[Amsn-devel] workarounds for bugs in used components
Hi, Often when we find a bug in some code we use, we work around it in amsn code, or have a local patches copy of the other project. (2 latest example are commits 8758 (DEMANDS_ATTENTION) and 8759 (snit tcl 8.5a6 ). It is absolutely very good that we do this, and we should continue doing so. However, these fixes are not necessary (anymore) when they are fixed by upstream projects. So I thought it might be good to have a page in the wiki for all workarounds we do, and the corresponding upstream bugreport. This way, if by any chance there is a regression because of a workaround, instead of working around the workaround we could check the upstream status, and if it is fixed there we could revert the workaround (if it would be easy for users to update the upstream code), or at least we know a bit what's going on. In the longer term this will lead to less workarounds, and less bugs in upstream software. What do you guys think? Sander - This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/ ___ Amsn-devel mailing list Amsn-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/amsn-devel
Re: [Amsn-devel] remember me bugfix
no problem. actually a CreateProfile creates the new profile, writes the new config and reactivates the last profile. i think this behavior is still from the time before you introduced the new login window, but that isn't really useful when you try to login with the new profile after that because the last one is reactivated. what i did is creating the new profile and let this profile active. so the newHOME isn't needed anymore, because you won't load the last profile again, so i used HOME to save memory by not using another useless temporary variable. the password and the remember me flag are the only temporary variables i need to save, because they are reset by the call of configDefaults. after creating the new, empty profile, i write them back to the config, save config and free the memory of the two unused variables. alright? everything clear? i hope i explained it understandable ... Greetings, Mirko 2007/5/25, Youness Alaoui [EMAIL PROTECTED]: Hi Mirko, thanks for the patch, but could you explain exactly what happens in the patch? why each line of the diff is there, why rename newHOME into HOME, why move the 'set oldpassword' a few lines before, etc... otherwise, we can't understand this... Thanks! KKRT On Fri, May 25, 2007 at 06:54:20AM +0200, square87 wrote: Hello all. Maybe Mirko are talking about this reported problem: http://www.amsn-project.net/forums/viewtopic.php?t=3267highlight= Square87 - This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/ ___ Amsn-devel mailing list Amsn-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/amsn-devel - This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/ ___ Amsn-devel mailing list Amsn-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/amsn-devel - This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/___ Amsn-devel mailing list Amsn-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/amsn-devel
Re: [Amsn-devel] remember me bugfix
Commited in r8762 on trunk Branch will be synced later Phil Mirko Hansen a écrit : no problem. actually a CreateProfile creates the new profile, writes the new config and reactivates the last profile. i think this behavior is still from the time before you introduced the new login window, but that isn't really useful when you try to login with the new profile after that because the last one is reactivated. what i did is creating the new profile and let this profile active. so the newHOME isn't needed anymore, because you won't load the last profile again, so i used HOME to save memory by not using another useless temporary variable. the password and the remember me flag are the only temporary variables i need to save, because they are reset by the call of configDefaults. after creating the new, empty profile, i write them back to the config, save config and free the memory of the two unused variables. alright? everything clear? i hope i explained it understandable ... Greetings, Mirko 2007/5/25, Youness Alaoui [EMAIL PROTECTED] mailto:[EMAIL PROTECTED]: Hi Mirko, thanks for the patch, but could you explain exactly what happens in the patch? why each line of the diff is there, why rename newHOME into HOME, why move the 'set oldpassword' a few lines before, etc... otherwise, we can't understand this... Thanks! KKRT On Fri, May 25, 2007 at 06:54:20AM +0200, square87 wrote: Hello all. Maybe Mirko are talking about this reported problem: http://www.amsn-project.net/forums/viewtopic.php?t=3267highlight= http://www.amsn-project.net/forums/viewtopic.php?t=3267highlight= Square87 - This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/ ___ Amsn-devel mailing list Amsn-devel@lists.sourceforge.net mailto:Amsn-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/amsn-devel https://lists.sourceforge.net/lists/listinfo/amsn-devel - This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/ ___ Amsn-devel mailing list Amsn-devel@lists.sourceforge.net mailto:Amsn-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/amsn-devel - This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/ ___ Amsn-devel mailing list Amsn-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/amsn-devel - This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/ ___ Amsn-devel mailing list Amsn-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/amsn-devel
Re: [Amsn-devel] workarounds for bugs in used components
I do think you're right and that it would be very ueful, unfortunately, I'm also pretty sure that in the end, we won't be able to do it. We can't even keep a TODO list in the wiki, so think about keeping that kindd of thing documented! We could try if you're willing to drive this, I wouldn't mind, I just think you shouldn't have high hopes over the success of this 'adventure'. p.s.: DEMANDS_ATTENTION is not only a workaround, but also a feature enhancement, but it needs to be well tested otherwise it could be considered a bug. I wanted to talk about it on the ML but I'm just far too busy lately... sorry. KKRT On Fri, May 25, 2007 at 09:39:44AM +0200, Sander Hoentjen wrote: Hi, Often when we find a bug in some code we use, we work around it in amsn code, or have a local patches copy of the other project. (2 latest example are commits 8758 (DEMANDS_ATTENTION) and 8759 (snit tcl 8.5a6 ). It is absolutely very good that we do this, and we should continue doing so. However, these fixes are not necessary (anymore) when they are fixed by upstream projects. So I thought it might be good to have a page in the wiki for all workarounds we do, and the corresponding upstream bugreport. This way, if by any chance there is a regression because of a workaround, instead of working around the workaround we could check the upstream status, and if it is fixed there we could revert the workaround (if it would be easy for users to update the upstream code), or at least we know a bit what's going on. In the longer term this will lead to less workarounds, and less bugs in upstream software. What do you guys think? Sander - This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/ ___ Amsn-devel mailing list Amsn-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/amsn-devel - This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/ ___ Amsn-devel mailing list Amsn-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/amsn-devel
Re: [Amsn-devel] workarounds for bugs in used components
why don't use something like this: http://wiki.tcl.tk/5598 you can mark workarounds directly on code, and use this tool to extract the information :) On 5/25/07, Youness Alaoui [EMAIL PROTECTED] wrote: I do think you're right and that it would be very ueful, unfortunately, I'm also pretty sure that in the end, we won't be able to do it. We can't even keep a TODO list in the wiki, so think about keeping that kindd of thing documented! We could try if you're willing to drive this, I wouldn't mind, I just think you shouldn't have high hopes over the success of this 'adventure'. p.s.: DEMANDS_ATTENTION is not only a workaround, but also a feature enhancement, but it needs to be well tested otherwise it could be considered a bug. I wanted to talk about it on the ML but I'm just far too busy lately... sorry. KKRT On Fri, May 25, 2007 at 09:39:44AM +0200, Sander Hoentjen wrote: Hi, Often when we find a bug in some code we use, we work around it in amsn code, or have a local patches copy of the other project. (2 latest example are commits 8758 (DEMANDS_ATTENTION) and 8759 (snit tcl 8.5a6 ). It is absolutely very good that we do this, and we should continue doing so. However, these fixes are not necessary (anymore) when they are fixed by upstream projects. So I thought it might be good to have a page in the wiki for all workarounds we do, and the corresponding upstream bugreport. This way, if by any chance there is a regression because of a workaround, instead of working around the workaround we could check the upstream status, and if it is fixed there we could revert the workaround (if it would be easy for users to update the upstream code), or at least we know a bit what's going on. In the longer term this will lead to less workarounds, and less bugs in upstream software. What do you guys think? Sander - This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/ ___ Amsn-devel mailing list Amsn-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/amsn-devel - This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/ ___ Amsn-devel mailing list Amsn-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/amsn-devel -- Bruno Costa - This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/___ Amsn-devel mailing list Amsn-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/amsn-devel
