Re: Automatic Debug Packages
On Fri, Aug 14, 2009 at 05:50:50PM -0700, Russ Allbery wrote: Peter Samuelson pe...@p12n.org writes: [Emilio Pozuelo Monfort] We haven't agreed on whether there should be one ddeb per source or per binary package, so I would leave this still opened. Maybe I'm losing track of things here, but it seems to me that everyone except you is saying one ddeb per binary. And then you say sure, we could do that if we need to. How many times has this happened so far in the thread? I haven't been keeping count. Joerg was also advocating one ddeb per source package in the summary message that he sent about the ftp-master approach, and Emilio has mentioned a few times that ftp-master needs to buy in on that decision (which I agree with). I'm not sure if I'm missing some concern from the ftp-master side. So if we have one ddeb per source package, which generates multiple binary .debs for different libraries --- say, libext2fs, libcom_err, and libss, to take a completely random example --- and the user installs different versions of said libraries coming from different versions of the source package, won't there be a problem if there is only a single ddeb per source package? I assume you can't install multiple ddebs coming from different source packages at the same time, since the pathnames would conflict, right? - Ted -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: net-tools future
On Sun, Mar 15, 2009 at 02:30:18PM -0300, Martín Ferrari wrote: About the wrapper scripts: * ipconfig, route: the most difficult ones, both can be replaced by calls to ip, maybe except for some obscure options. Suggestion about the wrapper scripts. It would be nice if they had a mode enabled by an environment variable or a command-line option which printed the equivalent ip commands which they issued, so people can learn the new ip interface if they are so interested. Also, I'd recommend making the scripts carefully cover 100% of the ifconfig, route, and netstat commands, since these commands have a very long history, and are extremely well known by many sysadmins, and used in many, *many* shell scripts. - Ted -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: -dbg packages; are they actually useful?
On Tue, Mar 03, 2009 at 10:12:22PM +, Steve McIntyre wrote: I'm looking at my local mirror (slowly) update at the moment, and I've got to wondering: are the large -dbg packages actually really useful to anybody? I can't imagine that more than a handful of users ever install (to pick an example) the amarok-dbg packages, but we have multiple copies of a 70MB-plus .deb taking up mirror space and bandwidth. I can understand this for library packages, maybe, but for applications? There are people working on ways of compressing the debuginfo information, and I've been told they might have results within a couple of months. Part of the problem is that depending on how the package is built, the -dbg packages can be huge, so it makes the cost/benefit ratio somewhat painful. If the -dbg files were more like these sizes: 224 e2fslibs-dbg_1.41.3-1_i386.deb 52 libss2-dbg_1.41.3-1_i386.deb 452 e2fsprogs-dbg_1.41.3-1_i386.deb48 libuuid1-dbg_1.41.3-1_i386.deb 76 libblkid1-dbg_1.41.3-1_i386.deb48 uuid-runtime-dbg_1.41.3-1_i386.deb 44 libcomerr2-dbg_1.41.3-1_i386.deb I doubt there's be too much concern - Ted -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: Forthcoming changes in kernel-package
On Fri, Feb 20, 2009 at 12:56:30PM -0600, Manoj Srivastava wrote: BTW, I have a set of patches you might want to consider. I'll file them in BTS if you're currently making make-kpkg. Please. I have been thinking about the request you made for debugging symbols being packaged, and now I do have some time to play with building kernels again, I would like to see that in Squeeze. Sorry for the delay; I've sent you the private patches I've been using for make-kpkg. Some of them are quite hackish, and some of they you may have fixed in other ways, so I won't feel bad at all if you need to significantly rework them before you can merge them into your master sources. The BTS bug numbers are #517290, #517291, #517292, and #517293. Best regards, - Ted -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: Is the FHS dead ?
On Tue, Feb 24, 2009 at 08:20:31AM -0600, Gunnar Wolf wrote: Interesting. And yes, illustrative of the historically (and, should I add, ridiculous? No, I'd better not ;-) ) rivality between Linux and the *BSDs, big egos included. Well, the last time we tried to make reasonable accomodations for *BSD's, some of the biggest biggest whiners^H^H^H^H^H^H^H complaints came from Debian. In fact, some later complaints from Debianites about the lack of /usr/libexec is largely the fault of Ian Jackson, who ***strenuously*** opposed /usr/libexec on the mail thread which I quoted. In fact, as I recall, he threatened to rally all of Debian NOT to support the FSSTND/FHS if we didn't drop /usr/libexec from the draft spec. Ah, history. The painful fact of the matter is that anytime a draft like the FHS forces any distribution or OS to change, there will be opposition. In some cases it will be principled and constructive. In other cases, it will question the spec writers' technical judgement, ethics, and even their paternity. However, Linux's position WRT the commercial Unixes has radically shifted in the last decade. Linux is no longer considered a toy, and is taken seriously into account. So, even with the big inertia that might hamper more than one initiative, perhaps the FHS could be pushed in collaboration with their respective companies? At least, I'd be surprised if -say- the Solaris or HPUX people weren't open to discussion leading to better interoperability. Last I heard, HPUX is on maintenance life-support, and they don't have enough engineers to make sure their userspace is uptodate. And as far as Solaris is concerned, they currently have a project to update to a 16-year-old shell (ksh93) in their distribution. Solaris has constraints on their filesystem so they can be compatible with the 1986-era SVID specification. Can you really see Linux distributions embracing installing a 3 decade-old awk in /bin/awk, a 2 decade-old awk in /usr/xpg4/bin/awk, with no GNU extensions allowed, just to be compatible with Solaris? I thought not. Look, the proprietary Unix systems as a whole are losing market share; Linux is gaining market share. The *only* proprietary Unix that gained market share last year was AIX, and it has a Linux compatibility subsystem, AIX 5L. Fundamentally, I just don't see Debian, Red Hat, SuSE, Ubuntu, et. al, all making changes in were things are installed just to accomodate Solaris --- or any other legacy Unix system. Do you? Especially given the flames raised last time by folks like Ian Jackson --- who is, I might remind you, still on the Debian Technical Committee and is still considered one of Debian's technical leaders? Maybe I'm being too cynical, but I just don't see it. - Ted -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: Is the FHS dead ?
On Thu, Feb 19, 2009 at 05:24:24PM +0200, Guillem Jover wrote: Reiviving the FHS is great! Something that is bothering me a bit, though, is that historically it seemed to try to cater to Unix in general, not only Linux, even if most of the participants were coming from the Linux world. So hosting it under the LSB auspices might deter other Unix vendors to consider it or get involved, which would seem like a regression. Maybe hosting it on a more neutral place would be better? Well, realistically we didn't have very good participation from anyone other than one or two *BSD folks, and at the time some of the changes that were made for compatibility with *BSD (and, to be fair, to be closer to the rest of the Unix world) caused no small amount of controversy. Consider the following thread from debian-devel approximately 8 years ago (which was about the last time we had any *BSD participation): http://www.mail-archive.com/debian-devel@lists.debian.org/msg11462.html Realistically, I think we will have a hard enough time dealing with any places where the various Linux distributions have chosen different pathname (i.e., differences between Debian and Red Hat, or Red Hat and Novell, et. al). If it turns out that all Linux distributions do it one way, and OpenBSD has chosen a different hierarchy --- let's be honest with ourselves, would we really try to engineer change at Debian, Ubuntu, Red Hat, SLES, etc., just as a peace offerring to keep Theo deRaadt from OpenBSD happy? I just don't see it And if it's not going to happen, we shouldn't try to set up expectations that we would try to change all of Linux just because a *BSD happened to have chosen a different pathname Regards, - Ted -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: Forthcoming changes in kernel-package
On Mon, Feb 09, 2009 at 12:14:49AM -0600, Manoj Srivastava wrote: Hi, This is a heads up for a major change in kernel-package, the tool to create user packaged kernel images and headers; which will make the make-kpkg script far less error prone, and far more deterministic. a. Every invocation of kernel-package will remove ./debian directory, and regenerate the changelog and control files. This will get rid of any remaining issues with the ./debian directory getting out of sync with the kernel sources; and will allow people to make small tweaks to the kernel sources and have make-kpkg reflect those changes. Is there going to be a way for people to replace the changelog with one that contains useful information in that case? I've been doing this by doing a make-kpkg configure and then editing the debian/changelog file afterwards... BTW, I have a set of patches you might want to consider. I'll file them in BTS if you're currently making make-kpkg. - Ted -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: Is the FHS dead ?
On Mon, Feb 16, 2009 at 06:08:17PM +0200, Teodor wrote: On Mon, Feb 16, 2009 at 5:14 PM, Josselin Mouette j...@debian.org wrote: Le lundi 16 février 2009 à 14:20 +, Matthew Johnson a écrit : the FHS should certainly continue to exist and be coordinated between distros though. I agree that if it needs taking over we should do so in cooperation with the other big distros. Certainly. It's just that someone needs to start the work. There is no need to create another standard, FHS is being continued in the LSB project at linuxfoundation.org / freestandards.org. FHS was the starting point for LSB. Even if the LSB project has been criticized by the Debian project, this seems to become the the facto file hierarchy standard for Linux. It is not perfect, but is being adopted by the majority of Linux distros. It's true that the FHS work group has been largely moribund. It's something I've been working for the last couple of weeks, actually. I've gotten agreement form Dan Quinlan (the former chair of the FHS workgroup) and Rusty Russell who had done most of the changes the last time it had been updated if they had any objections having the LSB working group take it over, and they had no problem with that. See: http://bugs.linuxbase.org/show_bug.cgi?id=2511 So the plan is FHS will be updated in the context of the LSB workgroup, since the FHS mailing list has largely been taken over by SPAM, and it seemed that most of the people who were interested in it were active LSB work group members. That being said, if there are things that the the folks on this thread are interested in working on in terms of updating the FHS (which we clearly badly in need of updating) I invite them to join us and make proposals on the lsb-disc...@lists.linuxfoundaiton.org mailing list. Indeed, getting more folks interested in the updating the FHS would be really great! Regards, - Ted P.S. Or feel free to submit bugs to: http://bugs.linuxfoundation.org, component FHS. It may take a while for us to get set up, but we will be paying attention to them. -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: Results for General Resolution: Lenny and resolving DFSG violations
On Sun, Dec 28, 2008 at 09:55:36PM -0800, Thomas Bushnell BSG wrote: I would prefer this. But I am afraid of it, and so I would vote against it. I am afraid that there are folks in the project who really don't care if Debian is 100% free--even as a goal. I think that Ted Tso is even one of them. Fear is a terrible thing to use as the basis of decisions and of votes; consider it was fear that drove many people to vote for Proposition 8 in California As I said in my recent blog entry[1], I believe that 100% free is a wonderful aspirational goal --- other things being equal. However, I don't believe it to be something that should be Debian's Object of Ultimate Concern; there are other things that need to be taken into consideration --- for example, allowing various machines owned by Debian to be able to use their network cards might be a nice touch. [1] http://thunk.org/tytso/blog/2008/12/28/debian-philosophy-and-people/ In other words, I believe in 100% Free as a goal; but I'm not a fundamentalist nor a fanatic about it. I wish we could have in the world of GNU/Linux one, just one, please--just one--distribution which really took free software as of cardinal importance. As others have pointed out, there is such a distribution, gNewSense; in fact, if you look at [2], you will find that there are five others, Ututu (the first fully free GNU/Linux distribution recognized by the FSF), Dynebolic, Musix GNU+Linux, BLAG, and Trisquel. So not only is there one such distribution that takes free software of cardinal importance, there are six in the world already. Does Debian really need to be the seventh such distribution? [2] http://www.gnu.org/links/links.html#FreeGNULinuxDistributions In my opinion, developers who are unwilling to abide by the Social Contract in their Debian work should resign. But they don't, and this is what has me afraid. That would be like saying that people who don't agree with Proposition Eight's amendment to the California constitution should leave the state, as opposed to working to change it. I prefer to stay within Debian in the hopes that I can help it change to something which I think is better; at the very release, reverting the 1.1 version of the Social Contract, and perhaps, clarifying it. I will note that Option 1, Reaffirm the Social Contract came in *dead* *last*: Option 1 2 3 4 5 6 7 === === === === === === === Option 1 4660727389 117 Option 2281 160 160 171 177 224 Option 325561 125 137 151 204 Option 4253 121 146 160 166 194 Option 5234 105 128 135 136 191 Option 6220 118 134 125 134 180 Option 7226 129 145 153 160 169 It was beaten by options 2 (281 - 46 = 235), 3 (255 - 60 = 195), 4 (253 - 72 = 181), 5 (234 - 73 = 161), 6 (220 - 89 = 131) and 7/FD (226 - 117 = 109). Put another way, _very_ few people are willing to take a fundamentalist interpretation of the Social contract (by AJ's calculation, 9.3%) ahead of delaying Lenny. I don't think encouraging 90% of the Debian Developers to resign would be a particularly constructive suggestion. Fixing the Social Contract so it reflects our common understanding of what's best for the Debian Community, both users and developers, is IMHO a better choice than striving to become the Seventh Fundamentalist Linux Distribution on the FSF's approved list. Best regards, - Ted -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: Results for General Resolution: Lenny and resolving DFSG violations
On Mon, Dec 29, 2008 at 03:02:41PM +1000, Anthony Towns wrote: Using the word software as the basis for the divide might be too much: we've already done a lot of work restricting main to DFSG-free docs, and I think it makes sense to keep that. Having main be a functioning bunch of free stuff with a minimal and decreasing amount of random non-free stuff we still need to support it works well, it seems to me. I'm not convinced that leaving important parts of Debian undocumented over doctrinal disputes over licensing terms is actually in the best interests of users, but I recognize that's a position that people of good will can (and have) disagreed upon. If it were up to me, I would have Debian work towards a system where packages could be tagged to allow enable common user preferences (we won't be able to make everyone happy) be enforced by what packages they can see/install. Some users are OK with GFDL documentation, others are not; some users are OK with non-free firmware, other are not. So why can't we tag packages appropriately, so that this can be reflected in a configuration file so that people who are passionate about some particular issue can decide what tradeoffs they are willing to make with respect to usability and/or documentation based on how Fundamentalistic they want to be with regards to the 100% Free goal/requirement? Separating packages into separate sections to support these sorts of policy preferences is a hack, and with appropriate tagging, in the long run we can allow users to be much more fined-grained about expressing their preferences --- which would be in line with our goal of being a Universal OS, I think. Back in the day, I tried writing a version of the SC that felt both inspiring and within the bounds of what we could actually meet. It looked like: I like this a lot. However, I do have a few nits... We, the members of the Debian project, make the following pledge: 1. We will build a free operating system We will create and provide an integrated system of free software that anyone can use. We will make all our work publically available as free software. Given how literalistic some members of our community can be about interpreting Foundation Documents, the second sentence is a little worrying. I can easily imagine a Free Software Fanatic using the second sentance as an argument that we must stop distributing the non-free section, since non-free is, by definition, not Free Software. And it could easily be argued that the work that Debian Developers to package non-free packages, which is after all distributed on the Debian FTP servers and via Debian Mirrors, would fall under the scope of All our work. I'm not sure what you were trying to state by the second sentence above; one approach might be to simply strike it from the draft. Or were you trying to add the constraint that any work authored by DD's on behalf of the Debian Project should be made available under a free software license, even if in combination with other software being packaged, the result is non-free? 2. We will build a superior operating system We will collect and distribute the best software available, and strive to continually improve it by making use of the best tools and techniques available. I'm worried about the first clause, because of the absolutist word best in best software available. Again, some literally minded DD's could view this as meaning that the best is the enemy of the good, and use this as bludgeon to say that since we have package X, we should not have packages Y or Z, because, X is the *best*. Again, I'm not sure what you intended to add by the first clause, so my first reaction would be to strike it and make it shorter/simpler: We will strive to continually improve the software we collect and distribute by making use of the best tools and techniques available. I don't think the community clause is terribly well worded, but that's what you get when you make stuff up out of whole cloth rather than building on previous attempts. It's not bad. The one thing that I noted was community wasn't terribly well defined. Do we mean the user community? The developer community? Upstream developers? All of the above? Adding an initial phrase or sentence that affirmed that everyone who touches Debian in some way (users, developers, upstream) are considered part of the community --- and then follow it with your formulation pledging that we will work to ensure that members of the community shall be treated with respect --- would be the way I would go. Anyway, given the last proposal I made [0] went nowhere, unless people want to come up with their own proposals, or want to second the above as a draft proposal to be improved and voted on, I suspect nothing much will change, and we'll have this discussion again in a few years when squeeze is looking like releasing. I would
Re: Results for General Resolution: Lenny and resolving DFSG violations
On Mon, Dec 29, 2008 at 04:38:25PM +0100, Romain Beauxis wrote: To me, the social contract is a very good compromise. It states first an idealist acheivement, but moderates it by some pragmatism concerning the users. unproductive discussions fall into the same category, when they do not end as flames or trolls. It's a claim which has never been true. Debian shall _remain_ 100% free? Remain implies that at one stage Debian had reached such a state of Nirvana. That has never been the case! The disputes that we have had, at each stable release since the 1.1 revision to the Social Contract, have been precisely because some large, and vocal, set of developers have not been willing to be pragmatic, but who have instead argued for a very literalistic reading of the Social Contract. That is mainly why I am against the notion of Code of Conduct. I don't see the connection which leads you to be against a Code of Conduct, but I will note that Ubuntu CoC does not use any absolute words. It merely asks participants to: * Be considerate * Be respectful * Be collaborative * When you disagree, consult others * When you are unsure, ask for help * Step down considerately These are idealistic goals --- by your own argument, what's wrong with having them? We need to moderate them by an understanding that human nature being what it is, we will occasionally fail at this ideal; and then ecourage and remind each other to try to strive for this as an ideal --- not throw people out of the project when they fail to live up to such a goal. Maybe you don't like the name Code of Conduct, because it implies a certain amount of inflexibility? If so, maybe a different name would make you more comfortable? Eventually, that is also the same vision that drives me in politics: an ideal goal moderated by pragmatism. Not the converse. That's my vision as well; we might disagree about how far our pragmatism might take us, but our ideals tell us which direction to go, even if we are far from it at the moment. The question is how much patience do we have, and should we have? I do feel quite strongly, that aspirational goals, if they are going to be in Foundation Documents, must be clearly *labelled* as aspirational goals, and not as inflexible mandates that _MUST_ be kept. In politics, can have aspirational ideals such as a chicken in every pot and two cars in every garage which get used in campaign slogans, but you don't put such things as a MUST in a country's constitution. - Ted -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: Results for General Resolution: Lenny and resolving DFSG violations
On Mon, Dec 29, 2008 at 12:48:24AM +, Simon Huggins wrote: I wonder how many DDs were ashamed to vote the titled Reaffirm the social contract lower than the choices that chose to release. I'm not ashamed at all; I joined before the 1.1 revision to the Debian Social Contract, which I objected to them, and I still object to now. If there was a GR which chainged the Debian Social contract which relaxed the first clause to only include __software__ running on the Host CPU, I would enthusiastically vote for such a measure. Also see: http://thunk.org/tytso/blog/2008/12/28/debian-philosophy-and-people - Ted -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: DFSG violations in Lenny: Summarizing the choices
On Sat, Nov 08, 2008 at 10:24:16PM -0800, Thomas Bushnell BSG wrote: Neither does it (currently) contain an exception for debian.org machines, or very popular Dell machines with Broadcom ethernet firmware. Great! Cut them off!! Let's see how quickly we can get users moving to non-official kernels and installers when the official ones don't work for them. Then we can stop fighting about it. The DFSG hard liners can go on using the DFSG free kernels, and everyone else can either move to another distribution or use an unofficially forked kernel package and installer. Why not just support it in non-free exactly the way we do other things? Because according to you, Debian isn't allowed to ship any non-free bits, right? I assume that includes in the installer CD-ROM itself. And if you need non-free bits in order to download from the non-free section of the archive, what do you do? From what I understand, the Lenny installer is currently includes the non-free archive, and automatically enable the non-free section in order to allow those users to win. Is that considered OK by you? If the proposal is to delay the release to make sure that **all** bits are moved into the non-free section, what about the debian installation CD itself? If it is true that __Debian__ never includes any DFSG bits, I would think that would have to include the installer CD/DVD image itself, no? - Ted -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
DFSG violations in Lenny: Summarizing the choices
On Fri, Nov 07, 2008 at 12:47:01PM +, David Given wrote: In which case things have changed within the past couple of years --- after all, the whole purpose of the Atheros HAL was to inforce those FCC limits. Do you have any references? Like, to an FCC statement of policy change? If so, it would be extremely useful to have. There are corporate lawyers who are very much afraid that the FCC could, if they were alerted to the fact that someone had figured out how to reverse engineer the HAL and/or the firmware to cause their WiFi unit to become a super radio that could transmit on any frequency, that the FCC could prohibit the *hardware* from being sold anywhere in the US. Given that the US is a rather large market, and that some of these providers sell a very large number of WiFi units in laptops (i.e., HP, Lenovo, Del, etc.), and where only a *small* percentage of said units will ever run Linux, and even smaller, vastly infintisimal percentage of those systems will run Debian, the reality is that you look at the downside risk of not being able to sell, say iwl4965 chipsets and having millions and millions of suddenly useless pieces of silicon become the governments stop allowing said unit from being sold, and weigh that against a very small number of Debian users not being able to use the wireless unit out of the box, it's really a no-brainer to guess how the WiFi manufacturers will react. So realistically, let's be honest with ourselves. Not supporting devices that require non-free firmwares is not going to help make the world a better place. What it will probably do is that users, once they find out that that a Debian install will result in various bits and pieces of their hardware being non-functional until they figure out how to download various magic firmware components, or manually configuring the non-free repository, will probably simply switch to another distribution, such as Fedora or Ubuntu. At which point there will be even *fewer* Debian users, and so Debian will have even *less* leverage. Now, if the majority will of Debian is that all bits distributed by the Debian distribution must be DFSG free, even if it doesn't run on host processor, and we should hold up the release until this can be accomplished, that's a legitimate choice. That choice will have consequences; in the meantime more users will simply switch to other distributions, and Debian can be the distribution with a tiny niche number of users, with developers shaking their fists about how they are Free, just as OpenBSD users can shake their fists about how they are Secure (but have almost no users). Another choice open to Debian is to make it easier for users to opt into downloading firmware --- perhaps by making very easy through the installer to select the non-free section. That choice also has consequences. For one, it won't help in the cases where the non-free firmware is needed for the system to boot, or to access the network in order to download the non-free .debs. (I'm assuming for the sake of argument that it would be considered verboten to ship non-free firmware in the Debian installer CD-ROM.) Fortunately for us, at the moment I am not aware of large numbers of highly popular laptops or servers for which non-free firmware is necessary before the firmware would be able to access the network. This could potentially happen in the future if there are netbooks that only have wifi networking, for example. Another consequence of making it easy for the users to add non-free to the repositories so they can download firmware necessary to make their hardware useful is that a huge number of users may end up enabling non-free just to make their hardware work, and then they may end up installing even more non-free packages on their system. It's much like the argument that the current copyright laws around downloading music is insane, because it increases the disrespect of all laws, and we are training an entire generation of users that breaking copyright law so they can download their favorite music or video torrents is OK. Yet another choice which Debian could choose is to create a new firmware section; this would allow users to only be able to select non-free firmware, without accidentally installing other non-free packages. This has the advantage of more fined-grained control of what users might want or not want to install on their systems. The firmware section would be just as non-free as the non-free section, but for people for whom the distinction of running on the host CPU or not has meaning, it gives them a way of allowing some non-free packages on their systems, but not others. For people who feel passionately that they will not abide any non-free software, they can choose not to install from the firmware and non-free sections. The final choice which Debian could make is to ignore the problem and punt making one of the above decisions for yet another release. This seems to be the path that the
Re: DFSG violations in Lenny: Summarizing the choices
On Sat, Nov 08, 2008 at 03:29:44PM -0800, Thomas Bushnell BSG wrote: On Sat, 2008-11-08 at 14:11 -0500, Theodore Tso wrote: There are corporate lawyers who are very much afraid that the FCC could, if they were alerted to the fact that someone had figured out how to reverse engineer the HAL and/or the firmware to cause their WiFi unit to become a super radio that could transmit on any frequency, that the FCC could prohibit the *hardware* from being sold anywhere in the US. I've heard this claim before. Can you substantiate it in some way? Private conversations from a representative of a company who genuinely whats to do the right thing, but who had to battle the lawyers who were afraid of losing millions and millions of dollars. I'm honor bound not release the name of the engineer(s) and companies involved, so you can take that as you will. It seems to me that, if this is really true, then the hardware manufacturers have been lying to the FCC for years, claiming that the user cannot reprogram the card, without explaining that, in fact, it's just that users may not know how to, but that they can do so without any hardware mucking. The FCC understands that you can't make it *impossible*. Even before software radios, it was understood that someone posessing the skills, say, of an amateur radio operator might be able to add a resistor or capacitor in parallel with an RC/LC tuning circuit, and modify the length of the antenna, etc., thus making a radio transmit outside of the band which it was type-certified to operate on. A radio manufacturer is not required to dunk the entire radio in epoxy, and make it utterly *impossible* for someone to modify the radio; on the other hand, if all it takes is clipping a jumper or cutting a trace on a board, the FCC does have the power to order that the radio not be sold in the US. So just as the GPL has never been tested on point about whether or not a program which dynamically links against a GPL'ed library would be infected by the GPL, and just as the FSF has appropriately pointed out that the court system does not operate algorithmically, but can make decisions based on intent --- similarly, the FCC has not ruled on point on implementations that rely on software radios. Most lawyers seem to agree that that documenting how to modify the firmware is roughly equivalent to providing a trace that if cut, would allow scanners to listen in on cell phone frequences, and then leaking tech sheets that would allow people to modify scanners to do something which the US Congress (rightly or wrongly) has declared to be illegal. There seems to be some disagreement about whether security by obscurity is sufficient for the FCC, or whether you have to implement hard crpyographic signing to prevent non-vendor-approved firmware from being used, but that's because there's no precedent. Given that it seems pretty clear that the FCC has never penalized a radio manufacturer if a skilled Ham Radio operator or Electrical Engineer reversed engineered an analog circuit, and then modified it to transmit on a band that the equipment wasn't type-certified, one could argue that security by obsecurity is considered permissible by the FCC. But until there's precedent, we won't know for sure --- just as we won't know whether or not a program which dynamically links against a GPL library is really bound by the GPL until a court rules on point on the issue --- and then we'll only know in that legal jurisdiction. Regardless, the DFSG doesn't say anything about unless the FCC has an annoying rule. We don't distribute non-free software in Debian. And that's not some sort of choice we might make--it's a a choice we have already made. And as I said, I think we should let the DFSG hard-liners win. Let's yank all of the binaries that require a firmware, and release Lenny as-is. If that causes some users switch to some fork that actually has a kernel that works for them, given their hardware, or said users switch to Ubuntu, then so be it. At least we'll stop flaming about the issue. - Ted -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: DFSG violations in Lenny: Summarizing the choices
On Sat, Nov 08, 2008 at 05:05:50PM -0800, Thomas Bushnell BSG wrote: But now we have this claim that the FCC's well-understood rule about hardware does not apply to software: that software modifications *are* traceable back to the manufacturer, even though hardware modifications are not. Oddly, however, in all these conversations, we've never seen any indication that this is really the FCC's policy. The analogy that has been made is between hardware modifications assisted by information about how a jumper or removing a trace is leaked by a manufacturer *is* traceable back to the manfacturer, and a manufacturer documenting their hardware (or releasing firmware source) thus enabling someone to modify said firmware. In both cases, the manufacturer is assisting by making the information available. In the former case, it was via nth generation fax/xerox copies that was passed under the table by retailers, but it was still held to be the manufacturers problem. So if people think that they are going to be able to get firmware in source form so that popular wireless chips can be driven using 100% DFSG pure firmware, I suspect they will have a very long wait ahead of them. The issue is that software controlled radios are cheaper, and that drives the mass market, so that will be what most manufacturers will use. And none of this is really relevent: the DFSG and the Social Contract do not contain an exception for dishonest or scared hardware manufacturers, or stupid FCC policies. Neither does it (currently) contain an exception for debian.org machines, or very popular Dell machines with Broadcom ethernet firmware. Great! Cut them off!! Let's see how quickly we can get users moving to non-official kernels and installers when the official ones don't work for them. Then we can stop fighting about it. The DFSG hard liners can go on using the DFSG free kernels, and everyone else can either move to another distribution or use an unofficially forked kernel package and installer. - Ted -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: DFSG violations in Lenny: Summarizing the choices
On Sun, Nov 09, 2008 at 12:21:26PM +0900, Paul Wise wrote: On Sun, Nov 9, 2008 at 4:11 AM, Theodore Tso [EMAIL PROTECTED] wrote: Another choice open to Debian is to make it easier for users to opt into downloading firmware --- perhaps by making very easy through the installer to select the non-free section. For machines where non-free firmware is required, lenny d-i defaults to adding non-free and installing that firmware. I found this out when installing on my laptop, which contains an Intel 3945 wireless chip. Oooh does that means Debian is distributing non-free bits? My suggestion is that we either change the DFSG, or drop it so that it complies with the DFSG, like the DFSG hard-liners want. - Ted -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: dhclient-script, hooks, and changing the environment
On Wed, Aug 06, 2008 at 09:10:56PM -0300, martin f krafft wrote: Anything else? Do you know of packages that rely on this functionality? Do you have scripts of your own which modify the environment? Would you please be so kind as to explain to me what they do, and help me figure out whether there isn't a better way for them? In the past I've used dhclient-enter-hooks.d to work around buggy hotel networks which advertise a gateway for the default route which is outside the local network. So I used a dhclient-enter-hooks.d script to hack the netmask returned by the DHCP server so that the default route would be accepted. (Unfortunately, Windows allows the default route to be outside the local network arrange of the ethernet interface, and apparently if there is only one ethernet interface, Windows will assume that instead of rejecting the route, or dropping packets on the floor, to do the convenient-but-wrong thing of assuming the packets should go out via the default route to the ethernet interface, even though the default route is an invalid non-local address for that interface. As a result, trashy hotel networks have no incentive to fix their DHCP servers, since they work just fine on Windows laptops - Ted -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: dhclient-script, hooks, and changing the environment
On Fri, Aug 08, 2008 at 03:29:43PM -0300, martin f krafft wrote: also sprach Theodore Tso [EMAIL PROTECTED] [2008.08.08.1453 -0300]: In the past I've used dhclient-enter-hooks.d to work around buggy hotel networks which advertise a gateway for the default route which is outside the local network. netconf could handle this internally, and it would be fair to include this. So assuming yo get a 192.168.0.0/24 address and the default gateway is 10.0.0.1, what is the best approach? Hacking the netmask seems awful. Adding a route that makes 10.0.0.1 link-local seems better, no? Yes, adding a specific link-local route is probably better. In practice, the various Awful Hotel Network implementations would give you an IP configuration that was something like this IP Addr: 10.0.2.67 Netmask: 255.255.255.0 Gateway: 10.0.8.1 So the simpler thing for me to do was to hack the netmask to be 255.255.0.0. But I agree, adding IP-specific route for the gateway out to the interface is the best thing to do. - Ted -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Shouldn't tar 1.20-1 be in testing?
On Mon, Aug 04, 2008 at 03:43:51PM +0200, Julien Cristau wrote: On Mon, Aug 4, 2008 at 15:34:40 +0200, Vincent Lefevre wrote: tar 1.20-1 entered unstable on 2008-04-17, so several months before the freeze. The essential toolchain was frozen before the rest of the archive. Yes, but there have been exceptions made; I'm not sure that lzma support would be good and sufficient reasons to sway the release-team, but Julien should try sending that note to debian-release, and see what they say. I'm guessing that perhaps since tar is probably used by d-i, the tar maintainer needed to explicitly request that tar get manually pushed to testing, and that didn't happen back in May/June? - Ted -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Policy or best practices for debug packages?
On Mon, Jul 07, 2008 at 02:41:50PM +0200, Mike Hommey wrote: *) Do we dump everything into /usr/lib/debug, i.e., /usr/lib/debug/sbin/e2fsck? Or should we put it in /usr/lib/debug/pkg, i.e., /usr/lib/debug/e2fsprogs/sbin/e2fsck? Most packages I've seen seem to be doing the former. /usr/lib/debug/$pathoforiginalfile This is where gdb is going to look for these debug info. Yep, that's what everyone else is doing, so that's what I did too with e2fsprogs. *) Is it OK to include the -dbg information in a library's -dev package? Or should it be separated out? Otherwise as more and more packages start creating -dbg packages, the number of packages we have in the debian archive could grow significantly. I think the problem is going to be with size way before becoming a problem with the number of packages. I tried putting them in -dev, and lintian complains with a warning. So I guess while it isn't official policy yet, it seems to be the general practice. The e2fsprogs source package just grew an extra 7 binary packages as a result for a total of 20 packages, though. :-) *) Red Hat includes source files in their debuginfo files, which means that their support people can take a core file and get instant feedback as to the line in the source where the crash happened. But that also means that their debuginfo packages are so huge they don't get included on any DVD's, but have to be downloaded from somebody's home directory at redhat.com. (which appears not to be published, but which is very easy to google for. :-) What do we want to do? You don't need source files in the debuginfo files to get line numbers. You only need line tracking information, and if you build with -g, that is already what you get. True, although it means there's a bit more work to actually install the source package, and then running ./debian/rules build in order to make sure the sources are unpacked and patches appropriately applied. With Red Hat all you have to do is unpack the debuginfo package, and the sources that were used to build the binaries are made available with no muss and no fuss in /usr/lib/debug/usr/src/pkgname. (And an obvious thing for Red Hat to have done is to hack gdb to automatically figure out the location of the source files, possibly by encoding it in the build-id --- although I don't know if they have done it.0 Is this worth the bloat in packages, especially since the -dbg packages are architecture specific and thus would be replicated N times? Probably not, but it's at least worth thinking about the functionality and deciding whether we want to replicate it. Speaking of the -g option, does anyone know off-hand whether or not it's worth it to build with -g3 (to get cpp macro definitions into the DWARF stubs)? Note that it would be better for our users if we could have a debug info server instead of having them install dbg packages, it could be nicer. Obviously, gdb would need to be able to talk to it. Yep, agreed. That would be nice. I will say that having started to build and installing the debug debug packages, for e2fsprogs, being able to run gdb on installed system binaries is *definitely* very nice. :-) - Ted -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Policy or best practices for debug packages?
On Mon, Jul 07, 2008 at 05:39:00PM +0200, Mike Hommey wrote: There are 3 kind of people who need -dbg packages. - Users, when they are asked to provide proper backtraces in bug reports - Developers, when they need to debug stuff - Maintainers Obviously, the latter will be able to get the sources themselves, so do the second, most of the time, though the debian/rules patch thing might be a problem, especially when you need to install cdbs or some other stuff to get it working (only to apply dumb patches, d'uh). Well, maybe the dpkg-source 3.0 package formats will make this easier. It should would be nice if there was a standardized debian/rules target which would unpack the source tarballs and apply any necessary patches, such that the sources were in a state usable by gdb, though. And there was a way that something like apt-get source would automatically run the rule, perhaps on an appropriate command-line option, even better. And she won't have a core for the previous crash because the default is not to core (and BTW, it's uselessly made difficult to override this, see #487879). I just looked at this bug, and are you sure this isn't because you have this configured in /etc/security/limits.conf? - Ted -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Policy or best practices for debug packages?
On Mon, Jul 07, 2008 at 05:39:05PM -0400, Daniel Jacobowitz wrote: Sorry, but this is either someone's uncontributed gcc patches, or (more likely) hearsay. The difference between -g (same as -g2) and -g3 is whether .debug_macinfo is generated - debug info for C/C++ preprocessor macros. It's off by default because the generated data is huge. Do programs like gdb take advantage of the .debug_macinfo in a useful way if it's there? (I guess I should try it and see how big the dbg packages get, and how useful it is for me in practice.) Thanks, regards, - Ted -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Policy or best practices for debug packages?
On Mon, Jul 07, 2008 at 05:42:47PM -0400, Daniel Jacobowitz wrote: I think they do this, using debugedit. We (CodeSourcery) do it for our libraries too. It's incredibly useful - but very spoiling; every time I'm without the automatic debug sources and source paths I get grumpy about it. WANT Where can you find debugedit? I did a google search and it looks like Gentoo as packaged it, and it looks like it might be an auxiliary program inside the rpm source packages? Is that what you were referring to? I wouldn't want them in the archive for everything, but it would be nice to be able to generate automatically usable source packages. Also debug packages without having to create them in debian/control and debian/rules. That would enable build daemons to generate and stash the packages somewhere if we decide to make them available. So in order to do this, we would need hack the debian/rules file to create a foo-dbgsrc package which contains a copy of the source tree after a successful build (but with all of the generated binary files stripped out). We would also need to agree on a standardized location where the *-dbgsrc files would install the source trees --- something like /usr/lib/debug/usr/src, perhaps? And then we would need to use the debugedit tool to edit the dbg files to point at the sources in /usr/src/lib/usr/src (or where we decide to have the *-dbgsrc packages install the source files). I guess depending on the package, the *-dbgsrc file could either be architecture independent, or if some of the generated source files contain arch-specific information, they could end up being architecture dependent files. And we definitely would want to serve these up in a different archive that our standard binary packages, since this would be a *lot* of data. It's probably less important that they get mirrored, though, since it's likely that far fewer users would actually need the -dbg and -dbgsrc packages. Am I missing something, or is that all that's necessary to bell the cat? :-) - Ted -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Policy or best practices for debug packages?
On Mon, Jul 07, 2008 at 06:16:22PM -0400, Daniel Jacobowitz wrote: For various reasons we don't use it at work - instead we added some GCC command line options to relocate the debug info at compile time. In the end, it comes down to the same result. Were these private hacks to GCC? I tried looking at the gcc info file, and I didn't see any options to force the debug info to a different pathname; maybe I missed it, or the info file I was looking for was too out of date (gcc 4.1.3). I think /usr/src/debian/ would be traditional. It really doesn't make a difference, though :-) Yeah, the only reason why I was hesitant about that is I was concerned that some users might already be using /usr/src/debian for their own purposes (in violation of the FHS), and would get annoyed if we started installing stuff there. Hence my suggestion of burying it in the /usr/lib/debug/usr/src hierarchy. But I don't really care what directory name we use, as long as we all agree on some pathname to start. I might start experimenting with with one of my packages, just to see how it works out. I suppose I should send mail to the ftp-masters directly and ask them this questions, but do people think there would be many objections if some -dbgsrc packages started appearing in the archive as an experiment? I could even imagine some hacks where the rules file does a diff of the build source directory against the orig.tar.gz file and the -dbgsrc package only contained the diffs, and the postinstall script searched for the orig.tar.gz file and pulled it off the network, and then applied the diffs, in order to keep the size of the -dbgsrc file small. That might answer the concerns about the size of the archive exploding, if this were to become popular. Regards, - Ted -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: How to handle Debian patches
On Fri, May 16, 2008 at 03:25:11PM -0700, Russ Allbery wrote: In fact, despite being one of the big quilt advocates in the last round of this discussion, I am at this point pretty much sold on using Git due to its merges and branch support and have started to switch my packages over. However, the one thing discussed on this thread is still the thing I don't know how to do easily in Git. I have each logical change on its own branch, so I can trivially generate patches to feed to upstream with git diff upstream..bug/foo, but I don't know how to maintain a detailed description and status other than keeping a separate file with that information somewhere independent of the branch, or some special file included in the branch. How often is a logical change more than just a single commit? Espeically in the context of packaging, usually the changes are pretty trivial, and don't require multiple patches. Sure, a few bugs may require some new infrastructure, or making changes that would be best done with 2-3 patches, but any more than that and you probably want to be consulting with upstream before submit any changes anyway? So normally I just keep those sorts of changes in the commit header, where it is easily and safely bundled with each patch. - Ted -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: git bikeshedding (Re: triggers in dpkg, and dpkg maintenance)
On Mon, Feb 25, 2008 at 12:19:33PM -0300, Otavio Salvador wrote: Robert Collins [EMAIL PROTECTED] writes: On Sun, 2008-02-24 at 16:46 -0300, Henrique de Moraes Holschuh wrote: Yet, rebasing is still routinely performed in the Linux kernel development. What I find interesting and rather amusing here is Linus talking negatively about rebase: in particular its propensity to turn tested code (what you actually committed) into untested code (what you committed + what someelse has done, in a version of a tree no human has ever evaluated for correctness). If people doesn't test and review the patches after rebasing, it looks right but everyone is suppose to test the changes after a merging (as for rebasing). I'll note that when I submit a branch, I prefer to do a rebase, and *then* do extensive testing. That's because for a new feature, I generally understand it better than the upstream maintaining, and *I* want to be the one doing the merge and testing after the fact, as opposed to assuming the upstream will do the appropriate merge fixups and testing. For big projects, this is essential, and Linus in fact does *not* test after doing a merge. (It doesn't scale for him to test after every single merge from his Lieutennants.) But for smaller projects, it should really be up to the submitter; I don't think there is any one Absolutely Right Way To Do It. If someone wants to rebase and then test before sending a pull request, I don't think there's anything wrong with that. Especially if the projects have a good regression test suite. (Both git and e2fsprogs have good regression tests, and that makes life *much* easier to test after doing a rebase or a merge; basically, I'll run the full regression test to make sure that anything unanticipated hasn't broken, and then do explicit testing about the feature being merged or rebased.) BTW, because of the regression test suite, and my general policy for e2fsprogs is that I want to make sure that make check has 0 failed tests after every single commit, rebasing to collapse and remove development history makes life easier to fulfill the fully git bisectable with 0 make check failures between every commit constraint. So as long as the person submitting the patch makes it clear that they have tested exactly what is being requested to be pulled, there's nothing wrong with whether or not they do the rebase right before sending the pull request. My preference is to do the rebase and test, and from a smaller project such as dpkg, I can't think of any good reason for the maintainer to force the submitters to follow one approach or another. Regards, - Ted -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: git bikeshedding (Re: triggers in dpkg, and dpkg maintenance)
On Fri, Feb 29, 2008 at 12:40:55PM +, Colin Watson wrote: That's why you should avoid using the branch as basis to others until it's clean and also avoid to make it public (without a reason) too. This makes it more difficult to ask for review while the branch is in progress, which is a valuable property. It is ridiculous to artificially avoid making branches public; a branch is a useful means of collaboration and we should take advantage of it as such. It's a bad idea to base work on a feature where the code is still being under review. Even if you keep all of the historical crap on the branch, to be preserved for ever, it's going to cause merge difficulties for people who base branches based on that patch which is under review. So you really, REALLY, don't want people basing work on code which is still being developed, since it may be that the review may include, why don't you totally refactoring the code *THIS* way, which will end up breaking everyone who depends on your new function interface anyway. So how to solve this problem? (a) Send patches via e-mail for review. This actually works better, because people can respond via e-mail much more easily than if it just shows up in a git repository. You can also send an explicit request for people to review the patch when it is sent via e-mail. (b) Put the patches on a git branch which is *guaranteed* to be constantly rewound, and is not to be used as a base for derived patches. By convention the 'pu' branch in the git (and e2fsprogs) source repository is declared to be one which is used only for people who want to test the latest bleeding edge code, but it should not be used as the basis of any derived or child branches. I have never once run into this problem with other revision control systems in which branching and merging are common. Somehow it just never seems to be a real issue. I contend that dpkg is not big enough for it to become a real issue. It's not a fatal issue, but in the long run, the code is more maintainable if the code revision history is clean. It's like having a few goto's in the code. Does that make the code unmaintainable? No. But it does make it worse. Or think about how much effort some of us spend to make the code gcc -Wall free of warnings. Does not doing it make the code fundamentally bad? No. Is it still worth doing? Many of us believe that it is worth doing, nevertheless. - Ted -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: the new style mass tirage of bugs
On Thu, Feb 21, 2008 at 02:15:10PM +0100, Mike Hommey wrote: Note that also doesn't indicate how many were actually fixed. We have nothing that look like bugzilla's NOTABUG or INVALID. It would be nice if we had this, actually, and it wouldn't be hard, right? Just define a convention for a new tag? - Td -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: How to cope with patches sanely
On Sat, Feb 02, 2008 at 12:04:16PM +, Roger Leigh wrote: While the time might not be yet, DVCS systems are getting to the point where they could make our lives all much simpler. Having all of Debian in git, where anyone can clone and hack would be (IMHO) a worthy goal to aim for. Currently, there are many packages I can't work on--simply because I am not intimately familiar with the patch-system-du-jour the maintainer chose, upstream-tarball-in-orig.tar.gz being my greatest bane. Having a single tool we all need to learn once would (again, IMHO) be useful in fixing this. While I'm a big fan of DVCS systems, and in fact use git all the time --- including git to manage quilt seriess --- I don't think DVCS systems would necessarily be right for Debian. The reason for that is because in the long-run, we do want to get our changes upstream, and not end up in a merge hell where Debian packages have diverged significantly from upstream and merging changes back is hard. The problem with DVCS tools is that they aren't necessarily well suited for that. It's too easy for people to just hack a few changes, then commit, then hack a few more changes, and commit, .etc. You can use tools such as git rebase --interactive to fold related patches and patches which fix bugs introduced in patches, but it's complicated, and not something a beginner DVCS user would find easy. If you force people to maintain patchsets, where each patch has a description describing *what* change was made, and why, it's much, MUCH more convenient for upstream to understand what you've done, and why. And if people want to use git, it's possible using git to take set of git comments and turn it into a patchset which is quilt-compatible. But let that be up to each maintainer. Different maintainers can use whatever tools they want, as long as the output format is a quilt-style patchset. What's far more important is that maintainers are strongly encouraged to maintain a high quality patchset which is suitable for acceptance by upstream. Otherwise, as upstream keeps moving, it will be harder and harder to forward patch our changes, and that way lies some of the headaches which the Ubuntu project has been facing (albeit for different reasons). - Ted -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: How to cope with patches sanely
On Sat, Feb 02, 2008 at 10:26:52PM +0100, Pierre Habouzit wrote: Bah that's the worst reason I ever seen to not use a DVCS in Debian. Please look at the debian packages with patches, and please tell me how many come with a comment about what the patch do. Please start with the glibc if you don't know where to start. Then please resume the discussion about how having patches series make them more commented. I was't arguing that everyone *shouldn't* be forced to use a DSCM. I was arguing against the proposal that everyone be *forced* to use a DSCM, and moreover a the same DSCM. Whether it is git or hg or bzr, forcing everyone to use the same DSCM is a *bad* idea because it makes it harder to do the right thing. Yes, it is quite possible to create patches that are horrible using quilt or dpatch. You can write Fortran code in any language. :-) But it requires more advanced skills using a DSCM to create good, upstreamable patches than it would using something simple like quilt. So again, lest I be misunderstood, using the quilt format is good. If people want to use a DSCM, great; they can use many different DSCM's many different ways to maintain a patch queue. We should not force people to use a DSCM, and not a specific DSCM, regardless of whether it is git, hg, bzr, or arch, just because we think DSCM's are cool. (They are, but that's not a good reason for force everyone to standardize on a single DSCM. :-) And I think it is a much better idea to encourage people to spend their time working on good, upstreamable patches, than to tell them that the need to learn some specific DSCM. - Ted -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: How to cope with patches sanely (Was: State of the project - input needed)
On Sat, Jan 26, 2008 at 11:39:32PM +0100, Pierre Habouzit wrote: I'm less and less sure that a git-based format is a brilliant idea. I like git more than a lot, but it's a poor idea to base source packages on them. That doesn't mean that we shouldn't be able one day to upload a signed git source url + sha1 signed with GPG and see DAK and other tools git pull from there to rebuild the source package using some normalized ways, but a source package should remain as simple as possible, using _textual_ interfaces. Something that we should be clear about is that there's a big difference between the *interface* and the tool. One of the reasons why the quilt *format* is so powerful is that it can be consumed by so many different tools. For example, the quilt format can also be consumed by guilt and stgit for git users, mq for mercurial users; not just by quilt. Just to give one example of how you can use a quilt *format* for a set of patches without actually using the quilt tool. For the ext4 kernel development, we use a svn-style central repository model, but using git at http://repo.or.cz. Multiple poeple have access to push into this central repository; it could have just as easily have been svn or cvs, but a number of us like the ability of being able to have off-line access to the repository. What we *store* in this repository is a quilt-style set of patches, where the series files indicate the base version of the kernel that the patches can be applied against. How people manipulate the quilt series of patches depends on the developer. Some developers use quilt, other (like me) use guilt. It would even be possible for people to write a set of shell scripts to parse the quilt series if they wanted; again, it is the format that is important, not the tool. One nice thing about the quilt format is they can preserve a large amount of state. For example, we use the following convention in the patch header: one line patch summary From: Author O' The Patch [EMAIL PROTECTED] Detailed patch description Signed-off-by: Theodore Ts'o [EMAIL PROTECTED] This is at the beginning of every single quilt patch, and because of this, we can easily important the patch into either a mercurial or git repository, while preserving the authorship information of the patch. So when it comes time to push patches to the upstream, it is much easier to do so, since all of the information is right there. Since we are keeping the patch series under source control, it allows us to refine patches, and keep a historical record of the improvements to a particular patch before it us pushed upstream. So for the people who were concerned because quilt is unmaintained, I wouldn't worry, because even if there is no one maintaining the official quilt sources, there are plenty of other quilt work-alikes that use the same format, and many of them are integrated into other SCM's. Andrew Morton actually maintains is own private collection of shell scripts to parse quilt patch series for his -mm kernel tree. Andrew's tools is what the current quilt system is modeled upon, but Andrew is still using his own set of shell scripts. So that's another example of Yet Another tool that all use the same quilt series file format. So maybe what we should do isn't trying to make the Debian source package a complex mix between a snapshot and a SCM, but rather let it be simple, and only a snapshot, and have ways to encode in it where to find (and how to use to some extent) the SCM that was used to generate it. Yes; all a Debian source package should be is a snapshot of a series of patches that can be applied against a source base. The series of patches can be maintained in an SCM, and that's a useful thing to do; but for the purposes of what you ship in a source package, the full SCM data itself is not strictly necessary. Regards, - Ted -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Bug#459403: libuuid1: missing depends on non-essential package passwd
On Sat, Jan 12, 2008 at 03:41:01AM -0800, Steve Langasek wrote: I don't think but we don't want to make adduser Priority: required is a good enough reason to add global static IDs; and passwd doesn't need to be made Essential just because an Essential package depends on it (Essential isn't closed under dependency). The use case here was that passwd would be a dependency of a pre-dependency of an essential package, which does promote it into the effective essential set. Well, I'm happy to make a request for a static ID, or not; do we have a consensus here whether I should do this? - Ted -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Bug#459403: libuuid1: missing depends on non-essential package passwd
On Mon, Jan 07, 2008 at 11:34:07AM +0100, Bastian Blank wrote: On Mon, Jan 07, 2008 at 03:23:23AM -0500, Theodore Tso wrote: So, I am doing so now. Any objections if I add a dependency on passwd for libuuid1? The aternative would be to roll-my-own useradd/adduser functionality, but that would be a real PITA There are several other possibilities: - Move the user creation. It is necessary for uuid-runtime (uuidd), not libuuid. It's actually needed in libuuid, because one of the modes how it can be used is a program can be setgid libuuid. If so, then when it calls libuuid, it can optionally save the clock sequence number in /var/lib/libuuid, and to detect circumstances when the clock goes backwards. This guarantees that uniqueness, and makes the UUID generation compliant with RFC 4122. Without libuuid being setgid, or without the uuidd package installed, the time-based UUID will *probably* be unique, but if the time gets reset, and you get very. Very unlucky, or if you have multiple SMP threads generating huge numbers of time-based UUID's, you could generate duplicate UUID's. What is the reason for this deamon anyway? It linearises the requests and limits the amount of available uuids. So there are some programs that like to use time-based UUID's because they tend to sort much better (particularly if you byte-reverse the UUID) when they are used as keys in a database. One such program generates a *huge* number of them while initializing its application database. It does so in multiple threads and multiple processes running in parallel, and it was running into problems because it could end up generating duplicate UUID's. The uuidd linearises the requests, yes, but in a very clever way where each particular thread can request a block of UUID's (with contiguous times) so it actually allows a much, MUCH larger number of UUID's to be generated, with *significantly* less CPU overhead. I know of only one application program that generates this many UUID's (as in tens of thousands per second). Unless you use this application, it's relatively unlikely that you need it; however, it is a matter of correctness as well --- if you want to guarantee uniqueness on an SMP system, and guarantee uniqueness in the face of an unreliable clock that might go backwards, implementing the requirements of RFC 4122, it's necessary. By default though most applications use random UUID's, for which none of this rigamarole is necessary. So without uuidd installed, the amount of time to generate 100,000 UUID's is 3.2 seconds. With uuidd installed, the amount of time to generate 100,000 UUID's drops down to 0.091 seconds. - Ted -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Bug#459403: libuuid1: missing depends on non-essential package passwd
On Sun, Jan 06, 2008 at 10:37:03AM +0100, Julien Cristau wrote: the libuuid1 postinst contains: groupadd -f -K GID_MIN=1 -K GID_MAX=999 libuuid if ! grep -q libuuid /etc/passwd; then useradd -d /var/lib/libuuid -K UID_MIN=1 -K UID_MAX=499 -g libuuid libuuid fi mkdir -p /var/lib/libuuid chown libuuid:libuuid /var/lib/libuuid chmod 2775 /var/lib/libuuid The groupadd and useradd commands come from passwd, which is not Essential: yes, so a Depends is needed. Moreover, the postinst succeeds even if any of these commands fail, because 'set -e' is missing at the top of the script. So e2fsprogs which is Essential: yes depends on libuuid1, so libuuid1 is effectively Essential: yes, right? So if I add a dependency on passwd, it will effectively make passwd Essential: yes, as well, and according to policy I should bring it up for comment on debian-policy. So, I am doing so now. Any objections if I add a dependency on passwd for libuuid1? The aternative would be to roll-my-own useradd/adduser functionality, but that would be a real PITA Thanks, regards, - Ted -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: List of packages which should probably be Architecture: all
On Thu, Jan 03, 2008 at 09:50:48AM +1100, Brian May wrote: Raphael == Raphael Geissert [EMAIL PROTECTED] writes: Raphael Brian May [EMAIL PROTECTED] Raphael dar-static Raphael Theodore Y. Ts'o [EMAIL PROTECTED] Raphael e2fsck-static Both of these (and maybe others) are false positives. Which would be caught if the test actually say, used the file command for any binary that had files installed in a .../bin directory or .../lib directory. BTW, I recently got a complaint from someone who was still using a 2.4 Woody system, and had been using e2fsck-static as a way of getting the latest e2fsprogs fixups for e2fsck, given that the woody backports effort had stopped a while ago. Apparently the latest glibc uses thread local storage in its locale code, so even linking statically against glibc will result in a binary that can't be used on a 2.4 kernel. Because I was a nice guy, I hacked up e2fsck-static to build against dietlibc instead, so it would work on ancient systems. Dropped the size of the binary from over a megabyte to about 330k, a savings of two-thirds - Ted -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: MySql broken on older 486 and other cpuid less CPUs. Does this qualify as RC?
On Fri, Apr 06, 2007 at 12:36:18PM +0200, Andreas Barth wrote: * Francesco P. Lovergine ([EMAIL PROTECTED]) [070406 12:33]: On Thu, Apr 05, 2007 at 04:10:19PM -0700, Steve Langasek wrote: Like I said, in practical terms, if a bug like this in a major server package goes unnoticed until 3 days before the release, we are not actually supporting 486. We support the i386 architecture quite well, but it seems only honest to admit that as a project, we don't care about 486 enough to even get 486-specific problems marked as RC in time to do anything about them for a release. That could be fixed in R1, isn't it? I see no major problems on those regards... Well, we can fix it - but are you sure that's the only package with an issue on 80486? I think we should put somewhere into Lenny that our code should still run on 80486, but it might not be QAed enough for all subarches (but we can discuss that later, don't need to reach a consensus now). The fact that we can't say for sure is a good reason to say that we don't support 486. That is to say, we can't guarantee that it will work, because we don't have enough people who are testing on that platform. There is a difference between we won't consciously break 486, and we'll apply patches when they are called to our attention, but we won't hold up an entire release for it, and support. For all that people like to beat up on Red Hat and Fedora for their supposed lack of quality, this is a concept which Red Hat and other commercially supported enterprise distributions understand. If they can't test on a platform, they don't call it supported. (And they do get help from major system vendors to do a lot of very serious regression testing before they say that it is supported on a particular platform.) If we are going to claim that Debian is stable enterprise system for servers, so stable in fact that we must use an obsolete version of glibc compared to RHEL and SLES (even though we are releasing later than RHEL and SLES), then it would be wise for us to use at _least_ as stringent set of QA gaurantees as Red Hat and Novell. And if that means that we can't say that we support the 486, then so be it. - Ted -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: On maintainers not responding to bugs
On Fri, Mar 02, 2007 at 06:57:01PM +0100, Pierre Habouzit wrote: You forgot a single damn point: in debian, like in many projects, the one that do things is often the guy that decide things because he's the one there. If you put people that work 5 times more as me because they have the time to do that, I will obviously feel they took my place. I'm not sure what I would do in those cases. Obviously not refusing the help and people that have the time to do this, but I would obviously lessen my implication and work for other teams where I've a single damn chance to see my contribution to be compareable to the others. The principle you stated obviously tends to be the case in volunteer organizations, true. It does not have to be the case of a paid employee, but yes, even if the maintainer team sets the general policy and gives direction to the employee, there will be a lot of hour-to-hour operational control which you will be ceding to the employee (unless you want to be one of those awful micro-managing managers --- but that's not a path to productivity, either for the manager or the managee! :-) I would feel bad to impose my views to a person that has huges amounts of time to work in the team. And necessarily (because of human nature) a decision will happen that I would not like or would have made differently, and at that point I guess that I would just leave. Well, anytime we have people working on a package with team maintenance, there are bound to be disagreements. If we all left the moment a decision was made that we didn't agree with, Debian would be empty. I can't read your mind, of course, but it may be that the harder psychological hurdle is the one where a DD realizes that (say) 10 hours a week of volunteer labor is no longer enough to be one of the primary contributors on a package team. That is a real issue, and perhaps maybe _the_ major issue. Whereas in balanced teams where every contributor has the same level of contribution, I would have argued my point, or tried to make the proposal better, or discussed it or... whichever adequate behaviour in a team where every single member is equal to the other. Although this may be a great platonic ideal, the reality is that no team is going to be completely balanced. First of all, not everyone does _can_ contribute at the same level. Some people have jobs that cause them to work very long hours, for example. (And of these, some of them would like to be able to help Debian, and one of the ways they might be willing and able do so is via contributing money, not time.) Secondly, even if assume that everyone could give the same number of hours of contribution, the reality is that different people have different levels of talent --- and it is still the case in the programming world that differences in talent can account for 2+ orders of magnitude in productivity. So in the end, talent will probably still dominate far more than whether someone can work 10 hours as a volunteer versus 40 hours as a paid employee. (This I think is one of the ways that an examples of paid versus volunteer firemen may not be applicable for Debian.) Money introduces bias. OK you were talking about bug triaging, and bug triaging is not necessarily a big decision making place, I agree. Though it will depend a lot of the kind of people you want to recruit: * if those are already contributors they will want to take more and more decisions, and won't only do bug triaging: if you do bug triaging you begin to know packages a lot, and become skilled enough to take decisions, and so on. Then commits rights are granted, and you take more and more responsibilities. That's good, it's indeed what is often suggested to newcomers. Though we end up in the not-so-nice situation I described. Money introduces bias; no question. But it also does introduce a certain amount of control. For example, suppose we only paid people to do bug triaging. If they want to do more, that's great but it will be on their own time. Would something like this magically make all of the problems go away? Of course not, but I think it shows that with the right amount of thoughtfulness, it's possible to make the benefits outweight the costs. but please, I'm not sure there is a damn single maintainer in a big team that will refuse help, paid or not. I don't really understand how that mythical maintainer in a big team that refuses help has emerged in the discussions, but I'd really like names here. In fact, that seems pretty contradictory with the very notion of a team. Of course, there is teams with 1 single member in it in debian, but that's not a large team and is out of the scope if I'm not mistaken. Well, Josselin has been very negative about the whole concept of paying volunteers, and given that he was asking for help, and saying that his GNOME team was drowning under bug reports, I couldn't help but reply that if he
Re: On maintainers not responding to bugs
On Tue, Feb 27, 2007 at 10:30:39AM +0100, Josselin Mouette wrote: Le mardi 27 février 2007 à 09:24 +0100, Eduard Bloch a écrit : And how do you help a maintainer that does not admit that he needs help? I can't believe people are thinking such crap. Please show me where a current maintainer of Mozilla, KDE, GNOME, the glibc, the kernel, X.org or any such big group of packages said he didn't need help for them. YES. WE NEED HELP. NOW. We are *all* *COMPLETELY UNDERSTAFFED*. We are drowning in bug reports and are not able to answer all of them, especially old ones dating from the pre-teams era. Who is not acknowledging such obvious things? So how do you help a maintainer who refuses help if it is paid? OK. The large teams are *COMPLETELY UNDERSTAFFED*. Volunteer labor is not able to keep up. Suppose we or some outside organization like dunc-tank raised money to pay someone who could afford to work full-time, 40 hours a week, doing bug triage for these large projects. Would those projects refuse help if the people doing the work happened if some of the people who showed up to help you from not drowning in bug reports just happened to be paid by Debian or by an outside group to do this work for which you have so eloquently said it's hard to get volunteers, and for which others have said is completely unfun, tedious work? Even if one or two people (for reasons that I don't understand) would stop spending maybe 10-12 hours a week on top of whatever they do during the day to earn money to feed his family because there is now some paid help, I would think that raising money to find someone to work 40 hours a week would be a Good Thing - Ted -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: On management
On Thu, Mar 01, 2007 at 03:28:50PM +0100, Turbo Fredriksson wrote: Quoting Josselin Mouette [EMAIL PROTECTED]: As of now, I see it as a failure of the project. But this is also nothing that can't be fixed. What do you people think could be done to bring the skills we are lacking to the project, with its current structure? Since I agree (well, more than agree - I think it's absolutly vital :), how about actually PAYING them? Get one or two professionals and pay them (halftime perhaps)... I have no idea how the economy looks, but would there be room for such a thing? A third party organization, dunc-tank, which included a number of prominent Debian Developers, including AJ, did pay for two of the release managers to work full-time for approximately a month at a time, at the end of 2006. This was actually highly contentious with some claiming that it delayed the release because it demotivated them. I know of now hard evidence to prove that the net result was negative, and certainly during the period when the two release managers were paid, the RC bug count did decline quite significantly, and you can see a significant difference in the slope and sign of the derivitive of the RC bug count before and after that period where two of the RM's were paid. Certainly the fact that we did pay them did cause a huge amount of flaming on various debian mailing lists, which perhaps might have delayed the release, but my personal opinion is that DD's being DD's, there would have found some other topic to flame about, whether it was license issues, or whether to expel a particular DD, or something else So I believe the paying RM's was a net positive, and there are those who would disagree with me. (And to be fair, I need to disclose that I was one of the people who helped to organize dunc-tank.) The harder problem, though is that finding a really good project manager. In my day job, I can tell you have as a technical architect, having a great project manager is like pure gold. My project manager defers to me on technical issues, but helps to coordinate all of the other technical teams so that we can make all of the schedules line up and release a coherent solution to the customer. Not to denigrate the efforts of the current release management team, but if we were to augment (NOT replace!) them with a good project manager, we could make them be far more effective. - Ted -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: /foo has been mounted xx times... check forced
On Tue, Feb 20, 2007 at 11:36:21AM -0800, John H. Robinson, IV wrote: Andrei Popescu wrote: On Mon, 19 Feb 2007 13:29:46 +0900 Charles Plessy [EMAIL PROTECTED] wrote: how about a I'm in a hurry boot option in GRUB, which would make the e2fscks skipped ? Too early. You might not know that a check is due. Perfect time: you already know you are in a hurry. It could be possible to use other tricks to shorten the boot cycle. I can't think of any at the moment, but that does not mean that they don't exist. Does XFS require fscks? Reiserfs does not. Maybe it is time to ditch ext3. You don't *have* to do the periodic checks. If you want you can disable it using tune2fs. tune2fs -c 0 -i 0 /dev/hdXX. The reason why ext3 has periodic checking is a *feature*, born out of the recognition that hardware is not perfect, and in fact, commodity class hardware can and does fail in various entertaining ways. By running e2fsck periodically, we hope to catch problems while they are small, instead of after massive data loss. But hey, if you know you have perfect hardware, and you do regular backups (YOU DO REGULAR BACKUPS, **RIGHT**?), hey, feel free to disable the periodic fsck's, or dial them back to a higher level. (For me, since I normally use suspend to disk/ram quite a lot on my laptop, the periodic check happens quite rarely --- except when I am rebooting a lot due to trying out lots of different kernels, but then I *want* to do the periodic checks just in case a kernel bug caused a filesystem corruption problem.) Finally, I will note that different filesystems generally get tuned to assume different use cases. XFS in particular fundamentally assumes that you are using drives (i.e., RAID at high levels) in data center conditions, and that you have a UPS to protect your system from power failures. (Yes it has a journal but the way it prevents security breaches if it's not sure the data block was written before the metadata was is to zero out the data block). Ext3 is more often used in cheap-*ss commodity equipment or for equipment with less-than-perfect drives (like laptop drives that tend to get banged around a lot when people shove the laptop into their knapsack and start walking off while the suspend-to-disk is in process), so it has a bit more paranoia about hardware designed into it. Regards, - Ted -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: /foo has been mounted xx times... check forced
On Tue, Feb 20, 2007 at 11:14:21PM +0100, Josselin Mouette wrote: Le mardi 20 février 2007 à 20:55 +0100, Mike Hommey a écrit : Does XFS require fscks? Reiserfs does not. Maybe it is time to ditch ext3. ReiserFS requires as much fsck as ext3. But it is much faster. In the worst case, when the filesystem is badly corrupted, ReiserFS will require reading every single data block off the disk, at which point it will look for every single block that *looks* like it might be part of an Reiserfs b-tree, and stich it together. The results if you have multiple Reiserfs filesystem images (for use by qemu, UML, Xen, VMware, etc.) in a resierfs filesystem, and the filesystem is badly corrupted, I will leave to you to imagine. (But a scene from from your favorite frankenstien movie might not be a bad place to start. :-) Also, reading every single data block from disk will almost certainly take longer than an ext3 filesystem check, which is one of the advantages of having a fixed inode table; ext3 knows where to start. - Ted -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: /foo has been mounted xx times... check forced
On Fri, Feb 09, 2007 at 10:55:49AM +0100, Enrico Zini wrote: Right. But would it actually be officially safe to interrupt with ^C ? That would give the user an opportunity to decide how in a hurry they are, and quickly get out of a difficult situation. If the answer is yes, ^C is officially safe, then I propose to add if in a hurry, interrupt with ^C to the check forced message. It's not a great idea to do this indefinitely, and it's a matter of whether or not you trust the person in front of the machine not to be in a hurry and to always type ^C all the time to avoid the e2fsck run. If the owner/administrator of the machine == the person who is normally in front of the console during the bootup (as is the case in a laptop and most single-owner machines), the obviously it should be up to the owner/admninistrator. At the moment, if you want ^C to interrupt the e2fsck and you want the boot to continue, you actually have to set the following in /etc/e2fsck.conf: [options] allow_cancellation = 1 See the e2fsck.conf(8) man page for more details. Regards, - Ted -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: /foo has been mounted xx times... check forced
On Tue, Feb 06, 2007 at 12:50:33PM +0100, Wouter Verhelst wrote: On Sat, Feb 03, 2007 at 09:39:24AM +, Enrico Zini wrote: Hello, the feature as in the subject is nice and makes me feel safe, but sometimes it hits on the laptop, when booting on batteries, with people watching. There actually is a feature in e2fsck to double the amount of mounts before an fsck is done if you're running on batteries; so unless you boot from batteries all the time, this shouldn't happen. See #205177 and #242136. You do need a mounted /proc at that time, though, which may be the reason it's not working for you. A mounted /proc and if ACPI has been built using modules, the ACPI battery module needs to be installed, since that's how we tell whether we are running on the AC mains or battery - tED -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: update on binary upload restrictions
On Thu, Jan 25, 2007 at 03:44:39AM -0500, Jaldhar H. Vyas wrote: On Thu, 25 Jan 2007, James Troup wrote: Summary Credit where credit is due. This is exactly the kind of informative explanation I have been looking for and I hope we'll see a lot of more of this sort of thing from the infrastructure teams in the future. Preferably before the flames start. Indeed, I have to second this, and state that for the next DPL elections, I will be looking to vote for DPL's that make as part of their campaign platform that Delegates will be appointed or removed based on not just how long they have been working on a task, or how good they are technically, but how well they can communicate with others, and how well they can send out reports and rationals to mailing lists (IRC isn't good enough!) **before** people start to flame and start talking about GR's to replace teams that can't seem to communicate with anyone else. (In fact, I'd prefer work getting delayed by 2-3 hours if that meant getting an adequate report sent out; Lord knows Debian has a whole lost a heck a lot more time than that due to the controversy caused by the buildd restrictions.) So again, I hope we can see a lot more reports like this, and not just in reaction to massive flame-fests on IRC and mailing lists and key individuals taking vacations out of sheer frustration. - Ted -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Bug#397939: Proposal: Packages must have a working clean target
On Sat, Nov 11, 2006 at 10:55:57PM -0600, Manoj Srivastava wrote: What you are saying, in essence, is that we have not been treating autoconf transitions with the care we devote to other transitions; and as a result people have started shipping intermediate files. While I recognize these past lapses, I am not sure why we should condone and in the future pander to them -- I am hoping that autotools are coming of age, and there shall be few major API changes in the future. Post etch, perhaps we can evaluate if overhauling our autotools usage in Debian to allow treating autoconf like we do lex and yacc -- and building from sources -- is feasible, or not. Why don't we wait and see if autoconf can manage to go through a half dozen or so releases without breaking backwards compatibility, first? And that's assuming we get some kind of commitment from the autoconf maintainers that they care about backwards compatibility in the first place. I just recently had to put in special case hacks in e2fsprogs so it could support both autoconf 2.59 and autoconf 2.60, and that's why e2fsprogs is still shipping the generated configure script in the upstream sources. Fundamentally, I still don't trust autoconf to be stable in terms of the configure.in constructs which it supports. - Ted -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Debian Installer etch RC1 released
On Tue, Nov 14, 2006 at 04:52:57PM +0100, Andreas Barth wrote: There are good chances that Etch will contain 2.6.18, but due to some open bugs the Release Candidate 1 of the installer has still 2.6.17. Because of the sysctl deprecation issue, it might be a good idea to either consider 2.6.19, or probably a better solution would be to backport the sysctl undeprecation patches into 2.6.18 - Ted -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: RFC: behaviour of bts show command with new BTS default behaviour
On Sun, Nov 12, 2006 at 01:02:06AM +, Julian Gilbey wrote: Thinking of changing the default behaviour of the devscripts bts show (aka bts bugs) command, and want to ask for opinions before I do so. The BTS behaviour of http://bugs.debian.org/package has recently changed. It used to resolve to: http://bugs.debian.org/cgi-bin/pkgreport.cgi?pkg=package which listed all open and closed bugs in the package, without any version tracking being taking into consideration in the listings. It now resolves to: http://bugs.debian.org/cgi-bin/pkgreport.cgi?pkg=package;dist=unstable which lists the status of all bugs in that package in respect to the version(s) of the package currently in unstable. I prefer the new behaviour, myself. If it must be changed, can it be configurable via some kind of .btsrc file? Also, can we please have this functionality for bts cache? It takes a *long* time to download a huge whackload of bugs, many/most of which are already fixed in unstable, and so don't matter to me at all. - Ted -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: default ext3 options
On Tue, Nov 14, 2006 at 12:35:30PM +, Sam Morris wrote: As far as I know, neither the resize_inode nore the dir_index ext3 option can be securely added after the file system is created. According to http://groups.google.co.uk/group/linux.debian.devel/msg/4d987ea414438e70, it should be perfectly safe to add dir_index to an existing filesystem. However, to get the benefit of the indexing for already-created directories, e2fsck -D should be run after dir_index has been added; therefore it's probably best to just document the procedure in the release notes. Correct, that it's perfectly safe to add dir_index to an existing filesystem. You can even do it to a mounted filesystem, and any new directories which are created and grow beyond a single block will use the directory indexing feature. (So yes, mkdir foo.new; mv foo/* foo.new; rmdir foo; mv foo.new foo will work module locking/race conditions with applications trying to read/write/create/remove files in the foo directory.) If you want to force all directories to be optimized, you can do an off-line (unmounted) e2fsck -fD command. The resize_inode however can not be cleanly to an existing filesystem. There is an ext2prepare command which can be used to do an offline add of resize_inode to an unmounted filesystem; it can be found in ext2resize package. The reason why that functionality hasn't been integrated into e2fsprogs is because I took one look at the source, and decided it had to be rewritten from scratch before I would trust it with my data and before I would be willing to be responsible for maintaining it; it's on my TODO list. That being said, I'm not aware of anyone who has lost data using ext2prepare. Each user will have to decide on their own whether or not you are comfortable using it. Regards, - Ted -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Proposed new POSIX sh policy
On Wed, Nov 15, 2006 at 11:36:44AM +0100, Gabor Gombas wrote: I'm not talking about the local admin. Right now Debian maintainer scripts are not allowed to use the enable command because that is a bashism, and more importantly there is _no reason_ to use the enable command because simply saying make /bin/sh point to dash if you want to go faster is much more effective and easier. What I'm saying is if you take away the freedom of allowing /bin/sh to point to dash, then people who care about shell performance will be forced to use other means _even in scripts shipped by Debian_ - and the enable command is a very powerful tool to achieve that. And at that moment you will have exactly the same builtins aliasing different external commands in different scripts problem as you have now when allowing different shells - so you gain nothing by restricting /bin/sh to bash. We don't have to require bash as the one and only /bin/sh in order to enable maintainer scripts to use enable. They could test to see if they are running under bash and only use enable bash is being used. If that isn't allowed by policy it should be. The rule should just be that the script *works* on non-bash shells (and I really like the proposal that says a script must work on bash, dash, and a shell to be named later), not that it be free of bashisms. - Ted -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Why does Ubuntu have all the ideas?
On Thu, Aug 24, 2006 at 11:56:21AM -0500, John Goerzen wrote: This sort of vague anecdotal evidence has been repeated over and over. It may be true, but as far as I know, nobody has yet to come forth with reporting specific problems in Debian, only x worked out of the box in ubuntu but not in Debian. OK, I have a brand-spanking new IBM/Lenovo T60p laptop with nice, fast SATA Drives, Intel Dual Core CPU's; 1600x1200 display --- sexy machine. Debian stable doesn't run on it. Ubuntu 6.06 LTS installed out of the box on it. So the laptop that I run when I give presentations at conferences says Ubuntu Breeze when I fire it up, and not Debian. My brand-spanking new home file server with a a real hardware RAID controller (Areca) and 16 hot-swap SATA drives (6 of them currently populated with 500 GB SATA II drives), with two dual-core Xeon chips is running Ubuntu Breezy 6.06 LTS, because Ubuntu supported it out of the box. Debian stable doesn't even have a chance of supporting this box. I'm not sure if Debian etch will support it, since the Areca RAID card is an out-of-tree (although GPL) device driver, but that's largely irrelevant, since I'm not going to run Debian unstable on a production file server! How many more concrete example would you like? - Ted P.S. So at the moment, I'm doing my debian development work using some crash-and-burn machines at home, and using some debian chroots created using debootstrap. It would be nice if I could help doing more dogfood testing on etch, and finding and submitting bug reports before etch shipped --- but life is short, and Ubuntu worked out of the box on my laptop, and it would have really difficult to figure out how to get Debian installed onto a laptop SATA drive. (And when I say work, I mean including making use of a built-in wireless on a PCI express bus that requires a driver that uses restricted firmware and a binary-only userspace daemon. So if people want to be pure, that's fine, but I'm not going give up using wireless just for ideological purity; I just won't be using a Debian supplied kernel, and once again, Ubuntu works out of the box. Oh well) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Why does Ubuntu have all the ideas?
On Sat, Aug 26, 2006 at 02:16:53AM -0500, John Goerzen wrote: OK, I have a brand-spanking new IBM/Lenovo T60p laptop with nice, fast SATA Drives, Intel Dual Core CPU's; 1600x1200 display --- sexy machine. Debian stable doesn't run on it. Ubuntu 6.06 LTS installed Out of curiousity, why not? No support for: (The * are critical) * SATA Hard Drives (*) * IPW3945 wireless (*) * Intel AD1981 HD Audio (*) * 3D Graphics support on the ATI FireGL V5200 card (propietary kernel module) * Verizon 1xEV-DO Pretty much all of the modern hardware on the T60 is completely unsupported by Debian; and most of the above is supported out of the box by Unbuntu. It sounds, though, that your problem could be solved if we revved the kernel in stable (and the installer) more often. See my message on -project about that. A lot of the problems would be solved with that, yes. Depending on backports so you don't have to depend on antique versions of OpenOffice, firefox, etc. Maybe the answer is getting modern kernels and modern installers should be adopted by backports.org. Or maybe backports should be considered a 1st class part of Debian; so in addition to old-stable, stable, testing, and unstable, we could add, stable-useful. The fact of the matter is that the stable distribution today is pretty much useless for desktop users, and useless for people who need to install on modern servers (i.e., anything sold in the last 6 months). But yet, we claim that our highest goals in our social contract is to serve our users. Sure. and anything that stable isn't useful for are simply not our users, but they become Ubuntu's users or Fedora's users instead. - Ted -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Why does Ubuntu have all the ideas?
On Sat, Aug 26, 2006 at 04:02:04PM +0200, Hendrik Sattler wrote: Am Samstag 26 August 2006 15:15 schrieb Theodore Tso: No support for: (The * are critical) * SATA Hard Drives (*) * Intel AD1981 HD Audio (*) This stuff did not even exist when Sarge was released. Half of userland would not fit this hardware, so who cares. Umm, the people owning this laptop who choose Ubuntu instead of Debian care. - installer did not read in the CDs for package lists and the GUI does not even support this (or for any other means of modifying /etc/apt/sources.list) From the menubar. System -- Administration -- Synaptic Package Manager Wait for the package manager to come up, click on Settings -- Respositories There is an Add CDROM button, and you just click on it. (No need to run vi, or emacs, or need to understand the /etc/apt/sources.list format.) Seems pretty user-friendly to me. - /etc/resolv.conf was not present but DHCP client complained about that Hmm, I didn't notice this problem. When the dhcp client started during the install process, it created the /etc/resolv.conf file for me, and subsequent dhcp clients updated the /etc/resolv.conf file information automatically from the DHCP serve. - the root has no password and you must use sudo sucks for many things as the access to root is not consistent (some invocation type can use su programs but those cannot work). That's a philosophical dispute, but it's easily fixed simply by setting a root password if you really want to use a root shell. (Or by just doing sudo bash, of course.) I happen to like having a root user with a password and to su to root, so I set up my system that way. However, I view that as an emacs vs. vi sort of religious dispute. - X ran with the wrong resolution (typical i915 problem) and with the wrong dpi setting Can't speak to that; my ATI Firegl video worked automatically out of the box --- with 3D accelerated graphics automatically. - /etc/network/interfaces listed non-existant devices and because of WPA, a manual setup of this file is needed I didn't notice that problem. - something useful like ifplugd was not installed and the user was puzzled by the fact that plugging in the network cable did not result in network access I agree that it would be nice if ifplugd or laptop-net were installed by default, but last I checked Debian didn't install either by default, either. So what's your point? - Ted -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Why does Ubuntu have all the ideas?
On Sat, Aug 26, 2006 at 03:26:12PM +, Sam Morris wrote: On Sat, 26 Aug 2006 16:02:04 +0200, Hendrik Sattler wrote: Am Samstag 26 August 2006 15:15 schrieb Theodore Tso: No support for: (The * are critical) * SATA Hard Drives (*) * Intel AD1981 HD Audio (*) This stuff did not even exist when Sarge was released. Half of userland would not fit this hardware, so who cares. How do other long-lived distributions handle this problem? How does one install RHEL 4 on such a machine? RHEL4 has updated kernels/installers that have additional device drivers added. And of course it helps that Red Hat has new releases somewhat more frequently than Debian does with its stable releases; but that's one of the downsides of relying on an all-volunteer engineering base. Things get done... whenever. - Ted -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Stuff the installer does which isn't done on upgrade....
On Tue, Aug 15, 2006 at 10:53:14AM +0200, Goswin von Brederlow wrote: Did they fix that? When I first looked into dir_index it was said that it would corrupt the directories since it would search the old linear dirs via hash and insert new entries by hash into linear dirs. Who said that? I was the one who merged the dir_index feature into Linux 2.5/2.6, and that was never the case; it was always safe to enable dir_index on the fly, by desgin. - Ted -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Stuff the installer does which isn't done on upgrade....
On Fri, Aug 04, 2006 at 03:02:34PM +0200, Goswin von Brederlow wrote: For dir_index you have to take the FS offline, tune2fs and fsck it or you totaly corrupt it. Actually, that's not true. It's perfectly safe to run tune2fs on a mounted volume to enable the dir_index feature. All directories from that point on which grow beyond a single disk block will use the hashed-tree optimization. e2fsck -fD run off-line is only needed in order to upgrade existing large directories to use dir_tree. (Of course, you can also do something like cd ~/Maildir; mkdir cur.new; mv cur/* cur.new; rmdir cur; mv cur.new cur if you want, as long as you are confident you don't screw up or confuse a currently running mail client.) - Ted -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Stuff the installer does which isn't done on upgrade....
On Sat, Aug 05, 2006 at 07:58:20PM +0200, Mike Hommey wrote: I don't know about the installer, but all filesystems I created with mke2fs recently also have resize_inode, which isn't even in the tune2fs manpage. It's not in the tune2fs man page because e2fsprogs doesn't currently support adding the resize_inode feature after the fact; this is one you have to do at mke2fs time. (There is an ext2prepare program that will do this, but the code has been too scary for me to just integrate into e2fsprogs with a complete rewrite, and I haven't had time to do this.) The e2fsprogs-udeb package does include the mke2fs.conf file, so if the installer is using the latest e2fsprogs-udeb, it should be creating filesystems with the dir_index and resize_inode features. - Ted -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Non-DD's in debian-legal
On Mon, Jun 12, 2006 at 09:35:32AM -0400, Jeremy Hankins wrote: This is one of the most common accusations leveled against d-l: that the membership of d-l is skewed and not representative of Debian as a whole. If that's true there's not much d-l can do about it, of course, and the whole process of license evaluation should perhaps be rethought. The simplest solution, though, is for those who think d-l skewed to start participating. I think what's concerning to most (it concerns me) is that people seem to be _avoiding_ d-l, presumably because they see it as invalid or corrupted by weirdos. That's indicative of a serious problem, because it means licensing issues aren't being discussed _at all_. As saddened as I would be if d-l went private, if doing so is the only way to solve that problem it's probably a good idea. The d-l list has a problem which is shared by many Debian mailing lists (including debian-vote and debian-devel, and I'm sure it's not limited to them) which is that far too many people subscribe to the last post wins school of debate. People don't listen, they just assert their point of view --- back and forth, back and forth. Foo! Bar! Foo! Bar! In addition, far too many people treat mailing lists like irc discussions, where one-line witty reparte's provide entertainment (perhaps) but do not necessarily further bringing an issue to closure. As a result, I have deliberately avoided d-l, because I have better things to do with my time. If you want to reform d-l, it's not enough to ask people to just participate. It's going to be necessary to enforce some cultural changes to how participants on Debian mailing lists behave. They need to be respectful of the other participant's time, and not just use the excuse of free speach to any kind of anti-social and self-centered behaviour. Unfortunately, the only thing I can think of that might be useful would be active moderation of the list, combined with summary of the opinions (with both majority and minority opinions) that is summarized by the moderator, and which when it is due, can be archived on some web site or wiki. Yes, that means that d-l won't be the home to free-spirited, free-ranging debate; instead, there might be structured discussion that actually leads to light being shed and work being accomplished in an efficient manner. But it does mean that a moderator has to be found that can declare certain discussions as ratholes, and be capable of fairly summarizing the positions being espoused by various camps on the list, and holding straw polls that are based on participants on the lists, and not by number of postings on the list (which unfortunately leads to the last post wins abuse and style of discourse that we see on all too many mailing lists). If everyone participating on the list were mature and grown-up, this wouldn't be necessary. And I would suspect that the call to restrict d-l to only DD's is a hope to exclude some of the more immature and less disciplined posters. But, as we all know, being a DD does not guarantee social maturity, so I don't believe that is necessarily the best way to do things. However, I *do* believe that d-l is a cesspit, and I for one am very glad that in its current incarnation, it is not at all binding and has no value other than being a debating socity --- a debating socity that I am very glad that I can avoid, thank you very much. - Ted -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: [Debconf-discuss] list of valid documents for KSPs
On Wed, May 31, 2006 at 02:48:13PM -0500, Manoj Srivastava wrote: The person who I thought was Marting has apparently revealed that the identity documents that were preseted to the key signing party participants were ones that did not come out of a trusted process. Typically, the identity papers are produced by official bodies, like governments, that have international treaties in place to assure a minimal conformance of identity checks. Wrong again. There are no international treaties, at least not general ones, that guarantee identity checks. (There may be specialized ones such as those that bind countries within the European Union, but not in general.) Does that mean that if someone shows up at an future keysigning party at OLS, for example, with an Transational Republic ID which has the name Manoj Srivastava, that everyone would be therefore be entitled to demand on debian-devel that all signatures for Manoj Srivastava should now be revoked? I would think that if an imposter was running around, and if people were no longer sure that such an imposter twas the one whose ID they had based their signatures on, HELL YES!!! So if someone purchases a fake ID for oh, $20 that appears to be a government issue ID, and successfully shows that at least one signature was signed using said fake, apparently government issued ID, you'd acquience to someone asserting that everyone should revoke their signatures on your key? I didn't say this; I think you were careless editing attributes in your note which replied to multiple e-mails Had Martin never mentioned this, it would have been a non-issue. There is no real damage. While signatures may have been based on a non-offical ID, Martin did indeed own the key in question, so the end harm is zero. But Martin decided to publish this experiment A security mechanism that only works in the non-presense of fraudsters is no security mechanism at all. A KSP that depends on there being any pre-existing trust to abuse is *completely worthless* as a KSP whether or not that trust is abused or not. You just dismissed signing PKA keys by individuals. There is no way that an individual with access to official records can determine if a particular passport is a test passport or not. And now you are sinking deeper and deeper into paranoia. Of course there is no way to tell whether or not a particular passport is real. Heck, do you know how easy it is forge even official government records? In most US states, it's trivial if you know what you are doing to get an official driver's license issued with a false identity. And it's only a little bit more work to get a passport with a false identity, if you are willing to be a bit dishonest about things. If absolute trust is the only thing you will accept, then you might as well withdraw from Debian project, and go hide in a hole with some paranoiod survivalists in Montana. We can't have absolute trust; it is impossible. And you seem to be the one demanding it, and if you can't have it, it's off with their signatures, or off with their key on the keyring! - Ted -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: [Debconf-discuss] list of valid documents for KSPs
On Tue, May 30, 2006 at 07:49:34AM -0500, Manoj Srivastava wrote: What Martin Krafft showed you was, How do I know that person actually was Martin Krafft? So if you have no idea whether or not someone was Martin Krafft, how can you ask everyone to revoke all signatures for Martin Krafft as you did earlier. That is really unreasonable. Does that mean that if someone shows up at an future keysigning party at OLS, for example, with an Transational Republic ID which has the name Manoj Srivastava, that everyone would be therefore be entitled to demand on debian-devel that all signatures for Manoj Srivastava should now be revoked? After all, we have no idea if anyone who might or might not have been Manoj Srivastava might or might not have produced an identification documents that may or may not have been false. We don't know! Do you see how rediculous this is? How irrational you are being? Let me try to spell it out another way. Either the entity at the the KSP who was allegedly Martin Krafft was indeed Martin Krafft, or he was not. It must be one or the other; you seem to be arguing things both ways, and you don't get to do that. If he was Martin Krafft, then he didn't carry out any attack! No identity was forged, and no harm was done. Maybe he presented identification that you wouldn't accept, but that is not intrinsically wrong! If the entity was indeed Martin Krafft, then that entity broke no criminal, civil, nor moral laws. If he was not Martin Krafft, then the real Martin Krafft was not culpable, and your arguments that the real Martin Krafft should therefore be censured in any way shape or form is not just. And as I've shown, if someone showing up with forged identity papers is enough to demand that all signatures on a key be revoked, it would be trivially easy for me or anyone else to arrange to have someone show up at OLS with forged identity papers with your name, and carry out a fairly devasting denial of service attack. I say people who try to trick me into signing a key based on an untrusted process of identity verification are evil doers. And I say, as have others, that untrusted process of identity verification is by definition not an absolute term. So how can you say that someone is an evil doer just because they present a form of identity which happens to be untrusted by *you*. What if someone presents an University ID? That isn't an government ID; does that mean they are evil? Quick, consign them to the Nineth Circle of Hell, reserved for traitors and people who commit treason! I say this is insanity. And obviously argument by assertion is a valid form of argument, since you seem to use it liberally. :-) A boss with no humor is like a job that's no fun. I guess you don't see how ironic your signature line is - Ted -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: [Debconf-discuss] Re: Please revoke your signatures from Martin Kraff's keys
On Thu, May 25, 2006 at 04:08:31PM -0400, Stephen Frost wrote: He didn't try to dupe people and this claim is getting rather old. Duping people would have actually been putting false information on the ID and generating a fake key and trying to get someone to sign off on the fake key based on completely false information. The contents of the ID were accurate, as was his key, there was no duping or lying. Whineing that he showed a non-government ID at a KSP and saying that's duping someone is more than a bit of a stretch, after all, I've got IDs issued by my company, my university, my state, my federal gov't, etc. Would I be 'duping' people if I showed them my company ID? What about my university ID? Would it have garnered this reaction? I doubt it. Indeed, duping people would have been if he had passed himself off as AJ, and managed to get people to sign a bogus key as belonging to the DPL. That would have been a demonstration that would have been really obnoxious, and would justify your reaction. In this particular case, he did not assert incorrect information, but rather (to use an X.509 analogy) used a Certificate signed by an untrusted Certification Authority. The fact that some people were willing to trust is about as surprising as the fact that many people click OK when they see a certificate signed by CA not in the browser's trusted list. But he didn't perpetrate fraud in any way. So this is not a surprise, and it's not what I would call an earth-shaking result. But nevertheless, Manoj, I think you are over-reacting. Chill. Relax. Have a alcoholic or non-acoholic beverage of your choice. :-) - Ted -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Bug#366780: ITP: summain -- compute and verify file checksums
On Fri, May 12, 2006 at 12:41:54AM +0300, Lars Wirzenius wrote: Apart from supporting more file formats, summain differs from the traditional md5sum and sha1sum utilities by providing progress reporting, and via convenience features such as automatic recursion into directories, and looking up files relative to the location of the checksum file, rather than the current working directory. Have you looked at the cfv program, which is already packaged for Debian? There are a huge number of checksum programs out there; it would be nice if there could be fewer with a greater concentration of features - Ted -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: 2.4.x Kernel, ECN And Problem Websites
On Wed, Apr 25, 2001 at 10:16:30PM +1000, Daniel Stone wrote: It may be a minor catch-22, but ECN is currently so broken, that only power users should be using it, as the rest will just continue flooding the netfilter list with Netfilter breaks all my websites!. [OK, ECN isn't broken, the routers are, I know, but same effect. ECN breaks stuff]. So, if you're smart enough to know that you want ECN, and smart enough to understand the consequences, you should be compiling your own kernel. Incorrect. ECN is not broken. The problem is there are broken firewalls and load balancing machines out there that incorrectly (violating the relevant RFC's) dropping packets with the ECN bit set, when they have no business doing that. (The RFC's indicate that the bit should be set to zero by the sender when it was previously undefined, but that receivers were supposed it ignore that bit. Be conservative in what you send, and liberal in what you accept.) The vendors who have broken hardware out there, such as the Cisco Load Director, have patches out there which fix the bug; they've had the bug fixes available for the better part of the year. The problem is that end-customers (i.e., sites like E-Trade) are being slow to install the patch. As to why install with ECN? That's because ECN is important in terms of helping the core internet routers deal with increasing amounts of load. ECN stands for Explicit Congestion Notification, and what it means is that routers can explicitly tell end-hosts to back off because of congestion in the internet core, as opposed to simply dropping packets on the floor. It improves the overall efficiency of the network, and in the future may be important in avoiding congestive collapse of overloaded links. Aside from being a real Linux kernel developer (sorry, couldn't resist :-), I also do quite a bit of work with the Internet Engineering Task Force, the standards body for the Internet where ECN originated, and my colleagues in this organization, which include Jamal Hadi Salim (one of the core Linux networking kernel developers who also works with the IETF), tell me that it's widely regarded that if it weren't for Linux, a lot of bleeding edge protocols that may ultimately become very important to the Internet either wouldn't have been widely adopted, or the adoption rate would have been much, much slower. So I think it's important that Linux distributions provide an easy way for sites to use ECN. Whether or not ECN should be enabled by default is a more difficult question, and really depends on what you think is more important. Do you turn off something that will ultimately be very beneficial to the entire Internet because there are some broken sites out there that are willfully refusing to apply a bug fix which Cisco and other vendors have had available for months, at the cost of inconveniencing some users until they can figure out how to disable ECN or lobby those sites to apply the bug fix? Or do you take the Microsoftian approach way out, and sacrifice the long-term good of the Internet in the name of user convenience? Ultimately, how you choose is a matter of your priorities. But please don't call ECN broken. It's not ECN's fault; it's the fault of those web sites that refuse to update their software with a bugfix release. - Ted