stevecrox

@stevecrox@kbin.run

This profile is from a federated server and may be incomplete. Browse more on the original instance.

Digital artist here, need help deciding on a beefy machine config, but i'm a little rusty on the IT side of things. (lemmy.world)

So, my budget is 3KGBP, my use case is i do 3d modelling work (have done for 25years now), and would like a better machine that i can do (light) GPU render tests on. It will likely also do short (few days) runs of CPU rendering too....

stevecrox,

When AMD launched Ryzen they deliberately offered way more I/O bandwidth than Intel.

The first generation Ryzen CPU's used RAM frequency that could cause performance issues if you used low frequency RAM. That got fixed in the 3000 series.

There are a small number of Ryzen CPU's which end with "3D," it means they had 3D Cache memory and its supposed to add rediculous performance in certain situations. Phoronix runs tons of benchmarks on CPU and GPU.

The only Intel instructions AMD haven't implemented is AVX-512 and AVX-10. No one uses AVX-512 as Intel CPU's get so hot they performance throttle so much its faster to not use the extension. AVX-10 is something new Intel released this year to get around that.

AMD does support AVX2 which a lot of Audio/Video products do use.

stevecrox,

I think they are saying most attempts at diversity come from middle aged white guys and just end up being poorly done and so detract from the game/story.

Similar to how 00's electronic companies just painted it pink to appeal to women or why South Park added Token.

So arguing for more diversity within the companies themselves

TIL about Sublinks, a Java-based alternative to Lemmy's backend (github.com)

Today I learned about Sublinks (here), an open-source project that aims to be a drop-in replacement for the backend of Lemmy, a federated link aggregator and microblogging platform. Sublinks is designed to be initially API-compatible with Lemmy, allowing existing Lemmy clients, such as Lemmy-UI, to integrate seamlessly....

stevecrox,

Every programming languages has communities built around them.

Its becoming clear Rust solves a lot of C/C++ type problems and the embedded communities are definitely shifting over.

Apache is the primary community for Java, a quick look at their project list shows it's entirely web servers, data engineering and clustered projects for distributed computing.

Personally if you asked me to solve this problem I would use Spring Boot with various Spring libraries for talking to the caddy, user control, etc... Looking at the project, its exactly what they have done

stevecrox,

I wouldn't get massively excited.

Python is a scripting language, its shines when you want to write a stand alone file which takes an input and performs a task. Scripting languages are great to learn as a first language and so python is wonderful for non developers.

The issue you hit is the build management solutions for Python are kind of broken and these help support and encourage good development practice so a lot of Python projects end up a collection of scripts rather than a mature project. You can have good projects but...

In raw benchmarks Java has 90% of the performance of C/C++, but in reality Java is more performant because developers get bogged down in memory management on C/C++ and they get more time to optomise in Java as a result. I'm not sure where Rust will come out to be honest.

Python benchmarks at 50% the performance of Java, in reality I've found code ends up slightly worse because Python is procedural, library support and streaming is poorly supported.

Take library support, Spring really rose to prominese because of 'hibernate' which was a way to abstract talking to different databases through objects, you could switch from PostgreSQL to Oracle through config. Spring data has dumbed this down so I define a plain old Java object and Spring will generate everything I need.

Python expects you to hand craft SQL statements and every database extends SQL slightly differently, so i need to write SQL for every operation and manage/own it. So the win in being able to quickly read/write to a database (since you don't have to learn anything about Spring) is quickly ruined because of the all the boilerplate and error handling you now have to write.

stevecrox, (edited )

Every big UK company I have worked for doesn't own its building. They will typically agree to rent a building for 5-20 years at a fixed rate (longer times if its being purpose built for them) .

So I would expect this is paying out the rest of the rental agreements for a building to escape the building lease.

It is to do with financial reporting and the way asset and operational costs are reported.

stevecrox, (edited )

See its the opposite in Linux land.

AMD open sourced their drivers so everything just works, while Nvidia drivers have to be built against your system and Nvidia refused to supply proper desktop drivers for years (EGLStreams vs GBM).

The downside of AMD's approach is it has to trickle down which depending on what distribution you use can take weeks to a year and it normally takes a couple iterations to get everything working nicely. Which basically expect the 6800 XT to work brilliantly but the 7300 to be flakey for a bit.

My favourite bit is I owned a few Athlon 5300 APU and 5 years after they were released AMD were still adding performance improvements to them.

Meta: How can we grow this community?

The title says it all: How can we grow the Rust community here on Lemmy? Many users fled Reddit or are here for different reasons. But compared to it’s commercial big brother, the Rust community here, feels more or less dead. I would like to discuss ideas, on how we can changes that and make Lemmy the default for Rust related...

stevecrox,

I believe this post would be better if it was rewritten in Rust it would allow more efficent. memory usage compared to; the dynamically typed English language which doesn't have the borrower checker. while allows you to detect when resources are no longer used unlike English's poorly performing 'grammar checking' tools

But seriously there has to be content to engage with and people who respond to the content. I've noticed this community has someone posting really high quality updates but the community appears to be that person.

Posting blogs, or asking questions, etc.. would be a good way to engage.

stevecrox, (edited )

If you click the link and follow it through 3 redirects you'll get to Stamets post.

Basically risa is listed as having "no rules" but the community admin is/was trying to moderate things to keep it "fun".

Stamets and others were upset at the contradiction and through there own posts generated alot of noise to get it resolved the startrek.website admin got sick of the drama.

To be honest the site admin should require mods to accurately explain their actions "Doesn't inspire Jahamora" isn't a great reason and reading stamets examples I think you could build rules fairly easily.

But I won't be going to Ten Forward, reading the post was exhausting.

stevecrox,

Are you saying if Israel competes the UK might get points this year?!?!?

stevecrox,

Immutable distributions won't solve the problem.

You have 3 types of testing unit (descrete part of code), integration (how a software piece works with others) and system testing (e.g. the software running in its environment). Modern software development has build chains to simplify testing all 3 levels.

Debian's change freeze effectively puts a known state of software through system testing. The downside its effecitvely 'free play' testing of the software so it requires a big pool of users and a lot of time to be effective. This means software in debian can use releases up to 3 years old.

Something like Fedora relies on the test packs built into the open source software, the issue here is testing in open source world is really variable in quality. So somethinng like Fedora can pull down broken code that passes its tests and compiles.

The immutable concept is about testing a core set of utilities so you can run the containers of software on top. You haven't stopped the code in the containers being released with bugs or breaking changes you've just given yourself a means to back out of it. It's a band aid to the actual problem.

The solution is to look at core parts of the software stack and look to improve the test infrastructure, phoronix manages to run the latest Kernel's on various types of hardware for benchmarking, why hasn't the Linux foundation set up a computing hall to compile and run system level testing for staged changes?

Similarly website's are largely developed with all 3 levels of testing, using things like Jest/Mocha/etc.. for Unit/Integration testing and Robots/Cypress/Selenium/Storybook/etc.. for system testing. While GTK and KDE apps all have unit/integration tests where are the system level test frameworks?

All this is kinda boring while 'containers!' is exciting new technology

stevecrox,

The shower before a pool is to ensure people aren't entering the pool coated in dirt (e.g. sweat, hair, dead skin, etc..).

The chemicals in a pool are designed to bind to that dirt and kill any bacteria introduced.

There is a limit to the chemicals you can add to a pool (before it hurts humans) and once the amount has activated you need to drain the pool and refill it.

Swimming pools hold crazy amounts of water which is also really expensive to heat up, so pools want to do that as little as possible.

Clothing interfers with cleaning your body, so people entering near fully clothed (e.g. like a Burkina) will likely introduce more dirt into the pool.

That translates into increased costs for swimming pools or pools which maintain the old schedule and just operate unsafely.

This is all based on owning a hot tub and learning how to maintain it.

Hopefully this also explains why it doesn't matter people enter the sea fully clothed

stevecrox,

You seem to be intentionally missing the point, but to reiterate..

You shower before entering a pool to wash the dirt from your body off (your cleaning yourself).

The more of your body covered the less effective that shower is.

Ideally everyone would be naked in the shower, but there are probably outfits which increasingly render the shower less and less effective (e.g. speedos are better than shorts, etc .).

It would not surprise me if a Burkina covered so much that the cleaning shower is rendered pointless

stevecrox, (edited )

The developer behind KBin seems to have issues delegating/accepting contributors.

If you look at the pull requests, most have been unreviewed for months and he tends to regularly push his branches once complete and just merge them in.

That behaviour drove the MBin fork, where 4-5 people were really keen to contribute but were frustrated.

To some extent that would be ok, its his project and if he doesn't want to encourage contributions that is his decision but...

KBin.social has gotten to the size where it really should have multiple admins (or a paid full time person). Which it doesn't have.

The developer has also told us he has gone through a divorce, moved into his own place, gotten a full time job and now had surgery.

Thats a lot for any normal person and he is going through that while trying to wear 2 hats (dev & ops) each of which would consume most of your free time.

Personally I moved to kbin.run which is run by one of the MBin devs

stevecrox,

MBin is a fork by a group who tried to push into KBin but couldn't. There seems to be at least 4 active committers and stuff gets merged.

You will see a number of the KBin instances moved over https://fedidb.org/software/mbin

stevecrox,

It does but for the 90's/00's a computer typically meant Windows.

The ops staff would all be 'Microsoft Certified Engineers', the project managers had heard of Microsoft FuD about open source and every graduate would have been taught programming via Visual Studio.

Then you have regulatory hurdles, for example in 2010 I was working on an 'embedded' platform on a first generation Intel Atom platform. Due to power constraints I suggested we use Linux. It worked brilliantly.

Government regulations required anti virus from an approved list and an OS that had been accredited by a specific body.

The only accredited OS's were Windows and the approved Anti Viruses only supported Windows. Which is how I got to spend 3 months learning how to cut XP embedded down to nothing.

stevecrox,

Technical Leads are not rational beings and lots of software is developed from an emotional stand point.

Engineering is trade offs, every technical decision you make has a pro/con.

What you should do is write out the core requirements/constraints.Then you weigh the choices to select the option that best meets it.

What actually happens is someone really likes X framework, Y programming language or Z methodology and so decides the solution and then looks for reasons to justify it.

Currently the obvious tell is if they pitch Rust. I am not saying Rust is bad, but you'll notice they will extoll the memory safety or performance and forget about the actual requirements of the project.

stevecrox,

The team/organisations knowledge is a huge factor but its easy to fall into a trap where no matter what the problem is the solution is X language.

If I have an organisation that knows C# and we need to build a Web Application. I would suggest we need to learn Node.js and Typescript and not invest in a solution that turns C# into web pages.

stevecrox,

You can't fix a people problem with process.

For example I've worked in DevSecOps for 10+ years, whenever consulting my first step is to implement a CI that picks up Pull Requests, builds them and runs a code analysis tools (e.g. pep8, spotbugs, eslint, etc..) and have the CI comment the Pull Request. The idea is to get an understanding of the projects technical debt and stop things getting worse and ensure the solution 'just works'.

Teams with huge amounts of technical debt will find a way to disable it when your not looking. They will develop all kinds of reasons and in reality the technical debt was created because of cultural issues in the team.

So I've learnt its important if you spot a team doing that, the solution isn't locking it down the solution so they can't disable it or more process. But forcing out the technical leader and sitting with the team and working out why each one is fighting the tool and not seeing them as an asset and teaching them to be better.

stevecrox,

DevSecOps is all about process, I simplified my answer.

At a high level in software there will always be a review phase, where code needs to be built and pass tests (as a minimum). With Git being used by every organisation I have been involved in you will find the organisation/team will claim to follow a variation of 'Feature branch Workflow' with review happening at a specific point (Pull Request).

For the last ~10 years every organisation/team I've worked in/with has claimed to use a CI to verify the source code as part of that review phase.

In most dysfunctional teams that review phase will be broken, the fastest win is to bring in the CI. Static analysis tools are also impartial in how they review and useful in teaching people how to review.

I don't say your project must build with no warnings, I say you project must build and I'll have the CI point out where you have added warnings as part of review. When people complain I'll point to their teams Ways of Working or an organisations Software Development Plan (or in one case a System Engineering Management Plan).

The sort of team that then chooses to disable this will do so because the leadership are undermining it (normally a team lead turns it off or tells them to just ignore it). There seem to be a few common reasons as to why team leads will do this but it isn't something you can rationally debate with. The only solution I've found is to sideline the problem, change team culture and identify a leader within the team and hand it over to them.

Your talking about teams which are failing due to the environment, those normally understand what is wrong and the best approach is to be a good scrum master (e.g. run retro sessions, identify issues and work to resolve the environment problems with them).

stevecrox,

There will always be someone who is beating you in a metric (buying houses, having kids, promotions, pay, relationships, etc..) fixating on it will drive you mad.

Instead you should compare your current status against where you were and appreciate how you are moving forward

As for age

During university my best mate was 27 who dropped out of his final year, grabbed a random job, then went to college to get a BTEC so they could start the degree.

It was similar in my graduate intake, we had a 26 year old who had been a brickie for 5 years before getting a comp sci degree.

The first person I line managed was a junior 15 years older than me, who had a completely different career stream. They had the house, kids, had managed big teams, etc.. honestly I learnt tons from them.

stevecrox,

Java's niche is middleware and 'big data', so choosing it for a Lemmy drop in replacement an excellent choice.

Programming goes through various fads where a particular language is everyones favourite for a while (see the plethora of this is X tool .. rewritten in Rust!).

You've had Scala, Node.JS, Go, Python and now Rust try to supplant Java in its middleware role. Each one has tried to highlight reasons why Java is bad to justify the switch.

stevecrox,

Have you considered migrating to a magazine on Kbin/MBin?

The daily updates are quite useful but currently my means of finding them is via your posts in Reddit. The post views in Lemmy discourage you to go into spaces and I don't think thread pinning translates accross instances.

KBin/Mbin support the short message format in Activity Pub so creating a thread on that platform should reflect into Mastodon and Mastodon comments will come through into the post. If you add a couple of hashtags it would also help discoverability.

Your posts are the most useful ones for daily summaries and probably the most useful contribution to the reddit starship dev thread.

stevecrox,

I shall have a think where it makes sense to create one. Lits were created a while back and seem to have been purged

Short Message Format = Tweet
Thread = post

KBin calls a community a "magazine" and Magazine mods can associate hashtags with the Magazine.

The idea is from your instance you can look at the thread view or microblog (tweet) view.

KBin was started by a person called Ernest, he doesn't seem particularly great at accepting outside help. A number of people who wanted to contribute but found their work stuck for review for months forked KBin and called it MBin.

  • All
  • Subscribed
  • Moderated
  • Favorites
  • JUstTest
  • ngwrru68w68
  • everett
  • InstantRegret
  • magazineikmin
  • thenastyranch
  • rosin
  • GTA5RPClips
  • Durango
  • Youngstown
  • slotface
  • khanakhh
  • kavyap
  • DreamBathrooms
  • provamag3
  • tacticalgear
  • osvaldo12
  • tester
  • cubers
  • cisconetworking
  • mdbf
  • ethstaker
  • modclub
  • Leos
  • anitta
  • normalnudes
  • megavids
  • lostlight
  • All magazines