Unattended Copy is a great idea. To really add the functionality implied we need some more options....
Could we change it to assume that unattended is the default for new objects once the first unattended function is started and prior to it's completion or have it set in the settings to default to one or the other? Also let's assume the error options stay the same as well so we don't keep having the pop-up.
Once the operation is started could it automatically minimize and stay minimized until user selected or until an error is encountered?
BTW, are the error options set for each set of objects added or are they changed for all objects when another option is chosen?
Something I find I need especially when copying large files is a resume or retry until successful completion. Or at least offer the option to retry forever unless some alternative option is used like for 1 hour or 10 attempts or whatever other options might make sense. Nothing worse than coming back a few hours later to see Skip, Retry, Ignore, Abort staring back at you.
There may be some other options that I've thought of but aren't coming to mind at the moment. I'll continue posting to this topic if you like or would you prefer a new one?
I hope you feel these are desirable options. I think most of the user base will agree that they are desirable.
You can create buttons which always copy in unattended mode using a command like Copy UNATTENDED=yes
Allowing operations to start with minimized progress bars via something similar to (but not identical to) @runmode:hide might make sense.
Regarding resuming and automatic retrying, except for the FTP/SSH functionality (which do things differently), Opus is designed around the assumption that your devices are generally reliable. What kind of thing are you doing where it is so slow and fails so frequently that you want automated retries/resumes? How reliably could you resume such a failure? (e.g. if it isn't clear how much of the file was transferred without padding or corruption, resume may not be a good idea) Would it make sense to use another protocol to transfer that data, instead? (e.g. FTP itself, or BitTorrent or similar. There are much better methods than standard file copy if you need to transfer large files across a a slow and unreliable network.)
Separate threads, please. There are already too many different ideas in this thread to be able to keep track of and talk about all of them in detail.
[quote="leo"]You can create buttons which always copy in unattended mode using a command like Copy UNATTENDED=yes
Allowing operations to start with minimized progress bars via something similar to (but not identical to) @runmode:hide might make sense.[/quote]
Where can I find more info on these "coding type" options? I have a few I've done but I've done them through VBS. I don't really understand the DO command language and I'm not familiar with where I can even find any info that explains the structure and syntax.
The problem is it's a wireless connection and while small transactions are very reliable long copies/transfers are not always. An automatic retry could go a long way to curing the copy problem just by giving the transaction some time for the wireless to reconnect.
It's only copies that end prematurely that are in error. If you click retry when you return to your computer your copy is not going to be corrupt. But that's the whole point of asking for the change. When I leave a process running I don't want to return only to see it sitting at the notorious prompt to abort, skip, retry, or ignore.
I'm a bit confused as to some of your reply. If this functionality is available through DO coding then why create the Unattended Copy function at all? To me it tells me that this is something new that has not been available that someone requested - I'm merely asking for options that make it truly unattended.
You used to have the main site where you could request enhancements, you've taken that away and seem to be pointing people here to make those requests. If I should be requesting these upgrades elsewhere then please let me know where.
You're telling me that WHENEXISTS=skip skips everything regardless of the difference between skip and skip identical? Because in my situation I would prefer to "skip identical" and not skip on identical names. If what you say is accurate then I want to "treat it as an error" or "rename existing" but I can't find what the syntax of those actions.
Shouldn't the list of all the options also include their syntax and maybe, depending on how complex a particular option may be, an actual example?
Access Denied would mean either the folder permissions don't allow the change (or they do, but need UAC elevation, and the UAC dialog was cancelled), or a program has the file open and is blocking it from being replaced.
I found it. But I can't explain it. It was a restricted permission on a share that only allowed read for everyone. But here's the rub - I could copy the files anyway except by using the newly created button. Is it possible that DO could be skipping security?
On the destination drive, I had full access but because I was using a share that restricted me to read-only I should've been restricted regardless of how I tried to copy the files to the destination and yet it only got blocked when I used that new button.
I didn't see a command line for "skip identical" either but shouldn't there be one?
What do I use for "treat as error" or "rename existing"? Do I type them as I did here with spaces or does it need quotes around it or do I skip the spaces?
No, security is enforced by the OS and filesystem. Opus could not bypass it, even on purpose. Especially when the destination is a network server, since the enforcement is happening on a completely separate machine.
There isn't one, but you were already using WHENEXISTS=skip, so you're already skipping conflicting identical files (as well as non-identical files) unless that has changed?
I think "treat as error" is the default, so you wouldn't specify the WHENEXISTS argument at all for that.
For "rename existing", WHENEXISTS=renameold is what the manual says to use.
I could use skip but I wouldn't know if it was truly and error or not until I compared the files manually and I'd have to cross directories and maybe even drives. If I use renameold at least the files will all be together in one directory.
My command now looks like this: Copy NOQUEUEWHENSAME QUEUE=quiet UNATTENDED=yes WHENEXISTS=renameold TO=ask$
The only access to the error log is by CSV? And all errors regardless of how many queues may be created will post to the error log? In other words the 2nd copy queue wouldn't overlay the error log? Will queues running in parallel write their errors to the log creating and interspersed error log?
I found it a bit weird that there's no destination listed in the error file. How will I know in a multiple queue operation where a particular file went? Also, will the queues run in sequence or if they have unique destinations will they run together like they do if you aren't using a button? I recall that queues with the same destination prompted for how you wanted them to execute, concurrently or one at a time. I presume that as I press the button that still will happen at point of creation?
How does the renameold do it's rename?
I came across this as I was looking through the options RENAMEWHENSAME If the source and destination are the same folder, the newly copied file will be automatically renamed to avoid a clash.
I'm not sure how this would ever be an option your saying a new file exists in the directory with the same name but since you're source and destination are the same directory this couldn't happen because it would already be a renaming problem to even have it in the directory.
When you say "treat as error" is the default you're saying I don't even need to specify a keyword, option or anything and it will write it to the log?
And lastly I kind of figured you were using the OS to enforce security but I wanted to be sure. I guess I'll try to pass it on to MS although they're a bit squirrely and don't have much of an understanding of their own product but it is a security problem so maybe it will get some attention.