Talk:Custom track database: Difference between revisions

From genomewiki
Jump to navigationJump to search
No edit summary
 
 
(4 intermediate revisions by the same user not shown)
Line 2: Line 2:


<ul>
<ul>
<li>2006-05-12 Have basic system in place.  Can load up bed types 3 through 12 into database and hgTracks can display them.  Existing system remains in place.  Have fixed the itemRgb business when more than 9 columns.  hgLoadBed needs to make a temp file with a name of something other than bed.tab since more than one instance may be running at the same time.  Need to figure out how to delete db tables older than N hours.  Looking at hgTables code, doesn't appear to be too much confusion in there.  The loading appears to be isolated to routines in custom.c</li>
<LI>2006-05-30 Now have wiggle track types functioning, hgc click through and table browser functioning.  Need to begin thorough testing.</LI>
<li>2006-05-24 Remember to get back to the offset lift for regular bed files as they go by since right now they are being sent on with no processing and thus are missing the lift.  Maybe that could be put into hgLoadBed ...  (This was taken care of, they are lifted as they are sent into the loader pipeline)</li>
<li>2006-05-23 Beginning to work with the other types beside bed.  The GFFs are unusual in that they must be read in completely as a gff structure before they can be converted to bed.  Thus their output accumulates in memory until the very end, and then they are output all at once.  Created the dbTrash command to do the cleaning of the customTrash database.</LI>
<li>2006-05-12 Have basic system in place.  Can load up bed types 3 through 12 into database and hgTracks can display them.  Existing system remains in place.  Have fixed the itemRgb business when more than 9 columns.  hgLoadBed needs to make a temp file with a name of something other than bed.tab since more than one instance may be running at the same time.  Need to figure out how to delete db tables older than N hours.  Looking at hgTables code, doesn't appear to be too much confusion in there.  The loading appears to be isolated to routines in custom.c,  In fact it appears to be trivial and is in a single location.  Function has been added to the table browser.</li>
</ul>
</ul>

Latest revision as of 22:54, 30 May 2006

Progress Report

  • 2006-05-30 Now have wiggle track types functioning, hgc click through and table browser functioning. Need to begin thorough testing.
  • 2006-05-24 Remember to get back to the offset lift for regular bed files as they go by since right now they are being sent on with no processing and thus are missing the lift. Maybe that could be put into hgLoadBed ... (This was taken care of, they are lifted as they are sent into the loader pipeline)
  • 2006-05-23 Beginning to work with the other types beside bed. The GFFs are unusual in that they must be read in completely as a gff structure before they can be converted to bed. Thus their output accumulates in memory until the very end, and then they are output all at once. Created the dbTrash command to do the cleaning of the customTrash database.
  • 2006-05-12 Have basic system in place. Can load up bed types 3 through 12 into database and hgTracks can display them. Existing system remains in place. Have fixed the itemRgb business when more than 9 columns. hgLoadBed needs to make a temp file with a name of something other than bed.tab since more than one instance may be running at the same time. Need to figure out how to delete db tables older than N hours. Looking at hgTables code, doesn't appear to be too much confusion in there. The loading appears to be isolated to routines in custom.c, In fact it appears to be trivial and is in a single location. Function has been added to the table browser.