Use Longtext rather than varchar

[bmschwa]

Queries get really big... was trying to get some data from langfuse & their api describes modifying the query parameters depending on what you want. So in curl, I ended up with something like this to get an overview of AI usage:

curl -vv -s -G "https://us.cloud.langfuse.com/api/public/v2/metrics" -u "xxx:yyy" --data-urlencode 'query={"view":"observations","dimensions":[{"field":"providedModelName"}],"metrics":[{"measure":"count","aggregation":"sum"},{"measure":"totalCost","aggregation":"sum"},{"measure":"totalTokens","aggregation":"sum"},{"measure":"inputTokens","aggregation":"sum"},{"measure":"outputTokens","aggregation":"sum"},{"aggregation":"avg", "measure":"latency"},{"aggregation":"p50", "measure":"latency"}],"fromTimestamp":"2026-08-04T00:00:00Z","toTimestamp":"2026-09-03T23:59:59Z","orderBy":[{"field":"sum_totalCost","direction":"desc"}],"config":{"row_limit":50}}'

which works in curl but this monstrosity that needs to be kept in the database is:

https://us.cloud.langfuse.com/api/public/v2/metrics?query=%7b%22view%22%3a%22observations%22%2c%22dimensions%22%3a%5b%7b%22field%22%3a%22providedModelName%22%7d%5d%2c%22metrics%22%3a%5b%7b%22measure%22%3a%22count%22%2c%22aggregation%22%3a%22sum%22%7d%2c%7b%22measure%22%3a%22totalCost%22%2c%22aggregation%22%3a%22sum%22%7d%2c%7b%22measure%22%3a%22totalTokens%22%2c%22aggregation%22%3a%22sum%22%7d%2c%7b%22measure%22%3a%22inputTokens%22%2c%22aggregation%22%3a%22sum%22%7d%2c%7b%22measure%22%3a%22outputTokens%22%2c%22aggregation%22%3a%22sum%22%7d%2c%7b%22aggregation%22%3a%22avg%22%2c+%22measure%22%3a%22latency%22%7d%2c%7b%22aggregation%22%3a%22p50%22%2c+%22measure%22%3a%22latency%22%7d%5d%2c%22fromTimestamp%22%3a%222026-08-04T00%3a00%3a00Z%22%2c%22toTimestamp%22%3a%222026-09-03T23%3a59%3a59Z%22%2c%22orderBy%22%3a%5b%7b%22field%22%3a%22sum_totalCost%22%2c%22direction%22%3a%22desc%22%7d%5d%2c%22config%22%3a%7b%22row_limit%22%3a50%7d%7d

[Rello]

nice suggestion. thanks