Foldr Cloud and appliances
The language is the same in both places. What differs is the limits you can change, the network a script may reach, and a few library objects. The Limits page has the numbers. This page lists the rest.
Limits and settings
Section titled “Limits and settings”| Foldr Cloud | Appliance | |
|---|---|---|
| Queued run (event, schedule, queue) | 60 seconds | 60 seconds |
| Editor Run, webhook, form lookup | 60 seconds | 300 seconds |
| Action-menu entry | 60 seconds | 60 seconds, or MASH_TIMEOUT |
MASH_TIMEOUT (up to 3600 seconds) | ignored | honoured |
MASH_MEMORY_LIMIT | ignored | honoured, can only raise |
MASH_SINGLE_INSTANCE (schedules and queued scripts only) | honoured | honoured |
| Run log kept | by plan: 500 runs and 7 days (Business), 5,000 and 30 days (Enterprise) | 5,000 runs per script and 30 days |
Reaching other systems
Section titled “Reaching other systems”- Network. On an appliance an administrator can allow web and database hosts that the network guard blocks, with a script setting named
MASH_NETWORK_WHITELIST. Foldr Cloud has no such override. On both, a script can’t reach Foldr’s own internal addresses. - Email. On Foldr Cloud,
mash.mailsends through the managed service and only to members of your organisation. To send anywhere else, usemail.client(dsn)with your own mail server. The recipient limit applies on Foldr Cloud only, so on an appliance a script can mail any address its mail server will accept. - SMS. Sending an SMS isn’t available on Foldr Cloud. It works on an appliance.
Licence and feature gates
Section titled “Licence and feature gates”- Every script needs the MaSH feature. Webhooks and the action menu check it on each call.
mash.capturneeds the Captur feature.mash.reviewsneeds the Reviews feature.- Calls to a model (
mash.grace,mash.ai) use Grace credits on Foldr Cloud. The allowance depends on the plan.
Each library page carries badges for these conditions. Check the badge before building on an object.
Servers
Section titled “Servers”On a cluster, a schedule fires from one server and an event runs on whichever worker picks it up. A script must not assume it runs on the same machine twice, so keep state in a file, a field or mash.store.