I'm thiking about creating UNIX accounts merely for students to upload
their homework through scp. But I have not solved all what I want to
solve. (Such students are expected to be technical enough to use scp.)

I'm thinking of scp because it's solid enough. It's simple. It works.
I end up having all of the work centralized in a UNIX system. I get
dates et cetera.

But what'd really like is to have a program receive each file transfer
and rename the file, say. In other words, I'd like to associate an scp
transfer to a script, much like e-mail can invoke a script on a UNIX
system.

Maybe SSH can sort of do this. I'll look into the documentation. Any
ideas? Thank you!

(*) My thought process leading up to this

Solution 1. Students could send me e-mails, but I have the massive work
of pulling files from my inbox into a UNIX system where I can look at
each one. E-mail does the job because it /sort of/ proves the delivery.

Solution 2. They could scp files, but this lets them overwrite files.
It'd be nice if each upload would turn the submission into a
next-version.

Meredith Montgomery <mmontgomery@levado.to> wrote:
>
>Solution 1. Students could send me e-mails, but I have the massive work
>of pulling files from my inbox into a UNIX system where I can look at
>each one. E-mail does the job because it /sort of/ proves the delivery.

Your unix system doesn't have postfix? Send email to the unix system,
then write a procmail script to sort it out into files. You get multiple
valid timestamps on each message. You get some reasonable but not
completely validated idea of who sent the message. You don't need to
give any other access to students. There is only a way for them to send
data to you, never to remove data.

>Solution 2. They could scp files, but this lets them overwrite files.
>It'd be nice if each upload would turn the submission into a
>next-version.

If you HAVE to do this, you can rsync the directory they send it to into
another place with timestamps added to the filenames. This way if student
A uploads a file and then student B uploads an invalid file with student A's
name on it in order to help the grade curve, it becomes very evident.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

kludge@panix.com (Scott Dorsey) writes:

> Meredith Montgomery <mmontgomery@levado.to> wrote:
>>
>>Solution 1. Students could send me e-mails, but I have the massive work
>>of pulling files from my inbox into a UNIX system where I can look at
>>each one. E-mail does the job because it /sort of/ proves the delivery.
>
> Your unix system doesn't have postfix? Send email to the unix system,
> then write a procmail script to sort it out into files. You get multiple
> valid timestamps on each message. You get some reasonable but not
> completely validated idea of who sent the message. You don't need to
> give any other access to students. There is only a way for them to send
> data to you, never to remove data.

That's really what I wanted to do, but I'm running a server with no
rights to send or get email (to and from the outside world). I'm
running a cheap server at Hetzner. (Hm, maybe I should think of a new
server just to do this because the mail solution is indeed attractive.)

>>Solution 2. They could scp files, but this lets them overwrite files.
>>It'd be nice if each upload would turn the submission into a
>>next-version.
>
> If you HAVE to do this, you can rsync the directory they send it to into
> another place with timestamps added to the filenames. This way if student
> A uploads a file and then student B uploads an invalid file with student A's
> name on it in order to help the grade curve, it becomes very evident.
> --scott

That's interesting. Students A and B would have only access to their
own directory, so they would not be able to step on their toes.
However, I would not have any power of naming their files and there are
more inconveniences with this approach, but I should perhaps not even
describe and steal your time reading them. (The mail solution seems
more attractive. Maybe I should take this more seriously and buy a new
server.)

Thank you for your thoughts.

On 01/09/2022 11:07, Meredith Montgomery wrote:
> I'm thiking about creating UNIX accounts merely for students to upload
> their homework through scp. But I have not solved all what I want to
> solve. (Such students are expected to be technical enough to use scp.)
>
> I'm thinking of scp because it's solid enough. It's simple. It works.
> I end up having all of the work centralized in a UNIX system. I get
> dates et cetera.
>
> But what'd really like is to have a program receive each file transfer
> and rename the file, say. In other words, I'd like to associate an scp
> transfer to a script, much like e-mail can invoke a script on a UNIX
> system.
>
> Maybe SSH can sort of do this. I'll look into the documentation. Any
> ideas? Thank you!
>
> (*) My thought process leading up to this
>
> Solution 1. Students could send me e-mails, but I have the massive work
> of pulling files from my inbox into a UNIX system where I can look at
> each one. E-mail does the job because it /sort of/ proves the delivery.
>
> Solution 2. They could scp files, but this lets them overwrite files.
> It'd be nice if each upload would turn the submission into a
> next-version.

Presuming you are on a UNIX(TM) or UNIX-like Operating System, have a
cron job scan their directories and copy the files elsewhere, changing
the name to include the time they were moved.

Do it every five minutes or so.

Wouldn't take much more than five minutes to set up. (But test it for
an hour or two. :-) )

Cheers,
Gary B-)

On 8/31/2022 7:07 PM, Meredith Montgomery wrote:
> I'm thiking about creating UNIX accounts merely for students to upload
> their homework through scp. But I have not solved all what I want to
> solve. (Such students are expected to be technical enough to use scp.)
>
> I'm thinking of scp because it's solid enough. It's simple. It works.
> I end up having all of the work centralized in a UNIX system. I get
> dates et cetera.
>
> But what'd really like is to have a program receive each file transfer
> and rename the file, say. In other words, I'd like to associate an scp
> transfer to a script, much like e-mail can invoke a script on a UNIX
> system.
>
> Maybe SSH can sort of do this. I'll look into the documentation. Any
> ideas? Thank you!
>
> (*) My thought process leading up to this
>
> Solution 1. Students could send me e-mails, but I have the massive work
> of pulling files from my inbox into a UNIX system where I can look at
> each one. E-mail does the job because it /sort of/ proves the delivery.
>
> Solution 2. They could scp files, but this lets them overwrite files.
> It'd be nice if each upload would turn the submission into a
> next-version.

You might be able to write a script for students to run that would scp a
file chosen by the student to a target file whose name includes the
output of 'whoami', possibly the student's hostname, and a timestamp
(possibly generated by piping the output of 'date' through a 'tr'
command that would translate spaces to underscores).

This would hopefully give you unique file names. Presumably, you'll be
able to read each file and figure out who submitted it.

Students would not be executing scp directly. If you'd rather they
didn't see the destination directory at all, you might be able to use
shc to turn your script into a binary executable that students could run
but which no normal person could read.

Louis

Meredith Montgomery <mmontgomery@levado.to> writes:

> I'm thiking about creating UNIX accounts merely for students to upload
> their homework through scp. But I have not solved all what I want to
> solve. (Such students are expected to be technical enough to use scp.)
>
> I'm thinking of scp because it's solid enough. It's simple. It works.
> I end up having all of the work centralized in a UNIX system. I get
> dates et cetera.
>
> But what'd really like is to have a program receive each file transfer
> and rename the file, say. In other words, I'd like to associate an scp
> transfer to a script, much like e-mail can invoke a script on a UNIX
> system.
>
> Maybe SSH can sort of do this. I'll look into the documentation. Any
> ideas? Thank you!
>
> (*) My thought process leading up to this
>
> Solution 1. Students could send me e-mails, but I have the massive work
> of pulling files from my inbox into a UNIX system where I can look at
> each one. E-mail does the job because it /sort of/ proves the
> delivery.

This is what I used to do. The email had to have the exercise number in
the subject. I used to use mh, so a simple script could find
coursework, extract the work and, if appropriate, compile and run tests
on it.

I wonder why getting mail onto a Unix system would be such hard work.
What's going on to make that the case?

> Solution 2. They could scp files, but this lets them overwrite files.
> It'd be nice if each upload would turn the submission into a
> next-version.

--
Ben.

On Wed, 31 Aug 2022 22:07:17 -0300, Meredith Montgomery wrote:
> I'm thiking about creating UNIX accounts merely for students to upload
> their homework through scp. But I have not solved all what I want to
> solve. (Such students are expected to be technical enough to use scp.)
>
> I'm thinking of scp because it's solid enough. It's simple. It works.

Or I being the student would prefer rsync

> I end up having all of the work centralized in a UNIX system.

Sounds like a perfect way to cheat and to have homework of other users
destroyed or modified. :(

> I get dates et cetera.
> But what'd really like is to have a program receive each file transfer
> and rename the file, say. In other words, I'd like to associate an scp
> transfer to a script, much like e-mail can invoke a script on a UNIX
> system.
>
> Maybe SSH can sort of do this. I'll look into the documentation. Any
> ideas? Thank you!
>
> (*) My thought process leading up to this
>
> Solution 1. Students could send me e-mails, but I have the massive work
> of pulling files from my inbox into a UNIX system where I can look at
> each one. E-mail does the job because it /sort of/ proves the delivery.

Yep, and I would not think it too hard to have you write some code to
automagically extract each student's email into a file. Go ahead and
send yourself a few emails then cat /var/mail/your_login_id

I always smile when I see a teacher/professor post these kinds of requests.
Seems to me it would be dead easy to have the Computer Science teacher
make it a homework assignment and the class pick the best code as the
next example for fault testing.

> Solution 2. They could scp files, but this lets them overwrite files.
> It'd be nice if each upload would turn the submission into a
> next-version.

My suggestion: each student creates a $HOME/teacher, sets sticky bit so
teacher always has access to files. Teacher then has a cron job to pull
files into desired location readable only by the teacher.

Students are told time collect_homework script runs and everyone is on the
same page. Might be nice if a naming convention is used so students can
easily find a particular file when desired.

In article <867d2nnbqy.fsf@levado.to>,
Meredith Montgomery <mmontgomery@levado.to> wrote:

> Maybe SSH can sort of do this. I'll look into the documentation. Any
> ideas? Thank you!

My first inclination would be to make an HTTP server with PHP
available. You can then copy from other people to make a simple
upload page. You can expand it as you are inclined to put course
materials, schedules, etc online.

Doing PHP upload from scratch is annoying but you can borrow from
many available examples. It will also be fairly secure if you
keep it simple.

--
:-<> Siri Seal of Disavowal #000-001. Disavowed. Denied. Deleted. @
'I desire mercy, not sacrifice.' /|\
Discordia: not just a religion but also a parody. This post / \
I am an Andrea Chen sockpuppet. insults Islam. Mohammed

Ben Bacarisse <ben.usenet@bsb.me.uk> wrote:
> I wonder why getting mail onto a Unix system would be such hard work.
> What's going on to make that the case?

One possiblity, given what little details have been provided, is a VM
host that blocks incoming and outgoing port 25 on their VM's.

Bit Twister <BitTwister@mouse-potato.com> wrote:
> Yep, and I would not think it too hard to have you write some code to
> automagically extract each student's email into a file. Go ahead and
> send yourself a few emails then cat /var/mail/your_login_id

If they setup their MDA to deliver to mail-dir style folders, they get
each email in a separate file from the start, removing the need to
first take apart an mbox style folder (not that doing so is hard in any
case).

Meredith Montgomery <mmontgomery@levado.to> writes:
> I'm thiking about creating UNIX accounts merely for students to upload
> their homework through scp. But I have not solved all what I want to
> solve. (Such students are expected to be technical enough to use scp.)

[...]

> But what'd really like is to have a program receive each file transfer
> and rename the file, say. In other words, I'd like to associate an scp
> transfer to a script, much like e-mail can invoke a script on a UNIX
> system.

[...]

> Solution 1. Students could send me e-mails, but I have the massive work
> of pulling files from my inbox into a UNIX system where I can look at
> each one. E-mail does the job because it /sort of/ proves the delivery.

If you can arrange for e-mails to be delivered to the UNIX system, you
can process them automatically, ie, with a script there. The obvious
traditional suggestion for this would be procmail.

On Wed, 31 Aug 2022 22:07:17 -0300, Meredith Montgomery wrote:

> I'm thiking about creating UNIX accounts merely for students to upload
> their homework through scp. But I have not solved all what I want to
> solve. (Such students are expected to be technical enough to use scp.)
>
> I'm thinking of scp because it's solid enough. It's simple. It works.
> I end up having all of the work centralized in a UNIX system. I get
> dates et cetera.
>
> But what'd really like is to have a program receive each file transfer
> and rename the file, say. In other words, I'd like to associate an scp
> transfer to a script, much like e-mail can invoke a script on a UNIX
> system.
[snip]
> Solution 2. They could scp files, but this lets them overwrite files.
> It'd be nice if each upload would turn the submission into a
> next-version.

As an alternative, you could look into one of the many "revision control"
systems. I'm thinking specifically of git or cvs, both of which can handle
networked commits.

In theory, your students would "commit" their homework to your repository,
and you could then "pull" the homework out to be evaluated. Such revision
control systems include user privilege management that you may be able
to use to limit a student's ability to, for instance, "commit" to another
student's homework, or "pull" another student's homework for copying.

It would take a bit of set-up, but would give you control over what the
students can access, and an audit trail of what they /have/ accessed.

HTH
--
Lew Pitcher
"In Skills, We Trust"

Ben Bacarisse <ben.usenet@bsb.me.uk> wrote:
>I wonder why getting mail onto a Unix system would be such hard work.
>What's going on to make that the case?

She's using a system hosted by incompetent idiots. Incompetent German
idiots, but incompetent nevertheless.

The whole concept of putting data into the cloud relies on the notion
that other people can run systems better than you can. Sometimes this
is the case. Sometimes it is not.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

Meredith Montgomery <mmontgomery@levado.to> writes:
> Solution 1. Students could send me e-mails, but I have the massive work
> of pulling files from my inbox into a UNIX system where I can look at
> each one. E-mail does the job because it /sort of/ proves the delivery.

It’s probably worth looking at why moving files from your mail to a Unix
host is ‘massive work’. It shouldn’t be more effort than saving an
attachment and then scping it yourself.

--
https://www.greenend.org.uk/rjk/

Meredith Montgomery <mmontgomery@levado.to> writes:
>I'm thiking about creating UNIX accounts merely for students to upload
>their homework through scp. But I have not solved all what I want to
>solve. (Such students are expected to be technical enough to use scp.)

>I'm thinking of scp because it's solid enough. It's simple. It works.
>I end up having all of the work centralized in a UNIX system. I get
>dates et cetera.

>But what'd really like is to have a program receive each file transfer
>and rename the file, say. In other words, I'd like to associate an scp
>transfer to a script, much like e-mail can invoke a script on a UNIX
>system.

I like using https://github.com/dutchcoders/transfer.sh/ alot.
You can upload the files with curl (or a web browser). It'll make a
new file for each upload.

You'd have to determine how students differentiate their uploads so
you don't end up with a random jumble of files. (ie. they must name
their files after themselves, or student ID or something).

On the backend, it just creates a random directory to put the file in,
and then it can't be touched (other than deleted if the delete token
is captured). You'll end up with all the files in subdirectories of
where you have the web server dropping files.

--
Doug McIntyre
doug@themcintyres.us

On 01/09/2022 02:07, Meredith Montgomery wrote:
> I'm thiking about creating UNIX accounts merely for students to upload
> their homework through scp. But I have not solved all what I want to
> solve. (Such students are expected to be technical enough to use scp.)
>
> I'm thinking of scp because it's solid enough. It's simple. It works.
> I end up having all of the work centralized in a UNIX system. I get
> dates et cetera.
>
> But what'd really like is to have a program receive each file transfer
> and rename the file, say. In other words, I'd like to associate an scp
> transfer to a script, much like e-mail can invoke a script on a UNIX
> system.
>
> Maybe SSH can sort of do this. I'll look into the documentation. Any
> ideas? Thank you!
>
> (*) My thought process leading up to this
>
> Solution 1. Students could send me e-mails, but I have the massive work
> of pulling files from my inbox into a UNIX system where I can look at
> each one. E-mail does the job because it /sort of/ proves the delivery.
>
> Solution 2. They could scp files, but this lets them overwrite files.
> It'd be nice if each upload would turn the submission into a
> next-version.

The suggestions to have your own Unix server receive email are all well
and good but setting it up *and securing it* is a PITA.

An alternative is to use Amazon's SES (Simple Email Service)
<https://aws.amazon.com/ses/details/>

Your volumes are probably low enough that the free tier will cover
everything, and it's easy to have a script run on receipt of an email.
Anti-virus scanning of the attachments is automatic.

--
Bruce Horrocks
Surrey, England

Supersedes: <mail-20220902130956@ram.dialup.fu-berlin.de>

Meredith Montgomery <mmontgomery@levado.to> writes:
>I'm thinking of scp because it's solid enough. It's simple. It works.
>I end up having all of the work centralized in a UNIX system. I get
>dates et cetera.

written earlier:

So you have a UNIX system that is online meaning students can
connect to it via SCP, but this UNIX system does neither have
a mailserver nor are you allowed to run a mailserver on it?

written later:

Sorry, I missed the post where you already answered this!

I would not say that you "have a server" if you can't run
a mail server program on it. - If it's just port 25 that's
blocked then you could use any other port and dispatch a
specific client to the students that connects to that port.

Ben Bacarisse <ben.usenet@bsb.me.uk> writes:

> Meredith Montgomery <mmontgomery@levado.to> writes:

[...]

>> (*) My thought process leading up to this
>>
>> Solution 1. Students could send me e-mails, but I have the massive work
>> of pulling files from my inbox into a UNIX system where I can look at
>> each one. E-mail does the job because it /sort of/ proves the
>> delivery.
>
> This is what I used to do. The email had to have the exercise number in
> the subject. I used to use mh, so a simple script could find
> coursework, extract the work and, if appropriate, compile and run tests
> on it.
>
> I wonder why getting mail onto a Unix system would be such hard work.
> What's going on to make that the case?

I can't run a MTA in this server because e-mail ports are blocked by the
ISP. I can pay more and get that out of the way, though.

But it seems you mean you had a script that would look into an inbox and
extract homework from messages. But my mail is not in the UNIX system.
I'd need a script to POP3-download or IMAP-download or something like
that. I wouldn't know how to quickly do something like that. I'd feel
better setting up a new server with an MTA and have mail piped to a
program.

E-mail has the downside that the mechanism doesn't report right away
that the student is using it wrong. A run of scp (if there's no way for
files to the wrong directory) would always be successful (as long as I
could name the files).

I think I'll write a web server that takes file uploads after
authentication (to identify students). Students would just drag their
homework and the web server would do the right thing. I'm looking for a
quicker solution, but perhaps I'll end up doing just that.

Lew Pitcher <lew.pitcher@digitalfreehold.ca> writes:

> On Wed, 31 Aug 2022 22:07:17 -0300, Meredith Montgomery wrote:

[...]

> As an alternative, you could look into one of the many "revision control"
> systems. I'm thinking specifically of git or cvs, both of which can handle
> networked commits.
>
> In theory, your students would "commit" their homework to your repository,
> and you could then "pull" the homework out to be evaluated. Such revision
> control systems include user privilege management that you may be able
> to use to limit a student's ability to, for instance, "commit" to another
> student's homework, or "pull" another student's homework for copying.
>
> It would take a bit of set-up, but would give you control over what the
> students can access, and an audit trail of what they /have/ accessed.

This idea seems very interesting. I think you're saying each student
would have their own repository and they would push their repository to
a server (from where I'd pull and look at).

Richard Kettlewell <invalid@invalid.invalid> writes:

> Meredith Montgomery <mmontgomery@levado.to> writes:
>> Solution 1. Students could send me e-mails, but I have the massive work
>> of pulling files from my inbox into a UNIX system where I can look at
>> each one. E-mail does the job because it /sort of/ proves the delivery.
>
> It’s probably worth looking at why moving files from your mail to a Unix
> host is ‘massive work’. It shouldn’t be more effort than saving an
> attachment and then scping it yourself.

My mail is stored at Gmail. I'd have to click on each of them, save
them locally and then eventually move it to a UNIX system. (Downloading
mail through Gmail's POP3 or IMAP backends doesn't seem like a good idea
for this. I need something simpler. My knowledge is also limited.)
I'd be better off clicking on each of them, downloading and looking at
it at my own workstation. But it's the clicking-to-download that is the
massive work. Too many clicks to make it fun.

Doug McIntyre <merlyn@dork.geeks.org> writes:

> Meredith Montgomery <mmontgomery@levado.to> writes:

[...]

>>But what'd really like is to have a program receive each file transfer
>>and rename the file, say. In other words, I'd like to associate an scp
>>transfer to a script, much like e-mail can invoke a script on a UNIX
>>system.
>
> I like using https://github.com/dutchcoders/transfer.sh/ alot.
> You can upload the files with curl (or a web browser). It'll make a
> new file for each upload.
>
> You'd have to determine how students differentiate their uploads so
> you don't end up with a random jumble of files. (ie. they must name
> their files after themselves, or student ID or something).
>
> On the backend, it just creates a random directory to put the file in,
> and then it can't be touched (other than deleted if the delete token
> is captured). You'll end up with all the files in subdirectories of
> where you have the web server dropping files.

This is interesting, but it's not satisfying the requirement of naming
files. Naming the file is kind of important to me for avoiding
conflicts and letting me know who did what. Asking students to put
their names on their homework is not infallible: authenticating students
resolves that: they can't upload until their identification is done.

Nevertheless, this gives me the idea that if I write a web server for
the job, we should have an endpoint that students can submit their work
through the command-line. (A nice interface for the more technical
people.)

(*) Tangent

I dislike asking human beings to pay attention or put things in a
certain format. This is for computers to do. While we can get a great
percentage of students to stick to a format, there's always a minor
percent that will never do it. Programmers should tell computers to
check for things and let the user know it's not in the right format.
Honestly, I feel embarassased asking students to name their files
properly. I'd rather force them to authenticate themselves. (For
instance, perhaps just asking their ID in a system is enough. No
password necessary. Unless we would find we have a problem due to
mallicious people trying to create havoc, which case we'd put passwords
in place.)

Bruce Horrocks <07.013@scorecrow.com> writes:

> On 01/09/2022 02:07, Meredith Montgomery wrote:
>> I'm thiking about creating UNIX accounts merely for students to upload
>> their homework through scp.

[...]

> The suggestions to have your own Unix server receive email are all
> well and good but setting it up *and securing it* is a PITA.

Good point.

> An alternative is to use Amazon's SES (Simple Email Service)
> <https://aws.amazon.com/ses/details/>
>
> Your volumes are probably low enough that the free tier will cover
> everything, and it's easy to have a script run on receipt of an
> email. Anti-virus scanning of the attachments is automatic.

That's interesting. I watched a video in which they set up an S3 bucket
and the mail is stored there. I guess I could then write a program to
download all files and extract attachments. I'd have to learn to use
the S3 API. This seems simple enough, elegant enough.

ram@zedat.fu-berlin.de (Stefan Ram) writes:

> Supersedes: <mail-20220902130956@ram.dialup.fu-berlin.de>
>
> Meredith Montgomery <mmontgomery@levado.to> writes:
>>I'm thinking of scp because it's solid enough. It's simple. It works.
>>I end up having all of the work centralized in a UNIX system. I get
>>dates et cetera.
>
> written earlier:
>
> So you have a UNIX system that is online meaning students can
> connect to it via SCP, but this UNIX system does neither have
> a mailserver nor are you allowed to run a mailserver on it?

Almost right. I can run a mail server, but port 25 is blocked. To have
it open, I'd need to pay more.

> written later:
>
> Sorry, I missed the post where you already answered this!

That's alright.

> I would not say that you "have a server" if you can't run
> a mail server program on it. - If it's just port 25 that's
> blocked then you could use any other port and dispatch a
> specific client to the students that connects to that port.

If I need to worry about a mail client, I'd just write a web server that
authenticate users and gets the files from them --- in a drag-and-drop
fashion, say.

On Sun, 04 Sep 2022 17:22:37 -0300, Meredith Montgomery wrote:

> Nevertheless, this gives me the idea that if I write a web server for
> the job, we should have an endpoint that students can submit their work
> through the command-line. (A nice interface for the more technical
> people.)

Have you looked at installing Moodle?

--
Using UNIX since v6 (1975)...

Use the BIG mirror service in the UK:
http://www.mirrorservice.org

Meredith Montgomery <mmontgomery@levado.to> wrote:
> Ben Bacarisse <ben.usenet@bsb.me.uk> writes:
>> Meredith Montgomery <mmontgomery@levado.to> writes:
>
> [...]
>
>>> (*) My thought process leading up to this
>>>
>>> Solution 1. Students could send me e-mails, but I have the massive work
>>> of pulling files from my inbox into a UNIX system where I can look at
>>> each one. E-mail does the job because it /sort of/ proves the
>>> delivery.
>>
>> This is what I used to do. The email had to have the exercise number in
>> the subject. I used to use mh, so a simple script could find
>> coursework, extract the work and, if appropriate, compile and run tests
>> on it.
>>
>> I wonder why getting mail onto a Unix system would be such hard work.
>> What's going on to make that the case?
>
> I can't run a MTA in this server because e-mail ports are blocked by the
> ISP. I can pay more and get that out of the way, though.
>
> But it seems you mean you had a script that would look into an inbox and
> extract homework from messages. But my mail is not in the UNIX system.
> I'd need a script to POP3-download or IMAP-download or something like
> that. I wouldn't know how to quickly do something like that.

It shouldn't be hard with Movemail from GNU Mailutils, or
alternatives like fdm:
https://www.mailutils.org/wiki/Fetching_Mail_with_Movemail

> I'd feel better setting up a new server with an MTA and have mail
> piped to a program.

Mailutils also has Sieve, which allows filtering messages from
specified mailboxes ("-f"), without involving an MDA.

> E-mail has the downside that the mechanism doesn't report right away
> that the student is using it wrong. A run of scp (if there's no way for
> files to the wrong directory) would always be successful (as long as I
> could name the files).

Well email sent should always end up in your mailbox (though
using Gmail puts this logic at risk because they like to reject
things that they falsely identify as malicious), so it seems the
same to me. With email you don't have to set up a separate user on
the system for each student in order to identify who uploaded what.

--
__ __
#_ < |\| |< _#