[Tutor] Design question: Web-based vs. desktop-based vs. desktop-based with data backed up to web server with a tablet thrown in for all cases?

[boB Stepp]

On Fri, Jul 17, 2015 at 11:18 PM, Laura Creighton <lac at openend.se> wrote:
> I have a plug in usb keyboard that I can just plug into my tablet.
> It  works great for data entry...

I suggested this to my wife and believe she is looking into it.
However, whenever she is walking about her classroom, she does not
want to be messing with her keyboard, just quickly tapping and
selecting predetermined skill assessment options (Student just
beginning to learn; practicing now; has achieved mastery of this
skill; etc.).

>
>>1) Find a way to always keep the laptop and tablet synchronized.
>
> All the cloud storage services want to sell you something that
> does this.  So does Dropbox.

My wife has a Dropbox account and mentioned this as a possibility.


>>Is this a sensible way of approaching at least this one project?  Of
>>course, I have not a clue as to how to do any of this (yet!).

[snip]

> The idea is to use as much off the shelf technology to do what you
> want, and not reinvent the wheel.  Especially when it comes to
> security issues that, worse case scenario, have all the parents
> suing you for not keeping things that are required to be private
> from being hacked.
>
> Before there was cloud computing, and smartphones, we wrote a client
> server app that handled the billing situation for the local energy
> company -- plus all their power outages, service apps etc, etc, etc.
> It took 4 of us nearly a year to get the permissions down right for
> all possible cases.
>
> Doing the same sort of thing for our mobile app using existing
> cloud libraries took about 3 weeks.  Only some of that speedup was
> due to the fact that we knew what we were doing a lot better the
> second time around, and we had really extensive unit tests from
> the old way of doingf that we could slightly refactor and then
> make sure they passed under trhe new regime.
>
> You haven't picked yourself an easy one here ...

Tell me about it!  Vonda (my honored spouse) has been hinting about
all of this for quite some time.  But she had not figured out her
exact needs.  She actually still hasn't, but has a broad idea of what
she initially would like to have now.  And she has a greatly
exaggerated idea of what I can accomplish in five weeks!

I suspect version 0.0.1 will be something along the lines of:
Everything basically exists on the tablet.  Get a keyboard that works
with the tablet.  Manually backup/sync tablet with laptop as needed.
Does the program have the basic functionality you *really* want as you
wander about the room?  Probably not.  Tell me what you now need.
Etc.  Only when I feel I have what she really wants nailed down will I
get into the rest of our discussion.  But I *do* want to design my
code with where the final version needs to go, so hopefully I won't
have any major rewrites in my future!

> One warning -- the trick to high speed performance for such things is
> to copy as little data as is needed from one device to another (and
> then let the devices regenerate a lot of what they need from the
> basic necessary set).  But the trick for _developing_ and
> debugging such things is to send more information, including a
> bunch of handshaking things like 'I am trying to open the MongoDB
> Database now.  Ok, that worked, now I am adding new job tickets ...
> and so on and so forth.
>
> If you do not do things like that, the time will come when you sit
> pointing at your server saying 'You aren't doing anything.  Why aren't
> you doing <whatever it is you want it to do>?"  Utterly frustrating.
> YOu need to build a verbose way into what you do so you can have your
> server tell you what it thinks it ought to be doing, and then a
> way to shut the server up when you go into production.

This makes a lot of sense!  Must remember...

> But I would go google for cloud services that have data synchronisation.
> (Which is pretty much all of them, that being the point, after all).
> You want to build your app out of some of these ready made parts, or
> risk that  your wife's students will all be in university before you are
> done with your app.

For sure!  Especially when I get into some student drills programs
that she will undoubtedly want the students to work on both in class
and at home with an option for the parents to view their child's
progress online.

But for this student assessment project, it is going to have to be
without all the desired bells and whistles to have something that will
be practically useful for her when school starts.  Especially when I
am certain Vonda is still figuring out what she *really* needs!



-- 
boB