rostock is retracting to many times

[Right to Make]

I had just did an update from matter control 1.3 to 1.4 and noticed that slic3r was not working, so then i rolled back to 1.3 and now i am getting weird prints, i use matter control exclusively.
Some prints act just like under extrusion, while other prints retract every change in direction. I am back to calibration cubes and matter slice wants to retract after each change in printing direction, on a small 1x1x1 cube it retracts a lot of times on one solid layer. then using Cura, the block comes out as under extrusion problems and it retracts a lot more than before. Slic3r retracts quite a bit to. I did one of the fan shrouds again as its a good test piece to run with all the contours and some overhangs and it was greatly under extruded.
I did not change any settings and i took a fine tooth comb through all settings and they are the exact same i had been making parts with last month. Is there a setting i missed, or bumped? Anyone else having this issue? I have ran these parts that i printed just fine before, but they print differently now. I almost want to re-flash and start over the Rambo! My nozzle is clean although when it was doing so many retractions it clogged once. It was weird and made me sick that i think i changed something and didn't mean to.

I had re flashed and now its doing a solid .75"x.75"x.5" cube and its not retracting so many times as i can see its only retracting for the next layer, which is progress. I went back to matter control 1.3 and redid a new printer bio so its all stock now.
I did a solid cube and its perfect. hardly any retraction. Now i am doing a spiral vase with matter slice and its retracting like crazy. Never did that before! every back and forth motion its retracting. Am i missing something? it never did that before. speed settings seem mucked up as well. the infill was fast now the outer perimeter is slow.

[Right to Make]

Ok so the .750x.750x.5 cube is way off of speck it measures .670x.670x.495. Seriously last week was running parts that were dead on to what the model called for. the print looks perfect, just undersized.

[Right to Make]

Update.
Something went terribly wrong. I am printing some more cubes and it seemed to be printing nicely then i heard a snap. it was some clips holding the glass bed on and the bed moved because the nozzle ran into the block. it was as if the nozzle was never moving upwards. I am going to have to do some fine tooth combing because i dont get what is wrong. was printing fine now its got some bugs. ill post an update when i have one.

[Right to Make]

Update. I loaded Repetier host and calibrated that and sliced with slicer and printed some parts, and all is working well, no retraction issues and part is actually to size. Had to change my printing radius but that is all. For now ill stick with rep. Any insight on what was making matter control act like that? anyone having issues since the update?

[Tincho85]

I'm glad it's working now. Rep Host ftw!
Sorry nobody gave you any help with MC.

[Right to Make]

I was a bit nervous for rep because all i knew was matter control. But i think it all boils down to personal preference. I like repetier host better than i thought, but mattercontrol is nice but again with thte personal preference. and oh ya that thing about learning curves.

[lightninjay]

Could just remove the learning curve of printer host software all the way and just start putting your sliced Gcode directly on your SD card! lol

[Right to Make]

[Xenocrates]

Could just remove the learning curve of printer host software all the way and just start putting your sliced Gcode directly on your SD card! lol

Or go even farther, remove the slicer learning curve and write your own G code by hand (Do not do this, it is tedious, error prone, time consuming, and G itself has a learning curve that's not great, and every firmware handles the advanced codes which are the only justification differently.)

For a mill, or a lathe, G by hand works, especially once you learn those advanced codes which let you define geometries and the controller will take care of movement to make them happen, including roughing/finishing passes. It's hard, and time consuming. For a printer, be prepared to spend forever copy-pasting co-ordinates to make layers. But it could be done, if you want the bragging rights or are sufficiently OCD about optimization.

[Tincho85]

Xenocrates I can't tell if your are joking or not

[Xenocrates]

Xenocrates I can't tell if your are joking or not

It's partially a joke. I did actually look at writing parts by hand, but I haven't printed any of them yet, as I haven't had a great desire to print random shapes, since I have useful things to print. in theory, it isn't hard, other than computing E. But I have dropped decimals often in other programs, which leads to badness, especially in G, as a number on it's own is usually considered to be thousandths, whereas a number with a decimal is full units.