chantepierre

To add to this topic, people who do not know about erlang's hot code loading should watch this talk : https://www.youtube.com/watch?v=pQ0CvjAJXz4

A multi-DC running cluster where parts are progressively swapped at runtime. No database, only OTP.

Personal habit, honestly. If I had to re-start I'd probably pick React today, but the Vue switch to a composition API make me like it enough when I was thinking of switching.

What I'm truly rooting for is a non-limited Elm.

I'm not using this to update the app itself - which is a docker container that gets updated when I push a new version. I'm simply using the BEAM's code loading capabilities to add client-specific parts to the app while it is running. They are part of the monorepo (and thus part of the app at dev time), but get stripped at build-time so they can be selectively loaded later.

Author here, thanks for sharing ! Happy to answer any questions.

I think the article outlines it, but I'm at very low scale, with custom development for every client. I mostly build mini-figmas, collaborative or not, that automate specific document pipelines on top of my software, backed by elixir+liveview (or elixir+vue+channels).

Still working on Alzo [1], my services startup for french architecture companies. These days I spend most of my time writing client-specific apps [2] that are hot loaded on demand into my Elixir monolith. So far I like this architecture a lot because it is hard to break anything in it.

[1] https://alzo.archi

[2] https://lucassifoni.info/blog/leveraging-hot-code-loading-fo...

I'll push him to tidy this up :)

Hey, I missed your comment. Nowadays the best source for optics is often a secondhand instrument where you scavenge the optics. If we specifically talk about a 6"f/5, most recent industrial tubes have really decent optics.

Buying new optics online, avoid any ad where the price seems surprisingly low. For example you can find 6"f/5 mirrors advertised as lambda/8 precision. There's a lot of abuse with this denomination : some sellers use surface precision, whereas we usually talk about reflected wavefront, so lambda/8 becomes lambda/4.

Then, there are a lot of ads for optics at great prices that aren't specified as parabolic (in the case of newtonian mirrors). Most often those are spherical, which is totally unsuitable except for longer focal ratios (and that quickly becomes impossible as diameter grows).

On the other hand, I recently ordered a 40$ "lambda/8" spherical mirror, and it was a great sphere, so a great starting point to polish it to a parabola.

You need smooth movement on altitude and azimuth axes, then the telescope must be still and vibration-free to enjoy the view. The other constraint is structural integrity to maintain alignment of the optics. Those two things are the basis of all builds (except for optics!).

I don't have enough experience with nylon to assess whether it would be suitable for a full telescope, sorry.

The 6" f/5 can often be found cheap and without a mount, but if you are beginning, absolutely find some secondhand dobson before going DIY.

Or better : attend a few nights at a nearby astronomy club, that will allow you to get a feel of the practice, what you can see, what you should expect...

Depending on your location (city? middle of nowhere?) and willingness to drive to darker skies some types of instruments and targets are better than others.

This kind of kit is focused on being compact and lightweight, but that can be an advantage or a non-goal depending on what you want to do.

It's a compromise between having a geometry that is rigid enough to ensure the bottom and top planes stay aligned while the telescope slews up and down, while having a practical shape at the bottom (a square) and avoiding struts going into the light path, which happens if you do an hexapod shape without making the base larger to compensate. Since his initial goal was to make this printable on an ender3 build plate, the smaller square shape was preferable.

Dobsonian mounts are made to feel great and work with friction between (here) plastic and teflon pads, both in the altitude and azimuth axis. The goal is to have your telescope properly balanced so there is no need to apply force to hold it in position, nor need more force to start a movement than to keep moving, which would lead to jerky motion. Properly balanced, everything must feel smooth and easy.

There are a bit more information on his website [1] and I've made a temporary grouped BOM [2] :

    [1] https://la3emedim.fr/assets/SMALLEST%20-%20README.pdf  
    [2] https://docs.google.com/spreadsheets/d/1XXZB4jRtvtlDIqlR69C0FRri8x8lZDpxO6ZGnSp3L2Y/view

Basically anything but PLA will work, PETG will be the best difficulty/cost compromise.

Yeah his site isn't currently the clearest for that.

Here's a direct link to his build instructions [1] and I've made a temporary by-item BOM [2] while he improves things.

PETG is great, ASA and ABS might be better if you have the capacity.

     [1] https://la3emedim.fr/assets/SMALLEST%20-%20README.pdf  
     [2] https://docs.google.com/spreadsheets/d/1XXZB4jRtvtlDIqlR69C0FRri8x8lZDpxO6ZGnSp3L2Y/view

For budget entry eyepieces, the generic "goldline" eyepieces have pretty good bang for their buck, especially the 9mm/6mm ones. You can find them unbranded online, or the "redline" equivalents.

I'd recommend a quality 20/25mm plossl rather than the gold/red line 15/20mm that display a lot of issues.

For better eyepieces, at f/5, the rubber-top generic eyepieces with color bands sometimes branded as "artesky" are great for their prices too.

If you have a bit more budget, Explore Scientific is generally the overall best quality/price ratio. Avoid baader hyperions, except for the 13mm which hits above its price (second-hand)

6" f/5 kits sold by GSO are okay, but the best choice is to find any second-hand skywatcher 6" f/5 tube on your leboncoin/craigslist equivalent and scavenge the optics. For industrial optics, they are on average exceptionnally smooth and have a good shape with few defects.

The 6" f/5 optics kit are a bit of an oddity, those we test with our interferometers are consistently good, whereas smaller or bigger diameters are subject to a lot more vatiations.

The rods are heatshrink tubing wrapped-aluminum. I personnally have built a few partially 3d printed, or full 3d printed (except for a few hardware items and carbon fiber rods) telescopes – it all depends on print preparation skill, correctly drawing for 3d printing, and wise choice of material and print conditions.

The Hadley is the perfect gateway to telescope building.

A 6" packs a lot more punch in a not too much bigger package, and 6" f/5 consumer mirror kits that we test often range from okay (0.77+ strehl) to actually quite good (around 0.92).

You'd get quite an improvement on planets !

Hello HN,

A friend released files for this 6" f/5 telescope conversion kit yesterday.

It is a synthesis of a lot of telescope builds he, I and a few friends made over the last years, as well as general ideas floating in the amateur telescope making community.

It is compatible with most existing 6" f/5 newtonian optics kits. I've tried his model IRL and was pleased with both the movements and the focuser, which is often lacking on commercial offerings (which are also heavier and bulkier).

Unmounted, it forms a quite compact package to hike with or take camping.

It is licensed under CC-BY-NC-SA as he would like to sell parts kit at some point, but the files found on Printables are fully functional.

I'm currently donating monthly (small amounts alas) to :

    - Hologram, an isomorphic web framework in Elixir by Bart Blast  
    - Torchlensmaker, differentiable optics with pyTorch by Victor Poughon

Both projects are to this day small and one-person but they have in common to be new takes at existing problems. Maybe those specific projects will not become the default, but I believe they will influence by showing alternative ways of doing things.

If you want a purely visual experience (which I recommend), a dobsonian telescope has the best capability VS price ratio. The collapsible 150mm dobsons are already very powerful and quite compact. Avoid anything on a tripod under 500$, you'd be paying the "looks like a telescope" tax and have something deceptive.

That sounds about in line with a microlithography vendor (UK) I ordered zoneplates from. 7-10µm features were accessible and going to 0.7µm was an order of magnitude higher (and went from film to chrome on glass IIRC). File size was also a factor for more complex patterns.

Do not hesitate to post results on cloudynights or your club's website, I'll refresh it from time to time ! I do not think I have enough diffraction experience to help though.

Best of luck for your endaevor.

That's awesome, are you using lithographic processes to produce the diffractive elements ?

It's not about price, but enjoying the process, or making instruments that just are not commercially available. Amateurs are constantly pushing the limits of optical designs.

Here is a recent groundbreaking example from Rik Ter Horst, a 10" f/20 kutter with toroidal secondary : https://www.cloudynights.com/topic/935825-a-250-mm-f20-kutte...

I did not do anything that impressive but enjoy my own set of non-standard scopes :

- A 8" f/3.5 hyperbolic primary + ross corrector

- a 6" f/2.8 with permanently mounted 4-element corrector

and am working on a 16.5" f/3.3 that will have a permanent Paracorr II lens group.

The only messy moment was my first mirror when I learned to cut pitch, but grit and glass dust itself are no problem thanks to the wet process.

You are right though on the last bits of figuring, it's a psychological challenge and my first mirror took 3 attempts over 18 months.

I'm truly sorry for your experience. A first mirror can be a lot of process and can indeed get a bit messy. With experience (or guidance, which is even better) you can work cleanly and avoid glass dust contamination of your workspace.

Telescope making is much alive and there are communities of people (even young people) making their first mirrors right now. Most find their entry in the hobby via the forums (CloudyNights's ATM, Optics & DIY forum, Stargazerslounge, Astrosurf, Astrotreff.de) and amateur mirror maker Discord channels are popping up.

I also recommend anyone wanting to grind their first mirror to read about modern ways of testing in addition to all the classic books (Texereau, Sam Brown, Lecleire) about mirror making.

Bath interferometers changed the game and allow to reach λ/10 wavefront with certainty and repeatability compared to Foucault testing. They are affordable and there's a healthy community around DFTFringe, the de-facto standard interferogram analysis software at interferometry.groups.io

You can also find a Foucault + Ronchi + Bath combo tester's plans on Printables.com and a companion three-axis-table, allowing great testing ergonomics for a low cost if you have access to 3D Printing.

The best resources on how to setup a Bath Inteferometer can be found on the GAP47's website (french, but machine translatable) and GR5's YouTube channel.

Have fun :)

There are dedicated "dirty" schedulers for long-running NIFs that by nature cannot be preempted, to avoid interfering with BEAM process scheduling : https://www.erlang.org/doc/apps/runtime_tools/scheduler.html

I love to see "well-known" people in the Elixir community endorsing and actively developing that kind of approach. Our VM and runtime does so much and is so well suited to orchestrating other languages and tech that it sometimes feels there's a standard track and an off-road track.

The difference between an off-road "sounds dangerous" idea and its safe execution is often only the quantity of work but our runtime encourages that. Here, it's a NIF so there's still a bit of risk, but it's always possible to spawn a separate BEAM instance and distribute yourself with it.

Toy example that illustrates it, first crashing with a NIF that is made to segfault :

  my_nif_app iex --name my_app@127.0.0.1 --cookie cookie -S mix
  iex(my_app@127.0.0.1)1> MyNifApp.crash
  [1]    97437 segmentation fault

In the second example, we have a "SafeNif" module that spawns another elixir node, connects to it, and runs the unsafe operation on it.

  my_nif_app iex --name my_app@127.0.0.1 --cookie cookie -S mix
  iex(my_app@127.0.0.1)1> MyNifApp.SafeNif.call(MyNifApp, :crash, [])
  Starting temporary node: safe_nif_4998973
  Starting node with: elixir --name safe_nif_4998973@127.0.0.1 --cookie :cookie --no-halt /tmp/safe_nif_4998973_init.exs
  Successfully connected to temporary node
  Calling MyNifApp.crash() on temporary node
  :error
  iex(my_app@127.0.0.1)2>

Thankfully Python, Zig and Rust should be good to go without that kind of dance :) .

Continuing bootstrapping my software+service company Alzo (https://alzo.archi/), an Elixir modular monolith.

The first clients are here and I am working on "darker tech" now, a single codebase injecting their data in both MS Office and Adobe's software suites. That's quite a change from Elixir.

Also working on the reverse feature, reading/writing MS Office files inside Alzo. For that, I'm writing a Java app behaving as an Erlang node, connected to my main app via Erlang distribution, to leverage existing rich Java libraries for office tech.

Working on Alzo ( https://alzo.archi ), a half-SAAS, half-service business giving knowledge management + call for offers / layout automation to french, mid-sized architecture agencies.

It's kind of niche but I'm doing more and more automation on my side, to efficiently build custom one-off layout editors for my clients. Some have surprisingly specific wants (for example : "multi-player live google-slides like editor, but constrained to brand guidelines and layout systems with block-level pick and place across all documents").

I am having lots of fun with that and am trying to move from one-off bespoke systems to a common base + client-specific code.

Edit : it's 90% Elixir and designed to run both on-prem or as a managed service.