6 ms·

I think that the advice, if taken literally, is a bad advice. But it does point to the fact that the system is not just the technology, but also the people aro

I think that the advice, if taken literally, is a bad advice. But it does point to the fact that the system is not just the technology, but also the people around it I think that a better way to go about it was to comment first with the proper people that you MIGHT be able to make it run faster before doing anything (all though you know that you already did it). And drop the case if they don't want it. I can see that somebody might want to move the drawing calculation process to a more powerful box serving more clients, and distribute the client with a different license... in other words, there could be a business case for the original configuration, and knowing it might help you to discover the negative stakeholders (the people interested in the status quo) Should you always stay out of others people code? I think not, I think that there is a place for it. But from there to modify and publicize the modifications there is a big distance. Some people treat their code as their baby, and no parent like their kid criticized.

DavidBradbury 12y ago
The point wasn't to stay out of people's code. He used that "advice" as a reminder to not be what those people were/are.
- jorgeleo 12y ago
  And what are "those" people?
  - andreasvc 12y ago
    People who get upset if you touch their code. I think in your other comment you missed the fact that there was already an ongoing debate about this particular part of the code, and it wasn't going anywhere. In such a case a demonstration that the other method is faster is more compelling than more fruitless debate. Speculations on other reasons why the code should've used 2 processes is useless, as it is clear from the article that this change made it possible to get things done faster, no need to argue with that.
    - jorgeleo 12y ago
      "People who get upset if you touch their code" which is what i mean when i said that some people treat their code as their baby. I wrote it before the debate began, it just got posted after it, that is not in my control. The speculation was to point out that he didn't managed the situation properly, he just dismissed the current implementation as idiotic and proceed to change it. But you missed my point too. Which is why in some scenarios run faster might not be the first priority.
      - andreasvc 12y ago
        With "the debate" I referred to what is mentioned in the article of the two departments who disagreed about the best strategy. I don't quite follow you. In the article it's made quite clear that people wanted the code to run faster because then they could get more work done. Yes there are conceivable scenarios in which faster code is not better, but the article made clear that it wasn't the case, so I think the speculation didn't add anything.
scotty79 12y ago
> But it does point to the fact that the system is not just the technology, but also the people around it People come and go. They don't die when they are sacked. They are absorbed by other companies to do other possibly more awesome stuff. There is no point of delaying progress of software because that might mildly inconvenience some people. If half of the Adobe got sacked over a shitstorm caused by some rogue programmer making Photoshop suck 20% less I'd say it was worth it. The superpowers of companies is that they can die (or get severely crippled) without utterly destroying of what they consist of.
- aaronem 12y ago
  (In the following, we will stipulate that your "rogue programmer disrupts Photoshop, resulting in a better product and getting half of Adobe fired" situation is plausible, for the sake of illustration.) The half of Adobe you posit getting sacked wouldn't say it was worth it, though, and from a certain perspective -- namely, their own -- they'd be right. After all, Photoshop is already the unquestioned leader in its field, and the closest competitor isn't even within shouting distance, so what good does it really do Adobe's business if Creative Cloud's Photoshop component sucks 20% less? (Yes, no one is happy with the change from Creative Suite to Creative Cloud; this I concede. But those who are unhappy with it aren't decamping to the GIMP or some other laughable knockoff; instead, they are buying CS6 licenses, and they aren't going to be convinced to replace those with Creative Cloud subscriptions by a mere 20% improvement in Photoshop.) Meanwhile, there are a whole lot of people earning livings as part of Adobe Inc., who, in your "rogue programmer disrupts everything" scenario, suddenly won't be. This, for good and necessary reasons, makes every one of those people a political enemy of your rogue programmer; after all, whether he means to do so or not, the effect of his actions is to critically imperil those people's ability to maintain themselves, their families, and the position in life to which all of same have grown accustomed. (This might seem to you like a mild inconvenience. If so, I can only assume either that you're not getting paid well, or that you're only so recently getting paid well that you've yet to accustom yourself to it, and in either case that you're also not carrying a meaningful amount of debt. Hell, I'm not paid all that well, and I carry less debt than anyone I know, and I can assure you that my recent layoff was nonetheless far more than a "mild inconvenience".) Getting back to our illustration: Of course, I'm sure our lovable rogue doesn't intend to get all those people laid off; indeed, given his laserlike focus on the quality of the product for which he considers himself responsible and his evident and total innocence of matters political, it probably never even occurred to him that such a thing could happen. This doesn't matter at all. Whether he intended it or not, he's put a huge chunk of the corporation which employs him in danger of being laid off. Expecting those people to react in any other way than by attempting to squash him, by whatever means prove necessary and lie within their grasp, would be foolish; when you threaten a man's life, or something which a man regards as vital, i.e., essential to life, it is only basic sense to expect that he will react as strongly as is within his power. Speculate all you like on why this should be the case; regardless, it is a fact of human existence, and to regard it in any other frame of mind save acceptance is to waste your time and mental effort on utopian fantasy. And half of a company the size of Adobe is a group of people with a very great deal of power, indeed. Even if we scale down from Adobe as a whole to just the Photoshop project team, at least some of those people are perforce going to be in the higher ranks of management. In fact, since our rogue is a Photoshop developer, at least some of those people are going to be in his direct line of command. No points for guessing what happens to our lovable rogue when, through his innocence of the political realities of any large human organization, he places at risk the jobs of one or more of his direct superiors. Even if, by some impossible miracle, he manages not to lose his own job, then he's certainly not long for the company, because among the people he's managed to alienate are many of his peers. This means he's got more people looking to put knives in his back than he could keep up with, even if he realized he needed to be looking over his shoulder. Of course, our lovable rogue being not entirely without some sort of rudimentary sense of interpersonal relationships, even if he isn't shuffled aside immediately and laid off as soon as practicable, he'll realize that his only sensible option is to resign, because there is no longer any chance that he'll be able to work productively with his erstwhile colleagues, who now consider him about as safe to be around as old dynamite that's sweating nitroglycerin. So our rogue will quit his job, lick his wounds, and find somewhere else to work, and odds are the lesson he'll take from his error is, from the perspective of his future prospects and his ability to advance his career, the single worst one he could possibly choose: specifically, that office politics are incomprehensible, deadly, and evil, that anyone who engages in them is therefore necessarily an awful person and best avoided, as are office politics in general, because an honest hacker can't understand the game well enough to play it, and it'll only end in tears. None of which, of course, is true. But all of it accords perfectly with the default attitude among software engineers toward office politics, and I'm starting to think that might have a lot to do with the way that the default position of software engineers, when it comes to dealing with management, is "screwed".

DavidBradbury 12y ago

The point wasn't to stay out of people's code. He used that "advice" as a reminder to not be what those people were/are.

jorgeleo 12y ago
And what are "those" people?
- andreasvc 12y ago
  People who get upset if you touch their code. I think in your other comment you missed the fact that there was already an ongoing debate about this particular part of the code, and it wasn't going anywhere. In such a case a demonstration that the other method is faster is more compelling than more fruitless debate. Speculations on other reasons why the code should've used 2 processes is useless, as it is clear from the article that this change made it possible to get things done faster, no need to argue with that.
  - jorgeleo 12y ago
    "People who get upset if you touch their code" which is what i mean when i said that some people treat their code as their baby. I wrote it before the debate began, it just got posted after it, that is not in my control. The speculation was to point out that he didn't managed the situation properly, he just dismissed the current implementation as idiotic and proceed to change it. But you missed my point too. Which is why in some scenarios run faster might not be the first priority.
    - andreasvc 12y ago
      With "the debate" I referred to what is mentioned in the article of the two departments who disagreed about the best strategy. I don't quite follow you. In the article it's made quite clear that people wanted the code to run faster because then they could get more work done. Yes there are conceivable scenarios in which faster code is not better, but the article made clear that it wasn't the case, so I think the speculation didn't add anything.

scotty79 12y ago

> But it does point to the fact that the system is not just the technology, but also the people around it People come and go. They don't die when they are sacked. They are absorbed by other companies to do other possibly more awesome stuff. There is no point of delaying progress of software because that might mildly inconvenience some people. If half of the Adobe got sacked over a shitstorm caused by some rogue programmer making Photoshop suck 20% less I'd say it was worth it. The superpowers of companies is that they can die (or get severely crippled) without utterly destroying of what they consist of.

aaronem 12y ago
(In the following, we will stipulate that your "rogue programmer disrupts Photoshop, resulting in a better product and getting half of Adobe fired" situation is plausible, for the sake of illustration.) The half of Adobe you posit getting sacked wouldn't say it was worth it, though, and from a certain perspective -- namely, their own -- they'd be right. After all, Photoshop is already the unquestioned leader in its field, and the closest competitor isn't even within shouting distance, so what good does it really do Adobe's business if Creative Cloud's Photoshop component sucks 20% less? (Yes, no one is happy with the change from Creative Suite to Creative Cloud; this I concede. But those who are unhappy with it aren't decamping to the GIMP or some other laughable knockoff; instead, they are buying CS6 licenses, and they aren't going to be convinced to replace those with Creative Cloud subscriptions by a mere 20% improvement in Photoshop.) Meanwhile, there are a whole lot of people earning livings as part of Adobe Inc., who, in your "rogue programmer disrupts everything" scenario, suddenly won't be. This, for good and necessary reasons, makes every one of those people a political enemy of your rogue programmer; after all, whether he means to do so or not, the effect of his actions is to critically imperil those people's ability to maintain themselves, their families, and the position in life to which all of same have grown accustomed. (This might seem to you like a mild inconvenience. If so, I can only assume either that you're not getting paid well, or that you're only so recently getting paid well that you've yet to accustom yourself to it, and in either case that you're also not carrying a meaningful amount of debt. Hell, I'm not paid all that well, and I carry less debt than anyone I know, and I can assure you that my recent layoff was nonetheless far more than a "mild inconvenience".) Getting back to our illustration: Of course, I'm sure our lovable rogue doesn't intend to get all those people laid off; indeed, given his laserlike focus on the quality of the product for which he considers himself responsible and his evident and total innocence of matters political, it probably never even occurred to him that such a thing could happen. This doesn't matter at all. Whether he intended it or not, he's put a huge chunk of the corporation which employs him in danger of being laid off. Expecting those people to react in any other way than by attempting to squash him, by whatever means prove necessary and lie within their grasp, would be foolish; when you threaten a man's life, or something which a man regards as vital, i.e., essential to life, it is only basic sense to expect that he will react as strongly as is within his power. Speculate all you like on why this should be the case; regardless, it is a fact of human existence, and to regard it in any other frame of mind save acceptance is to waste your time and mental effort on utopian fantasy. And half of a company the size of Adobe is a group of people with a very great deal of power, indeed. Even if we scale down from Adobe as a whole to just the Photoshop project team, at least some of those people are perforce going to be in the higher ranks of management. In fact, since our rogue is a Photoshop developer, at least some of those people are going to be in his direct line of command. No points for guessing what happens to our lovable rogue when, through his innocence of the political realities of any large human organization, he places at risk the jobs of one or more of his direct superiors. Even if, by some impossible miracle, he manages not to lose his own job, then he's certainly not long for the company, because among the people he's managed to alienate are many of his peers. This means he's got more people looking to put knives in his back than he could keep up with, even if he realized he needed to be looking over his shoulder. Of course, our lovable rogue being not entirely without some sort of rudimentary sense of interpersonal relationships, even if he isn't shuffled aside immediately and laid off as soon as practicable, he'll realize that his only sensible option is to resign, because there is no longer any chance that he'll be able to work productively with his erstwhile colleagues, who now consider him about as safe to be around as old dynamite that's sweating nitroglycerin. So our rogue will quit his job, lick his wounds, and find somewhere else to work, and odds are the lesson he'll take from his error is, from the perspective of his future prospects and his ability to advance his career, the single worst one he could possibly choose: specifically, that office politics are incomprehensible, deadly, and evil, that anyone who engages in them is therefore necessarily an awful person and best avoided, as are office politics in general, because an honest hacker can't understand the game well enough to play it, and it'll only end in tears. None of which, of course, is true. But all of it accords perfectly with the default attitude among software engineers toward office politics, and I'm starting to think that might have a lot to do with the way that the default position of software engineers, when it comes to dealing with management, is "screwed".