Reputation
517
Top tag
Next privilege 1,000 Rep.
See votes, expandable usercard
Badges
4 12
Newest
 Enlightened
Impact
~12k people reached

  • 0 posts edited
  • 0 helpful flags
  • 77 votes cast
Aug
14
comment C: “this” vs “<instance name>”
@cmaster Good for you. Life is a journey of self-discovery.
Aug
14
comment C: “this” vs “<instance name>”
A suggestion for you for future reference - concise answers that assume that the OP knows what they are talking about and what they need, instead of attempting to question every premise, are time well-spent. Anything else is not. Refer to rwong's answer as an example of this.
Aug
14
comment C: “this” vs “<instance name>”
I'm sorry, but I fundamentally disagree with most of your assumptions, or find them trite in this context, and in some cases, such as your mention of an IDE, you are ignoring the specified parameters in my question. The whole idea of contextual variable names is old hat, the way we've all been doing things for many years. I am looking beyond that. The assumption that this (or similar) is the only way to do things is tiresome, and it does not address the idea of creating uniformity through the codebase.
Aug
14
comment C: “this” vs “<instance name>”
Another reason against using this in C code pretty much topples it for me. Thanks rwong. I will wait a while for other answers but this is looking likely.
Aug
14
comment C: “this” vs “<instance name>”
"Your are the same however people call you." No: There is one name on your birth certificate, for good reason: so that others can keep track of you, and not a thousand different aliases of you. This makes things simpler for everybody - and simplicity is what this question is about. Sometimes you have no choice but to change a name slightly for a given role ("Smith" is "Mr Smith" to his colleagues) but calling "hey you" ("this") can get you the unwanted attention of an entire room full of people. Always use the one true name if you can.
Aug
14
comment C: “this” vs “<instance name>”
"If you use indirect names like this, you actually do not need to search and replace all instances: this never needs to be replaced because it represents the current thing, whatever it is." No: If I want to find all instances of engine->window then I do not want to have to search for both engine->window AND this->window. Re coupling an instance to its type name, this is less of a problem than having to do multiple find-replaces. Many of these objects are singletons (as is typical in engines & frameworks) meaning that the only sensible instance name you can give is the name of the type!
Jul
1
comment How to be a zero-bug programmer?
+1. In the words of Twisted Sister, What you don't know sure can hurt you / What you can't see makes you scream.
Jan
29
comment How can we inspire “hacker” culture?
I... really don't see the connection between commoditisation and hacking, I'm sorry. I doubt others will, either, because it's not clear what goal your question has. Why would you want to "inspire" hacker culture? Your tastes are your tastes, not all of us want to be hackers!
Jan
26
comment In a legacy codebase, how do I quickly find out what is being used and what isn't?
Yeah, I don't think it's as clear cut as that, though that book looks worth reading. It does depend very much on the size/complexity of the codebase, and warm bodies available to do the work.
Jan
26
comment In a legacy codebase, how do I quickly find out what is being used and what isn't?
+1 This is a fantastic answer. Where's that +5 button gotten to...
Jan
26
comment In a legacy codebase, how do I quickly find out what is being used and what isn't?
+1. In response to Joel's comment, "It bloody well shouldn't be." Because I don't see the problem as inherent. I see it as being partly the fact that many people write shoddy code and don't care, while many others write reasonably good code but live by the "self-documenting code" concept... which is just plain BS: One may flatter one's own coding style all one wishes in privacy, but when it comes to public codebases just spawn comments like there's no tomorrow. Doesn't hurt. And finally there are people who have to get things working in a legacy codebases, on a tight time budget.
Jan
26
comment In a legacy codebase, how do I quickly find out what is being used and what isn't?
@Oded Rename is definitely easier than trial-and-error deletion! Good thinking there. That's one more tool in the box.
Jan
26
comment In a legacy codebase, how do I quickly find out what is being used and what isn't?
Thanks. Although your answer's somewhat Java specific, it's interesting to see your layered approach... peeling the onion, so to speak. Something to think about.
Jan
26
comment In a legacy codebase, how do I quickly find out what is being used and what isn't?
@jcmeloni No, I'm not being paid for the evaluation. But in my experience, and from small things I have picked up in the last couple of days, they don't have anyone else at the table right now. My skillset is fairly unusual, so I'm even more at ease that they don't have anyone else competing for it, based on the quote. The actual quote in question is from my client-to-be to their client, who is planning to re-award them the contract. Really from my end, I am meant to assist them in providing said quote. HTH.
Jan
26
comment In a legacy codebase, how do I quickly find out what is being used and what isn't?
Ordinarily I'd be 100% with you on the toss-and-rewrite approach. But in this instance (and at least for now), I am to be paid just for work to maintain the site, rather than a more extensive overhaul which would take several weeks. Also, even if I wanted to right now, I couldn't keep up with doing that and holding down the other contracts I have on the go, since my weekly availability for this is explicitly limited -- my primary contract must be fulfilled to its 40 hour weekly minimum.
Dec
15
comment Do TODO comments make sense?
Yep, given a listing in your IDE, they are helpful. I would say they're of very limited use otherwise, since the codebase may be enormous.
Dec
14
comment How do quick & dirty programmers know they got it right?
This is definitely the most insightful answer here. Thanks for sharing.
Sep
2
comment Would you put your stackoverflow profile link on your CV / Resume?
Agreed. I actually looked up this question because I was considering it myself. I've since gone ahead and put my gamedev.stackexchange profile on my resume (I'm in the top 30 or so users on that subsite). Re recruiters, you're spot on, but I've enough breadth of skills that my CV is likely to wind up on someone's desk, and that someone is likely to be a tech lead of some sort. I put it at the top of the CV, with eg. my skype and LinkedIn. The contract I've just signed isn't games-related, but I don't doubt it will carry some serious weight for the next such contract that comes up.
Sep
2
comment Would you put your stackoverflow profile link on your CV / Resume?
Yeah, that's what I'm saying. Using him as an example of the value of stackoverflow participation probably doesn't make sense in relation to the rest of us. The point is he probably probably produces more value when he breaks wind than the rest of us do in 6 months of crunch time. Tech directors likely worship statuettes of him, carved from precious materials, in dark corners of company HQ. He almost certainly groks at least 80% of the inner workings of the universe. And still has time to spend on stackoverflow.
Aug
29
comment What are the most common algorithms in the field of computational geometry?
Since you were the only one who answered this question before it got closed, guess you are the winner!