← Twelve Ways to Redesign a Blog
A Mark for a Broken Name
The site has a wordmark and nothing else. Ten logo directions, each drawn as real geometry and then put through the only test that matters: does it still mean something at sixteen pixels, on white and on ink, in one colour.
What is there today
Two words in Mluvka 600 at 24px. It is a good wordmark and it is doing every job on its own: masthead, favicon, avatar, share card. There is no mark to hand to any of them.
How to read a direction
Each one shows the mark alone, the same mark dropped into the real masthead, and then the proof strip: 32 and 16 pixels on white, the same two on ink, and one rendering with the fire removed. If a mark dies in that strip, it is not a logo, it is an illustration.
01Offset
In the masthead
Does it survive
The idea. The site’s own broken-rule primitive, promoted from a category mark to the logo. One continuous line, interrupted, the far half stepped down and set in fire.
Where it is weak. At 16px it is two dashes, and two dashes already mean “menu” or “equals” to anyone scanning a tab strip. It also competes directly with the 23 post marks, which are built from the same line.
02Fault
In the masthead
Does it survive
The idea. A solid block sheared along one steep diagonal and slid apart. The logo is the white seam, not the mass around it, and mass is what survives being shrunk.
Where it is weak. A sheared block is close to generic geology, and the fire half can be misread as a highlight rather than a piece that moved. It says broken, it does not say rebuilt.
03Kerf
In the masthead
Does it survive
The idea. One disc, one straight cut driven in from the rim on an off-axis angle. Nothing moves, nothing is added. The disc is pure mass, so the silhouette holds at any size and only the cut has to be defended.
Where it is weak. Widen the cut by two units and it becomes a pie chart; narrow it by two and the favicon is just a dot. It is also the only direction with no fire in it, which cuts it loose from the rest of the site.
04Bracket
In the masthead
Does it survive
The idea. Brackets are what a framework looks like on the page. The opening bracket is whole; the closing one has lost its bottom return, so the frame never quite closes.
Where it is weak. It is literal, and to any developer a missing bracket is a syntax error, which is exactly the geeky register the brief rules out. Two brackets also crowd fast at 16px.
05Seam
In the masthead
Does it survive
The idea. The repair instead of the damage. Two halves of one block, held by a fire spine that runs past the block at both ends, the way a splint is longer than the break.
Where it is weak. Two bars either side of a vertical line is one nudge away from a pause button. The two-unit white gaps that let it work in a single colour are also the first thing to close up below 20px.
06Fracture
In the masthead
Does it survive
The idea. A geometric B split along one horizontal fault, the top half displaced, a fire line left in the seam. One letter you cannot un-see once it is pointed out, carrying the wordmark’s own weight.
Where it is weak. Displaced type is a well-worn move, and the fire seam is a stroke away from reading as a strikethrough. A lone B in a browser tab also says nothing to a stranger, which is most of this audience.
07Collision
In the masthead
Does it survive
The idea. The monogram. Both initials set in the site’s own face and driven into each other, and wherever they overlap, both letters are knocked out to white. The break is a collision, not a cut.
Where it is weak. SB is a crowded pair of initials, and this is a lockup rather than a drawing: it exists only as long as Mluvka 600 does, so it cannot be redrawn at will or handed to anyone to cut in another medium.
08The Rule
In the masthead
Does it survive
The idea. No mark at all. A fire hairline stands in for the space between the two words, so the break lives inside the name rather than beside it. The wordmark is the logo.
Where it is weak. There is nothing to put in a tab, an avatar or a favicon except the two initials, and a vertical rule is a thin thing to own. Anyone can add a line between two words.
09The Step
In the masthead
Does it survive
The idea. Also no mark. Two lines set uppercase and tracked open, the second dropped and indented by one character, so the type block itself is faulted. Structure carries the idea.
Where it is weak. A two-line lockup fights a single-line masthead and every social profile field. Tracked-open caps stop being readable the moment the lockup goes below about 24px.
10Caret
In the masthead
Does it survive
The idea. The proofreader’s mark for something missing. Ink rule, ink gap, one fire caret pointing up into it. Fire is used exactly once, and it is the only colour on the page.
Where it is weak. The proofreader’s caret is not common knowledge, and to everyone else a chevron on the web means expand, play or next. Sitting between two dashes it can also simply read as a capital A.
If it were mine, two of these get built out.
02Fault
It is the only mark here made of mass rather than line, which is exactly why it still holds at 16 pixels while most of the others thin out. It also stays clear of the 23 post marks, which are all hairline drawings: the logo should not look like one more category. And the gesture is the name. Something solid, one clean break, the two pieces no longer aligned.
07Collision
The surprise of building this page. It is drawn in Mluvka, so it is already family with the wordmark, it hands over an instant favicon and avatar, and it is the only direction where the break is something the two halves did to each other rather than something done to a shape. And it is the direction I most expected to fail the size test: the knockout is still open at 16 pixels.
Everything above is one colour plus fire, no gradients, no rounded softening, drawn on the same 64-unit grid the post marks use. Any of the ten can be handed to a favicon, an avatar and an Open Graph card without redrawing. The real fork is not which drawing wins, it is whether this site wants a mark at all. Directions 08 and 09 say no: tighten the wordmark, let the favicon be two letters, and spend nothing on a symbol readers were never going to learn.