HN user

ahmadassaf

128 karma

VP of AI @beamery, PhD in Semantic Web and Information Retrieval. Knowledge seeker and food lover

Posts81
Comments44
View on HN
www.flowstate.inc 3mo ago

Know where your AI budget is going

ahmadassaf
1pts0
www.deepmyst.com 3mo ago

One API for every AI model. Lower costs on autopilot

ahmadassaf
1pts0
github.com 3mo ago

A cross-shell framework for managing aliases, plugins, completions, and themes

ahmadassaf
2pts1
www.wiz.io 1y ago

GitHub Action tj-actions/changed-files supply chain attack

ahmadassaf
1pts3
github.com 1y ago

Show HN: Craft Your Perfect Dev Environment

ahmadassaf
1pts0
github.com 1y ago

Show HN: A responsive, extensible and customisable menubar for OS X

ahmadassaf
1pts0
medium.com 4y ago

An Intro to Vertex AI Features Store

ahmadassaf
2pts0
medium.com 4y ago

Mastering Apache Kafka on Kubernetes

ahmadassaf
2pts0
medium.com 4y ago

Debunking the myth: “Data Scientists can’t write production-ready code”

ahmadassaf
3pts0
medium.com 5y ago

Deploy Quickly Everywhere

ahmadassaf
3pts0
medium.com 5y ago

Event Data at Scale

ahmadassaf
4pts0
news.ycombinator.com 5y ago

Ask HN: Would you rather optimise for quality or delivery?

ahmadassaf
2pts1
github.com 6y ago

Show HN: A simple Alfred 4 Fantastical workflow

ahmadassaf
1pts0
medium.com 6y ago

Torture the data, and it will confess to anything

ahmadassaf
3pts0
medium.com 6y ago

Free Springer e-books for Self Development

ahmadassaf
6pts0
twitter.com 6y ago

Bias and discrimination in Sort Algorithms

ahmadassaf
2pts0
github.com 6y ago

A List of Freely Available Springer E-Books

ahmadassaf
3pts0
medium.com 6y ago

Two years of micro-front ends: A retrospective

ahmadassaf
2pts0
forge.medium.com 6y ago

Productivity Culture Has an Empathy Problem

ahmadassaf
3pts0
news.ycombinator.com 6y ago

The nightmare that is selling on eBay

ahmadassaf
21pts46
medium.com 6y ago

Finding an optimal Knowledge Graph solution

ahmadassaf
4pts0
www.theatlantic.com 6y ago

Singles and Couples Are More Divided Than Ever

ahmadassaf
1pts0
www.nytimes.com 6y ago

Toronto Was Obeying Social Distance Rules. Then Came Adorable Baby Foxes

ahmadassaf
2pts0
news.ycombinator.com 6y ago

Ask HN: How you measure time to ramp-up for new engineers as a manager?

ahmadassaf
1pts1
towardsdatascience.com 6y ago

Extracting knowledge from knowledge graphs using Facebook Pytorch BigGraph

ahmadassaf
2pts0
syntagnet.org 6y ago

SyntagNet, a resource with 88,000 lexical-semantic combinations, is now out

ahmadassaf
3pts0
news.ycombinator.com 6y ago

Ask HN: How would you measure time to ramp-up for new engineers

ahmadassaf
1pts0
github.com 6y ago

Show HN: Responsive, extensible menubar with extensive lib of widgets for macOS

ahmadassaf
2pts1
github.com 6y ago

A responsive, extensible menubar for MacOS with an extensive set of widgets

ahmadassaf
1pts0
github.com 6y ago

Show HN: Responsive OSX geek bar with an extensive set of pluggable widgets

ahmadassaf
2pts2

Years ago I forked bash-it into gaudi-bash and rewrote large parts of it, cleaner architecture, async theme rendering, better component management, and a curated set of plugins that actually work well together. It became my daily driver for bash.

Then I started using zsh more. And I hit a gap.

The zsh ecosystem has plenty of frameworks and plugin managers oh-my-zsh, prezto, antidote, zinit, zplug, sheldon, zgenom, antigen, among others. But they all solve a different problem: they manage zsh-specific plugin installation and loading. None of them give you the workflow I was used to; an intuitive enable plugin fzf / disable alias git CLI, symlink-based component management, profiles to save and switch configurations, or a single search command to find what's available across aliases, plugins, and completions. And critically, none of them work across shells.

On the bash side, the options are even more limited. bash-it (which gaudi-bash forked from) is essentially the only framework with a comparable component management approach.

I found myself context-switching between two completely different mental models depending on which shell I opened. Different config patterns, different commands, different plugin ecosystems.

gaudi-shell closes that gap. One framework, one CLI, one set of components across bash and zsh.

I'd be happy to accept the return if it was within reason, but this claim seem to be outrageous, and more on Apple than eBay as why they claim their batteries can last for 10 hours.

I feel I'd be obliged to accept the return but my fear is getting back a damaged or beyond laptop, and all in all its just an unpleasant experience

I agree you with on this .. its not about making money at all and would accept a return within reason. It seems that eventually I'd have to, and I wonder what the laptop would look like when its back

1) For those who never put an ad on eBay before, when you finalizing your listing, you have this "FREE" word floating around the submit button, making you think that listing on eBay is free and how do they get money you say? well thats when you want to purchase addons or bump your add (or so the naive person inside me thought). The laptop was sold within hours of the listing, and the usual payment went through Paypal on hold until the shipment is sent. This is when you first realise that Paypal (which is owned by eBay) takes some fees to process your payment .. i was ok with that .. it was less than 60 USD on the transaction

2) Few days ago, I received a bill from eBay telling me that I have to pay close to 200 USD as ad fees, as it turns out that eBay takes a percentage on every transaction that happens.. I swallowed that too and made the payment .. in the end its on me not reading properly (but the devil in me completely blames eBay for not being transparent on this .. a simple fee calculator before you submit the ad is really minimal work)

Do you rely more on heirarchy or labels/tags and rely on the powerful search of confluence? Also have you tried Confluence for personal use or only professionally (I am thinking of moving away from Evernote but am afraid Confluence is an overkill)

There are many tools (paid and free) that assist with managing multiple kubernetes clusters but we still need to make it easy for human intervention. When an issue is happening with your kubernetes cluster and you need address it immediately, having the information that you need already at your finger tips is insanely useful. (3am me is going to appreciate it)

However, the problem of managing multiple kubernetes clusters isn't an easy problem to solve but a good problem to have. It creates a space open for new ideas on how best we make this part of the industry better and kubernetes itself has looked at this problem and suggested a solution with federated cluster. I don't have a straight forward answer, but I am excited to be part of the solution.

similar to this have you checked out https://www.zenhub.io .. those guys built an agile board on top of Github where you can have a PR and then link that PR to various issues.

However, something i see that can be done with the current script is to automate the process of opening a PR that cross multiple repos. A command that similar to `git flow` will flip various branches in various repos that match maybe a certain regex .. and then allow user to input the PR details (maybe based on a PR template as well) and then submit those to Github. On success, it will again add comment on each PR with a link to the other issues so that you have the Github issue reference link in there.

I am not yet sure if that is feasible but it seems an interesting thing to do and worth looking at

not really .. i believe it depends on what you really define as microservice. If we think at microservices as completely independent services that a change in them should not affect the system as a whole then i agree with you. However, very often, a microservice is independent but still tightly coupled in terms of logic with the rest of the ecosystem, so a change in that microservice out format for example will require other services that communicate with it to adapt to this change.

How to handle PRs/forks when you work merges/combines more than one original repo ?

What i meant here .. is that you might end up merging code from multiple repos together. To elaborate more, your work is based on two repos X,Y .. you might end up committing code from Y into X and code from X into Y with your edits in those, which will result in license conflicts that maybe the original repo owners do not want to adhere to .. so maybe a fork would be more suitable here as you will be the one responsible for that. I hope i was clear

i actually ignore a bunch of file extensions

   const BAD_EXTENSIONS = ["!*.jpg", "!*.jpeg", "!*.mov", "!*.mp3", "!*.gif", "!*.png", "!*.log", "!*.bin", "!*.psd", "!*.swf", "!*.fla", "!*.ico"];
I was wondering why `.png` was scanned. However, your comment makes so much sense and it will be wise to add default folders like .git as well

Thanks a lot for that, ill be looking forward to what defaults people also want to have :)