Graduates can’t get a job because they don’t have experience and they can’t get experience because they don’t have a job.

Nice show guys.

Yes, I have seen http://rosalind.info. We discussed it a little bit back in episode 210 and it’s really cool. I’m planning on experimenting with those problems in Catalyst once the kids are advanced enough.

1. You could probably go through the full set of courses without learning any practical skill which could help you with getting a programming-related job. I’ve seen people during the last year getting very confused by the idea of using any version control and being close to failing a fizzbuzz test.

2. This produces a lot of people who know the theory, but can’t actually create anything / be useful without a full training. I worked with people both straight out of unis and during their year out – even the good ones were a serious time sink for the existing staff, so I’m not surprised companies without existing, extensive training programs aren’t keen to hire fresh graduates.

3. The most experienced people I worked with either didn’t bother with the studies, or got offers so good they dropped out after a year or two. I managed to work part time while going to the university, but I’d probably just work full time if that was the only option.

CS seems to be also a completely different area compared to some other career paths in that table. The “best” subjects are “medicine”, “education” and “law”, where you’re usually required by law to have a relevant degree before you’re hired.

In short – there are so many reasons this result could be skewed, I wouldn’t take it seriously. Also it may actually answer a different question than people expect (precisely: how many graduates got hired, rather than: how many open positions are there for good CS people)

———

On a (almost) completely different topic – have you seen http://rosalind.info yet? I think it’s an awesome bioinformatics introduction. It may not contain long detailed descriptions of what it teaches, but the practical approach really impressed me. I got to the apprentice level now and the information contained in the assignments + wikipedia where needed were enough to let me understand exactly what I’m doing and why.

Justin, if you still want random queries its still possible even if you don’t have a sequential id column (and use php’s rand to pick some number), what you do is something like the following,

SELECT * FROM `table`
WHERE rand() > 0.99
ORDER BY rand()

The trick is to make the 0.99 number capture a small amount of the total rows in the table. So for 4 million rows use something like 0.99995 which returns about 200 rows. You still have to pass over the table once but its a hell of a lot faster then making a copy of the table.

I tried it on a table with 500,000 rows once and it took over 10 minutes to run with the normal way, and less then half a second with the above. I had another go then on a table with 4 million rows and it returned in less then 3 seconds. This is on a virtual machine as well so on any server it should be much faster.

Just thought I would throw this out there. Someone may find it useful.

@Justin I used New Relic, it did great job telling me if something was wrong with my databases, CPU and etc.

log_slow_queries = /var/log/mysql/mysql-slow.log
long_query_time = 1

Then setup a cron job to email you, or a script to tail the file then SMS/email you.