I have a project-specific prompt saved as a text file. It is very basic, just focusing on the app's most important security issues. I kept it broad, so as not to over-specify.
Something along the lines of: "Please run a full security analysis on the entire project. Make sure user documents are secure."
Just something like that prompt found a vector in my web app's MCP server that I never would have considered. It was very much an edge case, but it did exist.
Being broad allows the model and harness to do the work. Giving too many instructions can apparently work against you in many cases.
Of course, when dealing with new PRs, I use the /security-review and /code-review skills.
It's rather broad right now. I started reverse engineering a mac app and it started reading some binary data and then quickly told me to switch to Opus 4.8 to continue, because the guardrails kicked in.
Always a joy to read those and the fact that the emulator keeps on improving. You can run it on SteamDeck nicely and play Gamecube games better than Switch 2. It's a shame nintendo has only released about 5 games thus far.. Great work Dolphin Devs! I still need to give RetroAchievements a try.
Whats even more interesting IMHO is that according to TFA the most popular platform in terms of usage is Android. This is probably related to the large amount of dedicated Android gaming devices like the Ayn Thor and similar.
... and the fact that (at least so far) android devices haven't 3x'ed in price due to AI bubbling.
And to add insult to injury: take a galaxy s23 ultra. Start a gamecube title. Connect USB-C to a monitor, display, TV, project: you get an image (and with a bluetooth gamepad it works from a distance). USB-C to HDMI converter? Also works! Going to a friend's house? Use a display at work? Just works!
Take either of Nintendo's 2 most recent hardware platforms: it doesn't effin work! Despite this being a major complaint on the previous hardware ...
> Take either of Nintendo's 2 most recent hardware platforms: it doesn't effin work!
Assuming this is the Switch and Switch 2 here, are you specifically talking about Gamecube titles? Because I've recorded video from my Switch over a USB-C->HDMI->USB-C dongle chain (Zelda:TOTK)
My mistake, it was the Switch 2. Using JINGDU "Switch HDMI cable"[0] feeding into a UGREEN capture dongle[1] which then fed into my MacBook for capture.
(Super handy when you want a bigger screen on the train or wherever.)
This is true for all Nintendo emulators afaik. I’m semi-involved in the Eden discord (the community successor fork of yuzu) and over half their users are on Android. I for one don’t get the appeal of emulating games on a phone but maybe I’m just old fashioned.
Most of my emulation game time is outside the house, 90% (got a miyoo mini plus for when I'm in a que) is done on Android and none of it is on a phone :) Just the idea of touchscreen gaming is giving me RSI pain.
I guess I kind of agree with your disagree, but disagree overall!
The UI in LibreOffice feels quite clunky and outdated and never seems to have been given any thought since the OpenOffice days. But Google Docs is so feature poor that I'd rather live with LibreOffice's UI. Especially as you can adjust to the latter after using the software for a while.
Have you tried LibreOffice's ribbon interface? It is similar to the one on Microsoft Office. You go to View > User Interface and choose "Tabbed". There are 7 different variants available.
Have you tried LibreOffice's ribbon interface? It is similar to the one on Microsoft Office. You go to View > User Interface and choose "Tabbed". There are 7 different variants available.
Microsoft ToDo has native apps for Mac and Windows, and a third-party one for Linux (https://itsfoss.com/kuro-to-do-app/). It has apps on iOS and Android and you can access it via the web.
It works very well. It even _finally_ (in 2023) made me switch from paper grocery lists to electronic ones.
M2 with 16GB: It's slow for me. ~13GB RAM usage, not locking up my mac, but took a very long time thinking and slowly outputting tokens.. I'd not consider this usable for everyday usage.
I can't really fully understand what "Microsoft 365" stands for. Microsoft years in day? Microsoft all year? hard to understand. Sounds like some kind of new super-term for their Office products. In that case, LibreOffice works absolutely fine inside my company. Element works alright for secure e2ee messaging. For storage and document collaboration we are using a self-hosted Nextcloud cluster. All of these solutions are free and have no upfront fee, but they have some maintenance costs of course, yet I think they are magnitudes lower that M$ could scam us for. The AI tooling you can definitely integrate in the workflow yourself if you need them (we don't). As non-us company inside security field, we have little trust towards microsoft to store the sensitive data on their cloud services, and generally the amount of measures taken against corporate espionage is higher – communications are taken through verified and trusted e2ee mediums, and it's common to host enterprise infrastructure on-premises – it's totally realistic to have your company data and communications never leave your virtual private network, and it's a good practice to do so in my opinion.
… except that Google took this as permission to make their previously optional AI bolt-on mandatory and hiked the price for every user in exactly the same way.
Complaining to their support was an utterly miserable fruitless exercise that has me wondering what alternatives there are to this madness for SMEs…
Interesting stuff! I've been trying ollama/gpt + continue.dev and copilot in VSCode for a bit now and the chat-style assistant is a huge plus. Especially in DevOps (my main work) the codegen is rather unhelpful, but the ability to let LLMs explain some sections and help in rubber-ducking is huge.
I see that a good output requires a good input (prompt). How does copilot workspace determine a good input for the prompt? I see that in the github repo there is already a bunch of "Tips and Tricks" to get better results. What is your experience so far? Should we change our way of creating issues (user-stories / bug-reports, change-requests) to a format that is better understood by AI/Copilot? (half-joking, half-serious).
Well, that's basically the heart of Copilot Workspace! The whole UX is structured to make it easy for the human to steer.
- Alter the "current" bullet points in the spec to correct the AI's understanding of the system today
- Alter the "proposed" bullet points in the spec to correct the AI's understanding of what SHOULD be
- Alter files/bullets in the plan in order to correct the AI's understanding of how to go from current to proposed.
That said, I think there's definitely a future where we might want to explore how we nudge humans into better issue-writing habits! A well-specified issue is as important to other humans as it is to AI. And "well-specified" is not about "more", it's about clarity. Having the right level of detail, clear articulation of what success means, etc.
you can generate serviceAccounts in google cloud that act as clients for your google workspace with no additional cost. It's a neat hidden trick to circumvent the google quotas, but you need to write your own clients/wrappers to utilize that properly.
