Posts made by StorageNinja

@scottalanmiller said in Ajit Pai wants to raise rural broadband speeds from 10Mbps to 25Mbps:

In your corner of Houston. Where we go in Houston, they can't get stable Internet and the cost is insane.

Houston or Friendswood? The burbs have a few factors.

1. NIMBY groups who tend to oppose ugly retransmit boxes.

2. Timing is EVERYTHING Burbs have infrastructure that ties to how old they were. Older burbs were built out with older gear. They oddly may have gotten DSLAM's or DSLAM upgrades LONG before the dense, urban area's that are the proper City of Houston (When I moved here, AT&T was Somalian speeds promising 256KB only inside the loop). As a lot of infrastructure gets done on "worst first" basis this explains why beyond raw profit motive (Urban area has denser opportunities and higher median home prices justifying build-out costs for GPON network). It's worth noting the new subdivisions in Katy, new multifamily, new North West build outs are all FTTH. AT&T isn't seriously wasting money or time running copper to new area's.

3. Comcast generally is offering 100Mbps plus DOCSIS 3.0 offerings. It's worth noting that TimeWarner Cable left us for dead on shitty DOCSIS 2.0 until Comcast traded them Houston for another city, and rebuilt it all from the ground up for DOCSIS 3. Comcast is a terrible company, but they saved us for an even shittier fate.

5G upgrades are going to cost 250 Billion. LTE+ upgrades are not going to be cheap (What will cover rural areas). We are looking at another 100 Billion in FIber also to backhaul this stuff. This stuff is going to up speeds in core urban area's, and hte LTE+ upgrades should make mobile broadband a realistic alternative for many rural area's.

As of year-end 2016, 92.3% of all Americans have access to fixed terrestrial broadband at speeds of 25 Mbps/3 Mbps.

While this isn't crazy fast, it's more than enough to watch Netflix (5Mbps for an HD stream). It's enough for a Video call (not HD outbound). This should be enough to do an online class (primary focus) and get some educational information. This is the federal governments minimum for subsidies.

I expect when we kill rural telephone mandatory subsidies they will spike it a bit more (and have LTE+ replace it). There's currently no serious business case to rebuild rural and low-density suburbs with FTTH. The money has to come from somewhere so we have two options...

1. It comes from taxes, largely of people who live in urban area's that it will not benefit.
2. It comes from cost sharing where the costs are pushed onto the services of people who live in urban areas.

Given that the US is becoming more urban, I'd say your best chance is to just... Move closer to civilization. I could have lived out in Friendswood. I could have saved a lot of money. But I bought land, that was in Houston proper and only after verifying that it had connectivity that would meet my requirements.

@jmoore said in Ajit Pai wants to raise rural broadband speeds from 10Mbps to 25Mbps:

@StorageNinja I live in Waco.

Parents live there, I went to school there. You should come down for the Texas Bowl. Baylor should destroy Vandi.

@scottalanmiller said in Ajit Pai wants to raise rural broadband speeds from 10Mbps to 25Mbps:

Or those who live in an area with AT&T, Comcast, Cox, Frontier, or any rural area and get zero of the innovation.
From Houston to rural NY, none of that stuff has existed and zero innovation or competition comes along. It's a rare, very unique market where those innovations have affected anyone for a long time.

I have two providers offering me Gigabit service in Houston. As 5G comes online I'll be looking at 3-4 providers with 500Mbps+ Speeds. Waco while not truly rural is is a 5G test site for AT&T.

Modulva is LTE only in major cities. Rural coverage is HSDPA primarily, and for small villages and rural area's, it's xDSL.

The Reality is I can stream 3D 4K video on my existing 120 meg down circuit. Really the draw of the newer stuff for me is lower latency, and distributed service meshes embedded in the network slices.

@Dashrender said in Ajit Pai wants to raise rural broadband speeds from 10Mbps to 25Mbps:

Unless another player enters the market and lays their own fiber, you're just not going to get the incumbents to move voluntarily. Another reason the exclusive contracts are so anti-consumer.

You don't need to run fiber to the home though with 5G. You can run Fiber down a few major streets and shoot from there. Also, telco's are gearing up to radically change their CPE gear. Imagine if every Docsis modem could also do 5G. Imagine if the AT&T Fiber handoff could backhaul 4G to another pop if there's a cut.

3rd parties (Crown Castle) are running Fiber and towers and making it available to multiple third parties.
Network Slicing is going to allow virtual overlays to explode. NFV, private transport end to end. Even if you get "net neutrality" this stuff is all before the PoP so it means nothing when this stuff can do paid prioritization of its slice anyways (The same way that MPLS and Point to Points fall today under regulation). Is going to allow mobile virtual network operator (MVNO) growth to explode.

There's a narrative that we need to nationalize telecom, or that there isn't innovation going on in the last mile and it's largely being pushed by people who are missing out on all the cool stuff going on right now out of sight.

While networks have been sold as "pipe of xxx size" for years quality of peering, jitter, latency etc have existed as differentiation between carriers people buying it just were not always aware. Internap provided a far better mix than Cogent (who delivered what I always called porn grade bandwidth given the questionable peering). This granularity that network slicing can deliver is critical to a shift to declarative policies for computing and distributed applications, services meshes etc.

@JaredBusch said in Ajit Pai wants to raise rural broadband speeds from 10Mbps to 25Mbps:

@StorageNinja said in Ajit Pai wants to raise rural broadband speeds from 10Mbps to 25Mbps:

@Obsolesce

Couple things....

1. Other countries don’t have the landmass to cover than we do. Show me a largely rural area in Australia and I’ll show you shitty coverage.

2. The carriers have been deploying traffic shaping to slow Down video for some time.

Both points are true, but even in urban areas the US lags behind.

We lag behind if you talk Wireless for T-Mobile or Sprint.

I got 25Meg down on 4G LTE to my house. (Verizon likely faster and rolling out 5G in a few neighborhoods with other carriers following next year).
My last house could hit 90Mbps down.
I get Comcast 100Meg service (Faster available, I just don't feel like paying for it).
AT&T is offering Gigabit service in my neighborhood (I'm switching to it once my Comcast new customer discount runs out).
Next year I'll likely be blending 5G and GigE fiber with a VeloCloud box.

Everyone always points to urban results in Asian megacities (where population density is insane) our countries with last mile monopolies allowed (Which the EU countries approach reminds me of their jump on GSM they got, that eventually led to them falling behind as we let the market win things out and LTE is based on CMDA's time slicing tech).

Cell phone service only sucks around my city in shitty suburbs where they don't let people put op towers. NIMBY prevention of towers is one of the biggest problems the US has that other countries don't fight.

@Obsolesce

Couple things....

1. Other countries don’t have the landmass to cover than we do. Show me a largely rural area in Australia and I’ll show you shitty coverage.

2. The carriers have been deploying traffic shaping to slow Down video for some time.

@Dashrender said in Is the Physical Thin Client Era Dead?:

Just wasn't expecting to need thin client devices as powerful or more so than desktop machines running XP to be required to get an as good experience

Zero Clients have less intelligence than a rock (It's an ASIC that gets it's firmware by PXE boot) and I can play Skyrim on them over PCoIP. Printer Redirection will not really work, and god help you with a IO USB device over the WAN but a Thin client doesn't need to be that powerful for graphics beyond 2D resolution support, and number of monitor support.

All the new Thin Clients protocols are based on H.265. That is decoded in cheap(ish) SOC. Even an old iPhone 5/iPad 4 support H.265. In this case the CPU load is zero for the graphics as it's fully offloaded end to end.

@scottalanmiller said in Is the Physical Thin Client Era Dead?:

Many thin clients don't run Windows and use a completely different RDP library. That is often the cause of issues.

Many thin clients don't use RDP. While the protocol wars were fun (ICA vs. PCoIP!) I'm seeing everyone consolidate on highly customized stacks that at their core for image processing leverage H.264, and H.265. This is because for mobile and SOC have hardware decoders for this. (Blast Extreme and Citrix's HDX whatever it's called now do this).

@Dashrender said in Nine Out of Every 10 Silicon Valley Jobs Pays Less Than In 1997, Report Finds:

Right - so the point is these RSU's and other comps aren't really important for someone making $125K... because if you're lucky you might get another 50% - again if lucky.. and now you're making the same as you were 20 years ago - assuming there were no RSUs 20 years ago for these same people.

It's not really lucky if it's a public company of a major company who has a proven track record.
Looking at some recent offer data.

LinkedIn - $250K -400K in RSU's for a Senior Software Engineer for the initial grant.
Boeing - $300K with refreshers of 40K

Additional yearly grant refreshers beyond the initial are all over the place but tied to how much a company wants to keep you but assuming it's 25% of your base salary (not uncommon) with some reasonable stock appreciation you could be looking at close to 300% of initial offer in TC by year 4 which gets back to why I said this article was crap. Wages != Total Compensation in Silicon Valley.

@scottalanmiller said in Nine Out of Every 10 Silicon Valley Jobs Pays Less Than In 1997, Report Finds:

I'm making 17% more than I was 10 years ago. According to Google, since 2006, inflation is 21.59%. So i'm making noticeably less than I was making 10 years ago.

Almost no one who stays in a single position sees more money over time, it is a super rare company or position that keeps up with inflation.

That's a 1.59% annual raise (assuming you started at 100K and ended at 117K).
Note, the Consumer Price Index reports $100 to $116.
So that's basically a net zero raise (note, COL is more than CPI, and it depends on your region this might be more or less).

Again, getting back to my comment on SMB IT. There is rarely enough flexibility to move "up" in position in a small shop. If the IT roll provides the same "value" to the business (or less as services or key systems become outsourced), why would pay go up for this position?

@tonyshowoff said in Nine Out of Every 10 Silicon Valley Jobs Pays Less Than In 1997, Report Finds:

If I point out wages are down in Dallas, you can't say "Well, here in San Francisco it's about blah blah blah" and ignore the point I'm not talking about an area so out of control that using at as anything more than a case study is a ridiculous idea.

I live in Houston (Just bought a house in near the heights. I was a hiring manager for a 3 year period and was locally employed for ~7 years in the industry, and consulted all over the state). Wages went up everywhere but the SMB sector. The SMB sector core IT infrastructure and end user support suffers from...

1. Low barrier to entry.

2. Employers having poor ability to value staff (so even if the staff is good, they don't know it, and they will sometimes hire crap staff because again, without a full department and regular hiring of IT staff it's hard to do).

3. Increasingly lower skill overlap with large enterprises (unless said SMB does development or is a technology-focused firm). If IT services were generic and not strategic wages could actually go lower (even without inflation).

4. MSP's offering economy of scale. SaaS vendors offloading the "keeping core applications lights on" and delegating the local staff to increasingly be End User Support.

There is a real secular change to the IT roles in SMB. Some are becoming more an extension of operations and delivering value by being closer to the business. These guys can get 100K. The guys who actually screw with getting GPO to install a printer are becoming a smaller (quantity) and smaller part of the market

I'll note as a market compensation in Houston is a bit odd is tied to rig counts in operation (not the price of the barrel as many mistakenly assume). Dallas and Austin tended to pay a bit less (was a glut of people wanting to live there) while Houston (in good times) and San Antonio tended to offer a better wage to COL ratio in my anecdotal opinion. It's not the 80's and there are other jobs in Houston besides upstream, but it does swing a bit.

@Dashrender said in Nine Out of Every 10 Silicon Valley Jobs Pays Less Than In 1997, Report Finds:

Definitely have a point - but waiting 3 years or more to get your hands on some cash seems boardline crazy risky to me. Those options ideas are likely only for those who can afford to wait on those funds in many cases.

So if you have a 4 year RSU package for 100K, you get 25% of that at 1 year, and each 90 days thereafter 1/12th of it. ESPP ties up your cash for 6 months at a shot and with a lot of plans, you can hit "abort" and get it back with no loss (just time value money loss). Also it's worth noting these type of compensation plans when stacked on even a low six-figure salary that's survivable is what happens. People don't get 400K in RSU's, on a 50K salary.

The only people who have variable compensation plans that look like that are Salespeople who have big OTE (On Target Earnings) salaries that are contingent on hitting thresholds. These sales guys make their real money on accelerators (IE ever $1 past your target of selling your 5 million quota earns you 2x commision, and 3x beyond 200% of quota target). Accelerators can get stupid when sales teams hit "that big one" deal out of the blue (called a bluebird).

But as mentioned - the tax issue and living expense issue has driven these small wages into a real problem.

@Dashrender said in Nine Out of Every 10 Silicon Valley Jobs Pays Less Than In 1997, Report Finds:

Live like it's never going to happen when you bank 2-5 years worth of it, then maybe up your living style - or as you mentioned, use it on some huge capital transaction.

One of my local friend's whos a Sales Rep closed a big deal and I was doing the back of the napkin math to realize he was going to net 70K from the deal. I looked at him and said "Boat?" and he responded "boat house".

The other attitude is some guys in Finance who believe "If you don't live beyond your means a little bit, you'll not stay hungry". Then again, that attitude leads to idiots like this guy who decided naked OTM options on natural gas when winter was coming.

@tonyshowoff said in Nine Out of Every 10 Silicon Valley Jobs Pays Less Than In 1997, Report Finds:

Except, again, I'm talking pre-inflation since then. Let me know when they're paying engineers $190K+ because that's what it would be now. It sounds like a lot now because they've kept wages down against inflation.

and again, I'm not talking base wages. The plans are so stock and bonus heavy now that TC is the number that should be compared not wages. Also, I don't know if H1B wages are a good indicator of a "good" median career engineer (Just looking for public data). Now I'd argue the bigger thing is inflation adjustment's kind of a mess because the real estate market there and tax climate has gotten so ridiculous that's the bigger reason TC has to be as high as it is. It's not about wages being held down, it's about local real COL going to the moon.

@scottalanmiller said in Nine Out of Every 10 Silicon Valley Jobs Pays Less Than In 1997, Report Finds:

I did this for my time in a SV firm. Worked from Spain and Panama while on a SV salary.

The real fun is getting them to pay for the travel

@tonyshowoff said in Nine Out of Every 10 Silicon Valley Jobs Pays Less Than In 1997, Report Finds:

A Good "Engineer" can still make 125K.

We pay H1B's 130K base is what I'm seeing for federal stats. A good engineer working for FAANG, or one of the other big players can break 300K TC. You want to see some crazy high offer letters go look on Blind in the spring when offers start going out to graduates.

Back to my original point, TC isn't Wages.

@scottalanmiller said in Nine Out of Every 10 Silicon Valley Jobs Pays Less Than In 1997, Report Finds:

It's rare to find established companies not paying "anything" and giving all stock. Heavy stock, maybe, but it's above a baseline.

Finance can get crazy on variable bonus, or "skin in the game" (I-Bankers, Private Equity).

Public companies only do it for CEO's as a gimmick "CEO makes $1"

@scottalanmiller said in Nine Out of Every 10 Silicon Valley Jobs Pays Less Than In 1997, Report Finds:

That's pretty common. You only take stock like that if you ...

Are desperate and are willing to take massive risk just to have a job.
Really believe in the company and are into being highly at risk.
Are in a position of serious control and think that you can make a difference.

99% of good people won't consider jobs like that, if the salary itself isn't enough, those "benefits" are considered to be worthless.
Look what happened to the people who took minimum wage and stocks in .... turns out, they were just minimum wage workers.

I'll agree with you mostly on Startups. I think the game is really rigged against the little guy more than ever in that regard.

Publically traded companies where the stock is liquid and risk is a 50x less? That's a different game. Personally, when I evaluated my offer letter I considered the stock compensation and variable bonus at 50% of face value. Now I got lucky and things went the other way but I was still well "in the money" as my new base was still bigger than my old base, but the stock stuff was a fun side gamble that paid off. Note, I also didn't spend more than 5K of my stock comp (vacation), and hoarded it all, diversified it and then used a good chunk of it to buy my new house. The key to variable income is live like you don't need it, and use it for capital purchases that the opex can be controlled.

@Dashrender said in Nine Out of Every 10 Silicon Valley Jobs Pays Less Than In 1997, Report Finds:

The stock options idea sounds good as long as the company's value keeps ratching up.

Stock options are different. Stock Options are generally given for private companies (like Uber, or Slack) and while there are secondary markets to sell these (or the company will generally buy them back at a valuation rate by a 3rd party, this is typically a LOT lower value than publically traded stock that has access to more investors and the liquidity of the HFT bots etc). Options requires some "risk" on the employee, as he can pay taxes on their current value (which is a lot lower) in advance so he doesn't have to pay the full tax value later (It's actually a bit more complicated and in theory you can do this for RSU's but given RSU's are a lot less likely to shoot up this is a lot it's 100x less common). Yes, in this case you could end up paying taxes on stock that never has a real value (the IRS always wins).

What I'm talking about are Restricted Share Units (RSU's). These are stock that is given at zero cost to the employee, and at vestment, if sold can be sold on the open market (I sell mine on E-Trade). If sold on the day that the shares are released they carry the same tax liability as regular income tax (In fact ESPP sold this way will show up on your W-2 and your 1099-C from your brokerage so if you are not careful you can end up paying double taxes).

@Dashrender said in Nine Out of Every 10 Silicon Valley Jobs Pays Less Than In 1997, Report Finds:

If it was a sure thing, why wouldn't the company just keep the stock and sell some when it needs cash.

First off for Options only a Qualified Investor (Someone with over xxx income, or xxx assets) is allowed to invest in a private company's funding rounds. Typically this is done with Venture Capital rounds (or early on with Angel Investors). If you want to see how these deals can go crazy Watch Shark Tank, and read up on what a "Ratchet Clause" is.

If it's a public company the RSU is a PROMISE of future shares assuming the employee stays. It does a few things that are better than paying straight line cash...

1. It lets them spend cash today on expansion (Again, Hypergrowth!) as Time Value of Money means a lot to a company growing triple or double digits YoY. Remember the first 1/4 of that 4 year grand isn't due until one year in.

2. If the Employee leaves the stock is absorbed back to the company. It's a form of "Golden Handcuff".

3. It costs them less than cash. Given my companies stock has gone from a low of ~43 to 149 (today's close) if I had RSU's issued at that dip (or near it) The company can basically have to pay 1/3 today to keep me because it made promises 3 years ago.

4. If you want to get uber machiavellian the company can easily track how much outstanding I have and use it to manipulate who stays and goes. The stock gets cut from 80 to 43? and we want to keep Bob? Let's give him some more shares on a two-year vestment to keep him around until it recovers. Want to get rid of Tom? Stop paying him a variable bonus, and stop re-arming RSU's. He'll leave on his own and not ding your unemployment. This actually makes it safer to "make it rain" for key rainmakers because you are not having to spend as much money, and if it turns out you don't need their skills in 4 years, cutting back on RSU's will claw back that ridiculous compensation package.

@Dashrender said in Nine Out of Every 10 Silicon Valley Jobs Pays Less Than In 1997, Report Finds:

but if it hickups or fails, the employee is boned

True, but if the alternative is 120K vs 100K + [0 to the Moon!] smart employees who can see the momentum of a company are going to bite on the roll of the dice.

@pmoncho said in Nine Out of Every 10 Silicon Valley Jobs Pays Less Than In 1997, Report Finds:

I just hope the IT in Silicone Valley start paying up for the Sanitation employees or they will be in a deep pile of S%&*! (pun intended)

Government employees in the bay make decent money, but it's the pensions that are bankrupting some of the smaller cities.