Re: Capturing the state of keyboard modifiers on startup

On Mar 8, 10:37 am, "Aleksey Gureev" <spyro...@xxxxxxxxx> wrote:
Sorry for a long reply.





It shouldn't because it's a service function that I, as a developer
and supporter, need to help a user in trouble to recover in a
relatively easy way instead of telling something, like "go to that
hidden folder in your home directory that Microsoft placed god knows
where and remove that file". Shit happens and we need some "easter
eggs" to help ourselves.

Well...what happens if the user happens to accidentally press your
"easter egg" key and does a database reset, resulting in them losing
all their data. (And, of course, they won't know what happened.) You
may say that will never happen, but then you would be underestimating
what user's will try or accidentally do. Also, purposely programming
in hidden features or an "easter egg" can really come back to bite you
later...especially if the application is not well document.

I agree with Joe here...I feel there's a more intuitive way to
implement this.

Right, it's dangerous to some extent. Hmmm

The configuration wizard does run during the initial installation. I
need to be able to re-run it at will without re-installing the
application.

Usually something like this is a trigger in a file. When the
configuration wizard runs the first time, it sets a parameter in the
config file and makes the "run_config" value a 0 instead of a 1,
meaning "don't run the configuration wizard again." If you need to
run it again, though, just change the value back.

Yeah, but I also think how am I supposed to explain to a user-in-
trouble where to find a config file (or a key in Win registry) and how
to edit it. Another pain in the neck, I would say.

Please stop discussing the idea; it's just not worth your time.

To be fair - if you post up how you're doing something on Usenet, you
can expect people (especially technical types) to analyze what you are
doing. With that said...

I didn't mean to stop this discussion at all. Let me explain. I'm
under the strict time limits (who isn't) and if I write something to
Usenet, I know I tried everything else, Googled, performed other
research etc. So I expect people to be concise and constructive.
Instead, after a day of waiting, I get criticism with absolutely no
answers to my questions. What would be your reaction to this?
(rhetorical question)

What's of real interest is the technical side of a problem.

I honestly don't think there's a way to capture key information before
the application launches (except in your "kludge" method). I wouldn't
put it in a splash screen, though. Have something that checks in
between the splash screen and the actual start-up of the program to
accomplish this (just don't show the component).

Right. I don't like it too.

Again, you may not find a lot of people who can help you with this
problem as it's not one that should/will generally be tackled due
other implementation possibilities.

Yeah. Still I don't see a user-safe, developer-friendly method of
calling the list of service functions during the startup

a. without editing a file
b. without any visible GUI commands / options

The special stress is on (b) because when the DB is corrupted and I
need to reset it, the application won't start to give me any GUI with
commands / options. That's why I need something that could be invoked
during the initial startup phase.

Anyway, thanks for all bright ideas voiced and sorry if I offended
anyone! I really appreciate any help.

- Al- Hide quoted text -

- Show quoted text -

(b) because when the DB is corrupted and I
need to reset it, the application won't start to give me any GUI with
commands / options.

Ooo...I see. That's crummy. So, basically, the GUI is dependent upon
the database being correct - whatever correct may be?

Another alternative - create a second "admin" program that does the
things you want it to do (and can be done in a graphical way).

Also, like I said in my original post, how hard would it be to have
the user provide command line arguments? Not only is this already
supported (via the main method entry point), it's also very easy to
code and will be there before any GUI initialization. Since it sounds
like what you are doing is kind of kludgy to begin with, this really
isn't any worse of an idea, and, again is pretty easy to implement.

Good luck.

.

Relevant Pages

  • Re: How to thread a progress bar in a separate dialog?
    ... > but you can run into race conditions by touching things in the GUI. ... Touching commands such as ... Ahh yes, race conditions, a very baaaaad thing. ...
    (microsoft.public.vc.mfc)
  • Re: The future of Linux
    ... Occasionally, I might use a spreadsheet, wordprocessor, or database. ... depending on the selection of OS your work uses. ... "bash history" that remembers your commands and allows you to search ... talking to remote databases without a GUI overhead across the network. ...
    (comp.os.linux.misc)
  • Re: python QT or python-GTK
    ... drag&drop GUI BUILDER. ... Toolkits are libraries defining an API (in the ... multi-thousand lines of config file in xt; the config file was nearly as ... You can design your GUI's in the designer, ...
    (comp.lang.python)
  • Re: How to thread a progress bar in a separate dialog?
    ... but you can run into race conditions by touching things in the GUI. ... Touching commands such as ... If you were to try to run StepIt() on your ProgressBar in the ...
    (microsoft.public.vc.mfc)
  • Re: Loading large data sets
    ... - You use the Tcl-binding of VTK to load big datasets. ... While loading your GUI should not freeze. ... as a pipe and configuring a fileevent whenever the pipe becomes ... commands) to it asynchronously. ...
    (comp.lang.tcl)