You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Originally posted by benajamin April 15, 2024
One of the difficulties new adopters of HDF5 come across is that there are many ways to write strings (fixed vs variable length, null terminated or not). Many tools don't clearly tell what they are reading or writing, and some programs will crash or malfunction if they don't get an expected string type. The hdfview utility is the one that I find most useful for debugging string issues and it would be great if the h5web viewer could also provide this detailed string info in the "inspect" mode and therefore help debug HDF5 string bugs.
Originally posted by benajamin April 20, 2024
The string padding info would be great to have but can be hidden behind the Inspect accordion since it is probably only interesting for debugging. I don't think that there is a definite right or wrong type of padding to use, I have just seen incompatibilities between programs reading and writing HDF5 that caused bugs that were difficult to diagnose because the critical padding information was not easy to expose.
The text was updated successfully, but these errors were encountered:
Discussed in #1612
Originally posted by benajamin April 15, 2024
One of the difficulties new adopters of HDF5 come across is that there are many ways to write strings (fixed vs variable length, null terminated or not). Many tools don't clearly tell what they are reading or writing, and some programs will crash or malfunction if they don't get an expected string type. The hdfview utility is the one that I find most useful for debugging string issues and it would be great if the h5web viewer could also provide this detailed string info in the "inspect" mode and therefore help debug HDF5 string bugs.
Originally posted by benajamin April 20, 2024
The string padding info would be great to have but can be hidden behind the Inspect accordion since it is probably only interesting for debugging. I don't think that there is a definite right or wrong type of padding to use, I have just seen incompatibilities between programs reading and writing HDF5 that caused bugs that were difficult to diagnose because the critical padding information was not easy to expose.
The text was updated successfully, but these errors were encountered: