<rss version="2.0" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Hacker News: mariusziskra</title><link>https://news.ycombinator.com/user?id=mariusziskra</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Thu, 19 Nov 2026 20:26:00 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=mariusziskra" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by mariusziskra in "Show HN: Agency – give your agents the computer you own"]]></title><description><![CDATA[
<p>Does it automatically mount HDD on boot, or are they isolated by default when running from a USB drive?</p>
]]></description><pubDate>Mon, 16 Nov 2026 08:11:10 +0000</pubDate><link>https://news.ycombinator.com/item?id=50515232</link><dc:creator>mariusziskra</dc:creator><comments>https://news.ycombinator.com/item?id=50515232</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50515232</guid></item><item><title><![CDATA[New comment by mariusziskra in "Show HN: Trovemarks – Turn your bookmarks into followable collections"]]></title><description><![CDATA[
<p>Is there a possibility of communication e.g. with ChatGPT or Claude using MCP? Thanks to this I could after a conversation with the chat pass a link for saving in the Trovemarks. And the same the other way I could from the Trovemarks download and search appropriate links directly from Claude.</p>
]]></description><pubDate>Fri, 13 Nov 2026 15:03:48 +0000</pubDate><link>https://news.ycombinator.com/item?id=50484591</link><dc:creator>mariusziskra</dc:creator><comments>https://news.ycombinator.com/item?id=50484591</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50484591</guid></item><item><title><![CDATA[New comment by mariusziskra in "Show HN: Turri – a coding agent with the power of regenerative software"]]></title><description><![CDATA[
<p>I went the other way: I do not generate code from scratch. In my workflow, the specification sis the source of truth. Every change goes in as a change request that becomes new tasks in the plan. Instead of recreating code, the agent modifies existing code. Migrations are strictly additive. Every schema change results in a new file, leaving previous files untouched. Each migration is backward-compatible. Destructive operations, such as DROP commands, are isolated into a separate, subsequent phase that I personally review.<p>The largest project I’ve managed this way has been in production for about six months and has already gone through 45 versions. So far I haven't had encountered no no data loss so far”.<p>Renaming is where I would be wary of generating code from scratch. Based solely on a schema diff, a rename appears as the deletion of one element followed by the addition of a new one. Does turri ask about what you meant in such cases, or is there a way to specify this directly within the specification?</p>
]]></description><pubDate>Fri, 13 Nov 2026 07:11:28 +0000</pubDate><link>https://news.ycombinator.com/item?id=50479507</link><dc:creator>mariusziskra</dc:creator><comments>https://news.ycombinator.com/item?id=50479507</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50479507</guid></item><item><title><![CDATA[New comment by mariusziskra in "Show HN: Turri – a coding agent with the power of regenerative software"]]></title><description><![CDATA[
<p>How does regeneration deal with persistent state? If a **plain change leads to a different DB schema, does turri generate migrations, or is the data layer treated as something that can’t just be thrown away? This has been the hardest part for me with spec-first generation.</p>
]]></description><pubDate>Thu, 12 Nov 2026 15:47:50 +0000</pubDate><link>https://news.ycombinator.com/item?id=50469347</link><dc:creator>mariusziskra</dc:creator><comments>https://news.ycombinator.com/item?id=50469347</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50469347</guid></item></channel></rss>