|
The LD data also has a key in the root dictionary, otherwise the files are identical. For example, you have a GB file that is served twice, one of which also contains approximately KB of metadata. Therefore, smaller files are completely useless. On the one hand, it is a strange situation, on the other hand, it is an unexpected surprise. The same vendor needlessly released the vast majority of its data last time we looked at its work, and now only half of it is released.
Author: Screenshot, Ondřej Kokeš So we already half of the dataset, so we Benin Mobile Number List download the entire JSON-LD dataset. The high compression ratio might surprise you at first glance (approximately: , i.e. the largest file is about GB compressed but over GB decompressed), but this is due to the very verbose JSON format. Normally this doesn't matter, it's common to process data iteratively, but solution providers make this option more difficult by choosing a format.
Rather than choosing a format that allows stream processing, the vendor chose a structure that encourages loading the entire file into memory (few implementations allow sequential processing of this structure), which is very difficult for GB files on common hardware . This may sound cliché, but it is a major drawback of open data in electronic collections. All other limitations can be solved in some way, but if the average user encounters such a huge amount of data, it will be very difficult to process it in any way.
|
|