Random Rants
-
I drank like an eejit yesterday, resulting in a bad hangover… It's fading as we speak, but I did decide to ask my collegue to cover for me tomorrow morning.
It's a rant, cause I don't like skipping a workday unless I'm almost dying... I'll do the hours next week but still... I'm a dumbarse bastard son of a limping penguin.
I had a horrible hangover yesterday. Throwing up and cold sweats before a cab ride to the airport and a delayed flight is not the way you should ever spend your time while hung over.
-
Throwing up: check. Cold sweats: check.
Beer now: check.
My GF's parents had a little dinner for me when I got to Toronto. They bought 3 bottles of wine for the night. I had literally one sip and couldn't drink anymore.
In a very thick Ukrainian accent, her mother asked me, "Why are you no drinking?" "Are you a hangover?"
Yes, I'm a hangover.
-
Hah
-
working as a teacher is a lot of work. the job doesn't stop once 3pm rolls around. That all being said, it's great! but FUCK!!!! i miss wearing denim daily. Soon i'll need to get me a denim jacket and wear it like i would wear jeans in my life as a student… I NEED SOME FADES IN MY LIFE!
-
Packed some beers for a trade with a yank… Bombproof packaging... Added a lil extra bottle... Hauled it to the postoffice in the pouring rain... Clerk puts it on the scale: 5.1 kgs... 0-5 kg = 32.8 €, 5-10 kg = 65 €.
Took it back home, opened it up, took out the extra, closed it again, back to the PO in a bit.
-
Today I got a letter from the local tax office. I go to school, don't have a real job, feck off. Dickheads.
-
I've had a first round interview with a company for an interesting sounding contract, but it's come back to the dreaded home test.
There are two parts: One is a set of questions asking how I'd set up a user test to see if a new system was better than the old one. They want two alternative sets of metrics, and a plan of how it would be approached. I don't have a clue how to do this, as I write code for a living, and am not generally part of the design or optimisation team. Besides, as a freelancer I almost always join a project when there's a pressing need to get it out of the door.
So far I've written about a thousand words on it. I'm halfway through, and probably need a couple more hours to complete it. But I'm really feeling like I can't be bothered.
The second is to write a native Android app to hit a Flickr API, pull down a list of images with their metadata, and display them with a slick UI. OK, so I need to roll a HTTP call, which is pretty simple. Then decode a piece of XML, that can be done in Jackson. The UI could be done pretty basically using using a list adapter or similar, but if you want slick then it gets to be a lot more work.
Normally these tests come with a clause that they must be production quality. So I suppose that I could hack up an uncommented mudball of spaghetti code. (If you're not a professional nerd, that means it barely works, is needlessly complex, and can't be understood by another programmer.) That's what a large chunk of systems are like in the real world. But if you're doing it properly, then it'll be fully commented, have decent test coverage, be tested, etc.
Getting something up and running will probably take a day. Maybe two or three if I want to make something slick, and ensure it's fully tested.
And that's the trouble. I don't mind spending an hour or so on a job application, but when you start needing to spend one or two days it gets to be too much. Particularly when there's no compensation, and in a lot of cases I don't get any feedback.
Plus I'm having a real depressive downswing right now, and getting anything done is a struggle.
-
Chin up old chap! The last few job applications that my wife has had to do have been absolute monsters and as I've said to her, once it's done it's done and typically these things are simply a barrier to entry designed to limit the number of applicants. Get it done if you think it's genuinely worth going for the job and then reward yourself with some fine purchases from our good bank account emptying brethren in Gosport…..
Sent from my iPhone using Tapatalk
-
if you are willing to take a small risk, complete part 1 as good as you can and write for part 2 only the needed steps in detail and the time frame you would need to set up a running, full working and completed tested app.
And just tell them, Part 1 was a small challange, since it is out of your normal work field…. Part 2 your daily business, no need to proof that one...
You would show that you have flexibility and willingness to work in, for you, new / uncommon fields, but also know exactly where your strong points are.
-
Hi Graeme,
Are they not after A/B Testing for part 1? Add Analytics/Click Throughs/Time on site and call it done.
Second bit, sounds yuck. Any fancy android SDK front ends? Pump it through that if so. That does sound rather annoying.
As Cutlass said, they're probably looking for people willing to go beyond. In my interviews if I get a question wrong, at the next interview I'm expected to have an extremely good answer to the same question again, showing my ability, care diligence and ability to rise above.
For the right role, I don't imagine you'll have a problem getting it done. Maybe you should have a few beers, and submit it as a joke, who cares right? If you're not serious about it a 1/2 assed enjoyable attempt might be all they/you need?
-
Shit they probably just want to rip off a free test plan and software…
-
Thanks. It's all very well expecting someone to go beyond, but if it gets to the point that you're having to do a couple of days work for a three month contract…
@Giles I'm meaning to make a run down to Gosvegas in the near future. If @Danny RnH doesn't bankrupt me first. :o
-
I've had a first round interview with a company for an interesting sounding contract, but it's come back to the dreaded home test.
There are two parts: One is a set of questions asking how I'd set up a user test to see if a new system was better than the old one. They want two alternative sets of metrics, and a plan of how it would be approached. I don't have a clue how to do this, as I write code for a living, and am not generally part of the design or optimisation team. Besides, as a freelancer I almost always join a project when there's a pressing need to get it out of the door.
So far I've written about a thousand words on it. I'm halfway through, and probably need a couple more hours to complete it. But I'm really feeling like I can't be bothered.
The second is to write a native Android app to hit a Flickr API, pull down a list of images with their metadata, and display them with a slick UI. OK, so I need to roll a HTTP call, which is pretty simple. Then decode a piece of XML, that can be done in Jackson. The UI could be done pretty basically using using a list adapter or similar, but if you want slick then it gets to be a lot more work.
Normally these tests come with a clause that they must be production quality. So I suppose that I could hack up an uncommented mudball of spaghetti code. (If you're not a professional nerd, that means it barely works, is needlessly complex, and can't be understood by another programmer.) That's what a large chunk of systems are like in the real world. But if you're doing it properly, then it'll be fully commented, have decent test coverage, be tested, etc.
Getting something up and running will probably take a day. Maybe two or three if I want to make something slick, and ensure it's fully tested.
And that's the trouble. I don't mind spending an hour or so on a job application, but when you start needing to spend one or two days it gets to be too much. Particularly when there's no compensation, and in a lot of cases I don't get any feedback.
Plus I'm having a real depressive downswing right now, and getting anything done is a struggle.
question 1: speed city. is it faster? memory footprint? cpu usage? i always work like i'm peeling an onion. start at the outer level. then go inwards. next thing you know you've written skeleton functions for 10 layers and you finally got to the 10 lines of critical code that at that point are VERY easy to write since you've been so dilligently writing earlier… basically how i design stuff. it tends to work well.
c++ only:
channel stroustrup: when in doubt, use a vector
when not channeling stroustrup, use a map
lists are bad unless you need iterator validation
never seen a deque used
sets are near the same as maps, except for not as cool, although there are sometimes when you just need a set
hashmaps are iffy unless you have a proprietary implementation
multimaps are iffy but sometime the right answerbasically just use a bunch of vectors and maps and dont iterate too many times and you're good to go
2nd step you're on you're own because i hate xml lol
-
They just dont fit
Freshly waxed, sales-thread will be online soon… -
What size? Still not sure if those Vibergs are for me or not
-
@Megatron1505 8.5E, but they fit very narrow, more like a 8.5D…
I like the Bergs on you.
