[ https://issues.apache.org/jira/browse/HDFS-14377?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16796309#comment-16796309 ]
Anu Engineer edited comment on HDFS-14377 at 3/19/19 5:32 PM: -------------------------------------------------------------- [~xkrogen]/ [~dannytbecker], Thanks for the links and the patch. I am ambivalent about this change. I debated whether I should reply and then I remembered that +friends don't let friends drive drunk.+ It is also quite obvious from this reply that I have nothing of real value to do on a Tuesday morning :). *First of all, I am a friend of Danny. He may not know me, but trust me on this.* >From my experiences, this is how this patch is going to evolve. Someone will look at this(at us assume it is me for time being) and will comment {noformat} Anu: Danny, Thanks for the patch. This looks Awesome, I have always wanted Hadoop to be ISO compliant. +1, Pending Jenkins. <A Few minutes later, another passerby committer, since we are friends, let us say Erik, comes by and says> Erik: Danny, this is so awesome, while you are at this, would you please fix the command line tool xyz, that still prints KB and MB instead of KiB and MiB. Danny being good-natured and nice (remember he is my friend) decides to make changes to the tool xyz and puts up patch v2. <In this particular story, the Evil Villian is also me, so let us call the second person Engineer for fun, it is an apt moniker in this story> Engineer: Btw, Danny this breaks the app-compact rules. You *cannot* make this change. This can *only* be done in a major revision, oh, btw, if you are planning to make this change only in the UI, that is now inconsistent. Some places of the HDFS speaks ISO and parts speak "custom" memory Units. That is a mess. I really think we should do this in the next release. <Few minutes later, Anu comes in > Anu: oh...I did not think of that, you have a point. @Erik, what do you think? Erik: Oh.. that is right, let us open a new branch, Hadoop 4.0 to commit this. Engineer: Perhaps we should start a discussion thread in the mailing list to open a new branch? Anu: good, Idea, let me fire off a thread. [100 reply thread ensues -- and finally, after 4 months, a new branch is opened, and Danny is sitting there wondering what the s$*t did I do? it was a one-line patch] Anu: @Danny, Could you please rebase this patch, and btw, we found three other tools that need the fix. Can you please take care of that while you are at this? Gandalf(some wise-committer in Hadoop): This change impacts Hadoop Common, that means this change also impacts, Spark, Sqoop, YARN, HBase and Kitchen Sink. You cannot make this change without considering the downstream impact. <At this point, Danny just flies away , Anu Open the next bottle of beer, Erik is just sad ...> {noformat} Danny being my friend and a new contributor, I was just trying to be nice and helpful and steering him gently away from the mine field was about to step in. I did not even start the discussion on whether we should be a "traditionalist" vs. "modernist" and use "KB" vs. "KiB". I am sure some other committer will add that perspective. Given all this,* I am +1 on this change*, I hope my parody of our lives will motivate us to stay away from a long discussion on the merits of this one line patch. Yesterday night, I was truly in a good mood, having just seen how humanity saves dragons and was generally feeling good and charitable. In that moment of weakness, I decided to be kind of Danny and save him some pain. Danny, I hope you see the wisdom in being my friend and hopefully you will be nice enough to buy me some beer when we finally meet. Ps. Truly, I have nothing better to do :( that is a sad state of my life :( I need to find something better to do than comment on random JIRAs. was (Author: anu): [~xkrogen]/ [~dannytbecker], Thanks for the links and the patch. I am ambivalent about this change. I debated whether I should reply and then I remembered that +friends don't let friends drive drunk.+ It is also quite obvious from this reply that I have nothing of real value to do on a Tuesday morning :). *First of all, I am a friend of Danny. He may not know me, but trust me on this.* >From my experiences, this is how this patch is going to evolve. Someone will look at this(at us assume it is me for time being) and will comment {noformat} Anu: Danny, Thanks for the patch. This looks Awesome, I have always wanted Hadoop to be ISO compliant. +1, Pending Jenkins. <A Few minutes later, another passerby committer, since we are friends, let us say Erik, comes by and says> Erik: Danny, this is so awesome, while you are at this, would you please fix the command line tool xyz, that still prints KB and MB instead of KiB and MiB. Danny being good-natured and nice (remember he is my friend) decides to make changes to the tool xyz and puts up patch v2. <In this particular story, the Evil Villian is also me, so let us call the second person Engineer for fun, it is an apt moniker in this story> Engineer: Btw, Danny this breaks the apt-compact rules. You *cannot* make this change. This can *only* be done in a major revision, oh, btw, if you are planning to make this change only in the UI, that is now inconsistent. Some places of the HDFS speaks ISO and parts speak "custom" memory Units. That is a mess. I really think we should do this in the next release. <Few minutes later, Anu comes in > Anu: oh...I did not think of that, you have a point. @Erik, what do you think? Erik: Oh.. that is right, let us open a new branch, Hadoop 4.0 to commit this. Engineer: Perhaps we should start a discussion thread in the mailing list to open a new branch? Anu: good, Idea, let me fire off a thread. [100 reply thread ensues -- and finally, after 4 months, a new branch is opened, and Danny is sitting there wondering what the s$*t did I do? it was a one-line patch] Anu: @Danny, Could you please rebase this patch, and btw, we found three other tools that need the fix. Can you please take care of that while you are at this? Gandalf(some wise-committer in Hadoop): This change impacts Hadoop Common, that means this change also impacts, Spark, Sqoop, YARN, HBase and Kitchen Sink. You cannot make this change without considering the downstream impact. <At this point, Danny just flies away , Anu Open the next bottle of beer, Erik is just sad ...> {noformat} Danny being my friend and a new contributor, I was just trying to be nice and helpful and steering him gently away from the mine field was about to step in. I did not even start the discussion on whether we should be a "traditionalist" vs. "modernist" and use "KB" vs. "KiB". I am sure some other committer will add that perspective. Given all this,* I am +1 on this change*, I hope my parody of our lives will motivate us to stay away from a long discussion on the merits of this one line patch. Yesterday night, I was truly in a good mood, having just seen how humanity saves dragons and was generally feeling good and charitable. In that moment of weakness, I decided to be kind of Danny and save him some pain. Danny, I hope you see the wisdom in being my friend and hopefully you will be nice enough to buy me some beer when we finally meet. Ps. Truly, I have nothing better to do :( that is a sad state of my life :( I need to find something better to do than comment on random JIRAs. > Incorrect unit abbreviations shown for fmt_bytes > ------------------------------------------------ > > Key: HDFS-14377 > URL: https://issues.apache.org/jira/browse/HDFS-14377 > Project: Hadoop HDFS > Issue Type: Improvement > Reporter: Danny Becker > Assignee: Danny Becker > Priority: Trivial > Attachments: HDFS-14377.000.patch > > > The function fmt_bytes show the abbreviations for Terabyte, Petabyte, etc. > the standard metric system units for data storage units. The function however > divides by a factor of 1024, which is the factor used for Pebibyte, Tebibyte, > etc. Change the abbreviations from TB, PB, etc to TiB, PiB, etc. -- This message was sent by Atlassian JIRA (v7.6.3#76005) --------------------------------------------------------------------- To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org