Very small and soft (in the head)

Some­times even I have to in­stall some­thing on an Win­dows serv­er. Late­ly one of these things re­quired .NET 3.5. OK, in­stall it via Serv­er Man­ag­er and you're done, right?

Wrong. This does in­stall .NET 3.5 – but with­out any se­cu­ri­ty fix­es. These will ar­rive by the next morn­ing, so the box was vul­ner­a­ble. The up­dates don't even force a restart, so the box will re­main vul­ner­a­ble un­til some­body restarts ei­ther the new ser­vice or the whole ma­chine. Man­u­al­ly.

Ma­jor facepalm, this.

Fun with InoReader+Tumblr

I'm us­ing In­oRe­ad­er. I al­so like to look at lots of pret­ty im­ages on Tum­blr. (No, not what you're think­ing of. Not just that, any­way.)

That so­lu­tion is nice, but I have a cou­ple of small prob­lem­s:

  • Tum­blr's RSS refers to small pre­view im­ages.

    I want the large ver­­sion.

  • I want to skip en­tries with­­out im­ages.

    In­­oRe­ad­er can do that. How­ev­er …

  • I want to skip small im­ages.

    In­­oRe­ad­er can't do that.

  • I want to skip GIF­s: they're smal­l, low-qual­i­­ty, and rarely in­­ter­est­ing.

    In­­oRe­ad­er can't do that ei­ther.

Thus, let's build a cou­ple of workaround­s!

I've parked the sources in a gist for you.

This small Python script scans one of my In­oRe­ad­er fold­er­sm mark­ing all en­tries as read which I want to skip.


This is a Grease­mon­key / Tam­per­mon­key scrip­t. It re­places all links to Tum­blr pre­view im­ages with the "large" ver­sion.

API keys

The whole thing needs a couple of API keys. There's no API to fetch these, so you need to do it yourself. The start of tells you where to find them.

Another new home page

This time for re­al.

Af­ter a few goes at tech overkil­l, this is go­ing to be a most­ly-stat­ic home­page / blog. No database, no frill­s. In­stead, con­tent. I'm do­ing enough stuff that would ben­e­fit from more vis­i­bil­i­ty.