3 ms·
The article convinces me on the failure of the ZIP method in at least 3 respects: 1) For efficient incremental updates. Especially when you're saving a large f
by knighthack 6y ago
The article convinces me on the failure of the ZIP method in at least 3 respects:
1) For efficient incremental updates. Especially when you're saving a large file often, making constant single changes to ZIP files is just inefficient.
2) The insane memory gulps that ZIP takes.
3) The pile of files method. The data in OpenDocument files could be much better sorted/accessed through SQLite tables, rather than through XML files in ZIPs.
I think SQLite makes far more sense for OpenDocument. Incidentally, SQLite has also made good arguments previously for its use to replace application file formats: https://www.sqlite.org/appfileformat.html https://www.sqlite.org/appfileformat.html