From an image preview to a printed sticker

One of my recent Ubuntu projects involved printing my FA artwork with an iiGenai S1 printer.

With help from ChatGPT, I developed iiGenai Print and worked through the image controls, previews, packaging, and tests. The practical objective was straightforward: produce a sticker I was satisfied with.

Getting there involved checking more than whether the application could send a print job.

Checking the preview

The application provides different processing options for photos and text or line art.

For my artwork, I needed to see how those options affected the lines and lettering before printing. Brightness, threshold, and the treatment of fine strokes all became part of that review.

The controls also needed predictable behavior when opening another image. A preview is useful only if I can understand which settings produced it.

Investigating a failed test

During acceptance testing, an automated check reported that opening another image did not restore the expected defaults.

The investigation found that the test requested the new image before the previous editor had finished closing. It then inspected the old controls.

Waiting for the editor’s confirmed lifecycle event resolved that test failure. The application did not need a correction for this particular issue.

That distinction mattered. Changing the application in response to a faulty test would have addressed the wrong problem.

Checking the actual print

I tested multiple physical prints, including the original artwork and a revised version with thicker horizontal lines beside fasouza.me.

The accepted result used Text / line art, brightness at zero, Fine strokes off, automatic threshold, and no cropping.

Those were the settings that worked for this artwork. I would still review the preview and result when printing a different image.

Keeping the result reproducible

Version 1.2.1 was subsequently installed as the stable production baseline. The installer, matching source, and recovery instructions were also backed up to my USB drive.

The physical-print acceptance belonged to the release candidate. The stable release received separate non-printing checks, and its normal desktop visual confirmation was still pending in the latest installation report.

Keeping those results separate makes the record more useful. I can say which output I accepted without suggesting that every later installation check included another physical print.