Re: [opensuse] How do I mount USB drive world writable using device notifier?
On 2/6/2011 2:17 PM, Anton Aylward wrote:
Marc Chamberlain said the following on 02/06/2011 04:22 PM:
I do NOT understand the reasoning behind having an inconsistent security model It is not inconsistent. One of the 'rules' of NFS is that it won't allow access across a mount point.
Why do you think I have so many entries in my server's "exports"? The server has a big disk, lots of (LVM) partitions.
The 'anton' archive there has mounted file systems for 'business', 'email', 'email-archive', 'development/programming' - each type, downloads, various non-business documents. All quite apart from the other mounted file systems such as /var, /usr, /usr/lib, /usr/share and /local.
So if I mount "Server:/home/anton" on "laptop:/mnt/server/anton" and access it via the ~anton/server symlink, and do
laptop:~> ls server/Mail or laptop:~> ls server/Documents
I will see nothing. The rule is you can't traverse a mount point with NFS.
So when you export "/" and mount it on you laptop and then try accessing a mounted USB
OF COURSE YOU CAN'T ACCESS IT.
I did say that exporting just root and just /home made no sense.
Then it appears that NFS is a dead end for me. I tried to export the mount point itself with the following added to my exports file - /media/My\040Passport 192.168.2.0/255.255.255.0(rw,root_squash,sync,no_subtree_check) which got a whine from the NFS server saying it was unsupported. rcnfsserver restart Shutting down kernel based NFS server: nfsd statd mountd idmapd done Starting kernel based NFS server: idmapdexportfs: Warning: /media/My Passport does not support NFS export. mountd statd nfsd sm-notify done And the export of /media/My Passport is not available to mount on another system. So perhaps you are correct, NFS may not be inconsistent, within its security framework, but this security rule is restricting the usability of our systems and apparently making Linux+NFS an unsuitable choice for our needs. It also breaks the model implied by using simple commands such as cp, mv, chmod, etc. I remain uncomprehending as to why, I as a user/systems admin, cannot use MY system(s) in such a way as to easily meet MY/OUR needs, and this is proving to be a case in point... I still believe that if I don't want all this security I should be able to easily turn it off, or at least have control over it and use easy to understand models and tools to configure it.. That is my choice and risk to take, especially in a small SOHO network on my own systems. I will use security, such as firewalls where needed, but I don't want it getting in my way elsewhere. Accessing a USB drive through a Samba share at least allows read access on a mounted USB drive, and that gets us halfway to what we want. We can see and pull files across the network from a USB drive when accessing it through a Samba share, just can't push files to a USB drive. Perhaps there is a way to get it to allow write access also, I dunno, but at this point I think I will go explore it further... Some questions I do have, does anyone know where HAL/UDEV does the actual mount of a USB drive when it is plugged in? Can those mount options be directly configured? I don't understand how the configuration for usbfs in fstab gets translated into mount options when the device notifier causes the actual mount to occur, (when a USB drive is plugged in) so perhaps pointers to documentation would help... (I have found some high level docs on the internet, but nothing specific yet.) Alternatively, is there a way to give a user, coming in to a system via a cd or mv command though a mount/share, from across a network, the same permissions/capabilities as a local logged in user has? Perhaps that is another approach I could take, if it is possible.. Thanks again in advance... Marc... -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
Marc Chamberlin said the following on 02/06/2011 10:29 PM:
Then it appears that NFS is a dead end for me. I tried to export the mount point itself with the following added to my exports file -
I did say that I thought with all your fiddling your configuration is buqqered up somewhere. The reason I think this is that "It works for me" I've done - albeit on the USB port on my server - what I've been telling you. And not doing the things I've been telling you NOT to do. I've been following the normal Linux conventions and not fighting them or bitching about them. Yes I can successfully mount USB storage devices on the server, export them via NFS, mount them via NFS on my laptop and access them I've tried this with the crappy little 16M and 64M things I've been given at trade shows and with a 2G USB Drive on which I have a bootable system. The drive, being a drive, is slower than the solid state devices. I've tried this on the USB1 and USB2.0 ports. I'm happy that what I've been telling you to do works 'cos "It works for me". My conclusion is: You've Got Gremlins But then I am exporting the device, not the mount point. When you mount the device is BELOW the mount point, and as I said and you have verified, NFS won't go past the mount point. OBTW, you are aware, I hope, that the BUSUID, DEVUID and all that refer to the points under /proc/bus/usb and not to /media/ The rules for UDEV are under /etc/udev/ The rules for HAL are under /etc/hal (All of which you could have found from the MAN pages) As I was saying, Linux is perfectly logical. And, as I said, the device notifier is a KDE applet that lets you deal with the mounted devices, it does not do the mount. All it does is let you run the file manager, or gwenview or something. As for the idea of giving a user coming in from across the network the ability to copy/move .... We suggested ways to do that originally and you rejected them. We - I recall I did - suggested that you use 'rsync'. Rsync has the advantage that it can keep both USB sticks - or any other object/target pair, in sync, only updating what has been changed. It can also be automated. That way whenever you wife makes a change it triggers the rsync to update your copy. And vice versa. Mind you, most of us would find it simpler to have one and only one shared copy (on a server) and copy that the (local to you each) USB devices when needed. All of which ensures you have one reference copy and avoid all the problems you have encountered. The consultants among us know that productivity and overcoming problems is more often effected by changes in process. Difficulties such as yours are usually a Big Red Flag that says "you're doing it wrong". We ask "what are you trying to achieve" rather than "what are you trying to do". Perhaps that's part of your problem. OBTW: Marc, it is considered impolite to reply on the forum to mail that you have been sent privately, off forum. -- Parkinson's Law: Never Interview Emus -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
Marc Chamberlin said the following on 02/06/2011 10:29 PM:
Then it appears that NFS is a dead end for me. I tried to export the mount point itself with the following added to my exports file -
I did say that I thought with all your fiddling your configuration is buqqered up somewhere.
The reason I think this is that "It works for me" I've done - albeit on the USB port on my server - what I've been telling you. And not doing the things I've been telling you NOT to do.
I've been following the normal Linux conventions and not fighting them or bitching about them.
Yes I can successfully mount USB storage devices on the server, export them via NFS, mount them via NFS on my laptop and access them
I've tried this with the crappy little 16M and 64M things I've been given at trade shows and with a 2G USB Drive on which I have a bootable system. The drive, being a drive, is slower than the solid state devices. I've tried this on the USB1 and USB2.0 ports.
I'm happy that what I've been telling you to do works 'cos "It works for me".
My conclusion is:
You've Got Gremlins Anton - Using such sarcasm and telling me your system works, therefore I have screwed something up, is NOT helpful. I STRONGLY disagree with you
But then I am exporting the device, not the mount point. When you mount the device is BELOW the mount point, and as I said and you have verified, NFS won't go past the mount point. This I do not understand and makes NO sense to me. I have tried to export /dev/sdc1, /media/MyPassport, and /mnt/usbdrive all to no avail. What am I trying to mount, in all three of these cases, the is "BELOW
On 2/7/2011 4:57 AM, Anton Aylward wrote: that I am trying to do something in a wrong (non-Linux way). Copying or moving files from one file system device to another should be easy, straightforward and just plain work. That IS a concept long understood by computer users. As for security and permissions, that too should be easy to configure so as to both accomplish safety, and maintain usability. IT DOES NOT MATTER WHAT MY APPLICATION IS, you nor anyone else has the right to force users to transfer files via some sort of staging process, use central repositories, servers etc. That is a decision the users should be able to make for themselves. If NFS cannot handle this, conceptually easy task, then NFS IS BROKEN OR BADLY DESIGNED. If I have somehow misconfigured NFS, autofs, or the nfsserver, then I need to understand how and why. And so far I have not succeeded... I have tried to expose everything I have done and you have not told me anything specific other than to try and convince me that my process is wrong. And I repeat, what I want to accomplish, transferring files across a network, directly to a mounted USB drive, is NOT wrong. If you disagree with that, then we must simply agree to disagree and move on. the mount point"? It has become clear that I cannot mount anything "ABOVE" the mount point, like / and then expect to be able to access /media/MyPassport, because of an NFS restriction. OK, but I am exporting THE mount point and that should be sufficient for NFS since I am giving my blessing to NFS to export THE mount point. What do you mean by "BELOW"? That just does not make any kind of sense! A far better answer would have given examples. Explain what would work, and perhaps what does not work with reasons why... Examples are a very good way to communicate ideas!
OBTW, you are aware, I hope, that the BUSUID, DEVUID and all that refer to the points under /proc/bus/usb and not to /media/
The rules for UDEV are under /etc/udev/ The rules for HAL are under /etc/hal
(All of which you could have found from the MAN pages) I have seen these, but not been able to grok. The model behind what is going on, which gives the reader the ability to understand the
As I was saying, Linux is perfectly logical. NO, it is NOT, it is extremely complex understand for a beginning/intermediate and I would dare say even advanced users. It comes with commands that have overwhelming numbers of parameters, multiple models for doing similar tasks that are inconsistent at best, gives keyhole views of the internals unless one wants to read code, with documentation that is difficult to comprehend and often outdated, and GUIs that are not well thought out to act as guides/teachers. The good
And, as I said, the device notifier is a KDE applet that lets you deal with the mounted devices, it does not do the mount. All it does is let you run the file manager, or gwenview or something. You may be technically correct, but to someone who did not develop that tool, that is NOT obvious. What a user see's is that it notifies you when you insert a USB device. It pops up dialog, and tells you about the device. Then the user can click on the device, within the notifier, and IT GETS MOUNTED. To an uninformed user, that sure as heck appears as if
Again, I don't fully follow. I tried exporting /proc/bus/usb and the nfsserver and complained about it saying it is not supported. Now you mention yet a new interesting point in the file system - /proc/bus/usb! I have tried to export /media/MyPassport, /dev/sdc1, done a separate mount to /mnt/usbdrive and exported it, and now /proc/bus/usb. NONE worked! In exports I have tried all four variants, such as /proc/bus/usb 192.168.2.0/255.255.255.0(rw,root_squash,sync,no_subtree_check) and I get something similar to this as a result - rcnfsserver restart Shutting down kernel based NFS server: nfsd done Starting kernel based NFS server: idmapdexportfs: Warning: /proc/bus/usb does not support NFS export. mountd statd nfsd sm-notify parameters and setting in the context of and association with that model, is NOT well explained in the man pages. I also believe that external documentation such as MAN pages is not a good idea. If commands, configuration, and GUI's are not intuitive and easy to comprehend, then they have been badly designed. Presentation should be self documenting, I wish more programmers would understand that and not duck their responsibilities by saying "RTFM". Man pages and external documentation is difficult to maintain, and rarely kept up to date. This is a failed model and we need to approach documentation in a whole new way... thing is Linux does not block the user from gaining access to internal information, it just overwhelms the user. That is NOT a logical way to design an OS for all levels of users. Sadly no one has done so yet and I keep hoping Linux, because it is open sourced and has a lot of good people working on it, will get out of this quagmire soon... the device notifier IS mounting the device or causing the device to be mounted. Split hairs all you want, if the device notifier is presenting a model that is wrong, don't blame the user for not understanding.
As for the idea of giving a user coming in from across the network the ability to copy/move ....
We suggested ways to do that originally and you rejected them. We - I recall I did - suggested that you use 'rsync'. Rsync has the advantage that it can keep both USB sticks - or any other object/target pair, in sync, only updating what has been changed.
Anton, in the beginning, you said Do not use Samba, use NFS instead. OK I tried NFS, and I have so far failed to get it to work properly. Now you tell me to go use yet another tool - rsync. And you don't think this is complicated? Robust? Looks like it is yet another client/server system... Ok, I will take a look at rsync, but wow, for what should have been an easy task, this one has turned out to be a nightmare! And yeah I know you are going to say "That is because you are trying to butt heads against Linux and not do things the Linux way" To which you can guess what my reply is... IF I can copy files from a Linux system, via a Samba share to a USB drive mounted on a Windoz system, or from a Windoz system to a shared USB drive mounted on some other Windoz system, why in the world can I NOT do this from a Linux system to a Linux system via the Linux tools for NSF, which as you so often have said, "is what makes Linux easy".... (nor can I copy files from a Windoz system, via either Samba or NFS to a USB drive mounted on a Linux system) To me this indicates that some security constraint is buggered somewhere or there is a serious design flaw. I would like to believe the former and get help in finding out what I have done wrong, the latter is far more worrisome.
It can also be automated. That way whenever you wife makes a change it triggers the rsync to update your copy. And vice versa.
Mind you, most of us would find it simpler to have one and only one shared copy (on a server) and copy that the (local to you each) USB devices when needed. All of which ensures you have one reference copy and avoid all the problems you have encountered.
The consultants among us know that productivity and overcoming problems is more often effected by changes in process. Difficulties such as yours are usually a Big Red Flag that says "you're doing it wrong". We ask "what are you trying to achieve" rather than "what are you trying to do".
Perhaps that's part of your problem. Or perhaps you are not listening, computers are meant to be flexible tools to help users get their jobs done in as easy of a fashion as
What a weird question!!! Achieve? v.s Do? Those words are synonyms! What am I trying to achieve? I am trying to achieve a direct copy or move of a file from one Linux computer to a USB drive mounted on another Linux computer! The WHY is IRRELEVANT! i.e - cp myFile /mnt/usbdrive/someDirectoryOnTheUSBDriveMountedOnSomeOtherMachineOnMyLAN or if I use the KDE/Hal supplied automated mount point - cp myFile /media/MyPassport/someDirectoryOnTheUSBDriveMountedOnSomeOtherMachineOnMyLAN Is that clear enough? possible. Not straight-jackets such as what Microcrap wants to impose on its users. And Anton, you telling me that I have to do things in some two staged Linux way, is exactly the same sort of message Microcrap tells it's users - "Do it my way or take the highway" approach... And I don't believe you for one second, Linux has been very flexible for the most part and there is a lot of reasons I like it better... But this apparently is NOT one of those reasons and someone somewhere is either designing things wrong or making mistakes in implementing some of these tools, or presenting me with tools and a model that has allowed and made it too easy for me to make a serious mistake, without any easy/useful means of diagnosing the problem and correcting it....
OBTW: Marc, it is considered impolite to reply on the forum to mail that you have been sent privately, off forum.
This I did not know, but will turn this around by saying that you did not ask me to communicate in private up front. I thought I was being nice and correcting a mistake you were making, when you replied to me directly, by responding via the openSuSE forum. I prefer that communications be done publicly, on this forum, so that others may follow in my footsteps, should they encounter the same issue. Unless you are going to share with me some configuration details, or other such information that needs to be kept private for security reasons, please reply via the forum. Otherwise please give me a heads up first, and I will honor your privacy and respond as appropriate. Marc.. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
It looks like a cut/paste error occurred on me and lost a bit of what I had written... This part should read as follows... (easy to do, my touchpad seems to be super sensitivity to slight unintended touches and can screw things up on me!) Marc.. On 2/9/2011 3:33 PM, Marc Chamberlin wrote:
But then I am exporting the device, not the mount point. When you mount the device is BELOW the mount point, and as I said and you have verified, NFS won't go past the mount point. This I do not understand and makes NO sense to me. I have tried to export /dev/sdc1, /media/MyPassport, and /mnt/usbdrive all to no avail. What am I trying to mount, in all three of these cases, is what I perceive to be THE mount point for the USB drive. To me some of this is extremely confusing also! If I mount a USB drive at some location, or let KDE/Hal do it for me, the locally I can do an ls on that mount point and see what is on my drive. That makes sense and intuitively that implies that these paths ARE the mount point. But doing an ls on /dev/sdc1, or /proc/bus/usb shows nothing. Ergo I would never have guess these to be mount points, but what the heck, I experimented and gave them a try also.. No joy... What do you mean by "the device is BELOW the mount point"? It has become clear that I cannot mount anything "ABOVE" the mount point, like / and then expect to be able to access /media/MyPassport, because of an NFS restriction. OK, but I am exporting what appears to me as THE mount point and that should be sufficient for NFS since I am giving my blessing to NFS to export THE mount point. What do you mean by "BELOW"? That just does not make any kind of sense!
-- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
* Marc Chamberlin
first, and I will honor your privacy and respond as appropriate.
much toooo much detail why don't you try scp or sftp? write a script to automate the transfer -- (paka)Patrick Shanahan Plainfield, Indiana, USA HOG # US1244711 http://wahoo.no-ip.org Photo Album: http://wahoo.no-ip.org/gallery2 Registered Linux User #207535 @ http://counter.li.org -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
Patrick Shanahan said the following on 02/09/2011 07:17 PM:
* Marc Chamberlin
[02-09-11 18:34]: first, and I will honor your privacy and respond as appropriate.
much toooo much detail
why don't you try scp or sftp? write a script to automate the transfer
They say in perl "There More than One Way To Do It". The same appies here. In my DatabaseofDotSigQuotes People who won't quit making the same mistake over and over are what we call conservatives. - Richard Ford, in his novel Independence Day -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
Marc Chamberlin said the following on 02/09/2011 07:00 PM:
What do you mean by
"the device is BELOW the mount point"? It has become clear that I cannot mount anything "ABOVE" the mount point, like / and then expect to be able to access /media/MyPassport, because of an NFS restriction.
Correct.
OK, but I am exporting what appears to me as THE mount point and that should be sufficient for NFS since I am giving my blessing to NFS to export THE mount point. What do you mean by "BELOW"? That just does not make any kind of sense!
If you read the NFS man page it is quite clear. You need to export the directory and mount point. So I've exported Server:/home/anton but there is another fs mounted on server:/home/anton/PDF so I have to export that AS WELL Not one or the other. BOTH And, again, read the manual: you need to tell the server that you don't want the second FS to be hidden. Its there in the manual. So you need something like /media \ 192.168.x.0/255.255.255.0(rw,async,no_root_squash,crossmnt) /media/sdb1 \ 192.168.x.0/255.255.255.0(rw,async,no_root_squash,crossmnt,nohide) RTFM As for permission, it clear that you are mounting the USB with the wrong permissions. Example 1 clear out the mount 2 chown marc.users /media/sdb1 That sets the ownership of the mount point 3 insert the usb and let the kernel mount it PLEASE DO NO USE THE AUTOMOUNTER MAPS THEY JUST CONFUSE THE MATTER 4 verify its mounted $ mount ... /dev/sdb1 on /media/sdb1 type vfat (rw,noexec,nosuid,nodev....) .... OR WHATEVER 5 now check ownership of the mount point ls -l /media/sdb1 If I'm right, that's no lover "marc.users" so all the files on the mounted VFAT aren't either Both Tajas and myself have suggested you experiment with manually mounting. I don't think you understood. The word experiment means try variations. You point out in mail to Tajas that you tried mount -t auto /dev/sdc1 /mnt/usbdrive -o rw,nodev,noexec BUT YOU DIDN"T SAY WHAT UID YOU EXECUTED that as. Of course if you did that as 'uid=marc' than the mount is owned by 'marc'. If you do it as root then its owned by root. Guess what the kernel event driven mechanisms out of HAL operates as? $ ps -ef |grep sdb root 1680 1405 0 18:27 ? 00:00:00 hald-addon-storage: polling /dev/sdb (every 16 sec) Right - ROOT !! So it gets mounted as root. Now as root you can FORCE a different UID mount -t auto /dev/sdc1 /mnt/usbdrive \ -o rw,nodev,noexec,uid=<marc>,gid=<users>,umask=007 Now you want to substitute in there the numeric values. I think you said that marc=10001 "users" ... look it up /etc/groups. Its should be 100 So you have -o rw,nodev,noexec,uid=1001,gid=100,umask=007 Do that mount command as root, just as HAL would, and see what you get. Personally I'd use umask=007,fmask=0117,dmask=0007 but then I believe in not fighting Linux access permissions. So you need to make sure that the USB is always mounted like that. So put those values in /etc/fstab. None of this is weird; this is how Linux works and its all documented. If you'd backed off and experimented and read the MAN pages you'd have found this. ITS ALL DOCUMENTED. But you seem intent on hammering at the same-old-same-old and making disparaging remarks about Linux and the way it works. Its odd how those of us who are more relaxed and believe that the people who made the design decisions did so for sensible reasons, and that mature products like NFS have been worked over by many minds, minds smarter than you or me, can get these things to work. Its sort of like Ju-Jitsu, don't resist, use the opponents strength .... go with the flow. You derogatory attitude towards Linux and its facilities does not encourage people to help you. Patrick just suggested alternative methods of achieving the end result. Others have been put forward. What we're saying boils down to If you're not getting anywhere banging your head against the wall, why not stop and walk around the end of it instead? Which is why, despite what you say, IT DOES NOT MATTER WHAT MY APPLICATION IS, the objective matters. Achieving the objective means that getting hung up on one method that isn't working is not productive. In fact even with NFS - or CIFS - working "properly", you are not doing things right. Copy overwrites. You want to keep both USBs in sync as changes are made. That is what RSYNC is for. Yes you can use RSYNC with NFS cross-mounted devices, but it also works in the absence of NFS or CIFS. A small script, python using gamin perhaps, that notices changes to the file (File Alteration Monitoring - do a search) http://en.wikipedia.org/wiki/File_alteration_monitor and then use RSYNC. This would be most efficient for a number of reasons, not least of all reducing latency and making the issue of NFS/CIFS irrelevant. No-one is forcing you to stop banging you head against a wall. That you won't consider alternatives indicates you META-process is wrong. NFS can handle it; it is neither broken not badly designed. As I keep saying, "It works for me". It is your antagonistic attitude that is your problem. -- "Objectives are not fate; they are direction. They are not commands; they are commitments. They do not determine the future; they are a means to mobilize resources and energies of the business for the making of the future." -- Peter F. Drucker. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
Marc Chamberlin said the following on 02/09/2011 06:33 PM:
And, as I said, the device notifier is a KDE applet that lets you deal with the mounted devices, it does not do the mount. All it does is let you run the file manager, or gwenview or something. You may be technically correct, but to someone who did not develop that tool, that is NOT obvious. What a user see's is that it notifies you when you insert a USB device. It pops up dialog, and tells you about the device. Then the user can click on the device, within the notifier, and IT GETS MOUNTED.
Not so. It offers the choice of that to do with an ALREADY MOUNTED device. I don't have to be the developer to determine this. Logically, it has to have mounted the device so it knows what it is, storage, camera, phone ..., before it can offer the choice of what to do with it, what applications to offer.
It pops up dialog, and tells you about the device
Right. So it must have seen that the device is mounted! Or, it you really want to be picky. Remove the notifier applet. Open a terminal window. Run "mount" in a loop to tell you when what is mounted changes Insert a device. See! It gets mounted by HAL NOT by the device notifier. If you read the man page on the HAL daemon, the same one that spawns the process that polls the USB ports and appears in the process table: root 1680 1405 0 18:27 ? 00:00:00 hald-addon-storage: polling /dev/sdb (every 16 sec) which is responsible for mounting the USB device, you will see it says DESCRIPTION hald is a daemon that maintains a database of the devices connected to the system system in real-time. The daemon connects to the D-Bus system message bus to provide an API that appli- cations can use to discover, monitor and invoke operations on devices. For more information about both the big picture and specific API details, refer to the HAL spec which can be found in /usr/share/doc/packages/hal/spec/hal-spec.html depending on the distribution. OH WOW! Its all documented. Non of this is the least bit obscure. I've not had to be the developer or even read the source code. I've not had to use obscure or sophisticated tools or write scripts, even in shell. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
Marc Chamberlin said the following on 02/09/2011 07:00 PM:
What do you mean by
"the device is BELOW the mount point"? It has become clear that I cannot mount anything "ABOVE" the mount point, like / and then expect to be able to access /media/MyPassport, because of an NFS restriction. Correct.
OK, but I am exporting what appears to me as THE mount point and that should be sufficient for NFS since I am giving my blessing to NFS to export THE mount point. What do you mean by "BELOW"? That just does not make any kind of sense! If you read the NFS man page it is quite clear. You need to export the directory and mount point.
So I've exported Server:/home/anton but there is another fs mounted on server:/home/anton/PDF so I have to export that AS WELL
Not one or the other. BOTH
And, again, read the manual: you need to tell the server that you don't want the second FS to be hidden. Its there in the manual.
So you need something like
/media \ 192.168.x.0/255.255.255.0(rw,async,no_root_squash,crossmnt) /media/sdb1 \ 192.168.x.0/255.255.255.0(rw,async,no_root_squash,crossmnt,nohide)
RTFM I think Anton, that YOU need to RTFM on the man pages for NFS before you make assumptions/claims that I have not done so. NOWHERE, in the man
As for permission, it clear that you are mounting the USB with the wrong permissions.
Example
1 clear out the mount 2 chown marc.users /media/sdb1 That sets the ownership of the mount point 3 insert the usb and let the kernel mount it PLEASE DO NO USE THE AUTOMOUNTER MAPS THEY JUST CONFUSE THE MATTER 4 verify its mounted $ mount ... /dev/sdb1 on /media/sdb1 type vfat (rw,noexec,nosuid,nodev....) .... OR WHATEVER 5 now check ownership of the mount point ls -l /media/sdb1
If I'm right, that's no lover "marc.users" so all the files on the mounted VFAT aren't either
Both Tajas and myself have suggested you experiment with manually mounting. I don't think you understood. The word experiment means try variations.
You point out in mail to Tajas that you tried
mount -t auto /dev/sdc1 /mnt/usbdrive -o rw,nodev,noexec
BUT YOU DIDN"T SAY WHAT UID YOU EXECUTED that as. Of course if you did that as 'uid=marc' than the mount is owned by 'marc'. If you do it as root then its owned by root. HUMP! I have NO IDEA what you are trying to say Anton. I CANNOT execute a mount command EXCEPT as ROOT! Trying to execute a mount command as me, user marc, is NOT allowed. I get told promptly that the mount command is ONLY allowed by ROOT if I try. Therefore, since there was no uid
On 2/9/2011 7:12 PM, Anton Aylward wrote: page for NFS does it say anything like what you are claiming it does. I HAVE read it, several times, top to bottom. NOWHERE. There is not one word saying that NFS by default, will not allow a client to traverse past a mount point that is below an exported mount point. There is not one word that says if you are going to export a mount point that you must also export the parent directory and that the client has to import both. There is NOT a single shred of documentation regarding why these particular limitations are in place. In fact, there is very little description of the model(s) behind NFS and I get the feeling that what is documented is just a few bits and pieces. Do a man nfs and READ IT YOURSELF. It comes up as nfs(5). And you wonder why I despise external documentation so much, and people who yell at me to RTFM first. Rarely is the documentation referred to in sync with the tools it describes, is complete, well organized, or up to date. These man pages on NFS are a mess, and as I said before, using/maintaining external documentation is a BROKEN process. Tools should present a self documenting interface to users, that is the ONLY way I know of to reliably keep documentation in sync with the tool. Command line interfaces can be designed to be self documenting if they are interactive, most are not. GUI's make the job even easier. Maybe the NFS subsystem is "mature" and has been around for awhile. Maybe a lot of Linux gurus have gotten information via the grapevine also on how to use it. Doesn't change the fact that NFS IS broken, one way or the other. Without proper documentation, a user cannot learn how to use a tool on his/her own, therefore must seek help, and DOES NOT NEED SOME "EXPERT" YELLING AT HIM TO RTFM with bad documentation. So I dug deeper, and took a hard look at the man page for exports. In particular at the nohide and crossmnt parameters. And there, buried, is a couple of offhand remarks about these limitations and what has to be done to get around it. How in the world is a user suppose to grasp what has gone wrong, when the NFS system is failing him, with no descriptions of underlying models, and be able to recognize that one, or the other, or both, of these parameters is needed? A user unfamiliar with NFS does not even have a clue where to begin diagnoses! Now if ONLY I could be so lucky as to be able to even get to the point where I can even use these parameters. Read on sir, before you reply. Even with using these parameters I cannot get NFS to work properly... parameter supplied in the command I gave, the uid of the mount point will be ROOT, that is the default behavior of the mount command, and I showed you exactly what I executed. So what in the world are you saying here?
Guess what the kernel event driven mechanisms out of HAL operates as?
$ ps -ef |grep sdb root 1680 1405 0 18:27 ? 00:00:00 hald-addon-storage: polling /dev/sdb (every 16 sec)
Right - ROOT !!
So it gets mounted as root.
Now as root you can FORCE a different UID
mount -t auto /dev/sdc1 /mnt/usbdrive \ -o rw,nodev,noexec,uid=<marc>,gid=<users>,umask=007
Now you want to substitute in there the numeric values. I think you said that marc=10001 "users" ... look it up /etc/groups. Its should be 100 So you have
-o rw,nodev,noexec,uid=1001,gid=100,umask=007
Do that mount command as root, just as HAL would, and see what you get.
Personally I'd use
umask=007,fmask=0117,dmask=0007
but then I believe in not fighting Linux access permissions. Let me try this again, even changing the uid and gid does not make any difference. Exporting the mount points. BOTH of em, or even all THREE still fails on me. Follow the dots..
So you need to make sure that the USB is always mounted like that. So put those values in /etc/fstab.
None of this is weird; this is how Linux works and its all documented. If you'd backed off and experimented and read the MAN pages you'd have found this. ITS ALL DOCUMENTED. Maybe.. maybe I don't have all the document packages installed or something.. What I have is certainly NOT clear and easy to understand. Maybe there are other obscure locations in the man pages that describe what is going on. But just because the documentation mentions these sort of details, somewhere within it, does NOT mean the documentation has been written in a clear concise, complete, well formatted manner. It is
But you seem intent on hammering at the same-old-same-old and making disparaging remarks about Linux and the way it works. Its odd how those of us who are more relaxed and believe that the people who made the design decisions did so for sensible reasons, and that mature products like NFS have been worked over by many minds, minds smarter than you or me, can get these things to work. Its sort of like Ju-Jitsu, don't resist, use the opponents strength .... go with the flow.
You derogatory attitude towards Linux and its facilities does not encourage people to help you. You keep getting me wrong about how I feel toward Linux. I feel it could stand a lot of improvement, if that is being derogatory towards Linux,
First, I ensured KDE's handling of removable storage media was disabled
(Configure Desktop > Advance > Removable Devices > Enable automatic
mounting of removable media) and got it completely out of the picture,
(as well as the device notifier). I believe this is what actually will
cause the auto mount of a plugged in device to take place, if enabled,
but it was initially turned off on my system. And without it being
active, despite what you say, NOTHING would appear under /media until I
click on the button to "turn on", "mount" whatever you want to call it,
for the identified device in the popup menu from the device notifier.
Once I click on it, THEN AND ONLY THEN do I get /media/My Passport and
I presume that a mount process has then occurred, but maybe there is
some other term for it. IF KDE is configured to automatically mount
removable storage media, then I do NOT have to click on anything in the
Device Notifier and the drive is mounted at /media/My Passport. Anywise,
I disabled both the device notifier AND KDE's configuration to
automatically handle removable storage media.
Now I plug in my USB drive... and nothing obvious happens any more. No
mounts under /media nor under /mnt. The USB drive, for all intents and
purposes, remain inaccessable to me, until I perform the mount manually.
Second, as root, I created the mount point directory as root -
cd /mnt
mkdir "My Passport"
and executed the following -
mount -t auto /dev/sdc1 /mnt/My\040Passport -o
rw,nodev,exec,uid=1000,gid=100,umask=000,fmask=0000,dmask=0000
And yes it is mounted with the owner being me... No problems so far, can
access it, do whatever... I even tried it with your flavor of umask,
fmask, and dmask, no difference in behavior with regards to the failure
that is occurring when I try to export it...
Third, I edited exports in /etc as follows
/
192.168.2.0/255.255.255.0(fsid=0,crossmnt,rw,no_root_squash,sync,no_subtree_check)
/home 192.168.2.0/255.255.255.0(rw,no_root_squash,sync,no_subtree_check)
/tmp 192.168.2.0/255.255.255.0(rw,no_root_squash,sync,no_subtree_check)
/mnt
192.168.2.0/255.255.255.0(rw,no_root_squash,sync,no_subtree_check,crossmnt)
/mnt/My\040Passport
192.168.2.0/255.255.255.0(rw,no_root_squash,sync,no_subtree_check,crossmnt,nohide)
I also tried this with and without the space in the name - My Passport.
I went back to using the space since that is what KDE/Device
Notifier/Hal was doing, and I ultimately would like to be able to do
this without having to manually mount the device. But for now will play
along with manually doing things..
and finally I restart the nfsserver
rcnfsserver restart
Shutting down kernel based NFS server: nfsd statd mountd
idmapd
done
Starting kernel based NFS server: idmapdexportfs: Warning: /mnt/My
Passport does not support NFS export.
mountd statd nfsd
sm-notify
done
and I get a warning. If I ignore the warning, (since it is not saying it
is an error...) and restart the autofs daemon on another system, with
the following map file -
slash -rw,soft,intr,timeo=50 marcslaptop:/
homes -rw,soft,intr,timeo=50 marcslaptop:/home
temp -rw,soft,intr,timeo=50 marcslaptop:/tmp
mnt -rw,soft,intr,timeo=50 marcslaptop:/mnt
MyPassport -rw,soft,intr,timeo=50 marcslaptop:/mnt/My\040Passport
then if I try to access ANY of the exported/mounted directories, the
system hangs and waits for a timeout... Nothing reported in log files
under /var/log. Removing the import of MyPassport, and rebooting will
restore the ability to access the other directories.
like looking for a needle in a haystack. As for the documentation
itself, I cannot find anything obvious that I missed, but then I am not
a Linux guru who knows what and where everything is stored in all those
update repositories either.
then that is YOUR opinion. Google around on the internet for awhile, you
will find LOTs of people struggling with things like NFS. Hint - that is
a clue it needs to be improved.
Well crap! Now this is the kind of bug that can be REALLY aggravating! I had a friend come over to take a look at this problem, with me, to see if two eyes would be better than one. And he decides to bring up the map file using Xemacs, and damned if there weren't some non-printing characters embedded in that file. Dunno how they got there, and kwrite was not showing them to me... Sigh... Removed em and all started working as expected... Ugh! Anywise, my comments about NFS's documentation still stands. but we have been chasing red herrings apparently... Now to see if using the KDE/Hal automount process will still work... I am worried about how permissions/ownership gets set, and how to configure them, if it prevents me from exporting and mounting the USB drive... Will report back soon... Marc... -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
What do you mean by "BELOW"? That just does not make any kind of sense!
I understand 'below' the nfs mount point like this (eg): I export /home from 192.168.1.1 to 192.168.1.3 It becomes /home on 192.168.1.3 A user on 192.168.1.3 would then describe /home/lynn as _below_ home. I as root on 192.168.1.1 could see _above_ the share. eg I could go to /etc but the user on the shared box could not see _my_ /etc. Only browse the /etc on her own box.the end. Saludos, Cheers. L x -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
Marc Chamberlin said the following on 02/10/2011 01:19 PM:
I think Anton, that YOU need to RTFM on the man pages for NFS before you make assumptions/claims that I have not done so. NOWHERE, in the man page for NFS does it say anything like what you are claiming it does.
Yes, but I read ALL the documentation on NFS, not just the NFS man page. Try the 'exports' man page. There you will find it says Normally, if a server exports two filesystems one of which is mounted on the other, then the client will have to mount both filesystems explicitly to get access to them. If it just mounts the parent, it will see an empty directory at the place where the other filesystem is mounted. That filesystem is "hidden". You will see in the example entry I sent for the entries in /etc/exports.
/media \ 192.168.x.0/255.255.255.0(rw,async,no_root_squash,crossmnt) /media/sdb1 \ 192.168.x.0/255.255.255.0(rw,async,no_root_squash,crossmnt,nohide)
As I said there and then
RTFM It explains how deal with the situation you are in.
I'm so glad you found it. I'm only disapointed that you rushed ahead without reading everything. I know man pages are tedious and dry, but they are important. But recall: I said "It works for me" early on. That was becuase I'd read the man page. Man pages are also NOT there to justify the design decisions. They explain what the functions and paramaters are. If you want the design documents, I suggest you lok forthem on-line. NFS is a SUN design. Your bombast is unfounded. You make various other assertions about what can't be done. I'm sorry, Marc, its back to "it works for me". I don't know what you are doing wrong. That's why I said you must be infested with gremlins. If it works when I do it (and others I've spoken here and other client sites who've tried it), and it doesn't work for you, either you're not doing the same thing or you've got gremlins. -- Getting into a patent battle with IBM is right up there with starting a land war in Asia -- http://slashdot.org/~Don+Negro -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
Marc Chamberlin said the following on 02/10/2011 01:19 PM: to say Anton. I CANNOT execute
a mount command EXCEPT as ROOT! Trying to execute a mount command as me, user marc, is NOT allowed. I get told promptly that the mount command is ONLY allowed by ROOT if I try. Therefore, since there was no uid parameter supplied in the command I gave, the uid of the mount point will be ROOT, that is the default behavior of the mount command, and I showed you exactly what I executed. So what in the world are you saying here?
I'm trying to show that if you let HAL - which runs as root - mount the decvice, it will be mounted as root. Its odd, but I can mount my USB devices as 'anton'. Its back to the MAN pages. In this case the MOUNT man page, which explains how non-root users can mount things: The non-superuser mounts. Normally, only the superuser can mount file systems. However, when fstab contains the user option on a line, anybody can mount the corresponding system. A little later it goes on to say For more details, see fstab(5). Which gets back to my point about reading other man pages well :-) Further, it says The owner option is similar to the user option, with the restriction that the user must be the owner of the special file. ... The group option is similar, with the restriction that the user must be member of the group of the special file. I expect you'll find /dev/sdb1 is root.floppy. If it was marc.users there would be no problem. Well, part of the "it works for me" was to alter the entry in /etc/fstab in accordance with the man page. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
* Marc Chamberlin
And he decides to bring up the map file using Xemacs, and damned if there weren't some non-printing characters embedded in that file. Dunno how they got there, and kwrite was not showing them to me... Sigh... Removed em and all started working as expected...
Hint, kwrite is not a :text-editor". Graphics/GUI is *not* always better (er, ever). -- (paka)Patrick Shanahan Plainfield, Indiana, USA HOG # US1244711 http://wahoo.no-ip.org Photo Album: http://wahoo.no-ip.org/gallery2 Registered Linux User #207535 @ http://counter.li.org -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
Patrick Shanahan said the following on 02/10/2011 03:26 PM:
* Marc Chamberlin
[02-10-11 14:01]: And he decides to bring up the map file using Xemacs, and damned if there weren't some non-printing characters embedded in that file. Dunno how they got there, and kwrite was not showing them to me... Sigh... Removed em and all started working as expected...
Hint, kwrite is not a :text-editor".
Graphics/GUI is *not* always better (er, ever).
ROTFL! So it wasn't Evil USBs after all ... http://www.guardian.co.uk/science/the-lay-scientist/2010/nov/15/3 -- "...Then anyone who leaves behind him a written manual, and likewise anyone who receives it, in the belief that such writing will be clear and certain, must be exceedingly simple-minded..." -- Plato, _Phaedrus_ -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On Thursday 10 February 2011 20:15:02 Anton Aylward wrote:
Its odd, but I can mount my USB devices as 'anton'. Its back to the MAN pages. In this case the MOUNT man page, which explains how non-root users can mount things:
The non-superuser mounts. Normally, only the superuser can mount file systems. However, when fstab contains the user option on a line, anybody can mount the corresponding system.
Yes, but that doesn't apply if you give more than one parameter to the mount command. As soon as you specify anything more than the device or mount point, the fstab entry is overridden, and you can no longer mount as user. halmount should work, but since hal is deprecated it won't be around much longer. I'm not sure what the replacement will be Anders -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
Anders Johansson said the following on 02/10/2011 06:31 PM:
On Thursday 10 February 2011 20:15:02 Anton Aylward wrote:
Its odd, but I can mount my USB devices as 'anton'. Its back to the MAN pages. In this case the MOUNT man page, which explains how non-root users can mount things:
The non-superuser mounts. Normally, only the superuser can mount file systems. However, when fstab contains the user option on a line, anybody can mount the corresponding system.
Yes, but that doesn't apply if you give more than one parameter to the mount command. As soon as you specify anything more than the device or mount point, the fstab entry is overridden, and you can no longer mount as user.
I don't see the problem. It just means you have to get the entry in the fstab right :-)
halmount should work, but since hal is deprecated it won't be around much longer. I'm not sure what the replacement will be
I can't recall if pmount (policy mount), which can get around all that (and hence the hal wrapper 'pmount-hal') is in the future or in the past. The sad thing about hal being depreciated (2001 jokes aside) begins with phrases such as "just when I've got it all figured out ...." -- Any man who afflicts the human race with ideas must be prepared to see them misunderstood. --Henry Louis Mencken -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
Anders Johansson said the following on 02/10/2011 06:31 PM:
halmount should work, but since hal is deprecated it won't be around much longer. I'm not sure what the replacement will be
The replacement will be PolicyKit, and it looks like a lot of PolicyKit is already there and working. I'm going to dig to see if hal is actually using PolicyKit, in which case a PolicyKit solution would work and would last. We already have the /etc/polkit-default-privs.* family. I;m sure an entry in /etc/polkit-default-privs.local would take care of it. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
* Marc Chamberlin
[02-10-11 14:01]: And he decides to bring up the map file using Xemacs, and damned if there weren't some non-printing characters embedded in that file. Dunno how they got there, and kwrite was not showing them to me... Sigh... Removed em and all started working as expected... Hint, kwrite is not a :text-editor". Then what is it? It presents a text file to a user to manipulate. It apparently has a fault in that it does not display non-printing characters. Why am I, as a user, to be belittled for having chosen what appeared to be an easy to use tool to edit the configuration files? Graphics/GUI is *not* always better (er, ever). I NEVER claimed GUIs based tools are inheritantly better than command
On 2/10/2011 12:26 PM, Patrick Shanahan wrote: line tools, both can be badly designed. What I did say is that GUI's have a far better POTENTIAL to teach and guide a user in a tools usage, and make it easy for them to reach a satisfactory solution. Non-interactive command lines with external documentation has repeatedly been shown over and over, again and again, to be a failed approach when it comes to teaching/guidance. GUIs offer a far richer medium, using pictures, diagrams, built in layers of organization, can present models via easy to understand graphical concepts, automate much of the underlying difficult parts, make transparent the connections between different subsystems etc. Another advantage of GUI driven software is that the presentation layer/documentation often stays in better sync with the underlying tools capabilities, it HAS to in order to work. Yes, many GUIs are poorly design and only present a subset of a tools capabilities, but that is not the fault of the GUI approach per say, that is the fault of the designer. What Anton, and many Linux zealots fail to understand, is that half the battle of designing a well design tool(s) is that it is the job of the designers and programmers to make it easy for newcomers to learn how to use it also, intuitive if possible, and how to integrate said tool into their processes/needs to solve a particular problem. Stay with your command lines exclusively if you like, along with your external difficult to understand, unorganized documentation. The rest of the world will choose the easiest tools that guide and help them to learn, and I, as a programmer myself, will continue to be a champion for them. Why do you think GUI based tools such as Turbotax and H&R Block are so damn successful in helping people file their taxes? They are GUIDES! Try doing that with your command lines and external documentation and see how successful you will be. The current NFS system FAILED me, and many many others because of it's lack of guidance, organization, and usage of overwhelming large amounts of difficult to understand documentation. Yes the answers may be in that heap somewhere, but I am tired of having to sift through so much documentation, parse and understand the obtuse semantics used, just to find some little gem of information buried on some document page somewhere. And I am really sick and tired of callous answer such as RTFM. The moment someone says that, I KNOW I am dealing with badly designed/written software. Implementing features is one thing, designing for usability is quite another. Until you learn the difference, and understand that there are multiple levels of target audiences competency levels, please stay out of software design and development. People who do not understand this, and are writing programs, are making a mess out of computers and ruining them. I got no problems with experts who want to use concise command lines to quickly accomplish some task, but I got one hell of a problem with those same experts who want to force newcomers into HAVING to quickly learn so much background/underlying information/concepts necessary, in order to accomplish the same task, because the only means available is via those same concise command lines. That is building a needless barrier against entry into the world of Linux. Marc... -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On 2/10/2011 12:53 PM, Anton Aylward wrote:
Patrick Shanahan said the following on 02/10/2011 03:26 PM:
* Marc Chamberlin
[02-10-11 14:01]: And he decides to bring up the map file using Xemacs, and damned if there weren't some non-printing characters embedded in that file. Dunno how they got there, and kwrite was not showing them to me... Sigh... Removed em and all started working as expected... Hint, kwrite is not a :text-editor".
Graphics/GUI is *not* always better (er, ever). ROTFL!
So it wasn't Evil USBs after all ... http://www.guardian.co.uk/science/the-lay-scientist/2010/nov/15/3
No, it wasn't an evil USB, it was poorly organized documentation, the nfsserver failing to guide or tell a user that he/she was trying to do something that would not work i.e. trying to mount a filesystem without also mounting the parent filesystem, intolerance of a newcomers inability to quickly comprehend a wide range of concepts and to rapidly sift though vast amounts of documentation in order to find a trinket of information necessary to solve this problem, an editor that failed to properly warn a user about embedded non-printing characters, a parser somewhere in autofs that failed to properly warn a user about embedded non-printing characters in a configuration file, a mount process that failed to tell the user that it could not perform it's task because it encountered those same non-printing characters, or perhaps because those non-printing characters were being interpreted as part of a file system name that was not being exported and by that same mount process simply hanging until a timeout occurred without any explanation whatsoever... or perhaps there are some gremlins embedded in openSuSE/Linux somewhere.... Marc... -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
Hello, On Feb 10 21:31 Marc Chamberlin wrote (excerpt):
What ... many Linux zealots fail to understand, is that half the battle of designing a well design tool(s) is that it is the job of the designers and programmers to make it easy for newcomers ...
Basically there are no extra designers for most Linux software. Basically it is the "Linux zealots" who make most Linux software. It is a contradiction in itself to demand that "Linux zealots" should make their software easy for newcomers because they cannot. They would need newcomers who are willing to spend time and energy to contribute real value and work together with them over a longer time and help them to understand how their software looks from a newcomer's point of view. Unfortunately there is not much of those contribution. Instead there are mostly only complaints when something does not work which does not really help the developers.
Implementing features is one thing, designing for usability is quite another.
Yes. I look forward to your valuable contribution regarding software design and usability.
Until you learn the difference, and understand that there are multiple levels of target audiences competency levels, please stay out of software design and development.
Wow! If I understand you correctly, you say that many free software developers should "stay out of software development". In this case I assume with this kind of mind-set nobody would like to listen to what you will contribute regarding software design and usability. Regards Johannes Meixner -- SUSE LINUX Products GmbH, Maxfeldstrasse 5, 90409 Nuernberg, Germany AG Nuernberg, HRB 16746, GF: Markus Rex -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
Marc Chamberlin said the following on 02/11/2011 12:56 AM:
So it wasn't Evil USBs after all ... http://www.guardian.co.uk/science/the-lay-scientist/2010/nov/15/3
No, it wasn't an evil USB, it was poorly organized documentation, the nfsserver failing to guide or tell a user that he/she was trying to do something that would not work i.e. trying to mount a filesystem without also mounting the parent filesystem, intolerance of a newcomers inability to quickly comprehend a wide range of concepts and to rapidly sift though vast amounts of documentation in order to find a trinket of information necessary to solve this problem, an editor that failed to properly warn a user about embedded non-printing characters, a parser somewhere in autofs that failed to properly warn a user about embedded non-printing characters in a configuration file, a mount process that failed to tell the user that it could not perform it's task because it encountered those same non-printing characters, or perhaps because those non-printing characters were being interpreted as part of a file system name that was not being exported and by that same mount process simply hanging until a timeout occurred without any explanation whatsoever...
So, as well as not being able to tell the difference between a "programmer's" text editor and a word processor, you can't recognise a joke either. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
Marc Chamberlin said the following on 02/11/2011 12:56 AM:
..... an editor that failed to properly warn a user about embedded non-printing characters, a parser somewhere in autofs that failed to properly warn a user about embedded non-printing characters in a configuration file, a mount process that failed to tell the user that it could not perform it's task because it encountered those same non-printing characters, .....
Well, you were using a GUI. The command line version has STDIN, STDOUT an STDERROR streams Applications do complain - to the STDERROR stream that has no counterpart when you are using a GUI. Unless of course the GUI designer chooses to make the GUI very heavy and monitor all the error codes. But then perhaps the GUI designer chose not to write a wrapper around the CLI and instead, being a programmer, wrote it all in C, and found that checking all the return codes and interpreting the results and writing all the handlers and branching code was just too tedious and boring. Marc, you say you are a programmer. Do you ALWAYS check the return codes when subroutine and procedures are called, do you ALWAYS write the branching code and handlers for the results for each and every one at each and every level? By using the CLI and seeing the errors I learnt quickly. Of course you could have monitored the various files in /var/log -- Getting into a patent battle with IBM is right up there with starting a land war in Asia -- http://slashdot.org/~Don+Negro -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On 2/11/2011 12:59 AM, Johannes Meixner wrote:
Hello,
On Feb 10 21:31 Marc Chamberlin wrote (excerpt):
What ... many Linux zealots fail to understand, is that half the battle of designing a well design tool(s) is that it is the job of the designers and programmers to make it easy for newcomers ...
Basically there are no extra designers for most Linux software. Basically it is the "Linux zealots" who make most Linux software.
It is a contradiction in itself to demand that "Linux zealots" should make their software easy for newcomers because they cannot.
They would need newcomers who are willing to spend time and energy to contribute real value and work together with them over a longer time and help them to understand how their software looks from a newcomer's point of view.
Unfortunately there is not much of those contribution.
Johannes - Your thoughts are very welcomed, but I am in strong disagreement with you also. Getting newcomers to provide feedback to developers can be easily accomplished if tools are designed and developed to MAKE it easy. Feedback is THE most important tool a developer can have. With it, he can improve his/her product immensely, and thus provide the multi level support needed for all competencies. How often have I wished to see a feedback button located right under the Help button! That alone shows the developer(s) care. (Command line designers/zealots could also put in a 'feedback' parameter for their tools which would solicit and forward comments to the designers. Makes wrapping such tools with thoughtful GUI's easier also.) I recognize that you will argue, that developers will fear being overwhelmed then. Maybe in the beginning, right after a tool is released, that will happen. But if developers take time to listen, adjust the product and correct the errors, that rate will taper off. Tools like GUI based Bugzilla can also mitigate some of this, via their voting process. Trouble is, most newcomers cannot even FIND, or even are aware of, a bugtracking tool that supports a particular tool, or a forum such as this one where help can be solicited. So, BUILD IT IN to the product! Put it in the HELP design and lead the user to it. In other words put it up front so the newcomers and others CAN discover/use these mechanisms. Your complaint about the lack of getting feedback and help from newcomers again boils down to how the tool and its support subsystem was designed. Usually badly....
Instead there are mostly only complaints when something does not work which does not really help the developers.
Oh YES, but complaints are extremely valuable, IF AND ONLY IF you take the time to LISTEN and understand where the problem really lies, and take the time to then correct it. Most complaints are very likely to boil down to the fact that the user does not understand how to use a tool. By correcting, I don't mean correct the user, but rather correct the tool. Answers, such as what Anton gave me - RTFM - is so wrong on so many levels. Anton complained that I was beating my head against a wall because I was unwilling to change my approach to a problem I was trying to solve. I disagree. NO, it is the developers who are beating their heads against a wall, insisting that users RTFMs, designing tools in a now 30 year old fashion, and not being willing to LISTEN to users and change their approach in software design and development. I repeat my earlier statement, if you want less complaints, design the tools to be teachers and guides, provide easy feedback systems that allow corrections to take place, as well as implementers of functionality.
Implementing features is one thing, designing for usability is quite another.
Yes.
I look forward to your valuable contribution regarding software design and usability.
And you are getting it, right now! Anton said that I, and users in general, have to learn how to do things the "Linux way" Fine. I just want that "Linux way" to include all levels of user competencies, not just experts who have a broad width and depth of understanding the Linux architecture and the associated tools limitations. That requirement becomes a barrier to entering the Linux world. Where are the "Linus Torvold's" who set the standards for the look and feel of Linux and it's tool sets? I know Linus has been one of the main gatekeepers for the kernel itself, but what about the rest of the system? Who make's the decision about what tools get included in a distribution's packages? At some level this has to be happening, and this is where a "standard" of acceptance should be developed, and tools judged on how well they meet those standards. Then roadmaps can be laid out for volunteers to pick and choose what to work on, and tool development guided towards becoming more and more usable for more and more users...
Until you learn the difference, and understand that there are multiple levels of target audiences competency levels, please stay out of software design and development.
Wow!
If I understand you correctly, you say that many free software developers should "stay out of software development".
In this case I assume with this kind of mind-set nobody would like to listen to what you will contribute regarding software design and usability.
No, you do not understand me correctly. I have nothing against free software developers, and much appreciate their time and efforts. But they need to be guided as well, and shown the way towards how to create excellent software. What I was really trying to say is that if you are unwilling to change your software development attitude, and design for usability instead of just functionality, then don't design at all. That attitude is doing far more harm than good for computers in general. I have taken a lot of time to try and explain, with examples, of what sort of thinking is required, (got off on a huge tangent here!) if Linux ever wants to evolve into a really superior OS and a cool environment for ALL of it's users. I am willing to help, I have been in this business for a long time now and seen some really neat examples of what works and what doesn't. I have also been a teacher, a very humbling experience, and the one thing I have learned is you cannot pigeon hole users and force them into fast learning processes that require a broad width of understanding of computer science. You cannot simply rely on requiring users to read lots of external dry documents in order to use a particular tool. You try and you will fail. That is archaic thinking and there are far better ways.... I hope a few developers/readers are really listening and doing some soul searching. I will summarize - Successful software development is going to be judged in terms of usability. Functionality is 10% of the battle, and currently gets 90% of many if not most software developer's focus. That focus HAS to change to be closer to 50/50. By functionality I am referring to those features/software elements that are focused on making the tool perform its basic job. By usability I am referring to those features/software elements that are focused on making it easier for a user to use the tool, provide feedback to developers, guide and teach painlessly, and finally make it a joy for the user to use said tool.
Regards Johannes Meixner
Regards also... Marc Chamberlin -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On 2/11/2011 4:53 AM, Anton Aylward wrote:
Marc Chamberlin said the following on 02/11/2011 12:56 AM:
So it wasn't Evil USBs after all ... http://www.guardian.co.uk/science/the-lay-scientist/2010/nov/15/3
No, it wasn't an evil USB, it was poorly organized documentation, the nfsserver failing to guide or tell a user that he/she was trying to do something that would not work i.e. trying to mount a filesystem without also mounting the parent filesystem, intolerance of a newcomers inability to quickly comprehend a wide range of concepts and to rapidly sift though vast amounts of documentation in order to find a trinket of information necessary to solve this problem, an editor that failed to properly warn a user about embedded non-printing characters, a parser somewhere in autofs that failed to properly warn a user about embedded non-printing characters in a configuration file, a mount process that failed to tell the user that it could not perform it's task because it encountered those same non-printing characters, or perhaps because those non-printing characters were being interpreted as part of a file system name that was not being exported and by that same mount process simply hanging until a timeout occurred without any explanation whatsoever... So, as well as not being able to tell the difference between a "programmer's" text editor and a word processor, you can't recognise a joke either. Attempting to ridicule me again Anton will NOT come across as helpful. As for jokes, when done at someone's expense, that is an attempt to be hurtful and bullying. No, I do not accept those and have told you that ridiculing me, a novice in regard to usage of the NFS tool set, is NOT helpful in the least. It only reflects badly on your own character... BTW I regret having chosen to use your own remarks about gremlins against you, and apologize. That was twisting the knife and not necessary to make my point. Your ridicule of me made me angry, but I should stay professional and not respond in kind.
As for using a "programmer's" text editor v.s a word processor? If a tool purports to handle a plain text file, I expect that tool to honor the syntax conventions associate with plain text files. I see nothing wrong with using any tool that will edit text files, conceptually, as a simple text file. By not displaying or warning about non-printable characters, other than CR or LF, kwrite has a bug in it. I use other text editors such as xemacs and vi all the time. Because I had approached these configuration files via Dolphin, kwrite was handy, so I used it instead. It bit me, and I learned of it's limitation. Won't make that mistake again... I also pointed out, that there are a number of issues with various different tools, that aligned to make the process of setting up an nfs server and an autofs client extremely difficult. You might want to focus on those instead of attacking me. Marc... -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
Marc Chamberlin said the following on 02/11/2011 02:43 PM:
As for using a "programmer's" text editor v.s a word processor? If a tool purports to handle a plain text file, I expect that tool to honor the syntax conventions associate with plain text files.
"Purports to handle" could mean anything. Word processors can READ IN plain text files. That could be shoe-horned into "handle", but the resulting file when the word processor has finished with it is a word processor file. It is, after all, a word processing tool. There are plenty of tools for editing pain text files. Some "programmers editors" for pain text files like Kate and GVIM have facilities for "syntax directed editing" and can make smart guesses at the 'language' in the file: C/C++, perl, php, as well as various types of config file. Kate also has a terminal emulator built in, so you could view it as a development environment.
I see nothing wrong with using any tool that will edit text files, conceptually, as a simple text file.
Indeed, but I'd rather have one that will edit text files ACTUALLY.
By not displaying or warning about non-printable characters, other than CR or LF, kwrite has a bug in it.
That makes no sense. Kwrite is a WORD PROCESSOR. It writes out word processing files. It is not a text editor any more than Quanta+, Inkscape or GIMP is a text editor. All those can manipulate text, but what they write out isn't a text file
Because I had approached these configuration files via Dolphin, kwrite was handy, so I used it instead. It bit me, and I learned of it's limitation. Won't make that mistake again...
This is where Patrick would step in and say that your mistake was using a GUI -- Dolphin -- and let it make decisions for you. I would add that GUIs in general let other people make decisions for you and limit what you can see and do to what they think you should be doing and seeing. You have illustrated this point exceedingly well.
I also pointed out, that there are a number of issues with various different tools, that aligned to make the process of setting up an nfs server and an autofs client extremely difficult. You might want to focus on those instead of attacking me.
You might want to use the right tools and in the right way and gain an understanding of how the tools and facilities work and interact BEFORE hacking away. I kept saying "It works for me" and looking back I'm of the opinion that it worked without problems such as you encountered because I chose the right tools and read up and planned. I was taught "The seven Ps", but it seems one of the Ps is considered a rude word by many Nanny Filters, so its "The Six Ps" you'll find discussed most on the 'Net http://jbdcolley.com/?p=184 Marc, I see from other people and some out-of-band that you are coming across as attacking Linux, previous version of UNIX, their designers and developers and in many ways the OpenSource movement. This is not appreciated. There's a saying by the old Roman poet Horace that sums things up well: If a better system is thine, impart it freely; If not, make use of mine. -- 'Faith' means not _wanting_ to know what is true. -- Nietzsche, Der Antichrist -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
Marc Chamberlin said the following on 02/11/2011 12:56 AM:
..... an editor that failed to properly warn a user about embedded non-printing characters, a parser somewhere in autofs that failed to properly warn a user about embedded non-printing characters in a configuration file, a mount process that failed to tell the user that it could not perform it's task because it encountered those same non-printing characters, ..... Well, you were using a GUI. The command line version has STDIN, STDOUT an STDERROR streams Applications do complain - to the STDERROR stream that has no counterpart when you are using a GUI.
Unless of course the GUI designer chooses to make the GUI very heavy and monitor all the error codes. But then perhaps the GUI designer chose not to write a wrapper around the CLI and instead, being a programmer, wrote it all in C, and found that checking all the return codes and interpreting the results and writing all the handlers and branching code was just too tedious and boring. o STDERROR messages can easily be directed to a log file and a GUI designer can easily make a user aware of that log file and point him to it. Or even open it up to him/her and display it for him/her. That is a far better approach than simply dropping them on the floor, which so easily can happen is using command lines... Marc, you say you are a programmer. Do you ALWAYS check the return codes when subroutine and procedures are called, do you ALWAYS write the branching code and handlers for the results for each and every one at each and every level? I tend to use OO design methodologies and design patterns to encapsulate error handling. I tend to design programs up front to handle error
By using the CLI and seeing the errors I learnt quickly. Which, as I have been trying to point out, is responding to error situations after the fact, and forcing the user to try and visualize what might have gone wrong that caused such an error. That requires a LOT of inside detailed knowledge, on the users part, in order to successfully grok a solution. Reverse your thinking when it comes to designing for usability and try to be more preemptive when designing a
On 2/11/2011 5:10 AM, Anton Aylward wrote: processing, not put it in as an afterthought. Allows for reusable components that way. Yeah there may be a performance hit, but with today's superfast processors, it usually is not a big factor. The main question I focus on is what in the heck is a user going to do should an error occur, and how am I, as a developer, going to guide him/her to a solution, or teach him/her why what he/she was trying to do, is wrong. Focusing on error handling design up front usually means I don't have to deal with each and every return code explicitly as I write code. I wrap those kind of things with an OO pattern and focus on other issues. program to be a guide for a user. Rather than just throw the error out to the user, which often is just about as useful as saying "I can't do that, and I can't help you Mr. User. Good Luck!", give him/her some options, or a simple explanation immediately and guide him/her. You may not get it right the first time you throw out your application for users to use, but if you have provided him/her with easy feedback capabilities, your application can grow and adjust quickly to handle the unforeseen events. With a good robust error handling mechanism in place, you probably will be lucky and be able to repair the damage much easier. If it was a learning issue, that is an opportunity for you to enhance your program as a teacher, so others won't fall into the same trap. Granted you have to gage, as a developer, what learning/functional problems most of your users are having, and focus on those first, no one is expecting an application to get it right the first time, but just having the ability to let your users provide you with feedback easily, WILL advance your program far faster than the current methods of using external documentation and difficult to find bug trackers, forums etc., allow. You can do all this with CLI also, but it requires some careful thinking about how you will achieve most of these usability goals. You have less tools to use to interface with your users, but it probably is doable in most situations where CLI makes sense. .
Of course you could have monitored the various files in /var/log Of course I look in /var/log! Don't insult my intelligence. I found nothing helpful, why do you think I came onto this forum to ask for help in the first place?
Marc... -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
Marc Chamberlin said the following on 02/11/2011 02:09 PM:
Instead there are mostly only complaints when something does not work which does not really help the developers.
Oh YES, but complaints are extremely valuable, IF AND ONLY IF you take the time to LISTEN and understand where the problem really lies, and take the time to then correct it.
I suspect it was punishment for some grave and heinous sin I had committed in a previous life, but I once had a job on the phone support desk for a TLA company that - at that time - was a major name in the industry. My how the mighty TLA companies of old are fallen.
Most complaints are very likely to boil down to the fact that the user does not understand how to use a tool.
Dead to rights, Marc. Nineteen of of twenty calls were like that.
By correcting, I don't mean correct the user, but rather correct the tool.
And here's where you are wrong. The users really were losers. The joke about users taking the floppy out of its casing and sliding it in the gap between drives on the front fascia ... is not a joke. We used to whisper "I can hear him taking the shrink-wrap off the manuals". We had people who in the past used RSTS, the DOS-before-Microsoft or the IBM idea of a "dataset" and were confounded with the idea of a directory hierarchy or that they didn't have to create a file and preallocate its space before using it. We also had some people who had been using VAX/VMS where there are many different types of text file and you have to tell the editor to use the same kind that the compiler (or whatever) wants to see. The idea of a file as an array of bytes was was beyond them. (And I can imagine UNIX users having the converse problem when it comes to using VMS!) Were they wrong? Yes, they were wrong in the same sense that Brits visiting France and driving on the left are wrong. Or more. France and England have many more differences in road rules and etiquette.
Answers, such as what Anton gave me - RTFM - is so wrong on so many levels. Anton complained that I was beating my head against a wall because I was unwilling to change my approach to a problem I was trying to solve.
Where are you writing from, Marc? USA? Continental Europe? Visit England and try getting by if you are unwilling to drive on the left. Its more than your head that you'll be banging into. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
Marc Chamberlin said the following on 02/11/2011 02:09 PM:
Where are the "Linus Torvold's" who set the standards for the look and feel of Linux and it's tool sets? I know Linus has been one of the main gatekeepers for the kernel itself, but what about the rest of the system? Who make's the decision about what tools get included in a distribution's packages? At some level this has to be happening, and this is where a "standard" of acceptance should be developed, and tools judged on how well they meet those standards. Then roadmaps can be laid out for volunteers to pick and choose what to work on, and tool development guided towards becoming more and more usable for more and more users...
All of which exists. Why not google for it? -- Agnosticism simply means that a man shall not say he knows or believes that for which he has no grounds for professing to believe. Thomas H. Huxley -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
Marc Chamberlin said the following on 02/11/2011 02:43 PM:
As for using a "programmer's" text editor v.s a word processor? If a tool purports to handle a plain text file, I expect that tool to honor the syntax conventions associate with plain text files. "Purports to handle" could mean anything. Word processors can READ IN plain text files. That could be shoe-horned into "handle", but the resulting file when the word processor has finished with it is a word processor file.
It is, after all, a word processing tool.
There are plenty of tools for editing pain text files. Some "programmers editors" for pain text files like Kate and GVIM have facilities for "syntax directed editing" and can make smart guesses at the 'language' in the file: C/C++, perl, php, as well as various types of config file.
Kate also has a terminal emulator built in, so you could view it as a development environment. But I am not editing a file that requires syntax directed editing. Why is Kate and GVIM relevant? If I had chosen them instead, to use as a text editor because they claim they can do so, I would expect exactly
On 2/11/2011 12:37 PM, Anton Aylward wrote: the same behavior as if I had chosen any other tool that claims to be able to edit text files and write said files back out as a plain text file.
I see nothing wrong with using any tool that will edit text files, conceptually, as a simple text file. Indeed, but I'd rather have one that will edit text files ACTUALLY.
By not displaying or warning about non-printable characters, other than CR or LF, kwrite has a bug in it. That makes no sense. Kwrite is a WORD PROCESSOR. It writes out word processing files. It is not a text editor any more than Quanta+, Inkscape or GIMP is a text editor. All those can manipulate text, but what they write out isn't a text file
Funny that the menu item for KDE/openSuSE applications labels Kwrite as a Text Editor! I am suppose to have groked this distinction how? Oh I see, guess I am suppose to go read another man page first... Hmmm funny, no man pages for Kwrite! Aww guess what! KWrite has a Help button, that leads me to a built in document!!! Oh now this is interesting - Right at the very top of this help document it says - "KWrite is a text editor for KDE allowing you to edit one file at the time per window." What part of "text editor" am I suppose to have interpreted as - not really! KWrite is really a "word processor"?
Because I had approached these configuration files via Dolphin, kwrite was handy, so I used it instead. It bit me, and I learned of it's limitation. Won't make that mistake again... This is where Patrick would step in and say that your mistake was using a GUI -- Dolphin -- and let it make decisions for you. I would add that GUIs in general let other people make decisions for you and limit what you can see and do to what they think you should be doing and seeing. You have illustrated this point exceedingly well.
Dolphin offered me a choice, it did not force me to chose KWrite, but it did make it a little easier... That is called guidance...
I also pointed out, that there are a number of issues with various different tools, that aligned to make the process of setting up an nfs server and an autofs client extremely difficult. You might want to focus on those instead of attacking me. You might want to use the right tools and in the right way and gain an understanding of how the tools and facilities work and interact BEFORE hacking away. I kept saying "It works for me" and looking back I'm of the opinion that it worked without problems such as you encountered because I chose the right tools and read up and planned
I was taught "The seven Ps", but it seems one of the Ps is considered a rude word by many Nanny Filters, so its "The Six Ps" you'll find discussed most on the 'Net http://jbdcolley.com/?p=184
Marc, I see from other people and some out-of-band that you are coming across as attacking Linux, previous version of UNIX, their designers and developers and in many ways the OpenSource movement. This is not appreciated. There's a saying by the old Roman poet Horace that sums things up well:
If a better system is thine, impart it freely; If not, make use of mine. Anton, I grow tired of fencing with you. You are so blinded by your
Yeah!? I guess I had better go off to some mountain top somewhere, spend a few years groking ALL the documentation, and ALL the tools in Linux, before I use ANY of them, so I will know up front what ALL the possibilities are, and ALL of the interactions between ALL of the tools. HUMP!! prejudices against GUI's that you cannot see their potential. You are so blinded by your love of Linux that you cannot see that there is room for improvement. Your comments such as "It works for me" or "RTFM" shows that you are incapable of learning from others and have no tolerance for newcomers and people struggling to learn. Instead you launch personal attacks against me repeatably, which is called bullying. NO I am NOT a Linux hater in any way shape or form, but you want to establish and live by such doctrines of "It's my way or the highway.", or "Either you are with Linux, or you are against Linux." Such elitist attitudes do no allow you to see the middle grounds, the gray areas, just black and white simplistic answers. You are incapable of being a good teacher Anton, or a guide, and I suggest you should not try. You lack empathy for those who are struggling, a necessary quality of a good leader, and instead mock, and laugh at them. I thank you for trying though, you did help although it has turned into a rocky road that we traveled. I pointed out some faults of Linux, and you then choose to label me antagonistic against Linux. Not as someone who is trying to be helpful by sharing the difficulties he was having, offering suggestions on how to make improvements, and hoping to open a pathway towards those improvements. In all of our diatribes, you have not once offered suggestions of your own, on how Linux might be improved, for newcomers. Instead you complain, attack and imply that you wish beginners would go away and leave Linux to the elite few gurus who like it the way it is.. I suspect you will continue you attacks against me for awhile. I will reply no further to you in this thread. This has gone far enough. Regards... Marc Chamberlin -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On Fri, 2011-02-11 at 15:37 -0500, Anton Aylward wrote:
That makes no sense. Kwrite is a WORD PROCESSOR.
I would actually have to beg to differ there. kword is a word processor, but kwrite should work passably as a text editor. I've used it to edit configuration files in the past and it doesn't have any word processing features worth mentioning. Certainly no strange file format. Showing unprintable characters is not a feature of all text editors. Anders -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
Putting all the head butting aside, I would have to say Marc is correct in one fundamental point: why isn't the help environment more integrated into the application? As an application developer, and thus as a user support participant, I know very well that most problems stem from lack of information to the user more than poor program design. (My education is in cognitive psychology and interface design - so don't assume ignorance in this claim.) Not to say that poor program design is not also a problem. But really, no one sits down and says "How can I freak out the most users? What will elicit the most hostile reactions from the users? To be blunt, how can I be the biggest pain in the ass?" The developers are, in the total absence of a detailed specification from the user (read that a few times - it is a core issue here), surely doing their best. Given that they are often working sans enumeration, how on earth can you complain? But complain we do (drat, I am on that list of complainers when as a developer I should know better...). Instead of bitching at the programmers, perhaps we should address the more general question of how we could improve the situation for everyone. Developers are not evil bastards bent on making a bad experience for the user. I warrant that they are in fact very interested in quite the opposite. When users complain I suspect the developers are unhappy. What is missing is an infrastructure that developers can include in their applications that facilitates a moderated dialog between them and the users. Of course individual apps have done this. But a 'libc' that facilitates user/developer interaction in a constructive and commonly available way simply does not exist. And as long as there is no easy avenue for dialog, this complaint will persist. So instead of bitching at the obviously evil developers with hearts of pure evil, perhaps a more constructive approach might be to see how the developers can best keep abreast of the undefined and surely constantly shifting demands of the users. So, users, what do you suggest? And here is the rub: suggest something concrete and that puts some of the responsibility on you to invent a way to clearly define what some widget should do. After all, to complain that some unpaid coder did not implement some undefined (by the user) functionality is lame at best. -- Roger Oberholtzer OPQ Systems / Ramböll RST Ramböll Sverige AB Krukmakargatan 21 P.O. Box 17009 SE-104 62 Stockholm, Sweden Office: Int +46 10-615 60 20 Mobile: Int +46 70-815 1696 SHAW'S PRINCIPAL Build a system that even a fool can use, and only a fool will want to use it. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
Hello, On Feb 11 11:09 Marc Chamberlin wrote (excerpt):
Johannes - Your thoughts are very welcomed, but I am in strong disagreement with you also.
I think actually there is not much disagreement.
Your complaint about the lack of getting feedback and help from newcomers again boils down to how the tool and its support subsystem was designed.
A longer time ago we had a feedback system for Suse Linux but it was shut down - I don't know the reason but I guess it was because the company which made Suse Linux could no longer justify to pay those who evaluate the tons of incomming feedback. If there was unlimited resouces...
... complaints are extremely valuable, IF AND ONLY IF you take the time to LISTEN and understand where the problem really lies, and take the time to then correct it.
I absolutely agree. But note your wording "take the time". If there was unlimited resouces...
Answers, such as ... - RTFM - is so wrong on so many levels.
I absolutely agree. If a normal end-user must read documentation for a tool which is meant to be used by a normal end-user, there is something wrong. Of course a perfectly designed tool which is meant to be used by a normal end-user would not need documentation how to use it. But documentation is a relatively cheap way to somehow inform the user how a tool should be used. If there was unlimited resouces...
... it is the developers who are beating their heads against a wall, insisting that users RTFMs, designing tools in a now 30 year old fashion, and not being willing to LISTEN to users and change their approach in software design and development.
Here I disagree. In general developers follow users. I think very most developers are willing to listen to their users and are listening to their users as far as possible for them. It would not make any sense if a free software developer would insist not to change something in his software when the change makes sense for the developer. Often free software developers are experts in a certain area or at least experienced users of certain tools who make the tools first and foremost according to their own needs. Then there could be a difference in what changes other users like to have and what changes make sense for the developer. Other users cannot command free software developers. Only if a user would pay his developers he could command them.
Successful software development is going to be judged in terms of usability. Functionality is 10% of the battle, and currently gets 90% of many if not most software developer's focus. That focus HAS to change to be closer to 50/50.
From time to time I use an older Linux system and when I compare
I absolutely agree. But I don't see how such a change could happen currently. Currently a Linux distribution is mainly the result of what free software developers implement according to their own needs. Unless many usability experts join free software development or unless there is much more payment for usability experts and non-free developers who implement for money what others want to get from them, the free software must be accepted "as is". I think because very most developers listen to their users there is progress over the time - only some patience is needed. this stuff with what we have nowadays I see a lot of progress. Regards Johannes Meixner -- SUSE LINUX Products GmbH, Maxfeldstrasse 5, 90409 Nuernberg, Germany AG Nuernberg, HRB 16746, GF: Markus Rex -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On 02/11/2011 07:36 PM, Roger Oberholtzer wrote:
Putting all the head butting aside, I would have to say Marc is correct in one fundamental point: why isn't the help environment more integrated into the application?
As an application developer, and thus as a user support participant, I know very well that most problems stem from lack of information to the user more than poor program design. (My education is in cognitive psychology and interface design - so don't assume ignorance in this claim.) Not to say that poor program design is not also a problem. But really, no one sits down and says "How can I freak out the most users? What will elicit the most hostile reactions from the users? To be blunt, how can I be the biggest pain in the ass?"
The developers are, in the total absence of a detailed specification from the user (read that a few times - it is a core issue here), surely doing their best. Given that they are often working sans enumeration, how on earth can you complain? ^^^^^^^^^^^ remuneration But complain we do (drat, I am on that list of complainers when as a developer I should know better...). Instead of bitching at the programmers, perhaps we should address the more general question of how we could improve the situation for everyone. Developers are not evil bastards bent on making a bad experience for the user. I warrant that they are in fact very interested in quite the opposite. When users complain I suspect the developers are unhappy.
What is missing is an infrastructure that developers can include in their applications that facilitates a moderated dialog between them and the users. Of course individual apps have done this. But a 'libc' that facilitates user/developer interaction in a constructive and commonly available way simply does not exist. And as long as there is no easy avenue for dialog, this complaint will persist.
So instead of bitching at the obviously evil developers with hearts of pure evil, perhaps a more constructive approach might be to see how the developers can best keep abreast of the undefined and surely constantly shifting demands of the users. So, users, what do you suggest? And here is the rub: suggest something concrete and that puts some of the responsibility on you to invent a way to clearly define what some widget should do. After all, to complain that some unpaid coder did not implement some undefined (by the user) functionality is lame at best.
You have make some very interesting points, and I don't believe any of the readers of "help" think of the coders as evil, rather that they are uninformed as to what the readers need. What we need most is a feedback mechanism, so that the reader can _easily_ get back to the developer with the question he needs answered. This is not to say that the coder should have to answer each reader individually--that's what these mailing lists tend to do, usually pretty well--but that periodically the "help" line should get updated, just like the code gets. I never saw such absolutely useless "help" as what Microsoft provides. In contrast, Linux usually provides _no_ help at all. When you snap on the "help" title, you get nothing. There must be a middle ground somewhere. (Exception: Altho not written as "help" titles on programs, Ubuntu has some excellent instructional material on the Internet. None of the other distros [to my knowledge] have anything like that.) Just my 2¢. --doug -- Blessed are the peacemakers...for they shall be shot at from both sides. --A. M. Greeley -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
participants (8)
-
Anders Johansson
-
Anton Aylward
-
Doug
-
Johannes Meixner
-
lynn
-
Marc Chamberlin
-
Patrick Shanahan
-
Roger Oberholtzer