[solved]packages and architectures

How do you make a package from a PKGBUILD that specifies "any" in the architecture field so that the same pkg tarball can be installed on any system. I.e. how do you get makepkg to create "any.pkg.tar.gz" instead of "local_architecture.pkg.tar.gz"?
I didn't find anything in the makepkg man page or the wiki, but I'm sure that I've seen some doc pkgs that had this in the pkg file name. If I've imagined this, how do you create an i686 pkg from an x86_64 machine and vice versa when the pkg itself is architecture-independent?
EDIT
Rephrased the question to make it clearer.
EDIT 2
The behavior is exactly as one would logically expect...
"arch=('any')" generates an "-any.pkg.tar.gz" pkg
"arch=('i686' 'x86_64')" generates an "-<architecture>.pkg.tar.gz" pkg
I was so sure that the PKGBUILD specified 'any' that I didn't bother to double-check, hence my confused and stupid question.
Last edited by Xyne (2008-10-24 22:57:39)

Xyne wrote:How do you make a package that supports "any" architecture, i.e. get a makepkg to create "pkg.any.tar.gz" instead of "pkg.local_architecture.tar.gz"?
You put
arch=('any')
in the PKGBUILD.
Xyne wrote:If I've imagined this, how do you create an i686 pkg from an x86_64 machine and vice versa when the pkg itself is architecture-independent?
The 'package' is just a tarred bunch of files. If they are arch independent, then it's just about the filename, and about the file with the name .PKGINFO in the root of the tarball - you can edit that and change the "arch= ..." line. So although it's possible, why would you want to do that?

Similar Messages

Welcome to the Solutions and Architectures Data Center & Virtualization Community

Welcome to the Solutions and Architectures Data Center & Virtualization Community. We encourage everyone to share their knowledge and start conversations related to Data Center and Virtualization Solutions and architectures.All topics are welcome, including Servers – Unified Computing, Data Center Security, Data Center Switching, Data Center Management and Automation, Storage Networking, Application Networking Services and solutions to solve business problems.
Remember, just like in the workplace, be courteous to your fellow forum participants. Please refrain from using disparaging or obscene language or posting advertisements.
Cheers,
Dan Bruhn

Hi,
I have a question...
I going to install two Nexus 7009 with three N7K-F248XP-25 modules on each one, I am planning to create 3 VDC, but at the initial configuration the system does not show the ethernets ports of these modules, even with the show inventory and show module I can see tah the modules are recognized and its status is OK. There is something that I have to do before start to configure these modules...? enable some feature or license in order to see the ports with show running CLI...?

Welcome to the Design and Architecture Community

Welcome to the Network Infrastructure Design and Architecture Community. We encourage everyone to share their knowledge and start conversations on design and overall network topology. All topics are welcome, including requirements and constraints of the network, tools and techniques, issues of hierarchy, leading practices, services, build strategies, architectures and solutions to solve business problems.
Remember, just like in the workplace, be courteous to your fellow forum participants. Please refrain from using disparaging or obscene language or posting advertisements.
Cheers,
Dan BruhnCSC Community Manager

been waiting to see a forum like this. This forum should amalgamate all the other areas like R&S, security, DC design, and provide more quality , proven solns and discuss problems/challenges as well. Will be following this forum regularly
thanks dan for creating this community
Regards Kishore

Welcome to Solutions and Architectures Borderless Networks Community

Welcome to the Solutions and Architectures Borderless Networks Community. We encourage everyone to share their knowledge and start conversations related to Borderless Solutions and architectures. All topics are welcome, including Switches, Routers, Security, Wireless, Cloud and System Management, WAN Optimization and solutions to solve business problems.
Remember, just like in the workplace, be courteous to your fellow forum participants. Please refrain from using disparaging or obscene language or posting advertisements.
Cheers,
Dan Bruhn

Hi,
I have a question...
I going to install two Nexus 7009 with three N7K-F248XP-25 modules on each one, I am planning to create 3 VDC, but at the initial configuration the system does not show the ethernets ports of these modules, even with the show inventory and show module I can see tah the modules are recognized and its status is OK. There is something that I have to do before start to configure these modules...? enable some feature or license in order to see the ports with show running CLI...?

Welcome to the Solutions and Architectures Collaboration Community

Welcome to the Solutions and Architectures Collaboration Community. We encourage everyone to share their knowledge and start conversations related to Collaboration Solutions and architectures.All topics are welcome, including Collaboration Applications, Customer Collaboration, Telepresence, Unified Communications and solutions to solve business problems.
Remember, just like in the workplace, be courteous to your fellow forum participants. Please refrain from using disparaging or obscene language or posting advertisements.
Cheers,
Dan Bruhn

Hi,
I have a question...
I going to install two Nexus 7009 with three N7K-F248XP-25 modules on each one, I am planning to create 3 VDC, but at the initial configuration the system does not show the ethernets ports of these modules, even with the show inventory and show module I can see tah the modules are recognized and its status is OK. There is something that I have to do before start to configure these modules...? enable some feature or license in order to see the ports with show running CLI...?

New Network Infrastructure Design and Architecture Community

All,
I'm pleased to announce our new Network Infrastructure Design and Architecture Community. In this community we encourage everyone to share their knowledge and start conversations on design and overall network topology. All topics are welcome, including requirements and constraints of the network, tools and techniques, issues of hierarchy, leading practices, services, build strategies, architectures and solutions to solve business problems.
Who will be the first to make it on that leaderboard?
Cheers,
Dan

Mike,
Thanks for the kind words about the community. It's always nice to hear the things we do make a difference.
Launching a scripting community has come up a few times in the past and I'm open to it. Anyone else have thoughts on this?
On a separate note, you'll be thrilled to hear what one of our CSC booth give-aways will be at Cisco Live San Diego next week. Your best friend the cross-over-cable    For those of you out there attending the event be sure and stop by the CSC booth.
Cheers,
Dan

Packages and packages

Hi
I was rereading Martin's principles of package architecture and was wondering how the concept of package in this context relates to the concept of a Java package
after thinking a bit, I got the feeling these two concepts have nothing to do with each other
I feel that Java packages deal with namespace and access control, while OO packages (the components of your program's binaries) would be, in the Java world, jar files containing classes from different packages (and not especially all classes from these (Java) packages), and constructed as much as possible in order to follow these supposedly good OO principles
Am I right ? partially ? totally ? totally not ? I need some insight on this

Kind of. But what would stop me to define the
namespace on a per-component base? I used structures
like com.mycompany.myapp.database (data tier),
com.mycompany.myapp.web (presentation tier) and the
likes. You could put classes (or packages) belonging
to a certain subsystem into a package dedicated to it.I did it for a long time too, traditionaly with webapps :
be.mycompany.myapp
dataaccess
business
web
and that worked fine... but that's not my concern
actually, what concerns me is the business package itself... I'm writing a very simple peer2peer file sharing tool for both fun & education, and the packaging becomes trickier, since the app is naturally distributed :
* file sharing client package
* user manager client package
* server package (handles login requests & user management requests)
thing is, I reuse many classes among these three "kinds of endpoint", which makes me think about properly dividing my app into small reusable, well-organised components... when java packages come into play, things are getting less clear as of how I should do...
how do you people do when it comes to structuring (packages AND deliverables of) your business layer ?

Packaging and Different Directories

Hi there,
I am trying to build an application that will store plugins in a separate directory to the main program. For example, if the main program is at c:/myapp then the plugins are at c:/myapp/plugins . The plugins are also java files.
I want to be able to access these plugins to use them in my program - is there any way this can be done WITHOUT modifying the classpath?
This is on a similar note, but I have tried to package the plugins up into a plugins package and then access it, however I get issues trying to talk between files in the same package (example, 2 plugins talking to each other). How can I get package files to talk to each other when they both co-exist in the exact same package. Shouldn't this already be acceptable communication if I have .; in my classpath?
Thank you for your efforts, I know I keep bringing up the hard questions :)
WATTO

OK, here comes the questions... :)
I have successfully been able to load class files from a different directory using the URLClassLoader, however when it loads it complains that it can't access a variable in the superclass. The Superclass is another class that I have made, which is located in the main program directory, but it is not in the same directory as the plugin. So, this tells me that it has loaded the plugin class using a separate classpath to that used by the program, otherwise it would have been easily able to find the superclass.
So, I have tried copying/moving the superclass into the same directory as the plugin, but still it says that it is unable to access the superclass.
The way I have been working this so far is that I compile the plugin in the main program directory, so that the superclass and other relationships are preserver, and then move the plugin class file into the plugins directory.
Can anyone offer any suggestions as to why this would not be working. Does anyone know how the URLClassLoader uses the classpath of the program? Basically, why is it not finding my superclass even if the superclass is in the same directory as the class I am loading, and how can I solve the problem?
Just to clarify, it IS finding the plugin, so I'm not getting those ClassNotFound exceptions.
Thanks for any help you can provide.
WATTO

App-V 5.0 Package and system folders

Hope someone can help me, having a weird issue with an App-V package that I've sequenced. I am sequencing an application on Windows 8 x64, and during install the application adds .dll files to the
c:\Windows\SysWOW64 folder. My sequenced machine is a VM with a saved snapshot, so for testing I revert back to the snapshot and expand the package to the local system. The program runs fine on the sequenced machine, and all the .dll files
appear correctly in the SysWOW64 folder.
When I go and deploy the package to a workstation (also running Win 8 x64) I get
missing .dll errors and the files required are not in the SysWOW64 folder. I tried editing the package and under
Package Files tab, changed the SystemX86
folder to Override Local Directory. Still this does not solve the problem.
What I don't understand is that it works perfectly on the sequenced machine, but does not for a workstation on our domain. Is there an issue with adding virtual files to protected folders? Or is there another way I can get this working?
If anyone could help me with this issue, I would be greatly appreciated. It's sending me mad!
Cheers!

I have the same problem with Casesoft TimeMap 4 and Notemap 2. No joke, cmen.dll is supposed to be in c:\windows\syswow64 but the virtual file system in AppV puts it in {systemx86}. I'm not sure what path that resolves to but these two programs
cannot find it where it's supposed to be. If I add a file from c:\windows\system32, the variable is {system} and if I put it in c:\windows\system, it puts it in Windows\system in the VFS.
I tried forcing the path to be Windows\SysWOW64 but appv auto creates the path when you select the file and won't let you edit it manually.
I ended up making a cmd copy installer since it has simple copy protection of a plain text registry key

Just updated my iMac to OS X Yosemite and my Photoshop will no longer open. Pop up message informing Some of the application components are missing from the Application directory. Please reinstall the application. I have the pay monthly package and don't

Just updated my iMac to OS X Yosemite and my Photoshop will no longer open. Pop up message informing, Some of the application components are missing from the Application directory, Please reinstall the application. I have the pay monthly package and don't know where to start.

Do as it says: Reinstall the program. Run the cleaner tool and let CC app do the magic.
Use the CC Cleaner Tool to solve installation problems | CC, CS3-CS6
Mylenium

Development objects save automatically with out asking package and TR

Hi Experts,
Development objects in R/3 dev server are saving automatically with out asking any package and TR .
May be the development objects are saving in local package.
How can we check in which package the objects are saving
Please guide me how to solve this issue.
Thanks&Regards,
narasimha.

Howdy,
In that case check the Object Directory Entry to ensure that the objects are going to the correct package - you can also check Version Management to see which transport the objects are getting saved to.
if the objects are already locked on a transport, when you make changes it won't ask for a new transport.
Also try creating a new object and see if you get prompted for a transport then.
Cheers
Alex

Package and User Objects of reports/macros/...

Hello,
in my program I need to find out the package of a report. So for example I have a report with an include or macro statement. I know which report is included, but not its package.
The table TRDIR seems to have no information about the containing package and TADIR contains user objects not reports.
For example a global class ZCL_CAR is realized through reports like ZCL_CAR===(...)===CP and so on. Is there a way to obtain the package of arbitrary reports and the object in TADIR that contains that report? And do you know in which cases a user object can contain other reports?
- thanks in advance

Hi clemens heppner,
Package is there in TADIR table itself
See the field DEVCLASS
Or elaborate more about ur requirement...
Also pls have a look on below tables
d010inc, trdir, eudb, title, d020t, trkey, tfdir, tmdir, tadir,
Hope it will solve your problem
Thanks & Regards
ilesh 24x7

Powershell, Deployment Packages and Software Updates OH MY!

Ok, so I snooped around and was not able to find an answer to my question so here goes.
I'm building an automated process where Orchestrator runbooks will automate the entire process of downloading updates, deploying to test groups, waiting for an approved RFC then deploying the updates. That's the 10k foot level view.
My question revolves around the process of downloading software updates and placing them into a package. It has been suggested that the Save-CMSoftwareUpdate does this, but reviewing the doc I am less inclined to believe this statement. Was hoping
to get lucky and have someone reply that they have done this and here are the interesting parts.
This will all be performed via powershell commands with no direct user interaction at all. So some powershell script that pops up the download software updates wizard will not work for me.
Thanks in advance for any pearls of wisdom around this,
Jim

Ok, I know that we are playing 20 questions here, but I needed the first part to scope the next set of questions.
You say that the approvals are coming from an external source...are you wanting to automate the parsing of the external source or are you planning on putting the approvals into a WSUS group for it to download the update content? Keep in mind here
that the UpdateID that I am referring to is a GUID and not something that someone is likely to give you in a human readable document. The problem with Bulletin IDs is that they will likely apply to a number of operating systems and architectures.
It would be nice if you could get them to populate the approvals into the computer group directly to alleviate the ambiguity. Of course you have to remember that someone has to keep it cleaned up.
You can query WSUS for all of the UpdateIDs for a given computer group.
Beyond this, you will now need to use the WSUS APIs and may want to take a look at the IUpdate interface:
http://msdn.microsoft.com/en-us/library/microsoft.updateservices.administration.iupdate_members(v=vs.85).aspx and call the GetInstallableItems method, which will return a list of InstallableItem:
http://msdn.microsoft.com/en-us/library/microsoft.updateservices.administration.installableitem_members(v=vs.85).aspx each item will have an ID, which is the ID that you will need to publish into your update package; it also has a Files property, which
is a list of files for the installable item, and is of type UpdateFile:
http://msdn.microsoft.com/en-us/library/microsoft.updateservices.administration.updatefile_members(v=vs.85).aspx. These are the files that will get published into the update package. Each update file will have a FileUri property, which will
provide you with a web link to download the file. If you are a bit more enterprising, you can mangle the FileUri into a UNC path, so that you can just copy the file directly.
You should now create the update list. In the process of creating the update list, you will want to create a table of UpdateID to CI...you will need this when you create your advertisement.
Once you have gathered all of the update content, and their associated InstallableItem identifiers, you are ready to tell configuration manager to populate the update package with the installable items. Interesting here is that you are dealing with
the installable items and not directly with the updates. Configuration manager will associate the InstallableItems with the appropriate updates to be shown in the interface.
Once you get the package populated, you will need to create the advertisements to the appropriate collection. By the way, this is done on an individual update basis. Remember that update lists, update packages and update advertisements have
no direct relationship to one another, although a stronger bond was created in 2012.
Simple stuff huh?!?
Remember that Configuration Manager will not pull metadata across for updates that are already superseded, so you may not always have a one to one reference.
Keeping your packages cleaned up, removing extraneous update content that is no longer needed, is a completely different subject.
It would be greatly appreciated if you would mark any helpful entries as helpful and if the entry answers your question, please mark it with the Answer link.

[PARTIALY SOLVED] Gpgsm and Kleopatra error on import p12 (X.509) cert

Hello everyone.
Currently I use KDE 4.9.2, and the system is fully updated. Since some time ago, I have been trying to import an X.509 certificate, that I need to use the mail without success.
When I try the gpgsm (gnupg 2.0.19-2) this happens:
gpgsm --import foo.p12
gpgsm: enabled debug flags: assuan
gpgsm: DBG: connection to agent established
gpgsm: gpgsm: GPG_TTY has not been set - using maybe bogus default
gpgsm: gpg-protect-tool: error while asking for the passphrase: End of file
gpgsm: error running `/usr/lib/gnupg/gpg-protect-tool': exit status 2
gpgsm: total number processed: 0
secmem usage: 0/16384 bytes in 0 blocks
If I try directly by Kleopatra (kdepim-kleopatra 4.9.2-1) appears the following error: "Decryption failed".
The gpg-agent is running and all tests done by Kleopatra gave ok.
I know there are already a few threads on this subject. However, it seems to me that in this case the problem is in the gpg-protect-tool, as it appears in the error made by gpgsm.
After some research I found this bug reported at the following site:
https://bugzilla.redhat.com/show_bug.cgi?id=548528
In comment # 43 was created a patch that seems to have solved the problem.
But I do not know anything about programming and do not know how I can use the patch to see if it solves the problem.
So, I ask your help to verifying if the problem is in gpg-protect-tool, and if so, how can I solve it?
Thank you.
Last edited by Andrenas (2012-10-11 12:41:48)

I apply the patch, indicated in the previous post, to gnupg (and rebuild the package) and now I can import the certificate.
So, the problem was even nominated.
However, now I can not do kleopatra trust the root certificate ...
Does anyone know how I can do?
Regards

[SOLVED] Package installed - Directory missing

Hi guys,
while I've been a passive reader for a long time, this is my first post.
I used the search function, so if my question was answered before, please forgive me.
Some Time ago, I installed the apache package and later uninstalled it again.
I deleted /srv/* afterwards, since it was just sitting there and wasn't really needed.
Now, when I try to reinstall apache, /srv/http for example is not being created during the install or afterwards.
It's simply not there.
I can't imagine this to be a serious problem, but I'd really appreciate it, if you could point me in the right direction.
Last edited by legi0n (2011-12-17 09:11:49)

Duh!
You're right.
I somehow expected it to be created.
/Solved.
Thanks again.

[solved]packages and architectures

Similar Messages

Maybe you are looking for