ABS as an universal package manager

I've just had a rather interesting thought. I'm lately entertaining (and discussing) the idea of universal package management, self contained packages etc. As I've recently switched to Arch again, I thought about ABS as a system which actually reduces dependence on a single repository, a single rule-set so to speak and adds a great amount of flexibility. It may very well be the best way of utilizing the one universal packaging format that there currently is: source tarballs.
What if other distros started packaging and maybe even pre-installing ABS, or something similar? Any user of any other distro would only need a tool like yaourt to check up a mirror of all PKGBUILD's available in the whole GNU/Linux universe and initiate an install where the system would just follow the instructions laid down in the PKGBUILD.
Now, considering the flexibility of PKGBUILD's, you don't even need to have them all compile a given program on the user's computer. Instead of fetching a source tarball it could just fetch a binary tarball. The dependencies are still all pretty much the same (and stated and taken care of by the PKGBUILD instructions), the biggest difference is just that instead of compiling the user gets a pre-compiled package and the whole process is done much faster.
Anyway, I wonder what you think and if anyone has discussed this before. I think the PKGBUILD's may very well contain one possible solution to coming up with universal package management system, at least as an add-on to a given distro's base packaging system. And even as an add on, and especially with the ability to install binary and not just source tarballs (much like pacman in fact) may add a great amount of flexibility to a given distro's user that they may currently be missing.
Cheers
Last edited by libervisco (2008-07-04 02:25:39)

Stythys wrote:what's the point of having different distros if they all start trying to be the same?
There's no point at all there because that's not my point. The idea of universal package management is merely meant to increase compatibility, to provide a way to easily install the same piece of software using the same process on any distro. Imagine the benefits this could have for newbie users once such an option was further developed and polished. You could tell someone how to install a given program and it would work no matter what distro you run. As it is right now you have specific instructions per distro, not to mention that one distro has some software which another just doesn't.
It is not meant to replace the existing package management in these distros, just add a way to easily install software even without it, when it doesn't have what you need or when it's simply simpler for you to do it this way.
Stythys wrote:Anyways I seriously doubt other distros are just gonna start using ABS if asked. If you want pacman/ABS on a different system, just compile it yourself
Well, I agree they wont just adopt it. It would all be gradual. Someone starts compiling it for other distros and distributing it through a web site or where possible through community contributed repositories. If people like what they see they start using it more and more until some distros decide to include it by default. Depending on its success the extreme end result would be for it to be a common feature of *every* GNU/Linux system.
But yes, it does start with someone compiling it.
Stythys wrote:but then why not just use arch
Why aren't all GNU/Linux users the users of Arch?
ABS is by far not the only thing which differentiates Arch from others so it's not like putting ABS on other distros would result in them becoming exactly like Arch. That said, people have different preferences. Not everyone wants nor is knowledgeable enough to go through Arch's install process, for example.
Cheers

Similar Messages

Unified Package Manager/Format

I know that this topic is not a new one, but I thought it might be interesting to bring it up once again.
The reason that I'm suddenly interested in a unified package manager/format, is that I just saw Bryan Lunduke rant about it, and I thought to my self: "Hey, that makes sense!"
Bryan Lunduke is not someone I usually hold in high regard, but I was intrigued and started googling. To my surprise I found that there wasn't much interest in the subject.
Now, to you that may make perfect sence, but to me it's absolutely mindboggling. After all, the only reason I'm still sticking with Arch Linux, is the ABS. Not that I don't care for Arch, I really do, but I spend far to much time configuring it, rather than using it.
So what are the arguments for and against?
The arguments against is easy if you ask me:
PRIDE
Yes I said it: "We're happy with out system and it works damn you!"
So, it works... Well, so does a bike, but quite often you'd probably like to take your car instead. Maybe it's raining, maybe the distance is great. Who cares, the point is that the fact that it works, doesn't mean that it could not potentially work quite a lot better.
Back on topic.
Arguments for, is a bit harder, not least because I'm not as much of an expert on the subject as I'd like to be, but I think they're plentiful.
I would assume that developers would appreciate this.
If a set of tools, like the ABS, was included by default, that would indeed make me happy.
Maintenance of packages for multiple platforms would become quite a lot easier, if not trouble free.
And so on and so forth.
I suspect you will be more than able to help filling out the list.
In the end though, my biggest problem is that people ask for a reason to begin work on such a thing. My counter argument of course it: "Why the hell not?!" Arch Linux already provide the perfect foundation, i.e. pacman and ABS. Of course for cross platform compatibility, some work has to be done, but the greatest problem, I believe, is convincing the community that this is a worth while effort.
Oh well, enough ranting I suppose. I really just wanted to revive this debate, as I personally find it rather important. I don't have all the answers, but I dare you to come up with a problem that I cannot provide a theoretical solution to. To me it all seem so simple, and it's a mystery that this feat has not been achieved many years ago.
Best regards.
NB:
I'm not so arrogant that I do not appreciate that I may be entirely on the wrong track, but please, if that is indeed the case, do try to enlighten me rather than something uncomfortable.

Trilby wrote:
You beat me to that XKCD comic. It's good for a chuckle, but hopefully that doesn't make anyone think that's all it's good for. That comic captures the primary (and perhaps only) reason I'd never have any hope in such a cause.
Just look at attempts an universal language. You'd be hard pressed to find a handful of people in your day to day interactions who ever really learned esperanto, for example (I think I know one, but I haven't seen him in years). Now we can set standards, though. The arch linux forums have selected English as our default language - and in a sense we force speakers of other languages to use English ... in a sense, but not in another sense. There are various arch linux online communities in other languages, where some other language is the default. So one community can set a standard and try to enforce it, but that can't stop another community from springing up and using something different.
So if we say Arch linux's package format is the one package format to rule them all, fine. But then we have to convince others. Even if we convinced a lot of debian and red hat people, for example, we'd probably have to compromise a bit and we'd end up with an arch-pm-deb package. I'm sure some debian and redhat developers would fork their project or start new distros that continued to use .deb/.rpm, and some archers would conclude that this compromise was just a failure and they'd go on using pkg.tar.gz. And then we have the exact situation of the XKCD comic.
Sure, the world would be much simpler if one person could make an arbitrary decision based on their own preference and have everyone else in the world follow it. But that's not our world - and I'm glad it's not.
I'm usually not much for compromises (more on that on another occasion), but in this case I'm all for it, as long as it's sensible compromises. I have no illusions that the Arch way is the perfect way, and in any case, certain changes would need to be make to make it work.
I would suggest that some may misunderstand my intentions. I see no reason my the average Debian, OpenSuse, Mint, *buntu, Fedora, Gentoo (etc.) user should experience any change. A Debian user would still use apt-get install package, or whatever graphical front end he or she desires, as with most other distributions, possibly excluding Gentoo and such.
My interests are limited to the back end system end the tools that are provided by default, like something similar to ABS.
Last edited by zacariaz (2013-08-17 15:09:31)

Suggestion for new package management scheme

As was talked about in the INCOMING thread, Arch is growing and the devs are having trouble keeping up with maintaining packages. Arch is becoming disorganized and messy.
Well, Xentac and I were discussing this for a while, and we both seem to like a voting system. Arch shouldn't try to maintain every package out there, it's simply a waste of effort. If only one or two people want a package, they can build it themselves, it's not hard with ABS.
Arch devs should focus on keeping packages high quality, up to date, etc. They can't do this if there's 10,000 packages. So the solution would be that a package won't be included in the official repos (current, extra, unstable) unless it gets a certain number of votes, showing that lots of people really want the package, and it's "essential".
I don't know how many votes a package would need, but that's a little thing. I would suggest all packages in Extra/Unstable initially go up for this voting (once we implement it) and we make the vote quota be a little lower for this initial slimming, and that'll cut down the repos right there. All those packages that no one uses (that wont even get one vote) can be cut away. People can always vote them back if they become popular.
I don't want to have to do things like KDE, GNOME, etc myself, but I really don't mind maintaining my own little customizations, like Rhythmbox w/ XINE support (instead of Gstreamer). This vote system simply makes sure that a package is worth the developers time.
This goes with my idea of Arch as the perfect "Base" distro. People can build what they want on top of it.
This idea would also remove STAGING, as Arch wouldn't be accepting outside packages (though you're free to run your own repo or post PKGBUILDs on the forum). TESTING would be still used of course.
Suggestions?

beniro wrote:
I haven't really seen a single post in this thread that is without some merit, or at least an attempt at input.
The great thing about not being a larger distro is that the situation isn't as pressurized. Still, though, we (meaning the community AND developers) still obviously want Arch to be the best it can be and that's what this thread is about.
So...are any of the ideas mentioned here feasible? What do you think would be the best solution to creating a package release system that can handle Arch's current growth?
What about implementing a vote-based system for everything not in current? Would this be done with a web-based interface by using "submitrequest/vote" or something...obviously only registered users should be voting, right?
Anyway, I think a great job is being done and this thread is simply a positive indication of the vitality of the community. I think most of us fully recognize the resources of our developers and are simply wanting to find a way for them to bring us the best possible product. Arch rules!
I agree. I think what we should do is just take this time that we have while Arch is still young and not as well known among the Linux community, and try to make it great, relieving the pressures that we'd get a lot of later (not to say that we wouldn't get any, just less, hopefully).
As for the package setup, isn't there some way that we could write some system to do this? I had mentioned it before. And, I just came up with a small system. It would work like this:
We get one main server, whose "job" would be to create and store packages, and users could download from that server to install, etc. Now, whenever someone on the package management team finds an updated version of a package, they could log in to a special area of the site meant specifically for those developers. Then, using the special program on the website/server, they could link to the .tar or .bz2 that contains the source and submit it, along with the file's information that would be included with the package when you install it (file name, version, etc.) into the web-based form. Then, they click submit, and the server processes that file after downloading it, runs a makepkg-type thing on it, and stores the produced package in a "Fresh" repository. Then, the developer can download the file and opt to test it. If they're sure that it's okay, then they can replace the new file over the old one.
Just an idea, tell me what you think... :?

Not a single package manager GUI?

There is not a single package manager GUI for archlinux. The one provided in community repo jacman is quite unstable and the gtkpacman in aur can even search packages.
How about porting http://www.packagekit.org/index.html to archlinux.

Phrodo_00 wrote:yes I was thinking about you and the couple of packagekit topics that have appeared... and not thinking as seroius intentions from arch, but from, you know, some guy, you seemed pretty decided in that post, looks like I was wrong then... anyway, now that I know you aren't really interested and I'm starting vacation at university (and starting to work in a job if I can find one) I might aswell start taking a look at libalpm (why is it called like that?) and packagekit and see if I come out with something.
From http://www.archlinux.org/pacman/ :
As of version 3.0, pacman has been lib-ified, with a backend library named 'libalpm' (Arch Linux Package Management). Speed in some cases has been improved, and this should allow us in the future to speed development of alternative front ends.
EDIT: however, some other guy nicknamed tradiaz said he was going to do it, not a lot about him has been heard though.
There are two recent commits from him, you can see them in gitweb for example :
http://gitweb.freedesktop.org/?p=packagekit.git
(look for alpm and the date of today)
But it looks far from complete if the table on the faq is uptodate:
http://www.packagekit.org/pk-faq.html

Visual Studio 2013 Nuget Package Manager error

Hello everyone,
I tried to add and use OpenCV library in my Visual Studio javascript windows application project, but I still get the same error.
This is my Package manager console:
PM> Install-Package OpenCV
Installing 'OpenCV 2.4.8'.
Successfully installed 'OpenCV 2.4.8'.
Adding 'OpenCV 2.4.8' to BlankSharedApp1.Windows.
Uninstalling 'OpenCV 2.4.8'.
Successfully uninstalled 'OpenCV 2.4.8'.
Install failed. Rolling back...
Install-Package : Could not install package 'OpenCV 2.4.8'. You are trying to install this package into a project that targets 'Windows,Version=v8.1', but the
package does not contain any assembly references or content files that are compatible with that framework. For more information, contact the package author.
At line:1 char:1
+ Install-Package OpenCV
+ ~~~~~~~~~~~~~~~~~~~~~~
+ CategoryInfo : NotSpecified: (:) [Install-Package], InvalidOperationException
+ FullyQualifiedErrorId : NuGetCmdletUnhandledException,NuGet.PowerShell.Commands.InstallPackageCommand
Thanks for answers.

Hi Fačko,
Normally, we won't get any errors if we create a new project. Since you get this error when using OpenCV library and it is a third-party product, I think you need to ask in
OpenCV forum for better response.
Thank you for your understanding!
Best regards,
We are trying to better understand customer views on social support experience, so your participation in this interview project would be greatly appreciated if you have time. Thanks for helping make community forums a great place.
Click
HERE to participate the survey.

System settings Package management part crashes on open

Hello all,
Whenever I open the system settings and then go to "package management" EVERY TIME the screen shows up and inmediatly closes and I get a KDE crash handler window.
Please note I am using KDEmod repos.
If I hover over the Package Management icon I get that it contains 2 items, which are "Aqpm global configuration" and "Aqpm databases configuration"
No idea of what could be happening, I checked and the package kdemod-aqpm is installed
Any more info you need just ask I will be around.
cheers

There'sa forum for kdemod here - you might get a quicker/better answer there.

Package Manager console failed to load

Last night I installed update 3 for Visual Studio 2012 professional (version 11.0.60610.01). This morning when I started the package manager console I received the following errors. I did a repair of VS but other than taking a long time
it did not fix anything. I also installed all the extension updates, no help either. The file mention below is where it should be and I can open it. Tried running in administrator mode, again no help. All three versions of power shell
seem to run just fine (64, 86, ISE). I am running Windows 8 on a 64 bit machine, all updates are installed.
Any suggestions on how to fix this one?
The following error occurred while loading the extended type data file: Microsoft.PowerShell.Core, C:\Windows\SysWOW64\WindowsPowerShell\v1.0\types.ps1xml(2977) : Error in type "System.Security.AccessControl.ObjectSecurity": Exception: The getter method should be public, non void, static, and have one parameter of type PSObject.
The following error occurred while loading the extended type data file: Microsoft.PowerShell.Core, C:\Windows\SysWOW64\WindowsPowerShell\v1.0\types.ps1xml(2984) : Error in type "System.Security.AccessControl.ObjectSecurity": Exception: The getter method should be public, non void, static, and have one parameter of type PSObject.
The following error occurred while loading the extended type data file: Microsoft.PowerShell.Core, C:\Windows\SysWOW64\WindowsPowerShell\v1.0\types.ps1xml(2991) : Error in type "System.Security.AccessControl.ObjectSecurity": Exception: The getter method should be public, non void, static, and have one parameter of type PSObject.
The following error occurred while loading the extended type data file: Microsoft.PowerShell.Core, C:\Windows\SysWOW64\WindowsPowerShell\v1.0\types.ps1xml(2998) : Error in type "System.Security.AccessControl.ObjectSecurity": Exception: The getter method should be public, non void, static, and have one parameter of type PSObject.
The following error occurred while loading the extended type data file: Microsoft.PowerShell.Core, C:\Windows\SysWOW64\WindowsPowerShell\v1.0\types.ps1xml(3005) : Error in type "System.Security.AccessControl.ObjectSecurity": Exception: The getter method should be public, non void, static, and have one parameter of type PSObject.
The term 'Get-ExecutionPolicy' is not recognized as the name of a cmdlet, function, script file, or operable program. Check the spelling of the name, or if a path was included, verify that the path is correct and try again.
John J. Hughes II
Ran Get-ExecutionPolicy from PowerShell (x86) it returns Restricted
Also looking around... I have on add-in installed or running
Also tried uninstalling nugget and reinstalling
I noticed on the NuGet website it calls for PowerShell 2.0. It seems I have PowerShell 3.0, could that be a problem? Can I install PowerShell 2.0/3.0 side by side, is there an update to NuGet?

Found the following link in stack overflow, added the suggested items to the VS exe config file and it works now without error.
http://stackoverflow.com/questions/12638289/embedding-powershell-v2-0-in-net-app-on-windows-8-rtm
Added the following to the devenv.exe.config file in the runtime section
<dependentAssembly>
<assemblyIdentity name="System.Management.Automation" publicKeyToken="31bf3856ad364e35" />
<publisherPolicy apply="no" />
</dependentAssembly>
<dependentAssembly>
<assemblyIdentity name="Microsoft.PowerShell.Commands.Utility" publicKeyToken="31bf3856ad364e35" />
<publisherPolicy apply="no" />
</dependentAssembly>
<dependentAssembly>
<assemblyIdentity name="Microsoft.PowerShell.ConsoleHost" publicKeyToken="31bf3856ad364e35" />
<publisherPolicy apply="no" />
</dependentAssembly>
<dependentAssembly>
<assemblyIdentity name="Microsoft.PowerShell.Commands.Management" publicKeyToken="31bf3856ad364e35" />
<publisherPolicy apply="no" />
</dependentAssembly>
<dependentAssembly>
<assemblyIdentity name="Microsoft.PowerShell.Security" publicKeyToken="31bf3856ad364e35" />
<publisherPolicy apply="no" />
</dependentAssembly>
<dependentAssembly>
<assemblyIdentity name="Microsoft.PowerShell.Commands.Diagnostics" publicKeyToken="31bf3856ad364e35" />
<publisherPolicy apply="no" />
</dependentAssembly>
John J. Hughes II

How to deploy EAR on CE server using Java Support Package Manager

Hello,
I am using CE server where I want to deploy EAR file by using JSPM. I run this program in Drive:\usr\sap\CE1\J00\j2ee\JSPM\go.bat
I have also put the EAR file into the inbox folder Drive:\usr\sap\CE1\SYS\EPS\in but still It does not show me any file to proceed further for deployment.
Does any one has idea?
Regards,
NK

Hi,
The Java Support Package Manager is not able to deploy a single .ear file.
If the *.EAR file is placed inside an SCA (Software Component Archive) then JSPM is able to deploy it.
You have to create a .SCA from an .EAR file
SAP note 1223957 contains an attachment (nwpacktool.zip) which can be used to create a .SCA file from the .EAR file.
Gerd

How to import all Support Packages simulteneily support Package Manager EBP

Can any body tel me how to apply all support package sumultanesoly by using support Package Manager EBP/CRM option in SPAM.
I could not able to select all pactche at a time.It selecting one patch at a time only
Regards
G Sasikanth Reddy

I solved my self by selecting highest level of patch so that its dependencies can also add to the queue.
Regards
Sasikanth Reddy

Support Package Manager for EBPCRM

Hi,
In CRM/EBP SAP systems there is used special 'Support Package Manager for EBPCRM' to implement Support Packages instead of standard transactions like SPAM/SAINT.
I'm trying to implement more then one Support Package into the system however this is probably not possible. It seems that it is not possible to create import queue with more then one Support Package? Importing SPs one by one is very time consuming task. SO that's why I'm wondering about this special tool.
Do you have similar functionality?
BR
m./

Hi Jose,
Yes I've put all the SPs into '\usr\sap\trans\Eps\in'. But 'Support Package Manager for EBPCRM' asks me for particular one while I'm going to import it. There is no such an functionality to use queue creation.
BTW: I'm on SAP CRM 4.0 on WIN/MSSQL platform.
BR
m./

Managing JAVA JDK and JRE with Package Management in Solaris 11.1

I want to manage Java JDK & JRE in an Solaris environment using Package Management I'm currently running Solaris 11.1 in a VMWare host. My current latest Java release is JDK1.7.0_07 and JRE1.7.0_07. I want to upgrade to JDK1.7.0_17 & JRE1.7.0_17 using Package Management. I've tried on several occasions to upgrade to latest release of Java but I have not been successful. Each time I try using the Package Management GUI or use command line they appear to be successful but the new release does not show.
What could I be doing wrong?
Thanks

I also learned that the READMEs for the following patches in My Oracle Support also describe the downgrave process:
Patch 18362676: JAVA 6 DOWNGRADE INCORPORATION FOR ORACLE SOLARIS 11.1.15.4.0 AND 11.1.16.5.0
Patch 18362686: JAVA 7 DOWNGRADE INCORPORATION FOR ORACLE SOLARIS 11.1.15.4.0 AND 11.1.16.5.0
-- Alan

Oracle Universal Records Management adapter with FAST ESP

This is my first post on the forums and I have limited Oracle exposure. So if I’m posting in the incorrect spot or there are some other guidelines I need to observer, please let me know.
I’m trying to figure out the possibility of leveraging the Oracle Universal Records Management adapter for SharePoint in Records Management with FAST ESP (a best of class enterprise search platform). From the technical documents I’ve read on the adapter, it looks like it heavily relies on SharePoint search as the “feed’ for content. Does this mean that the SharePoint Search is required for the SharePoint adapter? As in environments that deploy FAST ESP also need to maintain SharePoint search for the adapter. Or can the adapter leverage FAST ESP?
Any insight would be appreciated.
TIA

anyhow i have tried ur tips as much i unterstand them (some of the command in cmd prompt and rest in sql plus)
it notifies an error of missing a datafile that i have corrected and finally the login page appeared in the em. also connection from sql plus is successful.
now while connecting to oracle through ORACLE DEVELOPER TOOL for .NET, i got this error:
"An error occured while opening the connection, connection must be open for this operation"
if i tried to connect through SYS as NORMAL it gave me the error to connect as SYSDBA or SYSOPER that is a valid error, but wat for the connection open error.
thanx.

Slackware user - Arch install process and package management questions

I currently am using Slackware 12 with updates. It has become increasingly cumbersome to compile new software given the number of dependencies that also have to be compiled, and have their own dependencies compiled, etc and so on.
Other than the fact that the adoption of new packages is somewhat glacial, and that there is no automatic dependency resolution, I really like Slackware. Its fast, relatively easy to configure, and easy to use. I'm conversant in commandline-ese.
I've been led to understand based on the Arch wiki that Arch is very similar to Slackware in its layout, but with bleeding edge packages and dependency resolution, so I would think it would be about perfect.
However, I also understand that the install is more difficult that Slackware. How much more so is my question. Is it impossibly confusing and unhelpful like Debian installation was, or just slightly more difficult than Slackware or FreeBSD (both of which have very smooth installation scripts)
Also, for packages that have to be compiled from source, are dependancies easy to satisfy from the arch repository, or do dependencies for new software have to be compiled with their dependencies.
I'd probably just be trashing everything but /home, and doing a clean reinstall.
Is Arch the right distro for me, or am I better off staying with Slackware? Like I said, I really like Slackware except for the package management issues.

I've been led to understand based on the Arch wiki that Arch is very similar to Slackware in its layout, but with bleeding edge packages and dependency resolution, so I would think it would be about perfect.
You are right about that.
However, I also understand that the install is more difficult that Slackware. How much more so is my question. Is it impossibly confusing and unhelpful like Debian installation was, or just slightly more difficult than Slackware or FreeBSD (both of which have very smooth installation scripts)
They are quite similar, but Slackware provides more help when doing tasks like - how can I partition my harddisk and so on. But If you're comfortable with Slack's install procedure then installing Arch is much easier, just read the install guide.
I think Arch is good for you, because currently I have both - Slackware and Arch and both suit me fine for my needs.
Also remove all *. entries in your /home

Oracle Universal Content Management 11gR1 'download file1' is corrupt

Hello,
I am at http://www.oracle.com/technetwork/middleware/content-management/downloads/index-085241.html and downloaded the part 1 and 2 of the Oracle Universal Content Management 11gR1 software. The part 2 looks good but part 1 is corupt (it is only 4mb). I tried this download from different computers and browsers which doesn't help
Does everyone have the same problem?
Thanks

Workflow for document approvals and routing is supplied out of the box without the need for separate software or installs.
What you can do is described here:
http://download.oracle.com/docs/cd/E10316_01/cs/cs_doc_10/documentation/admin/workflow_guide_10gr3en.pdf
The more complex integrated processes can use BPEL Manager which does require addditional component and software
Tim

Oracle UCM (Universal content management)

Hi Friends,
I would like to know the Details on Oracle UCM (Universal content management)
Please tell me the different versions and the installation and configuration documents.
Regards

Hi Eduardo:
I'd would like to know about ORACLE UCM, can you please tell me where can I find a tutorial about this. I've been read the manuals of Oracle page.
Do you know if there are spanish tutorials?
thanks,
Vilma

ABS as an universal package manager

Similar Messages

Maybe you are looking for