Hello, I submitted an update of numlock for mga 1; it should fix numlock under X (bug 1136). Packages list: --------------- x86_64 : numlock-2.1.2-5.1.mga1.x86_64.rpm i586 : numlock-2.1.2-5.1.mga1.i586.rpm How to test: ------------ on a system with numpad !! BEFORE with numlock-2.1.2-5.mga1, enable_X11_numlock doesn't work correctly; if it is called many times, enable_X11_numlock switches the numlock (on, then off, then on, ...) instead of turns it on. So that numlock doesn't work with KDE out of the box. AFTER with numlock-2.1.2-5.1.mga1, numlockx replaces enable_X11_numlock. If you launch many times numlockx, numlock should stay on. Numlock should work correctly with KDE. Suggested advisory: ------------------- Numlock doesn't work correctly under X on Mageia 1. This update fix this issue. regards, Luc
I see no update for Mageia 2. Does it mean that numlock is fixed already in Mageia 2? I can't reproduce the problem with numlock on my system, but I can test the update candidate to see there is no regression. I have one problem with that update candidate: what if users or programs call enable_X11_numlock in their scripts? Won't the update break those scripts because of the script rename? I think a symlink should be provided for backward compatibility, do you agree? Apart from that, the update candidate seems to work on Mageia 1 32 bits.
CC: (none) => stormi
(In reply to comment #1) > I see no update for Mageia 2. Does it mean that numlock is fixed already in > Mageia 2? yes, mga 2 already uses numlockx (since 2012-01-03). https://bugs.mageia.org/show_bug.cgi?id=1136#c15 http://svnweb.mageia.org/packages?view=revision&revision=189843 > > I can't reproduce the problem with numlock on my system, but I can test the > update candidate to see there is no regression. > > I have one problem with that update candidate: what if users or programs call > enable_X11_numlock in their scripts? Won't the update break those scripts > because of the script rename? I think a symlink should be provided for backward > compatibility, do you agree? not necessary, mga 2 uses numlockx, like most others distros.
(In reply to comment #2) > (In reply to comment #1) > > I have one problem with that update candidate: what if users or programs call > > enable_X11_numlock in their scripts? Won't the update break those scripts > > because of the script rename? I think a symlink should be provided for backward > > compatibility, do you agree? > > not necessary, mga 2 uses numlockx, like most others distros. Ok for mga2, but this update candidate is for mga1, and an update must usually not change the way a tool is used (ie, IINM, no binary name change without providing a symlink for compatibility). As it is, this update could potentially break something for users. A symlink would prevent that.
numlockx uses the same default (on, to turn numlock on in X) but doesn't have exactly the same syntax than enable_X11_numlock to toggle the numlock on and off: numlockx uses toggle, enable_X11_numlock uses switch, so I won't add a symlink. numlockx --help Usage : numlockx [on|off] on - turns NumLock on in X ( default ) off - turns NumLock off in X toggle - toggles the NumLock on and off in X enable_X11_numlock --help Usage : enable_X11_numlock [on|off|switch] on - sets NumLock on in X (default) off - sets NumLock off in X switch - changes NumLock state in X if this update is not valid, sorry, but personally I won't spend and lose more time on this bug only for mga 1.
(In reply to comment #4) > if this update is not valid, sorry, but personally I won't spend and lose more > time on this bug only for mga 1. I understand "spend" but not "lose" here. Since when is fixing bugs for supported Mageia releases a loss of time? However, I haven't said that the update is not valid. I tried to see if we could do it better. What could help decide if the update is valid or not is some insight about the chance there is to break something for users with the replacement of enable_X11_numlock with numlockx. I personnally don't know.
(In reply to comment #5) > (In reply to comment #4) > > if this update is not valid, sorry, but personally I won't spend and lose more > > time on this bug only for mga 1. > > I understand "spend" but not "lose" here. Since when is fixing bugs for > supported Mageia releases a loss of time? ie I won't try to patch numlockx so that it has exactly the same syntax as enable_X11_numloack, it would be a loss of time. > > However, I haven't said that the update is not valid. I tried to see if we > could do it better. What could help decide if the update is valid or not is > some insight about the chance there is to break something for users with the > replacement of enable_X11_numlock with numlockx. I personnally don't know. IMHO the risks are limited, but it's difficult to evaluate the benefit/risk ratio. Perhaps most of affected users already use a workaround for bug 1136, and an update is useless. Personally I had this bug on all my systems with numpad, but I'm not sure that this bug concerns all KDE users, or users of others DE.
Let's push it, I think the risks to break something for a user are very limited indeed, and many users experienced the numlock bug (I can't reproduce it on this computer but I had it on my work computer before it got upgraded to Mageia 2) Testing complete on Mageia 1 32 and 64 bits. Update validated. No linking required. see comment #0 for packages and advisory.
Keywords: (none) => validated_updateCC: (none) => sysadmin-bugsWhiteboard: (none) => has_procedure MGA1-32-OK MGA1-64-OK
Update pushed: https://wiki.mageia.org/en/Support/Advisories/MGAA-2012-0141
Status: NEW => RESOLVEDCC: (none) => tmbResolution: (none) => FIXED