Dear Python



  • It's the year 2012. Why the hell are you still trying to install yourself to C:\python32 ?



  • I installed a program for recording video from DirectX games, named FRAPS, that tries to install itself in C:\Fraps. Even more annoying, I changed the install directory to the correct one, and the program works flawlessly, even in Windows 7 with UAC on-- so it uses the wrong path for literally no reason.



  • @blakeyrat said:

    I installed a program .... that tries to install itself in .. it uses the wrong path for literally no reason.
     

    What you mean by "wrong path"?  If a program works properly no matter where it is installed, how can it be "wrong"?

    I guess maybe I'm commenting more on the view of "wrong" here - is being constrained to put programs in some specific location really "correct"? 

    Note I agree that it is more convenient - sometimes - and provides a standard interface, but is it "correct"?

    I don't even think I have a world-view statement that can even evaluate the "correctness" of a particular storage paradigm...


  • ♿ (Parody)

    @too_many_usernames said:

    What you mean by "wrong path"?  If a program works properly no matter where it is installed, how can it be "wrong"?

    It goes against the standards of the operating system.

    @too_many_usernames said:

    I don't even think I have a world-view statement that can even evaluate the "correctness" of a particular storage paradigm...

    Uh, so, the author of the OS cannot create a standard? Is that really what you're trying to say?



  • @too_many_usernames said:

    @blakeyrat said:
    I installed a program .... that tries to install itself in .. it uses the wrong path for literally no reason.
    What you mean by "wrong path"? If a program works properly no matter where it is installed, how can it be "wrong"?

    If I, the user, has told the operating system to put all program files in "E:\Programs" and Fraps installs to "C:\Fraps", then it is wrong. I don't see how there is any doubt about this. For all Fraps knows, the reason it's told not to install on C: is that C: has no space. Or is a read-only volume. Or any of a hundred things it shouldn't be second-guessing.

    The reason I said it's doubly-stupid is that if I move Fraps to "E:\Programs\Fraps" it still works correctly. Meaning, Fraps' installer is ignoring the OS's program path for absolutely no reason.



  • @tOmcOlins said:

    It's the year 2012. Why the hell are you still trying to install yourself to C:\python32 ?

    Real wtf is using Python on windows. Be a man and use the Mac Book


  • ♿ (Parody)

    @Nagesh said:

    @tOmcOlins said:
    It's the year 2012. Why the hell are you still trying to install yourself to C:\python32 ?

    Real wtf is using Python on windows. Be a man and use the Mac Book

    How do you use the Mac Book on Windows?



  • @boomzilla said:

    Uh, so, the author of the OS cannot create a standard? Is that really what you're trying to say?
     

    Not quite: I was trying to say that just because something is a standard doesn't necessarily mean it's "correct."  So breaking the standard doesn't necessarily mean you're "wrong" even though it does mean "you're not in conformance with the standard."  Granted, in the case where you base "correct" and "wrong" solely on the standard then the statement holds, but I was being more philosophical.

    @blakeyrat said:

    If I, the user, has told the operating system to put all program files in "E:\Programs" and Fraps installs to "C:\Fraps", then it is wrong. I don't see how there is any doubt about this. For all Fraps knows, the reason it's told not to install on C: is that C: has no space. Or is a read-only volume. Or any of a hundred things it shouldn't be second-guessing.

    The reason I said it's doubly-stupid is that if I move Fraps to "E:\Programs\Fraps" it still works correctly. Meaning, Fraps' installer is ignoring the OS's program path for absolutely no reason.

    I understand your sentiment, but I will pick nits here:  what isn't honoring your request to install all programs to the location you specify?  I would argue that the OS failed, because why should the installer program have the freedom to do its own thing? Isn't the OS supposed to enforce things like that since, after all, the job of the OS is to manage resources?

    As for the installer, I would say it's a bit of a WTF if it doesn't at least tell you where the default location is and give you an option to change it.  If the default is "wrong" but allows you to change it, that seems like a minor annoyance to me.

     



  • @boomzilla said:

    @Nagesh said:
    @tOmcOlins said:
    It's the year 2012. Why the hell are you still trying to install yourself to C:\python32 ?

    Real wtf is using Python on windows. Be a man and use the Mac Book

    How do you use the Mac Book on Windows?

    No windows in your house? lolololol

    I meant - do python development on apple mac instead of windows


  • ♿ (Parody)

    @too_many_usernames said:

    Not quite: I was trying to say that just because something is a standard doesn't necessarily mean it's "correct."  So breaking the standard doesn't necessarily mean you're "wrong" even though it does mean "you're not in conformance with the standard."  Granted, in the case where you base "correct" and "wrong" solely on the standard then the statement holds, but I was being more philosophical.

    Only if by philosophical you mean wrong, or in a postmodern sense. But I repeat myself.

    @too_many_usernames said:

    I understand your sentiment, but I will pick nits here:  what isn't honoring your request to install all programs to the location you specify?  I would argue that the OS failed, because why should the installer program have the freedom to do its own thing? Isn't the OS supposed to enforce things like that since, after all, the job of the OS is to manage resources?

    No, that's not the job of Windows. There may be some OS where this sort of thing is enforced, but that OS ain't Windows. You may argue about some theoretical OS, but that's not what we're talking about, and isn't actually you're saying.



  • @too_many_usernames said:

    Not quite: I was trying to say that just because something is a standard doesn't necessarily mean it's "correct."

    You're objectively wrong.

    @too_many_usernames said:

    Granted, in the case where you base "correct" and "wrong" solely on the standard then the statement holds, but I was being more philosophical.

    I was going to type "if by 'philosophical' you mean 'wrong'", but alas Boomzilla beat me to it.

    @too_many_usernames said:

    I understand your sentiment, but I will pick nits here:

    It's always a good idea to engage me by saying, basically, "my point relies on me being a pedantic dickweed", because I'm sure you know how much I love people being pedantic dickweeds.

    @too_many_usernames said:

    what isn't honoring your request to install all programs to the location you specify?

    Fraps. More specifically, Fraps' installer.

    @too_many_usernames said:

    I would argue that the OS failed, because why should the installer program have the freedom to do its own thing?

    Because Windows is a general purpose operating system.

    @too_many_usernames said:

    Isn't the OS supposed to enforce things like that since, after all, the job of the OS is to manage resources?

    It does. It does this in two ways:

    1) It enforces some things technically (UAC is a good example-- a program can't write data to C:\Windows without triggering a UAC warning)

    2) It enforces other things by contract (using named folders correctly is a good example-- programmers have a contract with the OS to do it correctly. The OS doesn't technically prevent them from doing it wrong, but programmers who do it wrong are breaking their contract and, in effect, breaking their program.)

    If you don't know that APIs have contracts, then you have failed as a programmer and you should quit and start digging ditches for a living. Because seriously man. This is Computer Science 101 shit here.

    @too_many_usernames said:

    If the default is "wrong" but allows you to change it, that seems like a minor annoyance to me.

    A bug is a bug. More egregious, the installer software they're using probably already defaulted to the correct path; I wager they had to go out of their way to break it.



  • @blakeyrat said:

    @too_many_usernames said:
    Not quite: I was trying to say that just because something is a standard doesn't necessarily mean it's "correct."

    You're objectively wrong.

     

    I'm going to create a "standard" that says the ratio of the circumference to diameter of a circle in base 10 arithmetic to infinite precision is the integer 3.  I have just created a standard that is "wrong."  This is what I meant by "A standard doesn't necessarily have to be 'correct'."

    @blakeyrat said:

    2) It enforces other things by contract

    Contracts never enforce anything; at best they provide instructions to an enforcing agent.  You have to have something (a military, laws of physics, an operating system, whatever) to actually enforce a contract.

    I also think you've blurred the line between "API" and "guidelines."  An API is something that gives you access to certain functionality, where guidelines suggest ways in which you should use that functionality.  In the case of install location, that's a guideline, not an API. If the only way to install a program was through an API call, then you could actually enforce the install location.

    Windows does not have an API that restrics program installation location. What it does have is an API of the form "write files to this location" and "hey I have installed executable things to this location."  Yes, this is because (as boomzilla said) Windows is general-purpose and does allow you the freedom to put your files anywhere. It seems like this is a subset of the much more general "with much freedom comes much responsibility" discussion: Windows gives programmers the freedom to write their installers however they want, but this means they are then responsible for the responses their decisions elicit.  I just don't think it makes them "wrong" for making a particular choice.

    I believe that it is an OSs job to control where you can install files. Windows does in fact provide this control, just with very few restrictions: it allows an installer the freedom to install its files in a large number of lcoations, limited only by "permission".  If a program installer chooses to install a program somewhere other than
    recommended by a guideline but still allowed by the API, I don't think
    it's "wrong." 

    Annoying? Perhaps. Wrong? I cannot support that view.




  • Why, may I ask, is it the job of the OS to prevent the user from installing files wherever he damn well pleases? A Windows PC is not an iPod...



  • @too_many_usernames said:

    @blakeyrat said:

    @too_many_usernames said:
    Not quite: I was trying to say that just because something is a standard doesn't necessarily mean it's "correct."
    You're objectively wrong.
     

    I'm going to create a "standard" that says the ratio of the circumference to diameter of a circle in base 10 arithmetic to infinite precision is the integer 3.  I have just created a standard that is "wrong."  This is what I meant by "A standard doesn't necessarily have to be 'correct'."

    Unless you're making an agreement with another developer to use that value, that's not a standard.


  • @too_many_usernames said:

    I'm going to create a "standard" that says the ratio of the circumference to diameter of a circle in base 10 arithmetic to infinite precision is the integer 3. I have just created a standard that is "wrong." This is what I meant by "A standard doesn't necessarily have to be 'correct'."

    Don't be stupid. Context matters.

    @too_many_usernames said:

    Contracts never enforce anything; at best they provide instructions to an enforcing agent.  You have to have something (a military, laws of physics, an operating system, whatever) to actually enforce a contract.

    Jesus, could you engage your brain for 13 seconds and think? Just stop and THINK.

    What, in Windows, enforces the API contract? Well, geewhiz professor, how about that whole "if you don't follow it your program is broken on many Windows configurations thing?" Or if you want to be literal, you can't get a Windows Logo certification without following it.

    @too_many_usernames said:

    I also think you've blurred the line between "API" and "guidelines." An API is something that gives you access to certain functionality, where guidelines suggest ways in which you should use that functionality.

    All APIs have contracts. Guidelines are not the same thing as contracts.

    Here, let's take the Twitter Streaming API:
    The contract states you may only have one stream open per API key at a time
    A guideline is that you should request only the users and keywords you absolutely need

    BOTH of those things are unenforced by Twitter's back-end. (Well, the first occasionally is. But anyway.) The difference is that the first is a HARD AND FAST LAW. You MAY NOT violate that law, the punishment is Twitter can block your API key. The second is a suggestion, more like, "hey our back-end can handle a lot of shit, but it would be nice if you didn't ask for more than you really need."

    Again, if you don't understand these very basic, very easy concepts you should not be writing software. At all.

    @too_many_usernames said:

    Yes, this is because (as boomzilla said) Windows is general-purpose and does allow you the freedom to put your files anywhere. It seems like this is a subset of the much more general "with much freedom comes much responsibility" discussion: Windows gives programmers the freedom to write their installers however they want, but this means they are then responsible for the responses their decisions elicit.  I just don't think it makes them "wrong" for making a particular choice.

    Even ignoring the fact that Fraps is violating the API contract, they're still selling a broken product.

    Now, it just so happens to work on your computer, but thats not by-design, that's just because the Fraps developers were lucky that you didn't happen to change the defaults. In other words, it only works at all by accident. THIS IS NOT GOOD SOFTWARE.

    The entire POINT of following the API contract is that you end up with software that's correct, even for configurations you can't test-- for example, Fraps is a small business, maybe they can't afford to set up a Active Directory environment to test their product in. If they followed the contract, it would work in Active Directory anyway. They get all kinds of QA benefit for free, just by following the contract. If Windows adds an accessibility feature it didn't have before-- for example when Windows 7 added universal speech recognition-- programs that followed the contract automatically benefited from it, and they didn't have to do squat.

    @too_many_usernames said:

    If a program installer chooses to install a program somewhere other than
    recommended by a guideline but still allowed by the API, I don't think
    it's "wrong."

    Yeah, you're objectively wrong on that point. Still.

    @too_many_usernames said:

    Annoying? Perhaps. Wrong? I cannot support that view.

    That's fine-- if you change oil for a living. If you're writing software and don't understand these basic concepts you are a shitty engineer who is creating a shitty product.



  • @ekolis said:

    Why, may I ask, is it the job of the OS to prevent the user from installing files wherever he damn well pleases? A Windows PC is not an iPod...
     

    It's not the job of the OS to "prevent the user from installing files wherever he pleases." The (well, one) job of the OS is to control the reading and writing of files.  If one OS chooses to limit where applications can be installed, so be it.  If a different OS doesn't have this restriction, good for it!

    @blakeyrat said:

    objectively wrong on that point
     

    I'm willing to admit that you're correct, provided you give me the objective metric by which to measure this.  Perhaps I'm just using an different (or even incorrect) metric.  (I know this sounds facetious, but I'm really curious the measure you're using, because it seems we have some kind of denotation/connotation issue going on here.)

    The denotation/connotation issue also applies to the "API" discussion as well - I think we mean different things when we use that acronym.  I'll give my definiton so you at least know the assumptions I have in my statements: The API is the software interface (header files, calling conventions, etc.) used by one piece of software to access functionality provided by another piece of software.  I do not include things like your Twitter example of "If you create more than one stream, we'll take action." as a part of the API, because it's not part of the software interface.  It's something "larger" than the API.  That's what I meant by "guidelines," and I would be willing to even include that in a type of "contract."  That is, "as long as you obey this rule, you're fine."  Perhaps the meaning I apply to "API" is antiquated.



  • @ekolis said:

    Why, may I ask, is it the job of the OS to prevent the user from installing files wherever he damn well pleases? A Windows PC is not an iPod...

    If you are not root on the OS, it may prevent you from going where you want to go.

    On down side, if you are root then it is like spiderman. you have freedom to do what you want.
    Also remember - with great freedom come great responsibility.



  • @ekolis said:

    Why, may I ask, is it the job of the OS to prevent the user from installing files wherever he damn well pleases? A Windows PC is not an iPod...

    Nothing should prevent the user. The argument is that something should (or should not, depending on side of argument) prevent the installer program from not-complying with the user's choice.



  • @too_many_usernames said:

    Not quite: I was trying to say that just because something is a standard doesn't necessarily mean it's "correct."  So breaking the standard doesn't necessarily mean you're "wrong" even though it does mean "you're not in conformance with the standard."  Granted, in the case where you base "correct" and "wrong" solely on the standard then the statement holds, but I was being more philosophical.
     

    In an environment where everything is malleable, is there any meaningful difference between "convenient" and "correct" ?

    @too_many_usernames said:

    If the default is "wrong" but allows you to change it, that seems like a minor annoyance to me.

    It's a symptom of lazy goddamn programmers.

     



  • @dhromed said:

    It's a symptom of lazy goddamn programmers.
     

    Definitely ignorant ones - those that can't (or won't) adhere to standards.

    Hmm.. does ignorance - as in "determined NOT to research", rather than "blissfully unaware" - stem from laziness, an unwillingness to put in the effort to discover and become better informed? Or from arrogance that current knowledge should suffice? Sure Dunning and Kruger have some insight on this, but I can't be bothered to look it up.



  •  Willful ignorance, I guess, yeah.


  • ♿ (Parody)

    @Cassidy said:

    Hmm.. does ignorance - as in "determined NOT to research", rather than "blissfully unaware" - stem from laziness, an unwillingness to put in the effort to discover and become better informed? Or from arrogance that current knowledge should suffice?

    There's certainly some of all of that going on. But I think it's also an issue of how much there is to know and how to learn about it. That is, how to detect the unknown unknowns. It's especially hard for things like this that are probably considered minor issues (at least until you come across a crazy configuration like blakey's frog overlords prefer and it won't work at all) as opposed to the core issues addressed by the application.



  • @boomzilla said:

    It's especially hard for things like this that are probably considered minor issues (at least until you come across a crazy configuration like blakey's frog overlords prefer and it won't work at all) as opposed to the core issues addressed by the application.

    We have different definitions of "minor issues" I guess. IMO, any problem that prevents the program from running at all on another person's system is not a "minor issue". And it's definitely not a "minor" if your program actually causes a denial of service attack on other software due to the issue-- that's malware behavior.

    Also, my work computer's configuration isn't really all that crazy at all. All of the named folders that needed lots of space had it-- the only named folder that was space-constrained was Roaming AppData. And even it had 200 MB free, when any given app shouldn't be putting more than, say, a single MB or so in there.

    I don't see any way you can classify that bug in Aptana Studio as "minor." And the only reason the same bug in NetBeans allows the program to work is that NetBeans just happens to write less stuff to %home%-- meaning it works by accident, not by design.



  • My edit timeout expired, I just wanted to add that for my last post I kind of misread what Boomzilla said. I think he means, "they consider it a minor issue UNTIL they encounter a system for which it's a fatal issue."

    Of course the whole point of following the contract in the first place is that you'll never encounter a machine for which it's a fatal issue-- because the OS itself follows the same damn contract (just from the other side).

    THERE WAS ALSO A TYPO IN THE THIRD SENTENCE get on me pedantic dickweeds!


  • ♿ (Parody)

    @blakeyrat said:

    We have different definitions of "minor issues" I guess. IMO, any problem that prevents the program from running at all on another person's system is not a "minor issue". And it's definitely not a "minor" if your program actually causes a denial of service attack on other software due to the issue-- that's malware behavior.

    I was really speaking from the POV of the developer. It's minor as far as he knows, because he's not aware that it's a real problem. I've found lots of bugs where I knew about them, but didn't comprehend the full scope of its effects, since you didn't always see the most catastrophic of them every time.

    And it may also be minor in the sense that it affects a minute portion of his user base. Your show stopper may be a non-issue to most other people, and in the relative ranking of issues, this may easily be quite minor from a project perspective. Before you imagine that I said something other than what I said, I'm not saying that it isn't wrong or that it can't be a show stopper or that it shouldn't be fixed.

    Unfortunately, stuff like installers get classified like documentation: stuff no one wants to do and that's often an afterthought.

    I think your classification of "malware behavior" is misleading and not very useful here. Malware can also cause things to crash or you to lose data or expose a security hole. So does lots of other buggy software. I'd probably use a word like dangerous or risky or shitty.

    @blakeyrat said:

    Also, my work computer's configuration isn't really all that crazy at all. All of the named folders that needed lots of space had it-- the only named folder that was space-constrained was Roaming AppData. And even it had 200 MB free, when any given app shouldn't be putting more than, say, a single MB or so in there.

    OK, maybe "crazy" was me being a little colorful with language. I think "rare" would be a better descriptor. I recall a discussion about how common it was to set up roaming profiles like that, and I think we all agreed that it was pretty uncommon.

    @blakeyrat said:

    I don't see any way you can classify that bug in Aptana Studio as "minor." And the only reason the same bug in NetBeans allows the program to work is that NetBeans just happens to write less stuff to %home%-- meaning it works by accident, not by design.

    Maybe we're confusing different categories (which lots of bug trackers do, too IMO). Obviously, its severity is pretty critical, since it prevents you from using it. But the priority as far as fixing it could be pretty low. So, I fully agree with you about the severity of it.


  • ♿ (Parody)

    @blakeyrat said:

    My edit timeout expired, I just wanted to add that for my last post I kind of misread what Boomzilla said. I think he means, "they consider it a minor issue UNTIL they encounter a system for which it's a fatal issue."

    Yes, that's a lot of what I was trying to say. I posted a reply before I saw this.

    @blakeyrat said:

    Of course the whole point of following the contract in the first place is that you'll never encounter a machine for which it's a fatal issue-- because the OS itself follows the same damn contract (just from the other side).

    I fully agree.



  • @boomzilla said:

    I think your classification of "malware behavior" is misleading and not very useful here.

    A reasonable anti-virus program (or the OS itself for that matter) wouldn't be able to tell Aptana's behavior apart from a malware program's behavior. That makes it just as bad as malware.

    We'll just have to agree to disagree on that point.

    I think it's worth noting that I normally focus on bugs and WTFs that have to do with the interface because, by and large, developers have the whole API contract all worked-out already. Coming across programs like Fraps (installs in the wrong directory), or Notepad++ (has broken pull-down menus) is actually pretty rare. And even software that does weird stuff, like Steam changing the permissions of its own installation folder, at least the weird stuff is justifiable and solves a problem the OS maker didn't anticipate.

    But obviously a bug that violates the API contract is a much, much, much more severe bug than the normal stuff I blog about.


  • ♿ (Parody)

    @blakeyrat said:

    @boomzilla said:
    I think your classification of "malware behavior" is misleading and not very useful here.

    A reasonable anti-virus program (or the OS itself for that matter) wouldn't be able to tell Aptana's behavior apart from a malware program's behavior. That makes it just as bad as malware.

    We'll just have to agree to disagree on that point.

    OK, just to backtrack a little, the DOS was caused by software writing data that went over the quota for the directory, right? So downloading something to that folder would flag your browser as malware. Saving a document to put you over the quota would cause whatever application you were using to be flagged as malware. Wouldn't it (assuming you're correct about this behavior)?

    Do any anti-virus programs flag this sort of thing*? This doesn't strike me as the sort of behavior that would send up a red flag. It's not trying to change any system files, or anything that obviously belongs to some other application or system, or requires elevated privileges, or makes an unreasonable number of requests like opening up too many file descriptors or sockets.

    Applying the term malware to this just seems to take most of the meaning out of the word to me. "Malware" has always included a connotation of intent, which was obviously absent in the case of Aptana.

    * Weaseling out with the "reasonable" qualifier and saying that there are no reasonable anti-virus programs is the same as admitting that you're wrong.



  • I'm not fucking responding to that because if I did you would, like you have so many times in the past, just ask pointless follow-up questions for WEEKS and annoy the fuck out of me. I'm not as stupid as you apparently think I am; I know exactly what you're doing here.


  • ♿ (Parody)

    @blakeyrat said:

    I'm not fucking responding to that because if I did you would, like you have so many times in the past, just ask pointless follow-up questions for WEEKS and annoy the fuck out of me. I'm not as stupid as you apparently think I am; I know exactly what you're doing here.

    WTF? You mean discussing something? Asking questions because you wrote something that didn't make sense? Asking you to defend your apparently nonsensical position? Misunderstanding what you meant because what you wrote was something else?



  • @boomzilla said:

    ...just to backtrack a little, the DOS was caused by software writing data that went over the quota for the directory, right?
     

    This touches on my comments from earlier: how the heck can a program write more data than its quota gives it?  Does the OS really have no control over the quota?  Or was the quota some kind of "gentleman's agreement API Contract"? (Yes, I admit it, this last bit is intentionally goading).



  • ♿ (Parody)

    @too_many_usernames said:

    @boomzilla said:
    ...just to backtrack a little, the DOS was caused by software writing data that went over the quota for the directory, right?

    This touches on my comments from earlier: how the heck can a program write more data than its quota gives it?  Does the OS really have no control over the quota?  Or was the quota some kind of "gentleman's agreement API Contract"? (Yes, I admit it, this last bit is intentionally goading).

    Uh, I think it just fills up the space and subsequent writes failed. However, other things relied on being able to write there, and so would start crashing or failing or whatever.

    An interesting question* might be how an application could detect such a thing. My understanding is that the directory could look just like a normal directory. So, perhaps you had checked for open space on the C: drive, found plenty there, but wandered into some directory that was so limited. Is it necessary to check each directory? A way to find all specially treated directories like this in advance?

    I'm assuming that this sort of policy could be applied in other places, which may be incorrect, and that there is a special policy specifically to limit the roaming profile directory, and it's the app's responsibility to know where it's writing.

    * This gets off the point of using the proper directories for the proper things, and addresses a more general point. If this is too much to handle, please skip.



  • @too_many_usernames said:

    This touches on my comments from earlier: how the heck can a program write more data than its quota gives it? Does the OS really have no control over the quota? Or was the quota some kind of "gentleman's agreement API Contract"?

    Just FYI, as a little hint: Windows is a multitasking OS. That means, shock and surprise, maybe Aptana Studio ain't the only process that needs space in Roaming AppData?

    @too_many_usernames said:

    (Yes, I admit it, this last bit is intentionally goading).

    Yeah, well, I mind less for two reasons:

    1) Your "intentional goading" here makes you look like a total retard, and I think you're kind of a total retard, so it fits

    2) You don't have Boomzilla's history



  • @boomzilla said:

    Uh, I think it just fills up the space and subsequent writes failed. However, other things relied on being able to write there, and so would start crashing or failing or whatever.
     

    Ah, ok - yes, that's definitely a problem with the application then; applications should be able to gracefully handle inability to write data (or if all its expected data is incomplete).

    That is probably one of the tough problems in OS development actually - do you provide a function to reserve a guaranteed amount of space for a write? Do you let apps check for space in a specific location or only on a (disk) global basis?

    I'll even agree with blakey and say that if the app checks for free space in the wrong location, that's also an application problem.


  • Discourse touched me in a no-no place

    @tOmcOlins said:

    It's the year 2012. Why the hell are you still trying to install yourself to C:\python32 ?
    Given all the crap on this thread, are you complaining about the default path that is supplied in the installer? You're complaining about the default path in the installer?



    To belabour the point - they give you a path (which you're free to change,) and you're complaining about the default they give you?


  • ♿ (Parody)

    @PJH said:

    @tOmcOlins said:
    It's the year 2012. Why the hell are you still trying to install yourself to C:\python32 ?
    Given all the crap on this thread, are you complaining about the default path that is supplied in the installer? You're complaining about the default path in the installer?



    To belabour the point - they give you a path (which you're free to change,) and you're complaining about the default they give you?

    No, I think it's clear that TRWTF is python32. C'mon, get with the friggen times and upgrade to 64 bits already!



  • @PJH said:

    @tOmcOlins said:
    It's the year 2012. Why the hell are you still trying to install yourself to C:\python32 ?
    Given all the crap on this thread, are you complaining about the default path that is supplied in the installer? You're complaining about the default path in the installer?



    To belabour the point - they give you a path (which you're free to change,) and you're complaining about the default they give you?

    I had to look for belabour.

    belabour US, belabor [bɪˈleɪbə]

    vb (tr)

    1. to beat severely; thrash

    2. to attack verbally; criticize harshly

    3. an obsolete word for labour



  • @PJH said:

    Given all the crap on this thread,
     

    Hey, it's all at least reasonably civil* discussion, loosely related to the original topic.

     

    *In the grand scheme of internet discussion boards.



  • @PJH said:

    @tOmcOlins said:
    It's the year 2012. Why the hell are you still trying to install yourself to C:\python32 ?
    Given all the crap on this thread, are you complaining about the *default* path that is supplied in the installer? You're complaining about the default path in the installer?

    To belabour the point - they give you a path (which you're free to change,) and you're complaining about the default they give you?

    Yes. The installer should use sensible (and correct) defaults. It should do the right thing without me having to tell it to. There's no reason any app written after 1995 should try to install itself to C:\

    @boomzilla said:

    No, I think it's clear that TRWTF is python32. C'mon, get with the friggen times and upgrade to 64 bits already!

    Not sure if trolling, joking, or...



  • @boomzilla said:

    @Nagesh said:
    @tOmcOlins said:
    It's the year 2012. Why the hell are you still trying to install yourself to C:\python32 ?

    Real wtf is using Python on windows. Be a man and use the Mac Book

    How do you use the Mac Book on Windows?

    I would like to try Macbook for Windows.  Where can I get this?

     



  • @PJH said:

    @tOmcOlins said:
    It's the year 2012. Why the hell are you still trying to install yourself to C:\python32 ?
    Given all the crap on this thread, are you complaining about the *default* path that is supplied in the installer? You're complaining about the default path in the installer?
    Well, yeah.  That's sort of the whole point of this thread.  Sure, in the grand scheme of things it's pretty minor due to the fact that you can change it to whatever you want, but, seriously, WTF.  Does the person who created that installer really not know that there has been a widely used standard for this sort of thing for  . . . what . . .almost 17 years now?



  • You could try iTunes or EA's DLC, as recommended elsewhere on this forum...



  • @tOmcOlins said:

    It's the year 2012. Why the hell are you still trying to install yourself to C:\python32 ?
    When I first saw the title of this thread I expected it to be something like:

     

    Dear Python,

     

    Fuck You.

     

    Sincerely,

    Everyone



  • @El_Heffe said:

    Dear Python,

    Fuck You.

    Sincerely,

    Everyone

    I should have been a pedantic dickweek and said "Dear Python Installer,". I like tolerate Python itself because it's probably the least wtf-filled scripting language I've used.



  • @tOmcOlins said:

    @El_Heffe said:

    Dear Python,

    Fuck You.

    Sincerely,

    Everyone

    I should have been a pedantic dickweek and said "Dear Python Installer,". I like tolerate Python itself because it's probably the least wtf-filled scripting language I've used.

    Every scripting language has it own set of short-comings, but that does not mean we should stop using them.
    They are very useful for various purposes.
    They are fun to learn
    They are easy to learn.
    Productivity goes up



  • @tOmcOlins said:

    I should have been a pedantic dickweek and said "Dear Python Installer,". I like tolerate Python itself because it's probably the least wtf-filled scripting language I've used.


    Also it has a very nice community of users. At least here in Argentina it rocks. And the PyCamp too, I hope to attend this year.


  • @tOmcOlins said:

    I like tolerate Python itself because it's probably the least wtf-filled scripting language I've used.

    I find python intentionally disallowing tail call optimisation to be a mega-wtf.
    Especially given that python supports features from functional languages, where TCO is normally rampant and useful.



  • @boomzilla said:

    Unfortunately, stuff like installers get classified like documentation: stuff no one wants to do and that's often an afterthought.


    Which is doubly a problem because installers have plenty of idiosyncrasies which take time to learn, and Microsoft's unit tests for installers (ICE) are slightly buggy, so you sometimes have to choose between doing it right or getting warnings.



  • @Nagesh said:

    I had to look for belabour.

    belabour US, belabor [bɪˈleɪbə]

    vb (tr)

    1. to beat severely; thrash

    2. to attack verbally; criticize harshly

    3. an obsolete word for labour

     

     

    In casual discourse, belabor usually means "to beat a dead horse with."



Log in to reply