crate / cloud-docs

Documentation for CrateDB Cloud.
https://crate.io/docs/cloud/
Apache License 2.0
0 stars 1 forks source link

Add documentation on S3 imports #34

Closed matkuliak closed 11 months ago

matkuliak commented 11 months ago

Summary of the changes / Why this is an improvement

Checklist

amotl commented 11 months ago

Hi Matej,

thanks for this patch.

I have a few suggestions about the screenshot images. Recently, I spent a few minutes on shrinking the screenshot file sizes, because I think 2.5 MB is way too large, and introduces recognizable delays when loading them, even on todays Internet's fast pipes.

I usually use something around like this, using a combination of ImageMagick and oxipng, but I will be happy to learn more how to eventually improve the procedure.

convert -resize 1024x -quality 100 cluster-import-tab-choice.png smaller.png
oxipng smaller.png
mv smaller.png cluster-import-tab-choice.png

Other than this, I think reading flow and information conveyance can be tremendously improved by "zooming in to the feature at hand". As a reader, you can't recognize anything detailed anyway when looking at the full screenshot. Or is it just my bad eyes?

On behalf a recent article, where I've demonstrated the Cloud Import feature, I just "zoomed in" to the import dialag, and I think it is much better like this.

https://github.com/crate/ml-sandbox/blob/main/timeseries-data/article/01-about-intro.md#importing-data

With kind regards, Andreas.

matkuliak commented 11 months ago

Hi Matej,

I have a few suggestions about the screenshot images

Hi @amotl. Thank you for the suggestions. You're right, the sizes could be improved. I applied oxipng on all the pictures for an average 75% size reduction, so now coming in at around 0.5MB. I will do this for all the pictures in a separate PR and add it to my workflow.

As for the Magick I'll go without it for now, hope that's okay. Zooming in on the pictures is for a debate, I considered this a couple of times previously. The issue, for me, is that in some cases it takes away some of the important context. E.g. you're not sure where to look for the specific field in the screenshot. And applying this case by case makes for a significantly worse visual of document, so I opted for universal fullscreen screenshots until now.