Hacker Newsnew | past | comments | ask | show | jobs | submit | more lmarcos's commentslogin

Same. I always work on any given (hard enough) problem twice or three times. The first time I come up with some kind of prototype that may work or not, but it gives me foundation to work on the second prototype. The second one tells me how a final working and performant solution could look like, so I discard it and start from scratch. At the end I come up with the best possible solution I can provide. Sometimes it takes fewer steps, sometimes more. And more often than not I let some time pass between attempts (maybe I go for a walk in between or I let a full day pass to work on something else).

At interviews they stop me at my first attempt.


The dragon book is a theoretical book. It's like any other (good) OS book out there: good source of information to know the foundations. After reading it, go out and read something more practical (which you probably won't have issues learning it since you now know the foundations)


We can use ChatGPT for that.


The problem, I think, is what kind of employees may leave. If it's the best of the cream, then the company is in trouble. And frankly, the best employees can allow themselves to say "I want to work from home. You don't allow? Then goodbye".


In my opinion, above a certain size, companies don't really care about employee quality all that much. So long as they meet the initial hire criteria it doesn't matter who's a 5x or 1x dev anymore. It's just one FTE anyway in the planning.


At this point the company has become a bullshit churning machine whose effectiveness is measured by its rate of churn. They milk cash cows that gained traction in decades past, while most new work is just part of the churn.


I think you are 100% correct that they don't care, that doesn't make it a smart decision. It's a good way to die a slow death.


dirty secret is, they'll just allow their best employees to work remote


Until this spreads and it backfires.

We had an example at our office several years ago where employee X was a great worker, nothing to say about that, that was having his own "working" rhythm (e.g., until late at night, mornings off, etc.), and was coming to the office only from time to time, just because the manager was really pushing hard, or for certain office events.

This spread and people started to ask - can I work from home? -> only 1 day Why? X does more than 1 day, and sometimes X is not available when I need.

... as a manager you can only "yes, but ..." that much. At a certain point people (who I would like to remind everyone again and again: most of the time are people with a high college degree, used to read books, papers, and do things, not 3 years old kids) connect the dots and say "alright, it's time to go". And before they go, they let the entire apple tree rot to hell. If that doesn't happen it might be because this spreads and goes to high management that asks the manager "WTF are you doing?"

That's when either more rights come in equally for everyone, or ... they need to let go of their best employee, or they need to find a compromise. Just having the nicely protected "best" employee is never a good strategy. Eventually people get pissed off and leave, and if they don't, congrats, you have just made the "non-best" employees even less motivated and less productive.


This was so before COVID. If you are good and productive you can often negotiate WFH like any other benefit.

Honestly it seems like it’s part of a tech career trajectory. Once you are very established and your skills are built up there are several tracks available to you. Two of those are WFH specialist or 5-10X experienced dev and WFH consultant.

The cube farm grind is something you go through to get your skills and network up.

I have seen the same in other knowledge areas: corporate law, finance, some types of editing and design, accounting, etc. WFH becomes one of the options at more senior levels.

The other reason it’s a more senior option is that senior people are seen as better able to self direct and thus requiring less micromanagement.


That seems to be true in my case, the first decade of my career was spent in cube farms and cramped offices. I learned a lot and made a few networks, and now I’m at a fortunate point where most jobs are available to me with a phone call if my current job sucks. Especiallly when I’m asked to RTO.


This was always so.

Pre-covid my entire team was office and I was the only remote. Without wanting to brag, it's factual: my Drupal knowledge made it worth for them to keep this up. Post-covid my entire team is scattered to the winds. You couldn't get them to come to the office because almost no one is within the same metro as the office.


It’s never bragging to say you work with Drupal ;)


Yeah, there is an example of OpenAI: Sam Altman says people should work in the office, however they are ready to hire exceptional talent anywhere.


The dirtier secret might be that the remote employees aren't necessary the best employees.


Yeah either they do that or they'll burn a lot of good will


They are going to quickly discover that this is the case. I left Lockheed right after they started forcing everyone back into offices that didn't even have enough desks. Those whom were actually creating value there all had plans to leave to. Plenty of them left as soon as they got the news lol. Sure it's a way to do layoffs without saying 'layoff', but the impact is going to be felt much harder down the line.


I'm not sure that there is a terrible large correlation between "best of the best" and "refuse to ever work in office." When I think about the engineers that I most respect and admire, they are very split on wfh vs wfo opinions.


[flagged]


And what exactly is taking Windows place on PCs? Microsoft’s market share hasn’t budged in decades.


I don't think this is correct. It is a little hard to tease out perfectly (e.g.: school computers having gone to chrome books in a big way might not be the statistic you want), but globally Windows marketshare has gone from 90% a decade ago to ~%75 now:

https://www.statista.com/statistics/218089/global-market-sha...

About half of that is to MacOS, but linux and ChromeOS have also made gains.


I'm not a part of this space personally so I can't say for certain, but one answer could be the not-PCs and cloud. Some people find a tablet with a keyboard perfectly adequate for their all-office needs. Even if the "sys-admins" still need PCs that is still a very substantial conversion of the market share away from PCs. If you reflexively scream at the notion of how horrible that would be for your working process, I do not blame you but note it could work for some.


Do you really need to be part of “this space” to know that Microsoft has a 90% share in PCs and most of corporate America and even personal computers are running Windows?

Even the people who like tablets are still buying Surface laptops - running Windows.

They aren’t buying iPads and definitely not Android tablets.

Even a lot of iPad aficionados are moving back to Macs now that they have a lot of the advantages of iPads - slim, lightweight and ridiculously long battery life.


[flagged]


You can't post like this to HN, regardless of how wrong another post is or you feel it is.

As I just mentioned at https://news.ycombinator.com/item?id=35965090, if you keep this up, we're going to have to ban you. Could you please review https://news.ycombinator.com/newsguidelines.html and stick to the rules?


it has indeed "budged" from 92% to 56%


AI should follow the same principle. It has the potential to become as widespread as the web... And yet here we are paying a fee per N tokens to use GPT.


Allowing suppliers to make money is a standard way to induce more supply.


Just hire hundreds of people to manually check the uploads. Easy. Sometimes, some things cannot be automated.


Is not ok, but Microsoft couldn't care less (because they are not going to get fined).


yes, because they don't indemnify their customers

anyone sensible should stay the hell away from copilot until the fair use question is settled


Looks like they do.

https://github.com/customer-terms/github-copilot-product-spe...

4. Defense of Third Party Claims. If your Agreement provides for the defense of third party claims, that provision will apply to your use of GitHub Copilot. Notwithstanding any other language in your Agreement, any GitHub defense obligations related to your use of GitHub Copilot do not apply if (i) the claim is based on Code that differs from a Suggestion provided by GitHub Copilot, or (ii) you have not enabled all filtering features available in GitHub Copilot.


interesting

> If your Agreement provides for the defense of third party claims

do any of them?

it also states:

> You retain all responsibility for Your Code, including Suggestions you include in Your Code or reference to develop Your Code. It is entirely your decision whether to use Suggestions generated by GitHub Copilot. If you use Suggestions, GitHub strongly recommends that you have reasonable policies and practices in place designed to prevent the use of a Suggestion in a way that may violate the rights of others. This includes, but is not limited to, using all filtering features available in GitHub Copilot.

(contra proferentem would apply though)


I think it's pretty clear. If you're not filtering, you're liable. If you are and something transpires, they'll fight your legal battle for you which is probably better than any monetary indemnity clause. I assume this is for enterprise users where it actually matters.


> I think it's pretty clear.

without an Agreement in sight, it's not, the two terms conflict with each other as there's no clear precedence

(I think it's likely that there's an indemnity clause in any Agreement though!)


Looks like copilot is pretty upfront:

Matched content:

n ; Ap = A->p ; Ai = A->i ; Ax = A->x ; for (j = 0 ; j < n ; j++) { for (p = Ap [j] ; p < Ap [j+1] ; p++) { y [Ai [p]] += Ax [p] * x [j]

License Summary

This snippet matches 500 references to public code. Below, you can find links to a sample of 50 of these references.

NOASSERTION (405)

MIT (26)

GPL-3.0 (19)

BSD-3-Clause (16)

GPL-2.0 (11)

Apache-2.0 (7)

BSD-2-Clause (7)

LGPL-3.0 (6)

LGPL-2.1 (3)

File References

Match Location Repo License

ChRis6/circuit-simulation Unknown license

AndySomogyi/SuiteSparse Unknown license

ru-wang/slam-plus-plus Unknown license

Cruvadio/invariant_measures Unknown license

nishant-sachdeva/rrc-g2o Unknown license

alecone/ROS_project Unknown license

gustavopr/HANK Unknown license

lcnbeapp/beapp Unknown license

imod-mirror/IMOD Unknown license

clach/MPM Unknown license

MagicPixel-Dev/cxsparse Unknown license

elshafeh/own Unknown license

squirrel-project/squirrel_nav Unknown license

lcnhappe/happe Unknown license

cix1/OpenSees Unknown license

pachamaltese/dulmagemendelsohn Unknown license

gina10287/Interactive-shape-manipulation-FinalProject Unknown license

GuillaumeFuchs/Ensimag Unknown license

w2fish/CSparse Unknown license

cffjiang/cis563-2019-assignment Unknown license

diesendruck/gp Unknown license

hendersk101401/jlabgroovy Unknown license

robhemsley/SuiteSparse Unknown license

Glaphy/Emission Unknown license

Glaphy/Emission Unknown license

daves-devel/ECE1387 Unknown license

anranknight/TE Unknown license

weigouheiniu/TE Unknown license

Datoow/fm Unknown license

chaoyan1037/openMVG_modified Unknown license

cran/igraph Unknown license

GHilmarG/UaSource Unknown license

ZhaoqunZhong/Kalibr-ubuntu18-ros-melodic Unknown license

hechzh/g2o Unknown license

Open-Systems-Pharmacology/OSPSuite.CPP-Toolbox Unknown license

elshafeh/own Unknown license

yizhang/riotstore Unknown license

sgalazka/porr_mtsp Unknown license

skydave/sandbox Unknown license

alitekin2fx/orb_slam2_android Unknown license

Tianhonghai/vslam14_note MIT

LRMPUT/PlaneSLAM MIT

albansouche/Open-GeoNabla GPL-3.0

khawajamechatronics/mrpt-1.5.3 BSD-3-Clause

3000huyang/suitesparse-metis-for-windows BSD-3-Clause

igraph/igraph GPL-2.0

LRMPUT/DiamentowyGrant Apache-2.0

kurshakuz/graduation-project BSD-2-Clause

ghorn/debian-casadi LGPL-3.0

rmcgibbo/tungsten LGPL-2.1

Looks like this code in https://github.com/ChRis6/circuit-simulation/blob/2e45c7db01... is older then the GPL code in question or provided by the example. Uh oh, did we discover something? Who actually owns this code because this code predates the code in question by a calendar year using git blame, by a different author, and with no license attached to the oldest code. Is it possible the code in the codeium.com example is relicensed and not GPL code at all?


Narrow perspective. What about all the other non-tech jobs?


Unemployment tends to be lower here than in say France. Don't get me wrong, it's not the best system here in the states ( mainly since healthcare is tied to your job), but it has its upsides .


Even though I know OP lives in US: in Europe you get written down the notice period on your contract. If you decide to quit or if your company decides to fire you, both parties have to honor the notice period (usually between 1-3 months).


I personally am okay with the American model tbf.

When I want to quit immediately, I don't want to stay miserable in the same job for 1-3 months. 3 months would be insane for me. As long as it works both ways (I can quit whenever, and you can fire me whenever) - it's fine.

Acceptable. Preferred even.


If you get fired in Switzerland, your employer gives you a notice period of 3 months as well. Which is nice because it's plenty of time to find another job.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: