2026/08/23
|
Sunday, August 23, 2026 (#235)
|
|
References |
Main Work
- Ferreteria FileSys stuff in order to support sftp in FTM
- Currently fighting with Streams. Everything is broken to various degrees...
It seems likely that some of the problem may (still) be the naming of process-related streams...
| current | in pipes | proposal |
|---|---|---|
| ExecLecture | OStream | ExecToApp |
| ExecRespond | IStream | AppToExec |
| ExecErrsBin | EStream | ExecOErrs |
| AppListen | IStream | AppFromExec |
| AppLecture | OStream | ExecFromApp |
| AppErrorBin | EStream | ? |
(Source: Session)
The proposed names may be less ambiguous, but I also kinda hate them.
I also need to get it clear in my head why we end up with two sets: why doesn't the app just write directly to ExecRespond?
I think I know why we need an AppListen that's separate from ExecLecture, though: if we want to monitor progress, then we need to be able to intercept data-packets as they happen. The same is presumably true for AppToExec and ExecFromApp when we're sending data.
...but maybe it would be better to hook things up directly, and use stream_notification_callback() to monitor progress?
Later: actually, it turned out to be a problem with the r1 qvar/qarray clades.