Talk:RPM Package Manager
|This is the talk page for discussing improvements to the RPM Package Manager article.|
|WikiProject Computing / Software||(Rated Start-class, Mid-importance)|
|WikiProject Free Software / Software / Computing||(Rated Start-class, High-importance)|
- 1 Red Hat Linux
- 2 Alien
- 3 Naming
- 4 Porting
- 5 MD5
- 6 Redundancy
- 7 Update
- 8 Recursive abbreviation
- 9 types of RPM's
- 10 Please review and delete the external link to "Who Maintains RPM?"
- 11 Disadvantages
- 12 When was first released
- 13 On Portal:Free software, RPM is currently the selected article
- 14 Supported Linux distributions??
- 15 Again I ask:
- 16 Advantages, Disadvantages
- 17 Tidy up
- 18 Examples of use
Red Hat Linux
I replaced Red Hat Linux for Red Hat Enterprise Linux, sine Red Hat Linux exists no more. It was replaced for Fedora Core and afterwards Fedora. One of the reasons was to avoid misunderstanding between Red Hat Linux and Red Hat Enterprise Linux, that's why even nowadays it's important to say RHEL or Red Hat Enterprise Linux when you're meaning it.
Shouldn't this article mention alien? Debian-based distros use alien to install rpms.
According to rpm.org, the name is "RPM Package Manager". For more discussion, see: http://www.wikipedia.org/wiki/Talk%3ARed_Hat_Package_Manager --Tero 20030806t1415 EET
Which other OSes has rpm been ported to? BL 14:07, May 2, 2004 (UTC)
I added a link to the page at RPM.org that lists the various operating systems to which it has been ported. Better this way than to have to continually update it whenever a new port is made. Avalyn 10:05, 6 May 2005 (UTC)
Use of the (now seriously broken) MD5 algorithm is listed as an advantage. Is this still true (in which case it should surely be listed as a disadvantage!), or has RPM transitioned to a more secure method already (in which case that should be mentioned in the article)? --220.127.116.11 12:26, 14 November 2005 (UTC)
Since rpm-4.1, rpm metadata content (like the MD5 file digest) usually comes from a signed header.
So the issue of whether MD5 is "broken" or not is moot, one has to break the digital signature on the header in order to change the file MD5 digest, and that requires breaking DSA or RSA, not MD5.
No. Since MD5 is broken, it means that the code could be modified in such a way as to produce the same MD5 value. It is certainly not a moot point. 18.104.22.168 (talk) 23:50, 27 April 2009 (UTC)
Furthermore, rpm-4.4.6 permits replacing the file MD5 with any modern digest up to and including SHA-512.
The "advantage" comes from verifying a digest while installing, and verifying the signature (if present) on the header (where the file digest tag resides) whenever desired, not just when the package is installed. This is a (at least) reasonably secure persistent verification mechanism when used correctly.
Isn't the "Package Manager" in "RPM Package Manager" redundant?
- That's the idea of recursive initialisms. Chris Cunningham 14:15, 31 August 2006 (UTC)
Can someone with the expert knowledge please update the description for rpm naming format for the release naming convention. The description is lacking for the .rf and .at, they seem to indicate packagers but i'm not certain enough make an update to the article. Thank you.
There are many release naming conventions that are currently deployed. No one scheme is dominant or accepted sufficiently widely that it needs documenting.
RPM is not a "recursive acronym". It is a "recursive abbreviation".
- To be precise, it's a recursive initialism. I've edited the article appropriately. Chris Cunningham 14:15, 31 August 2006 (UTC)
- No it doesn't. Oh, and please get an account, and sign your posts. EdC (talk) 13:13, 3 May 2008 (UTC)
- It's not recursive. It would be if it stood for "RPM is a Package Manager." When people say "RPM Package Manager" the "Package Manager" part is redundant, because the last 2 letters already say that it's a package manager. It's kind of like how people say "CIA Agent." Central Intelligence Agency Agent? KenFehling (talk) 10:14, 16 January 2009 (UTC)
Fist: It IS an Acronym.
Third: it is recursiv for fakt (Recursive acronym). Just read it again: RPM Package Manager
Here is a historical source:
Copyright © 1999 by Red Hat, Inc.
RPM is the RPM Package Manager.
The italic letters are part of the quote. Well, and it is from 1999, Copyright Red Hat.
types of RPM's
I don't mean to bother anyone, but shouldn't it be prudent to make seperate sections for the existing types of RPM's:
- Normal RPMs
- Source RPMs
- Delta RPMs
Binary (not "normal") and source RPM's have a handful of explicit differences, the most important of which is that binary *.rpm's have ptr's to the source rpm from which they were built, and *.src.rpm payloads do not specify any hierarchical directory structure.
Patch RPM's were an an interesting patch from SuSE that is largely historical. The idea behind patch RPM's was a delivery mechanism for partial packages that would be merged, rather than installed, with whatever content had previously been installed. The changes to the rpm install state machine, and the difficulty of creating patch packages (i.e. patches for all possible released software need to be produced) have made the idea of patch packages unworkable outside of a narrow vendor-specific distribution context.
SuSE today prefers Delta RPM's instead of Patch RPM's afaik.
There are also ""Repackaged RPM's" that can be produced before erasing the older package when upgrading to a newer package. Repackaged RPM's are a best effort (i.e. content on the file system may have been modified or deleted) attempt to recreate the original package from what had been installed on the file system.
Well, I fixed something that was inaccurate as I can understand , i.e. that RPMs have sources, only SRPMs have source.
Sebelk 17:26, 23 January 2007 (UTC)
(apologies for being a wikipaedia idiot, I don't know the correct way to make this request)
While overall the RPM page is quite objective and sufficiently complete, I am personally concerned by the external link to "Who Maintains RPM?" since it mentions me by name. I have gone to great lengths to preserve my privacy and anonymity (most of RPM was written by me) Much of the content about RPM has been supplied by me, e.g. to "The Red Hat RPM Guide" by Eric Foster-Johnson, and to LSB only under conditions of "anonymous contribution".
However, as the subject of the published article, I'm clearly biased. I will leave it to someone else to make the call whether an external link to "Who Maintains RPM?" is relevant or not. I am personally threatened by an ongoing and organized campaign (there is a slander of me at kuro5hin published in May 2006, and these people have appeared multiple times on mailing lists with aggresssive and hostile attacks that are coupled to my name) that has continued over several years that is ultimately based on a bugzilla report #119185 from the year 2004.
Meanwhile, I'll be happy to supply any information regarding RPM that you want (as long as my name is "anonymous")
Jeff Johnson <firstname.lastname@example.org>
Someone has removed this section. If they have a valid reason and read this, could they explain why? I feel that this section should remain, any one else agree? 22.214.171.124 05:24, 15 December 2006 (UTC)
When was first released
On Portal:Free software, RPM is currently the selected article
(2007-01-23) Just to let you know. The purpose of selecting an article is both to point readers to the article and to highlight it to potential contributors. It will remain on the portal for a week or so. The previous selected article was Advanced Packaging Tool. Gronky 14:50, 23 January 2007 (UTC)
Supported Linux distributions??
Just a minor niggle, but the section titled "Supported Linux distributions" seems to imply that RPM would have to explicitly support a distribution. I believe the intent of the section is something more like "Linux Distributions with RPM support". —Preceding unsigned comment added by 126.96.36.199 (talk) 17:22, 12 December 2007 (UTC)
I've tried cleaning up the controversy section slightly. Aside from the added reference to the LWN article http://lwn.net/Articles/196523/, one might want to link to the rpm.org FAQ entry at http://wiki.rpm.org/Docs/RpmOrgFAQ#head-bfc9aeacc86d4eec06dc6f559f4881d7428ca24b. This seems to paint a slightly more balanced picture: rpm.org acknowledges that rpm development languished after Jeff Johnson left Redhat, and that the rpm5.org fork actually started before Redhat launched rpm.org. The FAQ page is an immutable wiki page, so should be suitable as a reference.
- Sorry for talking to myself. I just found Jeff Johnson's request to not link to the LWN article above. I'm unsure what should be done here -- if you read about the fork and want to find out what it's about, you'll inevitably find the LWN article and also the controversial bug report in the end. I'm not sure there's any benefit to making everybody do the research on his own. Maybe somebody could rewrite the section to be less blunt (Johnson left Redhat instead of terminated) and balance the controversy with the fact that the parties involved seem to be on good terms (witness the rpm.org FAQ)? I don't feel up to the task. 188.8.131.52 (talk) 09:42, 10 January 2008 (UTC)
- Another good reference for a rewrite: http://trainofthoughts.org/blog/2008/01/06/rpm5-vs-rpm/. 184.108.40.206 (talk) 09:50, 10 January 2008 (UTC)
- The section was less blunt up until 4 days ago when someone, apparently with an IP address registered to Red Hat, edited the page to read that Jeff had been "fired", and deleting other content from the section with no explanation in the edit summary or here on Talk. I reverted this edit a few hours later, noting the potential conflict of interest.
- Today, someone with an IP address in Raleigh, NC (home town of Red Hat) re-did the previous edits, again deleting links to rpm5.org and oldrpm.org, but this time claiming that Jeff had been "terminated", which sounds pretty harsh even for Red Hat. Could all this courageously anonymous mudslinging possibly have anything to do with Jeff and the rpm5 team releasing the new RPM 5.0.0 stable on 5 Jan, the day before the allegations starting being posted here?
- Thanks for re-writing that last part. Red Hat people, give the personal attacks a rest, please. Technobadger (talk) 10:55, 10 January 2008 (UTC)
- I've arrived on this article like probably many people looking for information about RPM. But after reading and rereading the controversy section, and this relevant discussion, I still find the section very hard to understand. Which is the "official" RPM page, and which is the fork? Thanks! --Vlad|-> 17:21, 29 January 2008 (UTC)
Again I ask:
Please remove my name everywhere on this page. I am tired of a controversy I did not create and want my private life back.
One does not have to got too much farther than http://rpm5.org/team.php to see many people, not just me, working on RPM 5. Your link is factually incorrect.
(aside) It is not very hard to see that the intent of the word spinning is to imply that me and me alone is responsible for RPM 5.0. Again, factually incorrect.
Finally, rpm5.org is not my home page, also factually incorrect.
By continuing to associate "Jeff Johnson" with RPM in WikiPaedia, the connection to Bugzilla #119185 and the kur5hin slander (which claimed that I bathe less often than once a month and find Jewesses "hot") keeps going on and on and on and on. Stop it now please.
Most of this section applies to using distributions using the rpm format, not to the rpm format itself. It is not a good introduction to the format itself, and if at all it should be moved way down the page. The advantage and disadvantes list and paragraph look biased regarding formatting and content.
Some advantages are outdated, or very biased. There are more .deb based distributions, and it clearly is very popular too. RPM packages do not conform to LSB automagically, if their content does not. Signing and verification isn't that special and other formats do it too. Most of the listed disadvantages are not a RPM format issue either, but of sloppy distribution maintainers.
IMHO this should be removed completely.
I have been through the article and attempted to tidy it up. Hopefully the result has better structure and consistency; it also happens to be somewhat shorter (a good thing in this context) by the removal of some oft-repeated information. More work is probably needed, but it's a start, at least. Feline Hymnic (talk) 16:22, 14 August 2010 (UTC)