Your analogy doesn't pass muster when one considers that sed is a text stream editor in its very nature, and only a stream editor, when the purpose of the question is to edit streams,
Perl contains stream editor capabilities, but is not a single-purpose stream editor.
A more accurate analogy, sed is a #2 philips screwdriver. Perl is a multitool. The job to be done requires a #3 philips screwdriver. Neither tool perfectly does the job, though both can be made to do the job.
That's the way it works. The more despised something is, the more annoying those that run against the grain are by embracing it.
The converse is not true though. As an example, Star Wars is very popular, but those who can cite everything to do with it including all of the Expanded Universe backstories for every single alien-extra that appeared in the Cantina or in Jabba the Hutt's lair are still incredibly annoying.
It's just a TV show dammit.. It's just a TV show...
Dad was a COBOL programmer. He really liked the language, felt it was a lot more of a natural-language flow than languages like C. We still have an AS/400 at work and the few times I've seen COBOL code I can't help but agree with him, the actual structure of the syntax is a lot easier to follow as a layman, especially when the variable names are kept sane.
I suppose this contributes to why I don't want to learn Perl; I don't program for a living but I do use various types of scripting for work, usually because I've had to reduce captured text down to something manageable, and if I find that I do the same replacement procedures in vi all of the time it makes sense to script it where I can do the processing from the command line without ever opening a text editor. When I look up how to do something with sed or other such tools I find a lot of other people that have asked questions for using a particular tool, only for answer to be, "you can do it in perl! Here's my script!" followed by a bunch of lines of code with no explanation for what they do.
The original asker didn't ask how to do it in any language or with any command, they asked how to do it with a specific command. It's like asking, "How many of you like chocolate ice cream?" for someone to reply, "Strawberry is great!" Not actually answering the question. The problem is that if a non-answer stands, then people may shy away from actually answering the question, or from explaining why said tool may have problems with the task at hand.
As an example, in vi I match patterns that span two lines, and it's easy to specify the linefeed as part of the string to search for, even when using the vi invocation of sed. By contrast the command line invocation of sed interprets the linefeeds differently than when used through an editor, so the commands that I tested in vi and tried to translate into shell-invoked didn't work. It took a lot of different search terms to finally find an explanation of why linefeeds were a problem, because I kept getting fanboy responses about perl.
A homburg hat might be obscure among the general population, but I would not be surprised if there's a disproportionate overlap between those that actually do know what a homburg hat is and have one featured in one's photos, and iPhone users that actually would seek to use these entirely unnecessary functions.
You've clearly never worked with 2000' of 62.5 OM1 originally designed for 10 Mbit over fiber trying to push Gigabit over fiber with LX optics and mode-conditioning cables...
I don't think that Nye is trying to get kids to grow up to get Doctorates. Nye himself does not have a Doctorate. Nye seems to be advocating for science education because it will allow students to better understand the natural world around them and to make decisions with that understanding. A byproduct of a good science education is the development of critical thinking skills, which further helps students to make good decisions.
At least with a farm it's straightforward. Trench in the right-of-way along the roadway or in the easements. Where trenching isn't practical like crossing asphalt or concrete paved roads, directional-bore. Drop Christy vaults at intervals or bigger manhole-access vaults if expansion is needed. Fiber can go miles and miles, so knowing the route to the customer, determine if it's necessary to build a local exchange with power that could potentially service multiple farms in an area, attempt to position it where it's the final powered-hop before the handoff to the customer. otherwise there's usually little if any infrastructure to have to worry about damaging along the way.
In a city there's trouble with easements, rights-of-way, working around existing infrastructure, working in congested areas, and the need to remedy problems as they occur, and that's before dealing with pathways in the building (who likes pulling fiber through the side of the elevator shaft?) and equipment space.
I doubt it. SF doesn't depend on manufacturing whose skillset is developed from initially unskilled labor through on-the-job training. The kind of employment driving the ridiculous housing prices requires the employee to bring skill in with them.
A major contribution to Detroit's decline was that it was not difficult to set-up shop elsewhere, then close the Detroit facility. The vast majority of the individual tasks for assembling cars are simple, so a new population that has never built cars can be trained to build cars without a lot of initial costs and without having to attract a lot of experienced workers to that new population center. On top of that the design work doesn't have to be done where the assembly is done either, so the skilled jobs could be in nicer places to live.
This is already a lot of the model for San Francisco. The big tech employers have most of their manufacturing elsewhere. The work that is done in SF and Silicon Valley requires one's brain much more than one's hands, and the companies set up shop there because it's generally a nice place to live, they want to live there despite the costs.
That plan is probably dictated by a combination of the edge service provider's connection to the backbone service provider, and the edge service provider's willingness to spend money on their distribution network in the form of switches and optics and the topology chosen for the fiber plant.
As a commercial customer, I've worked with both Centurylink and COX for fiber networks. CL uses a hub-and-spoke topology in my market that they inherited when they took over the old phone company. Each of my sites has a twelve-strand OM1 cable with two strands used, CL provides a handoff with their Metro Optical switch, and I then take service from that to my service entrance switch. CL needs four rack units, three for the LIU, one for the switch, which also acts as their demarcation point. Upsides, I'm the only customer on this spoke, so their equipment at my premises only has to meet the needs of the service I have subscribed to, and arguably if their MPLS network were configured for it they wouldn't even have to put a switch in my sites, just handing off at the coupler panel. Downside, running a spoke network is expensive, from every CO or NX they have to have a dedicated pair of strands to my site, either physically splicing in from backbone fiber running through the city, or else having a dedicated cable all of the way.
COX uses a ring-topology with mixed customers on the rings. They have something like 24 strands in, 24 strands out. They're using some kind of PON variant even for their own 2U switch, so they have an LIU for the raw fiber, then they pass through some kind of PON filter in what looks like a second LIU that directs the specific wavelength to their switch to the two 40G optics for the ring, but lets the rest of the wavelengths pass through. Also because of the topology they have around 8U of battery. Their switches are DC, so the rectifier powers the batteries and then the switches are powered off of that setup. Upside, the topology is a lot simpler, they don't have to run as much fiber except where they have to cut-in from the trunk cables to enter the customer premises, but the downsides are they use ~12U of rack space and that the rings can be broken and lead to isolated segments if customers premises issues at multiple locations impact the rings. It also requires us to provide site access to their techs to diagnose ring problems, so we have to do off-hours support a lot more than with CL.
From a workflow point of view I like the Centurylink setup better. Three rack units is easy to find for them, and since they're just running an off-the-shelf Metro Optical switch, it gets plugged into the same power redundancy as my own equipment. By contrast the COX setup uses a lot of room, allows for the sites to be taken out by something as simple as one customer with multiple sites downing the equipment, requires expensive switches and optics for the 40G shared backbone everywhere they have a service-entrance, and since those switches are DC-only, if the battery and rectifier fail (which they have) I can't bypass around them to power up the service entrance switch until a tech can replace the defective unit.
Not knowing the engineering choices made in your municipal network I cannot say what they've done. Personally I'd rather pay more for the fiber now and less for the equipment, knowing that the equipment will have to be changed-out on a somewhat regular basis as demand for more throughput compels it.
One of the biggest stumbling blocks are the landlords and what undoubtedly will be objections to losing profit-making square footage to telecommunications rooms that inevitably will be required in buildings.
I've had to do work in residential copper phone rooms. It was bad when it was just the telephone company. It got worse when the cable TV providers got space. For old, sufficiently large buildings it will be even worse, as the equipment for even a PON system requires space. That's not accounting for the fiber equivalent of a neighborhood exchange or central office either.
I found the people in Paris were a lot worse than the people in San Francisco.
As for getting from Point-A to Point-B, the subway is fairly effective, and the actual city is so geographically small that walking works fairly well too.
If you have a friend that lives in SF, then as long as the friend is on the western side of the city, Oakland. You'll need to set up a microwave link though, and hope that it isn't too foggy any given day.
If the algorithms that have come to these conclusions are based on analyzing public data from the Internet, then if an AI decides that any particular characteristic is negative, it's because it reflects the sentiments of those who bother to post opinions.
Most people that do not themselves exhibit the trait that's being argued-against by the noisy minority don't usually express opinions on it, so they're a hole in the data that needs to be accounted for. Unfortunately it's a lot easier to interpret based on what has been said than what has not been said.
Yes, but it will probably become more profitable because now operating costs will come down even further. All of the local on-air staff, gone. Local staff that manage programming and other content decisions, almost all gone if not all gone. Engineering staff, no longer need to staff enough to work the studios as there are no studios, probably half or more gone, just limited to staff to maintain a bit of local equipment. Even that might be reduced since it may be possible to outsource that, and multiple stations could end up with the same maintenance partners.
What's funny is that journalism schools already annually graduate more students with journalism or communications degrees than there are jobs in the whole profession. Now that all of your local on-air talent is essentially gone, in my market that could be 50 people, and while my market is a bit larger than most I could reasonably expect any city over 100,000 people to to have at least 25 on-air, with more than 300 metro areas over 100,000, that's 7500 on-air staff losing jobs, plus all of the rest of the support staff.
Frankly it's dangerous for so few people to own all of the press. We're already into an era of Yellow Journalism, and it's only going to get much worse and to polarize people far more. Honestly it could lead to outright civil war when those people that control the media push peoples' buttons in order to drive ratings.
It could also be that the Musk-involved companies are doing a lot more numbers-crunching to determine where weaknesses in the companies are.
In the case of manufacturing and a recall, if the manufacturer tracks serial numbers for the particular part numbers that are installed in cars built during particular shifts when particular employees were installing the parts on the cars, and if it turns out that the particular employee working these shifts did work that resulted in recall, but the employees in other shifts with the same job did not result in recall, then perhaps the particular employee caused the problem that led to that recall. Granted, this is a specific case, but if they can establish a limited set of employees whose work efforts align very specifically with the recalled vehicles, then perhaps those employees should be subject to review and if it turns out that gross, willful negligence was involved, then they should be let go.
In the case of a sales and installation company, sales people that are not strictly commission probably shouldn't be retained if they are not able to generate sales. Installers whose work results in lots of return visits to correct may also find that their jobs are in-peril. Having been on the customer side in a lot of huge projects at work, when the installer does a bad job once we're willing to play East-German-judge and nitpick until they fix the resulting problems, but if they chronically do poor work requiring an inordinate amount of our time to manage them then we do not keep them around. We cease doing business with them after we've found a new vendor, and when we document the poor performance, it's very unlikely that they'll win future open bids even if they submit because past performance counts in that process.
As far as I'm concerned, if they really did terminate employees that were severely underperforming then there's nothing wrong with that. Whether or not that's the real reason for the firings, no idea.
Unions can be beneficial when key managerial employees are jerks too. They serve two purposes in these cases, first, to help reduce unevenness in assignments within a given labor pool, such that the boss doesn't get to play favorites as much as the boss might like, and second, to act as an advocate when a boss might be seeking uneven punitive action against an employee when that employee's behaviors are not statistically different than the rest of the workers.
Now, this doesn't mean that the unions always do a good job of this. Sometimes unions end up protecting terrible employees. Sometimes the presence of a union coupled with lazy managers that aren't willing to give regular performance reviews or otherwise document behaviors or results means that bad employees that really should be taken to task are not.
Seriously, the author is worried about spending a tiny bit more at Walmart because school kids in China now get cleaner air?
I infer that perhaps it was only especially cheap to manufacture in China because they were willing to fuck their environment in order to do it. Once industry is required to both pay for new emissions controls and to pay to clean up the existing damage, it will not be nearly so cheap to manufacture in China.
Throw nice helpings of industrial espionage and lack of respect for patents and trademarks, and it's even less cheap to manufacture there long-term. But then, corporate officers and boards of directors don't seem to be interested in long-term profitability anyway, instead awarding themselves golden parachutes in the tens of millions of dollars for pumping up companies for short-term profitability that falls apart as they're leaving.
He got his message across. It wasn't the words coming out of his mouth in scripted campaign rallies, but the message has apparently come across loud and clear.
To the likes of the entities that funded them that $25,000,000 is chump-change. The roll-of-the-dice is worth it to them if the technology pans-out. I also expect that those companies that survived the dotcom bubble may themselves be slightly better at evaluating if a given startup has a better chance of actually providing a return, so they sponsor fewer losers relative to the winners.
Because I have a 27" monitor with me at all times for when I get unexpectedly called because the network is down somewhere and I have to help restore it...
Slashdot Mirror
Your analogy doesn't pass muster when one considers that sed is a text stream editor in its very nature, and only a stream editor, when the purpose of the question is to edit streams,
Perl contains stream editor capabilities, but is not a single-purpose stream editor.
A more accurate analogy, sed is a #2 philips screwdriver. Perl is a multitool. The job to be done requires a #3 philips screwdriver. Neither tool perfectly does the job, though both can be made to do the job.
That's the way it works. The more despised something is, the more annoying those that run against the grain are by embracing it.
The converse is not true though. As an example, Star Wars is very popular, but those who can cite everything to do with it including all of the Expanded Universe backstories for every single alien-extra that appeared in the Cantina or in Jabba the Hutt's lair are still incredibly annoying.
It's just a TV show dammit.. It's just a TV show...
Dad was a COBOL programmer. He really liked the language, felt it was a lot more of a natural-language flow than languages like C. We still have an AS/400 at work and the few times I've seen COBOL code I can't help but agree with him, the actual structure of the syntax is a lot easier to follow as a layman, especially when the variable names are kept sane.
I suppose this contributes to why I don't want to learn Perl; I don't program for a living but I do use various types of scripting for work, usually because I've had to reduce captured text down to something manageable, and if I find that I do the same replacement procedures in vi all of the time it makes sense to script it where I can do the processing from the command line without ever opening a text editor. When I look up how to do something with sed or other such tools I find a lot of other people that have asked questions for using a particular tool, only for answer to be, "you can do it in perl! Here's my script!" followed by a bunch of lines of code with no explanation for what they do.
The original asker didn't ask how to do it in any language or with any command, they asked how to do it with a specific command. It's like asking, "How many of you like chocolate ice cream?" for someone to reply, "Strawberry is great!" Not actually answering the question. The problem is that if a non-answer stands, then people may shy away from actually answering the question, or from explaining why said tool may have problems with the task at hand.
As an example, in vi I match patterns that span two lines, and it's easy to specify the linefeed as part of the string to search for, even when using the vi invocation of sed. By contrast the command line invocation of sed interprets the linefeeds differently than when used through an editor, so the commands that I tested in vi and tried to translate into shell-invoked didn't work. It took a lot of different search terms to finally find an explanation of why linefeeds were a problem, because I kept getting fanboy responses about perl.
Everyone is of negotiable virtue.
For a scant few, the price is to serve a life-sentence.
A homburg hat might be obscure among the general population, but I would not be surprised if there's a disproportionate overlap between those that actually do know what a homburg hat is and have one featured in one's photos, and iPhone users that actually would seek to use these entirely unnecessary functions.
Optical: expensive jacks, cheap cable, decisive worx/dontworx.
You've clearly never worked with 2000' of 62.5 OM1 originally designed for 10 Mbit over fiber trying to push Gigabit over fiber with LX optics and mode-conditioning cables...
I don't think that Nye is trying to get kids to grow up to get Doctorates. Nye himself does not have a Doctorate. Nye seems to be advocating for science education because it will allow students to better understand the natural world around them and to make decisions with that understanding. A byproduct of a good science education is the development of critical thinking skills, which further helps students to make good decisions.
At least with a farm it's straightforward. Trench in the right-of-way along the roadway or in the easements. Where trenching isn't practical like crossing asphalt or concrete paved roads, directional-bore. Drop Christy vaults at intervals or bigger manhole-access vaults if expansion is needed. Fiber can go miles and miles, so knowing the route to the customer, determine if it's necessary to build a local exchange with power that could potentially service multiple farms in an area, attempt to position it where it's the final powered-hop before the handoff to the customer. otherwise there's usually little if any infrastructure to have to worry about damaging along the way.
In a city there's trouble with easements, rights-of-way, working around existing infrastructure, working in congested areas, and the need to remedy problems as they occur, and that's before dealing with pathways in the building (who likes pulling fiber through the side of the elevator shaft?) and equipment space.
I doubt it. SF doesn't depend on manufacturing whose skillset is developed from initially unskilled labor through on-the-job training. The kind of employment driving the ridiculous housing prices requires the employee to bring skill in with them.
A major contribution to Detroit's decline was that it was not difficult to set-up shop elsewhere, then close the Detroit facility. The vast majority of the individual tasks for assembling cars are simple, so a new population that has never built cars can be trained to build cars without a lot of initial costs and without having to attract a lot of experienced workers to that new population center. On top of that the design work doesn't have to be done where the assembly is done either, so the skilled jobs could be in nicer places to live.
This is already a lot of the model for San Francisco. The big tech employers have most of their manufacturing elsewhere. The work that is done in SF and Silicon Valley requires one's brain much more than one's hands, and the companies set up shop there because it's generally a nice place to live, they want to live there despite the costs.
So like, in theory there's no difference between theory and reality?
That plan is probably dictated by a combination of the edge service provider's connection to the backbone service provider, and the edge service provider's willingness to spend money on their distribution network in the form of switches and optics and the topology chosen for the fiber plant.
As a commercial customer, I've worked with both Centurylink and COX for fiber networks. CL uses a hub-and-spoke topology in my market that they inherited when they took over the old phone company. Each of my sites has a twelve-strand OM1 cable with two strands used, CL provides a handoff with their Metro Optical switch, and I then take service from that to my service entrance switch. CL needs four rack units, three for the LIU, one for the switch, which also acts as their demarcation point. Upsides, I'm the only customer on this spoke, so their equipment at my premises only has to meet the needs of the service I have subscribed to, and arguably if their MPLS network were configured for it they wouldn't even have to put a switch in my sites, just handing off at the coupler panel. Downside, running a spoke network is expensive, from every CO or NX they have to have a dedicated pair of strands to my site, either physically splicing in from backbone fiber running through the city, or else having a dedicated cable all of the way.
COX uses a ring-topology with mixed customers on the rings. They have something like 24 strands in, 24 strands out. They're using some kind of PON variant even for their own 2U switch, so they have an LIU for the raw fiber, then they pass through some kind of PON filter in what looks like a second LIU that directs the specific wavelength to their switch to the two 40G optics for the ring, but lets the rest of the wavelengths pass through. Also because of the topology they have around 8U of battery. Their switches are DC, so the rectifier powers the batteries and then the switches are powered off of that setup. Upside, the topology is a lot simpler, they don't have to run as much fiber except where they have to cut-in from the trunk cables to enter the customer premises, but the downsides are they use ~12U of rack space and that the rings can be broken and lead to isolated segments if customers premises issues at multiple locations impact the rings. It also requires us to provide site access to their techs to diagnose ring problems, so we have to do off-hours support a lot more than with CL.
From a workflow point of view I like the Centurylink setup better. Three rack units is easy to find for them, and since they're just running an off-the-shelf Metro Optical switch, it gets plugged into the same power redundancy as my own equipment. By contrast the COX setup uses a lot of room, allows for the sites to be taken out by something as simple as one customer with multiple sites downing the equipment, requires expensive switches and optics for the 40G shared backbone everywhere they have a service-entrance, and since those switches are DC-only, if the battery and rectifier fail (which they have) I can't bypass around them to power up the service entrance switch until a tech can replace the defective unit.
Not knowing the engineering choices made in your municipal network I cannot say what they've done. Personally I'd rather pay more for the fiber now and less for the equipment, knowing that the equipment will have to be changed-out on a somewhat regular basis as demand for more throughput compels it.
One of the biggest stumbling blocks are the landlords and what undoubtedly will be objections to losing profit-making square footage to telecommunications rooms that inevitably will be required in buildings.
I've had to do work in residential copper phone rooms. It was bad when it was just the telephone company. It got worse when the cable TV providers got space. For old, sufficiently large buildings it will be even worse, as the equipment for even a PON system requires space. That's not accounting for the fiber equivalent of a neighborhood exchange or central office either.
I found the people in Paris were a lot worse than the people in San Francisco.
As for getting from Point-A to Point-B, the subway is fairly effective, and the actual city is so geographically small that walking works fairly well too.
Before that it was AOL, this is nothing new.
If you have a friend that lives in SF, then as long as the friend is on the western side of the city, Oakland. You'll need to set up a microwave link though, and hope that it isn't too foggy any given day.
If the algorithms that have come to these conclusions are based on analyzing public data from the Internet, then if an AI decides that any particular characteristic is negative, it's because it reflects the sentiments of those who bother to post opinions.
Most people that do not themselves exhibit the trait that's being argued-against by the noisy minority don't usually express opinions on it, so they're a hole in the data that needs to be accounted for. Unfortunately it's a lot easier to interpret based on what has been said than what has not been said.
Yes, but it will probably become more profitable because now operating costs will come down even further. All of the local on-air staff, gone. Local staff that manage programming and other content decisions, almost all gone if not all gone. Engineering staff, no longer need to staff enough to work the studios as there are no studios, probably half or more gone, just limited to staff to maintain a bit of local equipment. Even that might be reduced since it may be possible to outsource that, and multiple stations could end up with the same maintenance partners.
What's funny is that journalism schools already annually graduate more students with journalism or communications degrees than there are jobs in the whole profession. Now that all of your local on-air talent is essentially gone, in my market that could be 50 people, and while my market is a bit larger than most I could reasonably expect any city over 100,000 people to to have at least 25 on-air, with more than 300 metro areas over 100,000, that's 7500 on-air staff losing jobs, plus all of the rest of the support staff.
Frankly it's dangerous for so few people to own all of the press. We're already into an era of Yellow Journalism, and it's only going to get much worse and to polarize people far more. Honestly it could lead to outright civil war when those people that control the media push peoples' buttons in order to drive ratings.
It could also be that the Musk-involved companies are doing a lot more numbers-crunching to determine where weaknesses in the companies are.
In the case of manufacturing and a recall, if the manufacturer tracks serial numbers for the particular part numbers that are installed in cars built during particular shifts when particular employees were installing the parts on the cars, and if it turns out that the particular employee working these shifts did work that resulted in recall, but the employees in other shifts with the same job did not result in recall, then perhaps the particular employee caused the problem that led to that recall. Granted, this is a specific case, but if they can establish a limited set of employees whose work efforts align very specifically with the recalled vehicles, then perhaps those employees should be subject to review and if it turns out that gross, willful negligence was involved, then they should be let go.
In the case of a sales and installation company, sales people that are not strictly commission probably shouldn't be retained if they are not able to generate sales. Installers whose work results in lots of return visits to correct may also find that their jobs are in-peril. Having been on the customer side in a lot of huge projects at work, when the installer does a bad job once we're willing to play East-German-judge and nitpick until they fix the resulting problems, but if they chronically do poor work requiring an inordinate amount of our time to manage them then we do not keep them around. We cease doing business with them after we've found a new vendor, and when we document the poor performance, it's very unlikely that they'll win future open bids even if they submit because past performance counts in that process.
As far as I'm concerned, if they really did terminate employees that were severely underperforming then there's nothing wrong with that. Whether or not that's the real reason for the firings, no idea.
Unions can be beneficial when key managerial employees are jerks too. They serve two purposes in these cases, first, to help reduce unevenness in assignments within a given labor pool, such that the boss doesn't get to play favorites as much as the boss might like, and second, to act as an advocate when a boss might be seeking uneven punitive action against an employee when that employee's behaviors are not statistically different than the rest of the workers.
Now, this doesn't mean that the unions always do a good job of this. Sometimes unions end up protecting terrible employees. Sometimes the presence of a union coupled with lazy managers that aren't willing to give regular performance reviews or otherwise document behaviors or results means that bad employees that really should be taken to task are not.
Seriously, the author is worried about spending a tiny bit more at Walmart because school kids in China now get cleaner air?
I infer that perhaps it was only especially cheap to manufacture in China because they were willing to fuck their environment in order to do it. Once industry is required to both pay for new emissions controls and to pay to clean up the existing damage, it will not be nearly so cheap to manufacture in China.
Throw nice helpings of industrial espionage and lack of respect for patents and trademarks, and it's even less cheap to manufacture there long-term. But then, corporate officers and boards of directors don't seem to be interested in long-term profitability anyway, instead awarding themselves golden parachutes in the tens of millions of dollars for pumping up companies for short-term profitability that falls apart as they're leaving.
He got his message across. It wasn't the words coming out of his mouth in scripted campaign rallies, but the message has apparently come across loud and clear.
Chinese Scientists Create Genetically Modified Low-Fat Pigs
But what are they for?
To the likes of the entities that funded them that $25,000,000 is chump-change. The roll-of-the-dice is worth it to them if the technology pans-out. I also expect that those companies that survived the dotcom bubble may themselves be slightly better at evaluating if a given startup has a better chance of actually providing a return, so they sponsor fewer losers relative to the winners.
Because I have a 27" monitor with me at all times for when I get unexpectedly called because the network is down somewhere and I have to help restore it...