bvrmn

23 ms·

361.
by bvrmn 3y ago
Awful solutions TBH. They are quite hard to implement (requirements are coupled and could not be composed, logic heavy details) very specific and non-extendable for requirements changes in any way. You don't want this kind of code in a
362.
by bvrmn 3y ago
> Yeah, the mental model of Mercurial is way better. It's more complex comparing with git naked graph. Contains more domain entities. But it could be more convenient for starters and SVN users.
363.
by bvrmn 3y ago
For me it was a quite important factor TBH in 2007 when I wanted to try something new outside of SVN. And it was my big argue point to force my company to switch to Git.
364.
by bvrmn 3y ago
> the osi model layers describe exactly what is going on in embedded systems Unless it doesn't. OSI describes non-existing stack from the long gone past.
365.
by bvrmn 3y ago
It would be wonderful to see your non-coupled design without middle layers knowing about top context to handle errors without bubbling it up. It already sounds like a failure.
366.
by bvrmn 3y ago
Because local objects names could be mapped freely to anything, it's only a convention to use <remote>/ as a prefix.
367.
by bvrmn 3y ago
From my teaching experience it could be a fear to do something wrong. I've come from SVN and ClearCase where doing wrong was quite a trivial thing. And Git relaxed me a lot. Knowing only push/pull makes actual changes and you have
368.
by bvrmn 3y ago
Fracturing a big commit to smaller parts doesn't help. Amount of work is the same, number of comments would be higher or same due to bigger context loss. GH PR UI is notoriously confusing while reviewing multi commit PRs.
369.
by bvrmn 3y ago
Changesets in Gerrit are more easier to manage from processes and UX standpoint. There is no branch/commit/comment/rebase/commit/resolve-staled-comment/fight-for-true-pr-merging-strategy dance. I'm doing a
370.
by bvrmn 3y ago
RIP. Vim is my main editor for 15 years already. I spend several hours in Vim everyday. It's so ingrained in my soul, I can't imagine how to work without Vim.
371.
by bvrmn 3y ago
A single Go thread still replaces 10 Python threads, speaking very roughly. It's a quite particular narrow set of problems noGIL would solve.
372.
by bvrmn 3y ago
I currently try to create a some middle ground. Basic idea: use eDSL to output drawio xml and use drawio to render a final diagram. drawio could back many complex things like arrow routing. WIP output example: https://docs.tuxcar
373.
by bvrmn 3y ago
There is no advantages if you want to splash node relations and get quick and ugly diagram :)).
374.
by bvrmn 3y ago
Pikchr is quite low-level and could require a considerable high amount of work to get a result.
375.
by bvrmn 3y ago
It's SQLite's locking implementation. In real code it includes some backoff but on long timeouts it reaches 3s. Jitter doesn't help BTW it's still unfair. On high contention each worker only has a slim chance to get a lo
376.
by bvrmn 3y ago
Yes. It's doesn't matter wal or default behavior.
377.
by bvrmn 3y ago
AFAIK SQLite has weird locks, half year ago they weren't fair and writers could get timeout easily. For workaround we implemented external lock and it have been worked quite smoothly since. SQLite's locking code looks like:
378.
by bvrmn 3y ago
Tinc is a perfect tool to make a VPN mesh across different clouds/hosters. Been using it for 5 years. It's so much easier in support comparing with ipsec madness.
379.
by bvrmn 3y ago
> Five years So basically 2 times less, nice.
380.
by bvrmn 3y ago
AFAIK, there is no rule RHEL is strictly based on Stream.
381.
by bvrmn 3y ago
Some reasons why Stream is not replacement for Centos alike (Alma, Rocky, ...). 1. Stream can't be used as base to build el-compatible packages. There is no any guarantees Stream doesn't break ABI compatibility. 2. Stream has no l
382.
by bvrmn 3y ago
And what the point to make `if err {return err}` in most of cases? I've read a lot of code and error handling logic almost never in caller location.
383.
by bvrmn 3y ago
In Rust you have convenient escape hatch with a default allocator. It's similar to say "you can do safe applications using C or C++".
384.
by bvrmn 3y ago
You can choose own color!
385.
by bvrmn 3y ago
I can clearly see a raw AST of imperative language behind html tags. It's not obvious why it's better than JS.
386.
by bvrmn 4y ago
Author literally gives a plot description. Review itself is a one paragraph. It's not a kind of review one can expect from a movie critic.
387.
by bvrmn 4y ago
I own Valve Index. State of the industry is not very bright. The best VR game so far is HL2 VR mod. Alyx the second. Looks like investors are pretty shy to put money in VR game studios to make big and story rich content.
388.
by bvrmn 4y ago
> When I came to the valley in 2012, I had a manager who wrote as much code and shipped as much as I did, and prevailing wisdom was that this was a good thing. My team lost manager almost 8 month ago. And I took additional management rol
389.
by bvrmn 4y ago
Looks like you need a monorepo.
390.
by bvrmn 4y ago
It's better to make exception handling explicit. It has the same semantic complexity comparing to passing possible good exception types into retry or attempt. Inline loop logic to decide if iteration is successful can be tricky.

More ›