tag:blogger.com,1999:blog-2248117304334958898.post7434653879075338756..comments2022-03-09T06:39:11.501-08:00Comments on Thoughts of a technocrat as letter sequences: The bioinformatics sysadmin craftmanship: The MRS v5 platform: Part 1kompioyterashttp://www.blogger.com/profile/07993947898374939412noreply@blogger.comBlogger2125tag:blogger.com,1999:blog-2248117304334958898.post-79513411909077468512011-12-26T07:01:39.695-08:002011-12-26T07:01:39.695-08:00George, all these are valid points and questions. ...George, all these are valid points and questions. I shall address them in part II, which explains the installation.<br /><br />Cheers,<br />GMkompioyterashttps://www.blogger.com/profile/07993947898374939412noreply@blogger.comtag:blogger.com,1999:blog-2248117304334958898.post-7267023438433476032011-12-26T06:00:08.194-08:002011-12-26T06:00:08.194-08:00here are a couple of observations for the .deb pac...here are a couple of observations for the .deb package of MRS v5 : <br /><br />1) ubuntu-specific dependancies in the .deb package: perlapi-5.10.1-ubuntu? ಠ_ಠ<br /><br />2) what is so special about 5.10 ? why can't we use 5.12? <br /><br />3) the /etc/init.d/mrs-ws script distributed with the .deb package does not a) check to see which user started it 2) use the start-stop-daemon to become the mrs-ws user. inst<br /><br />5) /var/lib/mrs : ug+rwx g=rx , even if you chmod the damn directory to u=rxw go=rx . which means there is a cronjob somewhere which makes sure it is 775. why on earth does the group need writing priviledges?<br /><br />also the author has not really specified if it is a "best practice" (bliah :P ) for people to log in as the mrs-ws user in order to do developmet. i got a couple of people who demand direct access to the shell in order to develop applications. and as the helpful guy i am , i provide the access for them...George Marselishttps://www.blogger.com/profile/05011444265945817497noreply@blogger.com