual stable rules, the kernel should be <br />relatively stable and well-tested giving the best possible basis for a <br />long-living series.<br /><br /><br />Q:<br />How long should this 2.6.16 series be maintained?<br /><br />A:<br />Time will tell, but if people use it I'd expect 2 or 3 years.<br /><br /><br />Q:<br />Stable API/ABI for external modules?<br /><br />A:<br />No.<br /><br /><br />Q:<br />Who will maintain this branch?<br /><br />A:<br />I could do it, but if someone more experienced wants to do it that would <br />be even better.<br />-<br />To unsubscribe from this list: send the line "unsubscribe linux-kernel" in<br />the body of a message to majordomo@vger.kernel.org<br />More majordomo info at <a href="http://vger.kernel.org/majordomo-info.html">http://vger.kernel.org/majordomo-info.html</a><br />Please read the FAQ at <a href="http://www.tux.org/lkml/">http://www.tux.org/lkml/</a><br /><br /></pre></td><td width="32" rowspan="2" class="c" valign="top"><img src="/images/icornerr.gif" width="32" height="32" alt="\" /></td></tr><tr><td align="right" valign="bottom"> 聽 </td></tr><tr><td align="right" valign="bottom">聽</td><td class="c" valign="bottom" style="padding-bottom: 0px"><img src="/images/bcornerl.gif" width="32" height="32" alt="\" /></td><td class="c">聽</td><td class="c" valign="bottom" style="padding-bottom: 0px"><img src="/images/bcornerr.gif" width="32" height="32" alt="/" /></td></tr><tr><td align="right" valign="top" colspan="2"> 聽 </td><td class="lm">Last update: 2005-12-03 14:58 聽聽 [from the cache]<br />漏2003-2020 <a href="http://blog.jasper.es/"><span itemprop="editor">Jasper Spaans</span></a>|hosted at <a href="https://www.digitalocean.com/?refcode=9a8e99d24cf9">Digital Ocean</a> and my Meterkast|<a href="http://blog.jasper.es/categories.html#lkml-ref">Read the blog</a></td><td>聽</td></tr></table><script language="javascript" src="/js/styleswitcher.js" type="text/javascript"></script></body></html>