It's nearly 2017, we still cannot into software
-
When I see job ads for programmers nowadays, they are blooming with quests for "data scientists", "big data engineers", and "machine/deep learning experts" like it was once with "Ruby on Rails ninjas" and "Web 2.0 gurus". While I appreciate that Amazon and similar cloud providers enabled us with sufficient CPU cycles to create big neural networks, feed them gobs of data and bend them over so they process cat pictures in funny ways, or present me with contextual ads for the shit I have purchased exactly 30 seconds ago, it irks me that since 1980s, we still:
- Struggle wherever there's a form to fill out because the forms are full of whooshy animations, but validation is fucked up.
- Struggle with data consistency, because data consistency is not webby and scaly enough to be cool. Sometimes, sophisticated processing in the application is involved where it perfectly sufficed to say CASCADE or LEFT OUTER JOIN a few years ago.
- Struggle with character encodings (I thought we already can into Unicode, seriously).
- Develop software that's supposed to work with money and sell around the world, but it is English-only and cannot accept user input with ANY kind of special characters for no good reason (see 1; the field for "name" is validated against [A-Za-z]+ and yours is RenΓ©).
- Produce more and more opinionated frameworks and weirdly limited libraries for one trendy language, so the developers have to dedicate lots of time just to keep up and have marketable skills, while having very little time to solve, you know, actual problems.
- Every large corporation, including government-owned ones, has adopted some mutated and mutilated form of "Agile", whether it contributes to better process or just makes micro-managers happy. Either way, developers lose every single fucking time.
- Produce software that demands you give your soul away into its cloud with "sophisticated algorithms" even before their authors grow enough wit to stop it crashing every other launch.
- Abolish some established and proven software development and release practices because they don't understand the reasoning behind them, only to painfully rediscover them (after probably having sunk a few products in between), give them a flashy name and think they're smart.
- Struggle when getting even 5-year-old laptop hardware to work in a remotely usable way. On each and every system, except probably macOS.
- Cannot squeeze decent battery life from a smartphone.
- Cannot produce a kiosk/vending machine that takes any valid user action sequence into account. I dread self-checkout at most supermarkets that have them (they mostly require an assistant to unfuck them every other customer), I managed to confuse train ticket machines (the fancier ones) and banking terminals. I cannot be having that anymore.
But I'm not yet old and grumpy enough, so I look forward to 2017. Maybe that year is going to produce something useful, usable, and reliable in the world of software.
Or 2018, for that matter.
...Or maybe the events transpire that even after Polska can into the space, software will still suck syphilitic camel's balls. :-(
-
People once said: "If you data can fit on single harddisk, you don't need big data".
I think this still hold true.
-
@cheong People once had 5 MB hard drives.
-
@wft said in It's nearly 2017, we still cannot into software:
Struggle wherever there's a form to fill out because the forms are full of whooshy animations, but validation is fucked up.
Pop quiz. What is the correct format for a Canadian postal code. Is it:
- A9A9A9
- A9A 9A9
- A9A-9A9
If you picked one, you are wrong and need to have your genitals soldered shut for the good of humanity. The correct answer is "All three ways are correct. The computer should only care that those 6 digits are present, and extract the data, regardless of human-readable format".
BUT WAIT! THERE'S MORE!
How do you store that data in the database? If you answered 1) you are correct. If you picked 1) in the "which is correct format" because that's the format your database, then again-- please report to the soldering station. Because shock of all shocks-- if you can validate that the data you want is present, you can extract it!
BUT WAIT! THERE'S EVEN MORE!
What letters are not allowed in the 1st, 3rd or 5th position? If you have an answer, then you're wrong, and you're validating against outdated information. You probably copy/pasta'd a validation regex from 2002 Stack Overflow, and never bothered to check with Canada Post, did you?
OH BUT WAIT! THERE IS EVEN MOOOOOOOOOOOORE!
Let's say for some reason you NEED to restrict the user input based on the seperater character, or letter restrictions based on position. The user enters A1B2W6, which fails your fuctarded regex because W isn't valid in position 5. Do you:
a) Accept the input, then like a fucking gooma give the user a generic error of "invalid data", not even telling them it's the postal code?
b) Accept the input, then like a shit-sucking asshole, give the user a generic error of "invalid postal code", but not tell them why?
c) Make the user enter A1B2W6, A1B-2W6 and A1B 2W6, as they try to guess which goddamn format you use, and each of them fails, and like a cock-stuffing blanket-eater, you just keep telling them "invalid postal code".
d) After they VIEW SOURCE and check the regex, they see what you're doing, and enter A1B2Z6, the have to email customer support to tell them the postal code is actualyl A1B2W6 and that their postal validation rules are fucking retarded?If you answered D the I hope someone sodomizes you with the soldering iron after sterilizing you, just to ensure that even if you think you're going to eventually get pleasure out of anal sex, that possibility is stripped from you and you forever know only terror and pain from any comforting contact.
Also
because we used to call those shit-puckers "Rock Stars" and FUCK THEM AHAHRLKGJLA:KSDJFLKJLDKJFIOPWEJ!!!!!!!
-
@El_Heffe said in It's nearly 2017, we still cannot into software:
@cheong People once had 5 MB hard drives.
I think by the time the term "Big Data" is created as the next bizzword to sell, the maximum harddisk size that money can buy is 4TB.
And I'm not talking about extended harddisk space by creating RAID partitions.
-
@wft said in It's nearly 2017, we still cannot into software:
- Cannot squeeze decent battery life from a smartphone.
There's the problem. If people would realize that they didn't need to run all those silly little micro-transaction games and apps and stuff, they could get a phone that is just a phone, with maybe texting and simple photo messaging capabilities, and the battery would last for several days. The only reason I personally would want to get a "smart" phone is for the GPS capabilities that is built into them.
More technically, the technology used in the phone's radio connection has a huge effect on the time the phone can run on a single charge. GSM phones are constantly emitting and receiving while they listen for their particular identifier to appear, meaning someone is calling or trying to make some other type of connection. CDMA phones only check every once in a while, which means the radio is off for most of the time, which saves the charge on the battery.
As an aside, the frequencies that CDMA uses also have longer range and better material penetration than the frequencies that GSM uses. The connection systems are also quite different: CDMA uses a round-robin queue for all the connections, which means a theoretically unlimited number of connections can be maintained (although each connection could therefore have to wait an eternity before its turn came up again). GSM divides the available frequency bandwidths into separate chunks and then assigns each connection to a dedicated slot. This allows a connection to continuously transmit data, which could often mean "faster" total download/upload speeds, but also means that if the bandwidth is completely assigned, no new connections can be made until an existing one is closed, which means there is a harder limit on the number of supported connections.
I don't really understand the push to get rid of CDMA. It's a big part of why the one telco provider in the USA promises better coverage than their competitors (and they do).
-
@cheong said in It's nearly 2017, we still cannot into software:
"If you data can fit on single harddisk, you don't need big data".
You don't even need to go that far. Most companies who think they have "big data" can fit all their data into main memory on a reasonable server. 64 GiB of RAM are pretty cheap.
-
@asdf said in It's nearly 2017, we still cannot into software:
@cheong said in It's nearly 2017, we still cannot into software:
"If you data can fit on single harddisk, you don't need big data".
You don't even need to go that far. Most companies who think they have "big data" can fit all their data into main memory on a reasonable server. 64 GiB of RAM are pretty cheap.
Holy shit, my steam library is big data?
-
@ben_lubar said in It's nearly 2017, we still cannot into software:
Holy shit, my steam library is big data?
According to companies like @Weng's employer, yes. I don't even want to know how many other companies out there set up Hadoop clusters for calculations that could easily be done twice as fast on a single machine.
-
@cheong Actually, I have been pricing a system with two 8TB HDDs for a friend recently (he needs a video-editing system for indie film work - and it also is going to have two 512GB SSDs, one for the boot drive and the other for fastloading the project files). The only reason it isn't 10TB is because those cost at minimum $450 USD apiece right now, and he can't wait until next spring for the price to drop. Check on Amazon if you don't believe me:
That's a single drive. Ten Terabytes. Holy fuck.
Filed Under: I remember when there wasn't ten terabytes on the entire Internet. Seeing this makes me feel old.
-
It's nearly 2018 and still canno
-
@djls45 said in It's nearly 2017, we still cannot into software:
More technically, the technology used in the phone's radio connection has a huge effect on the time the phone can run on a single charge. GSM phones are constantly emitting and receiving while they listen for their particular identifier to appear, meaning someone is calling or trying to make some other type of connection. CDMA phones only check every once in a while, which means the radio is off for most of the time, which saves the charge on the battery.
Yes, technically. But the power savings of using a given technology over another is nanoscale compared to what the display uses. The display typically uses the same amount of juice in an hour of being on as a typical radio in standby (or equivalent) for a day.
-
@asdf said in It's nearly 2017, we still cannot into software:
@cheong said in It's nearly 2017, we still cannot into software:
"If you data can fit on single harddisk, you don't need big data".
You don't even need to go that far. Most companies who think they have "big data" can fit all their data into main memory on a reasonable server. 64 GiB of RAM are pretty cheap.
Can you donate some cheap RAM? I tried downloading it, but it didn't work!
-
@ScholRLEA said in It's nearly 2017, we still cannot into software:
@cheong Actually, I have been pricing a system with two 8TB HDDs for a friend recently (he needs a video-editing system for indie film work - and it also is going to have two 512GB SSDs, one for the boot drive and the other for fastloading the project files). The only reason it isn't 10TB is because those cost at minimum $450 USD apiece right now, and he can't wait until next spring for the price to drop. Check on Amazon if you don't believe me:
That's a single drive. Ten Terabytes. Holy fuck.
Filed Under: I remember when there wasn't ten terabytes on the entire Internet. Seeing this makes me feel old.
With disk this large, my primary concern would be heat generation and average seek time. Price like USD450 would be little concern for a company that needs data storage of this size.
-
@Lorne-Kates said in It's nearly 2017, we still cannot into software:
Pop quiz. What is the correct format for a Canadian postal code. Is it:
- A9A9A9
- A9A 9A9
- A9A-9A9
If you picked one, you are wrong and need to have your genitals soldered shut for the good of humanity. The correct answer is "All three ways are correct. The computer should only care that those 6 digits are present,
I see only 3. Or 9, if you want to look at it that way. 12 if you include the numbered bulletpoints.
This post was brought to you by the letter 9 and the number A.
-
@cheong said in It's nearly 2017, we still cannot into software:
With disk this large, my primary concern would be heat generation and average seek time.
Why would you expect either of those numbers to be worse than for the same capacity implemented as an array of smaller drives?
My primary concern would be the universe noticing WTF is going on and putting a stop to it. That's some scary areal data density right there; the mechanical tolerances inside those machines must be utterly insanely tight.
-
@flabdablet said in It's nearly 2017, we still cannot into software:
@cheong said in It's nearly 2017, we still cannot into software:
With disk this large, my primary concern would be heat generation and average seek time.
Why would you expect either of those numbers to be worse than for the same capacity implemented as an array of smaller drives?
My primary concern would be the universe noticing WTF is going on and putting a stop to it. That's some scary areal data density right there; the mechanical tolerances inside those machines must be utterly insanely tight.
You will need to concern because companies need drive of this size will eventally install disk array of similar drives.
From their data sheet, the 10TB model have lower power consumption than 8TB one, and by the form factor, 10TB model most likely employs one more disk plate on it, so even though BTU for the harddisk is not listed in the data sheet, you can almost assume the heat generation will be lower than that of 8TB model, and likely will have faster seek time (because of more heads). Ignoring the price factor I would always recommand the 10TB one over 8TB when judging solely with data from data sheet.
Regarding concern on mechanical tolerances, since they dare to put a 5 year warranty on it, I'd expect it to be okay. Or they won't extend the warranty beyond normal 2-3 years that you see from domestic ones. Afterall it'll harm the company's ability to profit if each of such expensive product would at least need one replacement before warranty expires.
-
@cheong According to these specs, the 8TB model has 6 platters and 12 heads, and the 10TB has 7 platters and 14 heads. 17% more surface isn't enough to account for 25% more capacity: the 10TB drive runs 867Gb/in2 compared to the 8TB's manifestly inadequate 732.
Say that out loud. Eight hundred and sixty seven billion bits per square inch.
Track density is given as 386 KPTI (average). That's three hundred and eighty-six tracks per thou.
Maximum bit density along the tracks is 2230 bits per thou.
The idea that mechanical precision as barking mad as this is still the cheapest way to store bulk bits just boggles my grizzled old mind.
-
@ScholRLEA said in It's nearly 2017, we still cannot into software:
That's a single drive. Ten Terabytes. Holy fuck.
Wanna hear a fun fact? DNA stores around 1 PB per cubic millimeter.
So that's the lower bound to how densely we can store things.
-
@wft You nailed it. Developing trendy new stuff is fun and cool, actually fixing those 20,000 bugs you have in your existing Java/C++ system is boring and expensive. And it takes so long, who knows how the computing world will be 5 or 10 years from now? Will people still be buying software for [OS of choice]?
The problem is that software is hard, so even if you actually follow good development practices (which, as we all know, nobody does) it still takes years of bug-fixing for it to mature and become decent, and with the collective obsession with remaking everything from scratch every 5 years (and customers being generally unwilling to pay for stuff, etc.) there's just no opportunity to do that.
-
@anonymous234 said in It's nearly 2017, we still cannot into software:
@ScholRLEA said in It's nearly 2017, we still cannot into software:
That's a single drive. Ten Terabytes. Holy fuck.
Wanna hear a fun fact? DNA stores around 1 PB per cubic millimeter.
So that's the lower bound to how densely we can store things.
But I heard the limit for magnetic media is much lower (both limit of storage of data and reading precision).
That's why they're always seeking for other kind of material / method for storage.
-
I was gonna post a rebuke last night, but nodebb was bugging out.
-
@Lorne-Kates said in It's nearly 2017, we still cannot into software:
What letters are not allowed in the 1st, 3rd or 5th position?
Whatever letters the Canada Post's address validation API refuses?
-
@asdf said in It's nearly 2017, we still cannot into software:
I don't even want to know how many other companies out there set up Hadoop clusters for calculations that could easily be done twice as fast on a single machine.
This, so much!
My company is dealing with a lot of data. Like, really a lot. To give you a hint, we regularly get emails about storage systems being too full and asking us to do some cleaning, the last one was saying, I quote, "In the last 7 days, 2.7Pb of data has been deleted". Yep, removing 3 PB (3000 TB!) is not enough to significantly get the usage stats down.
And, my point is, we don't fucking yak about big data and Hadoop and trendy stuff. I can reasonably say that we do big data, but like with many things, there are those that do and those that talk about it.
-
@remi out of interest, what stack do you use to process and store that data?
-
@ScholRLEA said in It's nearly 2017, we still cannot into software:
The only reason it isn't 10TB is because those cost at minimum $450 USD apiece right now, and he can't wait until next spring for the price to drop.
When I built a couple of RAID arrays last year, the acceptable price point was at about the 6TB mark; it was the 8TB drives that were stupid expensive.
Moore'sKryder's Law is awesome.
-
@wft That is just the large difference between what we can do with the best of our knowledge, and what is usually done because of it's cost/benefit.
-
@djls45 said in It's nearly 2017, we still cannot into software:
If people would realize that they didn't need to run all those silly little micro-transaction games and apps and stuff, they could get a phone that is just a phone, with maybe texting and simple photo messaging capabilities, and the battery would last for several days.
If I could trust a any local phone operator to send a SMS
-
@wft I probably can't say much, mostly because I'm just a lowly dev working on a system that is not actually related to much of this (and also because I'm not sure how soon that would get into confidential info...), but essentially, we have huge Linux clusters with lots and lots of layers of custom code for parallel processing. And some computations run for weeks at a time, keeping 100's of nodes busy full time...
The key here is that we've been in this business for about as long as the company existed (well, actually, the other way round, and without bragging I think that this company is one of those that invented this particular business...). So we've had huge mainframes in the '70s and '80s, moved to clusters in the '90s (various unices before ending on Linux), and we had to more or less invent everything as we went along (that was ages before I joined, in any case). Data in these days was not obviously on disk but on magnetic tapes so we had complex interfaces with storage robots handling the tapes to try and optimize that, and that is still felt in the storage architecture where there are several arbitrary layers based on the latency time to access the data (which is less relevant today but the distinction still exists).
We tried about each and every new technology when it came along, so we have bits that are just Linux nodes talking to each other through the network, we have custom Linux kernel modules, we have Hadoop clusters, we have GPU clusters...
-
@wft said in It's nearly 2017, we still cannot into software:
But I'm not yet old and grumpy enough, so I look forward to 2017. Maybe that year is going to produce something useful, usable, and reliable in the world of software.
-
-
@Jaloopa said in It's nearly 2017, we still cannot into software:
Inexpensive Redundant RAID arrays of Disks?
HTH.
-
This post is deleted!
-
@remi Probably that, I was interested mostly whether you tend more to invent your own solutions vs using 3rd party stuff. I reckon that as soon as the amount of data grows over some threshold, it starts to matter a huge lot what kind of data it actually is, and it starts to make sense to custom tailor software for just that kind of data (blue elephant guns FTW!), as opposed to generic solutions.
-
@Lorne-Kates said in It's nearly 2017, we still cannot into software:
@wft said in It's nearly 2017, we still cannot into software:
Struggle wherever there's a form to fill out because the forms are full of whooshy animations, but validation is fucked up.
Pop quiz. What is the correct format for a Canadian postal code. Is it:
- A9A9A9
- A9A 9A9
- A9A-9A9
If you picked one, you are wrong and need to have your genitals soldered shut for the good of humanity. The correct answer is "All three ways are correct. The computer should only care that those 6 digits are present, and extract the data, regardless of human-readable format".
BUT WAIT! THERE'S MORE!
How do you store that data in the database? If you answered 1) you are correct. If you picked 1) in the "which is correct format" because that's the format your database, then again-- please report to the soldering station. Because shock of all shocks-- if you can validate that the data you want is present, you can extract it!
BUT WAIT! THERE'S EVEN MORE!
What letters are not allowed in the 1st, 3rd or 5th position? If you have an answer, then you're wrong, and you're validating against outdated information. You probably copy/pasta'd a validation regex from 2002 Stack Overflow, and never bothered to check with Canada Post, did you?
OH BUT WAIT! THERE IS EVEN MOOOOOOOOOOOORE!
Let's say for some reason you NEED to restrict the user input based on the seperater character, or letter restrictions based on position. The user enters A1B2W6, which fails your fuctarded regex because W isn't valid in position 5. Do you:
a) Accept the input, then like a fucking gooma give the user a generic error of "invalid data", not even telling them it's the postal code?
b) Accept the input, then like a shit-sucking asshole, give the user a generic error of "invalid postal code", but not tell them why?
c) Make the user enter A1B2W6, A1B-2W6 and A1B 2W6, as they try to guess which goddamn format you use, and each of them fails, and like a cock-stuffing blanket-eater, you just keep telling them "invalid postal code".
d) After they VIEW SOURCE and check the regex, they see what you're doing, and enter A1B2Z6, the have to email customer support to tell them the postal code is actualyl A1B2W6 and that their postal validation rules are fucking retarded?If you answered D the I hope someone sodomizes you with the soldering iron after sterilizing you, just to ensure that even if you think you're going to eventually get pleasure out of anal sex, that possibility is stripped from you and you forever know only terror and pain from any comforting contact.
Also
because we used to call those shit-puckers "Rock Stars" and FUCK THEM AHAHRLKGJLA:KSDJFLKJLDKJFIOPWEJ!!!!!!!In you i sense much anger.....
I too have been bit by bad postal code storage practices.
in the US postal codes are 5 numeric digits.
outside of newengland none of the zip codes begin with
0so badly written software writtin in california (
9as the first postal code digit) has MORE THAN ONCE sent very expensive packages to me with a zipcode of4101. Usually USPS figures out that the zipcode intended was04101and the package gets to me anyway, but sometimes it doesn't and sends the package to41010instead and i have to fight with the vendor that fucked up and stored their zip codes as an INT for a refund, chew them out for storing their zipcodes wrong, sourcing the product from a different vendor, and blacklisting them until they fix how they store their data.
-
@wft I think that the generic answer is that we try 3rd party stuff (and there are a lot of 3rd party vendors of interest, some specializing in what we do, that's a big industry...) and in many cases we end up rolling out our own solution because the 3rd party stuff is not perfectly tailored to our needs.
Some years ago, I know some colleagues who were working very closely with NVidia to test out GPU computing (when it was a new thing). I certainly would not claim that we had any influence on the development of their GPUs, but we had a closer relationship than even beta-testing a new card 2 months before release.
-
@cheong said in It's nearly 2017, we still cannot into software:
@anonymous234 said in It's nearly 2017, we still cannot into software:
@ScholRLEA said in It's nearly 2017, we still cannot into software:
That's a single drive. Ten Terabytes. Holy fuck.
Wanna hear a fun fact? DNA stores around 1 PB per cubic millimeter.
So that's the lower bound to how densely we can store things.
But I heard the limit for magnetic media is much lower (both limit of storage of data and reading precision).
That's why they're always seeking for other kind of material / method for storage.
I imagine we're pretty close to the limit for magnetic storage. If magnetic domains get too small, room temperature is enough to begin flipping bits. But I guess we already have helium-filled drives, why not liquid-helium-filled drives?
I think solid-state storage (not necessarily flash) will completely overtake magnetic disks in density within the next decade, and hard disks will only remain in niche roles like tape drives are today.
-
@PJH said in It's nearly 2017, we still cannot into software:
@Jaloopa said in It's nearly 2017, we still cannot into software:
Inexpensive Redundant RAID arrays of Disks?
HTH.
In this case, probably:
Redundant array of expensive disk
So what do we have now? RAED?
-
@accalia said in It's nearly 2017, we still cannot into software:
@Lorne-Kates said in It's nearly 2017, we still cannot into software:
@wft said in It's nearly 2017, we still cannot into software:
Struggle wherever there's a form to fill out because the forms are full of whooshy animations, but validation is fucked up.
Pop quiz. What is the correct format for a Canadian postal code. Is it:
- A9A9A9
- A9A 9A9
- A9A-9A9
If you picked one, you are wrong and need to have your genitals soldered shut for the good of humanity. The correct answer is "All three ways are correct. The computer should only care that those 6 digits are present, and extract the data, regardless of human-readable format".
BUT WAIT! THERE'S MORE!
How do you store that data in the database? If you answered 1) you are correct. If you picked 1) in the "which is correct format" because that's the format your database, then again-- please report to the soldering station. Because shock of all shocks-- if you can validate that the data you want is present, you can extract it!
BUT WAIT! THERE'S EVEN MORE!
What letters are not allowed in the 1st, 3rd or 5th position? If you have an answer, then you're wrong, and you're validating against outdated information. You probably copy/pasta'd a validation regex from 2002 Stack Overflow, and never bothered to check with Canada Post, did you?
OH BUT WAIT! THERE IS EVEN MOOOOOOOOOOOORE!
Let's say for some reason you NEED to restrict the user input based on the seperater character, or letter restrictions based on position. The user enters A1B2W6, which fails your fuctarded regex because W isn't valid in position 5. Do you:
a) Accept the input, then like a fucking gooma give the user a generic error of "invalid data", not even telling them it's the postal code?
b) Accept the input, then like a shit-sucking asshole, give the user a generic error of "invalid postal code", but not tell them why?
c) Make the user enter A1B2W6, A1B-2W6 and A1B 2W6, as they try to guess which goddamn format you use, and each of them fails, and like a cock-stuffing blanket-eater, you just keep telling them "invalid postal code".
d) After they VIEW SOURCE and check the regex, they see what you're doing, and enter A1B2Z6, the have to email customer support to tell them the postal code is actualyl A1B2W6 and that their postal validation rules are fucking retarded?If you answered D the I hope someone sodomizes you with the soldering iron after sterilizing you, just to ensure that even if you think you're going to eventually get pleasure out of anal sex, that possibility is stripped from you and you forever know only terror and pain from any comforting contact.
Also
because we used to call those shit-puckers "Rock Stars" and FUCK THEM AHAHRLKGJLA:KSDJFLKJLDKJFIOPWEJ!!!!!!!In you i sense much anger.....
I too have been bit by bad postal code storage practices.
in the US postal codes are 5 numeric digits.
outside of newengland none of the zip codes begin with
0so badly written software writtin in california (
9as the first postal code digit) has MORE THAN ONCE sent very expensive packages to me with a zipcode of4101. Usually USPS figures out that the zipcode intended was04101and the package gets to me anyway, but sometimes it doesn't and sends the package to41010instead and i have to fight with the vendor that fucked up and stored their zip codes as an INT for a refund, chew them out for storing their zipcodes wrong, sourcing the product from a different vendor, and blacklisting them until they fix how they store their data.Sounds like someone took their input validation routines from Excel.
Fuck you Excel.
-
@Lorne-Kates But why stop there?
Why limit it to 6 characters, when they could decide next month that the postal codes are 7 characters?
Why exclude "-" if that might be a valid character next year.
And if you say, "but it's postal codes", then I'll say, "but what if it's license plates? because those change on a goddamned whim around here."
-
@Tsaukpaetra said in It's nearly 2017, we still cannot into software:
@djls45 said in It's nearly 2017, we still cannot into software:
More technically, the technology used in the phone's radio connection has a huge effect on the time the phone can run on a single charge. GSM phones are constantly emitting and receiving while they listen for their particular identifier to appear, meaning someone is calling or trying to make some other type of connection. CDMA phones only check every once in a while, which means the radio is off for most of the time, which saves the charge on the battery.
Yes, technically. But the power savings of using a given technology over another is nanoscale compared to what the display uses. The display typically uses the same amount of juice in an hour of being on as a typical radio in standby (or equivalent) for a day.
So turn the brightness down, or just turn it off!
The radio does draw a fair amount of power, but I expect most phones wouldn't show that in the battery usage screen because that would be considered just part of having the phone on in the first place. The WiFi radio is also a significant battery drain. But you're right; the display is usually the biggest drain on a battery charge, which is why requiring the user to press a certain combination of buttons on the outside of the phone should be required in order to turn on the screen, and not just tapping it once; because with only a touch requirement on a touch-screen, a little bumping while in a pocket can activate the screen when it's not useful.
-
@fbmac said in It's nearly 2017, we still cannot into software:
@djls45 said in It's nearly 2017, we still cannot into software:
If people would realize that they didn't need to run all those silly little micro-transaction games and apps and stuff, they could get a phone that is just a phone, with maybe texting and simple photo messaging capabilities, and the battery would last for several days.
If I could trust a any local phone operator to send a SMS
And if their phone even lets them any more! I get MMSes all the time from people whose smartphones automatically turn emoticons into animations, when the entirety of the message should be text-only. It's not currently a cost problem for me, because my plan includes a certain number of each type of message, and I don't get close to using it all, but if I used my phone a lot more, it definitely could become one.
-
@djls45 All 3G standards use some form of CDMA. You must be talking about the standard popularly known as CDMA.
-
@djls45 Android does:
-
@djls45 said in It's nearly 2017, we still cannot into software:
It's not currently a cost problem for me, because
my plan includes a certain number of each type of messagereceiving messages doesn't cost moneyFTFCC
-
@remi said in It's nearly 2017, we still cannot into software:
Data in these days was not obviously on disk but on magnetic tapes so we had complex interfaces with storage robots handling the tapes to try and optimize that
Ah yes, storage robots.
Command 72: Mount Careless Visitor's Face on First Free Drive.
-
@PJH said in It's nearly 2017, we still cannot into software:
12 if you include the numbered bulletpoints.
13 if you can see the
up your ass.(And you can, because head up ass joke)
-
@cartman82 said in It's nearly 2017, we still cannot into software:
I was gonna post a rebuke last night, but nodebb was bugging out.
It's 2017, and we still cannot into rebuke
-
@Yamikuronue said in It's nearly 2017, we still cannot into software:
@Lorne-Kates said in It's nearly 2017, we still cannot into software:
What letters are not allowed in the 1st, 3rd or 5th position?
Whatever letters the Canada Post's address validation API refuses?
You think we have an Address Validation API?
HAHAHAHAHAHAHAHAHAHAHAHAAH!!!!!!!!!!!
Hahahahahah MAPLE LEAF HAHAHAHAHAHAHAHAH!
ha
No, Canada Post doesn't offer that. So someone made a crowd-sourced version of the postal code / reverse lookup database. So Canada Post sued him.
fake edit 1 Looks like after 4 years the case was settled out of court-- reading between the lines looks like Canada Post realized you can't sue someone for crowdsourcing information, and dropped the suit.
http://geocoder.ca/?sued=1"Canada Post commenced court proceedings in 2012 against Geolytica Inc. for copyright infringement in relation to Geolytica Inc.'s Canadian Postal Code Geocoded Dataset and related services offered on its website at geocoder.ca. The parties have now settled their dispute and Canada Post will discontinue the court proceedings. The postal codes returned by various geocoder interface APIs and downloadable on geocoder.ca, are estimated via a crowdsourcing process. They are not licensed by geocoder.ca from Canada Post, the entity responsible for assigning postal codes to street addresses. Geolytica continues to offer its products and services, using the postal code data it has collected via a crowdsourcing process which it created."
So please everyone, use GeoCoder!
fake edit Looks like Canada Post offers some very very limited web services-- but they're all around rate quotes.
-
@accalia said in It's nearly 2017, we still cannot into software:
chew them out for storing their zipcodes wrong
but numbers!
Amazon.com
Canada Post Developer Program β Code Samples for Web Services
