Page 1 of 2

a software problem

Posted: Sat Oct 09, 2004 12:31 pm
by rich_m
here's a small incident that led me into writing this post:

Our head informed us that our "course registration" was being computerized (for the first time) and handed us leaflets with the subjects we could choose from. He then warned us to be very careful while entering our choice because

:-\The software will accept a subject even its time clashes with another subject, and so if this kind of input is given we would have to find a way to attend two subjects in two different buildings at the same time.
:-\We cannot edit our choice's
:-\incase we tap the esc-key by mistake only the subjects we had chosen till then will be recorded we will not be able to register any more.
******************
:-[well this sure gave me a fright, but i think it can't be helped for this semester. anyhow I'm determined to make a better software for the next....
question is how??

can someone give me ideas...etc on the 'php-concept'

Re:a software problem

Posted: Mon Oct 11, 2004 2:47 am
by df
sounds like pure crap to me! how can they use such software?

i used to write school timetabling software. there is some mojo science underneath for timetabling without clashes. (it is not trivial OR easy).

Re:a software problem

Posted: Mon Oct 11, 2004 11:19 am
by rich_m
sounds like pure crap to me! how can they use such software?

i used to write school timetabling software. there is some mojo science underneath for timetabling without clashes. (it is not trivial OR easy).
you see thats exactly why they made that kind of software..
they feel its easier if the students r careful than breking thier heads with the mojo science. :-\

btw what did u use to build that software..

Re:a software problem

Posted: Tue Oct 12, 2004 2:49 am
by df
originally it was all in C for dos. we rewrote and redesigned the whole thing into vb6 with full business logic frontend/backend Ntier stuff. it worked out very well. (vb6 is quite capable if you use it right). we were behind the ball. the company had not bothered to upgrade to a windows version so when the dos version started failing (nobody had made any fixes to this version for years, it just worked...).. the compilers no longer worked under windows etc.

so boss said, rewrite. you have n-1 months! :) sheesh.

anyway, it turned out very well.

Re:a software problem

Posted: Tue Oct 12, 2004 7:33 am
by Guerst
any hope you might share the logic behind it?

Re:a software problem

Posted: Tue Oct 12, 2004 9:37 am
by df
i havnt worked there for more than 4 years.

Re:a software problem

Posted: Tue Oct 12, 2004 6:55 pm
by Andrew_Baker
I'd first build a database to hold all of the course schedules. Student Name, Id, chosen courses, etc.

Another component would be a database of courses and times.

The client program should take the student's choices and hold them in a buffer. It should then check for overlaps against the database of courses.

If there are no conflicts, commit the buffer to the student database.

If there are conflicts, either eliminate the violations or even better, flag them to show the student what's up.

Re:a software problem

Posted: Wed Oct 13, 2004 2:46 am
by df
hmm.

list of rooms
list of teachers
list of subjects
list of students
list of what subjects students are doing.

map fixed room/courses first (ie: IT into the IT room, metalwork into the metalshop room, etc. ones you cant move., no point doing it in the gym!). allocate the teachers + students. remove form lits.
now map the freeforum stuff to any other rooms.

the idea is to have the least amount of clashses (ie, a teacher doing two subjects at the same time etc). and then manually move things around for the end product.

Re:a software problem

Posted: Wed Oct 13, 2004 1:29 pm
by Curufir
I don't think what he's asking for is as difficult as you're making out df.

He's not, AFAIK, trying to schedule the classes themselves, but student registration to the classes. Room/Teacher scheduling is a PITA, but as that scheduling has already been done (And the data must be available or the students couldn't currently do course registration) it should be pretty simple to let students choose classes that don't clash.

In effect all the hard work has already been done. A database of course scheduling data and some simple clash detection is all that's needed IMHO.

Re:a software problem

Posted: Thu Oct 14, 2004 2:06 pm
by Andrew_Baker
Hell, even clashing would be tolerable if you could at least edit your entries before committing them.

Re:a software problem

Posted: Thu Oct 14, 2004 4:11 pm
by mystran
In a sub-University I could accept that overlaps aren't possible, but in a University system I would kill the developers personally, using my bare hands, if someone tried to tell me that I'm not allowed to take two course simply because they happen to share times for some lectures.

In any case, the best way to counter these kinds of systems is to organize a boycott. If sufficiently large number of students simply hit "esc" as the first thing they do, there's not much they can do, except either (1) accept the old method instead, or (2) fire everyone participating in the boycott... yeah.. right..

Re:a software problem

Posted: Thu Oct 14, 2004 5:26 pm
by Schol-R-LEA
From what I understand about it, US colleges and universities are very different from those in Europe and the UK, especially at state universities. They are often just an extension of the sort of classes they took in high schools, with attendance records and homework and so on, especially at the freshman and sophomore levels.

You don't really see anything like what I understand the Continental universities are like until your graduate studies, and since most US college students are there to earn a practical degree, rather than for scholarship, the majority of grad students are in specialized fields such as engineering, medicine, or law. As a WAG, perhaps only half of all US college students ever actually graduate (most end up side-tracked into a low-end career, even those who end up graduating later), and of those, less than a tenth take any further study in their major field afterwards, except as required to remain employed.

Re:a software problem

Posted: Thu Oct 14, 2004 5:29 pm
by mystran
I knew US was doomed, but I didn't know it was that bad already.

Re:a software problem

Posted: Thu Oct 14, 2004 5:46 pm
by Schol-R-LEA
Already? Trust me, from what I have heard, this is an improvement over how it was sixty years ago. Prior to WWII and the GI Bill (a scholarship program set up for returning veterans), only the brightest and wealthiest went to college; today it is expected, literally as a continuation of high school, and even the poorest student (in either sense of the word) can usually qualify for a grant, loan or scholarship of some sort.

But that doesn't mean scholarship itself is respected; as with everything else in the US, it all comes down to the bottom line. Only business, law and medical degrees are truly valued, and the latter only because of the potential earnings of a medical practice; many med schools actively discourage students whose primary interest is in helping their patients, and courses in insurance procedures are at least as important as those in medical procedures.

Physical Education is also honored, not for it's own sake but because it is a stepping stone to professional sports, and because successful teams bring in alumni donations for the schools.

Science, mathematics and engineering get some grudging respect, but only to the degree that they have practical results; anyone seeking a degree in a pure science, especially field biology or abstract mathematics, is likely to be dismissed as having their heads in the clouds (trust me on this; my original major specialization was microbial ecology). This can be seen in Computer Science quite clearly: a few years ago, it was the darling of the schools and quite heavily supported, but once the Internet bubble burst, interest and funding dropped off almost overnight.

It's even worse in the Humanities than in the Sciences; there is a widespread contempt for anyone whose field of study is not immediately practical. It is an often-repeated observation that, as far as one's career prospects in the US are concerned, a Liberal Arts degree is often worse than no degree at all.

Teaching is the field which gets the least respect of all; it is widely assumed that anyone going into teaching simply isn't capable of doing anything productive - which teaching is not considered as being.

This is perhaps too jaundiced a view; my own experience has left me rather bitter, obviously. There are pure scholars in the US, but they are few and far between, and they invariably were either prodigies who got full scholarships through grad school and beyond, or else from wealthy families who were willing to support their 'eccentricity'. Even then, they are under a tremendous pressure to produce new scholarship, regardless of quality - the term for this is 'publish or perish', and even tenured professors have trouble avoiding getting their funding cut if they don't emit a constant stream of new papers and theses. As a result, senior professors are often too busy to actually teach, and the low-level courses are often actually covered by assistant professors or even grad students, the latter of whom are often virtual slaves to their advisors (fanfiction writer Plaid Adder had some choice things to say about this, in her satirical 'critical analysis' of the Sith Academy fanfiction stories).

The oddest thing about this manic practicality is that it is not so much about money (though that is a strong motivator) as much as about productivity fnord. In a society where nearly everybody is a religious fanatic fnord (especially the atheists fnord, as odd as it might sound), Puritan Work Ethic is a more powerful factor than greed alone. For much of US society, your work is your life fnord, and everything else is secondary to it (including family and friends).

Re:a software problem

Posted: Sat Oct 16, 2004 12:32 pm
by rich_m
btw coming back to the topic
I don't think what he's asking for is as difficult as you're making out df.
He's not, AFAIK, trying to schedule the classes themselves, but student registration to the classes. Room/Teacher scheduling is a PITA, but as that scheduling has already been done (And the data must be available or the students couldn't currently do course registration) it should be pretty simple to let students choose classes that don't clash.
In effect all the hard work has already been done. A database of course scheduling data and some simple clash detection is all that's needed IMHO.
well thats exactly the case here.Although i think i was not specific bout it in the beginning & what does AFAIK-PITA-IMHO stand for??