I have to say, if we ignore Firefox/Linux/other ridiculously huge codes, the people having hours of compile time must be doing something wrong. The 30 000 line Fortran code I'm working on now takes
$ time make clean optim_ifort ...
make clean optim_ifort 50.16s user 0.86s system 98% cpu 51.731 total
in serial, and it's roughly 3x faster for the parallell cmake build. This is with -O3 and inter-procedural optimizations turned on, generating AVX-tuned code. If I have only edited a file or two, the compilation takes less time than a sip of coffee.
Q0. What is the Single UNIX Specification?
The Single UNIX Specification is a set of open, consensus specifications that define the requirements for a conformant UNIX system. The standardized programming environment provides a broad-based functional set of interfaces to support the porting of existing UNIX applications and the development of new applications. The environment also supports a rich set of tools for application development.
Could you please explain to me again how failing to include a C compiler fits with the above statement? I realize that OSX has gotten the stamp of approval (TM) of the Open-group, and that it is technically a UNIX. But it doesn't fit with the common expectation of what a UNIX does.
It's not a UNIX if it doesn't ship with a C compiler. End of story. I mean, you can take a motorcycle and add a roof and some gyro-stabilizing stuff and have it certified by the NTSB/whatever as a car, but it doesn't meet people's expectation of what a car is, and that's the only definition that matters in the real world.
Add in the mess that is ports and the hours you have to spend to get a decent environment for almost any programming language, and it's pretty far off fitting my definition of a UNIX at least. Let's see, on my computer, the time it takes to install python 3, including downloading and me answering "yes, I want to install it":
# time pacman -S python ...
pacman -S python 2.19s user 0.16s system 51% cpu 4.539 total
Nowadays we have buttons to tap, incrementing the count by 1 each time. Tap 50 times to set the minute display, and if you go too far you have to go all the way around again.
We do? That sounds horrible. Just about anything I've ever used that has you set large sets of numbers just lets you type them directly.
I think that GP isn't talking about smartphones/etc here, but rather stupid idiot shit like microwaves, alarm clocks and "induction"* oven tops with touch controls. I am now in the process of buying a new oven top for the place we just moved in to, not because the old one is broken, but because the user interface is mind-numbingly stupid. It looks near -impossible to find an induction oven with good-ole fashioned physical switches. Grrr.
* they should really be called hysteresis ovens, since they produce > 90% of the heat by ferromagnetic hysteresis. This has always annoyed me. If it were actually an induction oven, it would work equally well with non-ferromagnetic pots (i.e. aluminium).
(noticed what you actually pay for groceries vs the official inflation rate for example?)
Amen. The CPI has risen 30% here in the past 15 years, while the price of milk has gone up 70%. Gasoline costs have more than doubled, electricity costs have almost tripled, housing costs have more than tripled. Go figure.
The inbuilt PDF viewer is indeed horrible; the only one I've seen that is worse is the American Chemical Society's "ACS ActiveView PDF" thingy that makes my 4-core i7 Ivy Bridge crawl to a halt.
But the real statistic is when you show the last 12 month.
Nope, the real statistic is when you shown the last 5 years. Then you see that after Win8 was released, WinXP usage has gone from "declining" to "stabilizing". That, I think, is saying something.
I guess the major difference in our approaches is the amount of math required. My meeting notes frequently have equations, or at the very least some greek symbols thrown around, so I always do them on paper during the meeting and then type them up afterwards.
Yeah, why can't we go back to this standard? Barrel plugs for power, micro usb for emergency power and for data. Then you could keep your 500 mA charge over standard micro usb 2.0, and get your 2 A fast charge over a barrel plug. The final version of the Nokia barrel plugs were thinner than micro usb, so it wouldn't take up much space at all on the phone. Even Apple could cooperate with that, as long as they got to keep their proprietary data cable.
Sure, diagrams are nice when describing the structure of something. I usually draw diagrams with pencil and paper in meetings, if the result is sufficient I would just scan it. If not, I would re-create it in diagramo (diagramo.com) which is a free, open-source HTML5 diagram editor. You just click-n-drag to create the diagram, then "right click-> save as" to get a PNG file.
If printing is done often, you provide a CSS sheet for printing. This is fairly trivial, and it can be done once and used on all documents. You can even google and download one.
And I would say HTML has almost as good support for diagrams/charts/whatever as MS Word. In Word, like in HTML, you do all these things in an external tool (called Excel), the only convenience is that it's inserted "interactively" and updates when you update the data (until the one time where it mysteriously doesn't and you waste half a day figuring out why).
I would also say, if your meeting summaries often/always require diagrams and charts, you're doing something wrong.
I realize that you're probably talking about the "average" user, but for someone like myself who has spent some time writing LaTeX, that sentence isn't just easily readable, it took me about 3 seconds to realize that it is stupidly written: \emph and \textit do the same thing unless the user has redefined emph to mean something other than italic. For this reason, you really shouldn't use \textit for markup, you'll break everything (e.g. citation styles) if you want to redefine it.
And by the way, all this mention of DVI makes me suspect you haven't used LaTeX in about 10 years. Everyone (sane) uses PDF these days via pdflatex.
I think plain HTML is the absolute best choice for this. The document will look the same everywhere, and it can be opened quickly(!) on any device (phone, tablet, etc). Furthermore, it supports hyperlinks both internally and to other documents/webpages. You need to find a good editor, which is not that easy, but otherwise it works very well in my experience.
The reason to use Word is actually just because it's convenient and supports change tracking and reviewing, so it's convenient for communicating copy edits and dev edits to the author, and allows the author to accept or reject changes proposed by the copy editor.
While it's true that Word has a "change tracking and reviewing" support, I've yet to encounter anyone using this in a sane manner. Having used both LaTeX with git/hg diffs and Word with "track changes", I'll take the former plus a root canal over the latter any day.
once you change something on page 1, things cascade down to page 100 - pictures and tables jump onto different pages, blank areas show up where none were before...
Sounds to me like you just described how MS Word functions.
I should also point out: LaTeX was designed to solve one problem, and solve it well: scientific author(s) wants to write a paper to be published in a journal. He does not need to produce a camera-ready copy; in fact, the journal probably does not want him to be able to do so. He only needs to produce a document that has all equations, figures and references looking good enough that other scientist can peer-review the paper and decide how to proceed from there. LaTeX is supposed to get the design out of the way and let the writer focus on the science. There are extensions that enable document design, but it gets very hard once you stray away from the article/report/thesis/book formats.
If you are interested in documents that are constructed with design as a focus from the ground up, I suggest that you look at ConTeXt; it resembles LaTeX, but was created at a professional publishing house to enable the production of quality documents.
And another scene shows a geometry teacher reassigning students' seating assignments based on their 'character strengths', moving a green-coded female student ('actively participates: 98%') next to a red-and-yellow coded boy ('shows enthusiasm: 67%').
-- And kids with vision problems are also moved to the front of the class. What the point?
Personally, one of the things I hated the most in school was being used like this to "help the teacher manage the unruly ones". Way to go, teacher, rewarding the students who do a good job by (implicitly) giving them a crappy job.
Statistics from 1 data point is a bad idea. If out of the first 10,000 Tesla S owners there were 3 terrible drivers who crashed&burned in the first few months, you would be concluding that the Tesla S has an energy-source fire rate which is almost 25x that of normal cars. Because of 3 idiots who biased your stats. We'll have to wait several years before we can say anything conclusive about an event which is this rare.
In Norway, we have a "level" system that is used in academia throughout the country. It is used for evaluating researchers and research groups when it comes to employment, tenure, funding etc. Your "point score" is summed up, 2 points for publication in a "level 2" journal, 1 point for "level 1".
A journal is either "level 2", "level 1" or "level 0". "level 2" is a selection of top journals from each field in science, 2000 in total (for all of science, from computational physics to the sociology of music). "level 1" means the remaining serious peer-reviewed journals. "level 0" either means "bullshit journal" or "journal that was founded just last year".
Researchers may nominate journals for a change in status, e.g. 2->1, 0->, etc. The decisions are made by a government-appointet body on a yearly basis. It's nowhere near perfect, but it's a lot better than nothing.
Slashdot Mirror
Free market competition has worked out how to put those legal requirements into a car without it looking like a 1970's Volvo.
I beg your pardon? You mean like this?
I have to say, if we ignore Firefox/Linux/other ridiculously huge codes, the people having hours of compile time must be doing something wrong. The 30 000 line Fortran code I'm working on now takes
...
$ time make clean optim_ifort
make clean optim_ifort 50.16s user 0.86s system 98% cpu 51.731 total
in serial, and it's roughly 3x faster for the parallell cmake build. This is with -O3 and inter-procedural optimizations turned on, generating AVX-tuned code. If I have only edited a file or two, the compilation takes less time than a sip of coffee.
And adding a compiler is just an install able package, technically a couple packages. For free, from the vendor.
Have you actually tried installing Xcode? A several-GB install that takes over an hour is not "just a couple packages".
I agree (and have said so several times) that the specification does not require a compiler.
I said python 3. Please RTFC before you reply.
I don't know about C in Solaris, but SunOS 4 did at least include a C compiler. Same with HPUX, AIX and z/OS last I checked.
Q0. What is the Single UNIX Specification?
The Single UNIX Specification is a set of open, consensus specifications that define the requirements for a conformant UNIX system. The standardized programming environment provides a broad-based functional set of interfaces to support the porting of existing UNIX applications and the development of new applications. The environment also supports a rich set of tools for application development.
Could you please explain to me again how failing to include a C compiler fits with the above statement? I realize that OSX has gotten the stamp of approval (TM) of the Open-group, and that it is technically a UNIX. But it doesn't fit with the common expectation of what a UNIX does.
It's not a UNIX if it doesn't ship with a C compiler. End of story. I mean, you can take a motorcycle and add a roof and some gyro-stabilizing stuff and have it certified by the NTSB/whatever as a car, but it doesn't meet people's expectation of what a car is, and that's the only definition that matters in the real world.
...
Add in the mess that is ports and the hours you have to spend to get a decent environment for almost any programming language, and it's pretty far off fitting my definition of a UNIX at least. Let's see, on my computer, the time it takes to install python 3, including downloading and me answering "yes, I want to install it":
# time pacman -S python
pacman -S python 2.19s user 0.16s system 51% cpu 4.539 total
Come back when your "UNIX" does that.
Sure physics can explain chemistry. I challenge you to provide one example of a chemistry phenomenon that physics can not explain.
Nowadays we have buttons to tap, incrementing the count by 1 each time. Tap 50 times to set the minute display, and if you go too far you have to go all the way around again.
We do? That sounds horrible. Just about anything I've ever used that has you set large sets of numbers just lets you type them directly.
I think that GP isn't talking about smartphones/etc here, but rather stupid idiot shit like microwaves, alarm clocks and "induction"* oven tops with touch controls. I am now in the process of buying a new oven top for the place we just moved in to, not because the old one is broken, but because the user interface is mind-numbingly stupid. It looks near -impossible to find an induction oven with good-ole fashioned physical switches. Grrr.
* they should really be called hysteresis ovens, since they produce > 90% of the heat by ferromagnetic hysteresis. This has always annoyed me. If it were actually an induction oven, it would work equally well with non-ferromagnetic pots (i.e. aluminium).
(noticed what you actually pay for groceries vs the official inflation rate for example?)
Amen. The CPI has risen 30% here in the past 15 years, while the price of milk has gone up 70%. Gasoline costs have more than doubled, electricity costs have almost tripled, housing costs have more than tripled. Go figure.
The inbuilt PDF viewer is indeed horrible; the only one I've seen that is worse is the American Chemical Society's "ACS ActiveView PDF" thingy that makes my 4-core i7 Ivy Bridge crawl to a halt.
But the real statistic is when you show the last 12 month.
Nope, the real statistic is when you shown the last 5 years. Then you see that after Win8 was released, WinXP usage has gone from "declining" to "stabilizing". That, I think, is saying something.
I guess the major difference in our approaches is the amount of math required. My meeting notes frequently have equations, or at the very least some greek symbols thrown around, so I always do them on paper during the meeting and then type them up afterwards.
Yeah, why can't we go back to this standard? Barrel plugs for power, micro usb for emergency power and for data. Then you could keep your 500 mA charge over standard micro usb 2.0, and get your 2 A fast charge over a barrel plug. The final version of the Nokia barrel plugs were thinner than micro usb, so it wouldn't take up much space at all on the phone. Even Apple could cooperate with that, as long as they got to keep their proprietary data cable.
It's actually not a passive cable, otherwise it wouldn't be rotation independent. It has an embedded chip that does all kinds of nasty things, including authorizing that the cable is made by Apple (or a licensed third party). Fortunately this chip can be bypassed: http://www.redmondpie.com/apple-lightning-authentication-chip-has-been-bypassed-third-party-lightning-cables-and-docks-on-their-way-video/
Sure, diagrams are nice when describing the structure of something. I usually draw diagrams with pencil and paper in meetings, if the result is sufficient I would just scan it. If not, I would re-create it in diagramo (diagramo.com) which is a free, open-source HTML5 diagram editor. You just click-n-drag to create the diagram, then "right click-> save as" to get a PNG file.
Please mod parent up. First post I've seen in a while that deserves both +1 Informative and +1 Funny.
If printing is done often, you provide a CSS sheet for printing. This is fairly trivial, and it can be done once and used on all documents. You can even google and download one.
And I would say HTML has almost as good support for diagrams/charts/whatever as MS Word. In Word, like in HTML, you do all these things in an external tool (called Excel), the only convenience is that it's inserted "interactively" and updates when you update the data (until the one time where it mysteriously doesn't and you waste half a day figuring out why).
I would also say, if your meeting summaries often/always require diagrams and charts, you're doing something wrong.
\textit{The cake was \emph{huge} for a cup cake}
I realize that you're probably talking about the "average" user, but for someone like myself who has spent some time writing LaTeX, that sentence isn't just easily readable, it took me about 3 seconds to realize that it is stupidly written: \emph and \textit do the same thing unless the user has redefined emph to mean something other than italic. For this reason, you really shouldn't use \textit for markup, you'll break everything (e.g. citation styles) if you want to redefine it.
And by the way, all this mention of DVI makes me suspect you haven't used LaTeX in about 10 years. Everyone (sane) uses PDF these days via pdflatex.
summaries of a meeting
I think plain HTML is the absolute best choice for this. The document will look the same everywhere, and it can be opened quickly(!) on any device (phone, tablet, etc). Furthermore, it supports hyperlinks both internally and to other documents/webpages. You need to find a good editor, which is not that easy, but otherwise it works very well in my experience.
The reason to use Word is actually just because it's convenient and supports change tracking and reviewing, so it's convenient for communicating copy edits and dev edits to the author, and allows the author to accept or reject changes proposed by the copy editor.
While it's true that Word has a "change tracking and reviewing" support, I've yet to encounter anyone using this in a sane manner. Having used both LaTeX with git/hg diffs and Word with "track changes", I'll take the former plus a root canal over the latter any day.
once you change something on page 1, things cascade down to page 100 - pictures and tables jump onto different pages, blank areas show up where none were before...
Sounds to me like you just described how MS Word functions.
I should also point out: LaTeX was designed to solve one problem, and solve it well: scientific author(s) wants to write a paper to be published in a journal. He does not need to produce a camera-ready copy; in fact, the journal probably does not want him to be able to do so. He only needs to produce a document that has all equations, figures and references looking good enough that other scientist can peer-review the paper and decide how to proceed from there. LaTeX is supposed to get the design out of the way and let the writer focus on the science. There are extensions that enable document design, but it gets very hard once you stray away from the article/report/thesis/book formats.
If you are interested in documents that are constructed with design as a focus from the ground up, I suggest that you look at ConTeXt; it resembles LaTeX, but was created at a professional publishing house to enable the production of quality documents.
And another scene shows a geometry teacher reassigning students' seating assignments based on their 'character strengths', moving a green-coded female student ('actively participates: 98%') next to a red-and-yellow coded boy ('shows enthusiasm: 67%').
-- And kids with vision problems are also moved to the front of the class. What the point?
Personally, one of the things I hated the most in school was being used like this to "help the teacher manage the unruly ones". Way to go, teacher, rewarding the students who do a good job by (implicitly) giving them a crappy job.
Statistics from 1 data point is a bad idea. If out of the first 10,000 Tesla S owners there were 3 terrible drivers who crashed&burned in the first few months, you would be concluding that the Tesla S has an energy-source fire rate which is almost 25x that of normal cars. Because of 3 idiots who biased your stats. We'll have to wait several years before we can say anything conclusive about an event which is this rare.
In Norway, we have a "level" system that is used in academia throughout the country. It is used for evaluating researchers and research groups when it comes to employment, tenure, funding etc. Your "point score" is summed up, 2 points for publication in a "level 2" journal, 1 point for "level 1".
A journal is either "level 2", "level 1" or "level 0". "level 2" is a selection of top journals from each field in science, 2000 in total (for all of science, from computational physics to the sociology of music). "level 1" means the remaining serious peer-reviewed journals. "level 0" either means "bullshit journal" or "journal that was founded just last year".
Researchers may nominate journals for a change in status, e.g. 2->1, 0->, etc. The decisions are made by a government-appointet body on a yearly basis. It's nowhere near perfect, but it's a lot better than nothing.
1. Walk around Philadelphia
2. Carry voice recorder in backpack
3. ???
4. PROFIT!!!