3 ms·

1- at this point, you're just arguing semantics. Whether the subsidy takes the form of a loan or an ETF, it's all the same. 2-you've managed 3 designers for an

by sami36 14y ago

1- at this point, you're just arguing semantics. Whether the subsidy takes the form of a loan or an ETF, it's all the same. 2-you've managed 3 designers for an app or a website & you extrapolate on the basis of that experience to ballpark the effort needed to build an Android skin ? I'm done arguing with you. Look, the fact that you've mentioned the 1 Mil $ number leaves me highly doubtful of any (meaningful) real life experience budgeting for software projects. the average corporate cost for a corp project I've worked on is in the double digits (Mil) & we're talking basic internal-facing enterprise apps. We were at, yes, the Nexus 4 is a great device. Can we at least agree about that ?

beagle3 14y ago
> We were at, yes, the Nexus 4 is a great device. Can we at least agree about that ? Yes. However ... 1 - No, you don't get to redefine terms how you want. You're entitled to your own opinion, but not to your own definitions or your own facts. Loss leader is mostly well defined (and no, it's not "7% profit" or whatever it is you asked). Similarly, subsidy is well defined - it is a lower price than cost. And there's a huge difference between an ETF and a loan. There is a very simple test: Get a two year plan at verizon (or at&t). Use phone. Note cost. Get same phone and equivalent plan without commitment (paying for phone up-front, and month-to-month). Use phone exactly the same. Note cost. The first one is always cheaper with at&t and verizon (though not with t-mobile) Do same in Europe, with e.g. vodaphone, three, etc. And note that - up to financing, costs are the same (and you can get comparable financing from a bank). 2 - I actually haven't used any android skin, because everyone I know uses either plain Android or CM (don't have android myself). But from looking it up in e.g. http://us.gizmodo.com/5921220/every-major-android-skin-compared http://us.gizmodo.com/5921220/every-major-android-skin-compa... , it seems that actually my experience (which was a Win95/Win98 client, with tens of thousands of installs, and pervasive skinning, back in 99-2001) is extremely relevant. Please enlighten me, because from what I can tell, an android skin is comparable to an enlightenment/gtk/thunderbird skin, plus 4-5 app customizations (people, calendar, notification, maybe a few more). That can totally be done in under $1M. Actually, under $200K if managed properly. And I'm totally unimpressed by the multimillion cost at enterprises; I know it's there, and I know I've replaced such multimillion dollar projects with hundreds-of-thousands dollar projects that performed better (and were still very lucrative). I've also bought better hardware at 10% of the price than some enterprises I've worked with. They do NOT optimize for cost; they optimize for CYA. I've also helped a friend run a VLSI design project that got working tape-outs of a designed-from-scratch processor, all costs included, for $200K in 2005. Everyone you'll ask will tell you that it's impossible. And they'd all be right in their respective frame of mind -- but would be wrong in the real world (http://www.reghardware.com/2012/05/03/unsung_heroes_of_tech_arm_creators_sophie_wilson_and_steve_furber/print.html http://www.reghardware.com/2012/05/03/unsung_heroes_of_tech_... is enlightening, although our project was not as commercially successful) You might have spent too much time in the enterprise business to realize just how inefficient it is on that front. I probably have more real life experience running and budgeting software (and other) projects than you do: I've done many scales and many environments, software and hardware, not just "corporate and double digit (mil)".

by sami36 14y ago

beagle3 14y ago

> We were at, yes, the Nexus 4 is a great device. Can we at least agree about that ? Yes. However ... 1 - No, you don't get to redefine terms how you want. You're entitled to your own opinion, but not to your own definitions or your own facts. Loss leader is mostly well defined (and no, it's not "7% profit" or whatever it is you asked). Similarly, subsidy is well defined - it is a lower price than cost. And there's a huge difference between an ETF and a loan. There is a very simple test: Get a two year plan at verizon (or at&t). Use phone. Note cost. Get same phone and equivalent plan without commitment (paying for phone up-front, and month-to-month). Use phone exactly the same. Note cost. The first one is always cheaper with at&t and verizon (though not with t-mobile) Do same in Europe, with e.g. vodaphone, three, etc. And note that - up to financing, costs are the same (and you can get comparable financing from a bank). 2 - I actually haven't used any android skin, because everyone I know uses either plain Android or CM (don't have android myself). But from looking it up in e.g. http://us.gizmodo.com/5921220/every-major-android-skin-compared http://us.gizmodo.com/5921220/every-major-android-skin-compa... , it seems that actually my experience (which was a Win95/Win98 client, with tens of thousands of installs, and pervasive skinning, back in 99-2001) is extremely relevant. Please enlighten me, because from what I can tell, an android skin is comparable to an enlightenment/gtk/thunderbird skin, plus 4-5 app customizations (people, calendar, notification, maybe a few more). That can totally be done in under $1M. Actually, under $200K if managed properly. And I'm totally unimpressed by the multimillion cost at enterprises; I know it's there, and I know I've replaced such multimillion dollar projects with hundreds-of-thousands dollar projects that performed better (and were still very lucrative). I've also bought better hardware at 10% of the price than some enterprises I've worked with. They do NOT optimize for cost; they optimize for CYA. I've also helped a friend run a VLSI design project that got working tape-outs of a designed-from-scratch processor, all costs included, for $200K in 2005. Everyone you'll ask will tell you that it's impossible. And they'd all be right in their respective frame of mind -- but would be wrong in the real world (http://www.reghardware.com/2012/05/03/unsung_heroes_of_tech_arm_creators_sophie_wilson_and_steve_furber/print.html http://www.reghardware.com/2012/05/03/unsung_heroes_of_tech_... is enlightening, although our project was not as commercially successful) You might have spent too much time in the enterprise business to realize just how inefficient it is on that front. I probably have more real life experience running and budgeting software (and other) projects than you do: I've done many scales and many environments, software and hardware, not just "corporate and double digit (mil)".