Austin tech firms head to California to recruit talent

[user_id]

Originally Posted by Austin97

Many kids that I have interviewed coming out of CS programs (duke, UIUC (my alma mater), Stanford MIT and UT) can't even properly manage a tree or a linked list.

This is hard to believe considering this is basic material, how exactly can someone get through the rest of the program without even a basic understanding of a data structure? I mean, explaining a tree or linked list takes what....a few minutes?

In terms of threads, that stuff isn't rocket science...its hard to see how only the top 1~2% effectively understands it...

Regardless, firms recruit nationally all the time. Not sure how this is even news.....

[Austin97]

Originally Posted by user_id

This is hard to believe considering this is basic material, how exactly can someone get through the rest of the program without even a basic understanding of a data structure? I mean, explaining a tree or linked list takes what....a few minutes?

In terms of threads, that stuff isn't rocket science...its hard to see how only the top 1~2% effectively understands it...

Regardless, firms recruit nationally all the time. Not sure how this is even news.....

They learned it well enough to pass the test then moved on. But they didnt get any deep understanding - including big O notation and algorithmic complexity.

[user_id]

Originally Posted by Austin97

They learned it well enough to pass the test then moved on. But they didnt get any deep understanding - including big O notation and algorithmic complexity.

This is just my point, knowledge of data-structures isn't something you can just forget after the exams and still do well in later courses. A rigorous course in data-structures is taken either in your freshman or sophomore year and the knowledge is utilized in pretty much in all later course work..

Also, exams play only a small role in grading at top universities. The meat is all in the assignments and the first assignment in a data-structure course at a top CS program involves topics that go well beyond a basic understanding of a tree or linked list.

Anyhow, what you're saying doesn't add up....

[colombiunpride]

Originally Posted by Austin97

They learned it well enough to pass the test then moved on. But they didnt get any deep understanding - including big O notation and algorithmic complexity.

I find that hard to believe you're saying the majority of engineering students from top schools don't know this.

True, I've seen my fare share of slackers at my school, but the average student I've come across at GT knows big O and data structures.

Now, if you were referring to "Computer Science" majors from the pool of universities that offer the degree to begin with, I'd be more inclined to agree.

[Austin97]

Originally Posted by colombiunpride

I find that hard to believe you're saying the majority of engineering students from top schools don't know this.

True, I've seen my fare share of slackers at my school, but the average student I've come across at GT knows big O and data structures.

Now, if you were referring to "Computer Science" majors from the pool of universities that offer the degree to begin with, I'd be more inclined to agree.

I dont have a statistically significant sample by any stretch. I have interviewed at most 30-40 CS students from top universities.

[austin-steve]

A bit off topic, but since there are so many expert programmers chiming in, I've wondered:

Why is it so hard to program algorithms to sync Contact databases? Google has yet to do it right with Google Apps, such that Contact Groups or records can be shared as easily as Calendars can be.

I've tested multiple 3rd party Google Apps that claim to sync via the Google API the Google contacts system, but none of them work properly. They either do only a one-way write, or a limited "sync", or, as has happened with a couple of them, something goes haywaire and cretaes duplicate records for the entire database. Why is this so danged hard? Just curious.

[austinnerd]

Originally Posted by austin-steve

A bit off topic, but since there are so many expert programmers chiming in, I've wondered:

Why is it so hard to program algorithms to sync Contact databases? Google has yet to do it right with Google Apps, such that Contact Groups or records can be shared as easily as Calendars can be.

I've tested multiple 3rd party Google Apps that claim to sync via the Google API the Google contacts system, but none of them work properly. They either do only a one-way write, or a limited "sync", or, as has happened with a couple of them, something goes haywaire and cretaes duplicate records for the entire database. Why is this so danged hard? Just curious.

There may be some with some specific knowledge, but here is an example of why transferring database information can sometimes be more difficult than you might realize:

Let's say you have contacts database A that four fields for the name:

Salutation: (Mr, Mrs, etc)
First Name:
Middle Initial:
Last Name:

Now let's say you have contacts database B that utilizes a single field for the name:

Name:

(note: while at first blush the latter way might seem more restrictive, but in some cases it's actually less restrictive (e.g. international names that don't follow our "given/middle/family name" conventions))

So let's say we want to import contacts from A to B. In this case we can simply concatenate all the fields into a single name (Mr. Austin - Steve). Now let's say we want to transfer contact the other way, well now we have some parsing to do, we have to break up that 1 field to try to figure out how to make it fit into those four fields. OK so "Mrs. Abigail R Jones" fits just fine, but what about "Jones, Abigail", ooh, now we have to have fancier rules. What about "Mary Ann Smith" where "Mary Ann" is the first name? You can see where this starts to get tricky.

Like I said, the above is a made up example, I don't know the Google contacts schema/API enough to know what sort of issues are involved, but if no one is doing it "right", then I assume it's a similar issue (mapping schemas). Either that or just crappy Google API code (which wouldn't shock me).