Hi,
I reported this bug a long time ago and it got fixed but it seems to have come back again.
When you have a lot of results and start scrolling downwards through the results in PMA 4.4.11 the column headers shift out of position. I have included two images, one showing the columns in the right location BEFORE scrolling, another showing the positions after scrolling. The after position can cause an end user confusion if they don't realise the column headers have moved.
Please fix!
The old BUG REPORT was in #4538
I can reproduce by setting $cfg['RepeatCells'] = 10; but what I don't understand in your screenshot is that the second row of headers is not formatted the same way as the first row. When testing with Chrome, my second row of headers has all the headers in blue and they are clickable.
Hi Marc,
The second row of "headers" is data, not headers, i put them in to show where the headers should be in relation to the data.
Dean
Hi Dean,
Ok, I believed it was a case of incorrect repeated headers. Could you attach a small export of your table's structure and your sample data where you inserted the text of headers?
Hi Marc,
I am a little reticent to export the data as it is confidential customer data, however i have noticed that the problem seems to be evident in tables where there is no ability to perform INLINE edit, IE no primary or unique key exists in the table. To test this i have copied a table that does not exhibit this column shifting issue into another table, I tested the new table and it doesn't shift columns when i scroll, if i remove the primary and unique keys out of the new table and no longer have the ability to inline edit and start scrolling through the same table, the columns shift as shown in the attached images.
It looks like a calculation issue when no primary key / inline edit ability exists.
I have attached two images to show what i mean (excuse my crudely drawn data to header lines).
Does that help?
Dean
Last edit: Dean 2015-07-15
Dean,
I only asked for an export containing the table's structure and sample fake data (where you duplicated the headers), no confidential data. This would help most.
I confirm this observation
HI Marc,
give this a shot and let me know if it works (i am using firefox).
I don't seem to be able to reply to this, however this problem now looks fixed in the latest version, 4.4.12.
Dean
Last edit: Dean 2015-08-03