diff --git a/src/base/type_system/base_toString.c b/src/base/type_system/base_toString.c
index f40c732..1ad15cb 100644
--- a/src/base/type_system/base_toString.c
+++ b/src/base/type_system/base_toString.c
@@ -61,28 +61,25 @@ char* toString_uint(uint64 n, bool withPostfix, bool uppercase){
     return cptr_copy((char*)str+i);
 }
 
-char* toString_float(float64 n, bool withPostfix, bool uppercase){
-    // int64 d=n;
-    // float64 r=n-d;
-    // char* strint=toString_int(d);
-    // char strfract[32];
-    // uint8 i=0;
-    // strfract[i++]='.';
-    // while(r!=0){
-    //     r*=10.0;
-    //     char fc=r;
-    //     strfract[i++]=fc;
-    //     r-=fc;
-    // }
-    // if(withPostfix)
-    //     strfract[i++]= uppercase ? 'F' : 'f';
-    // strfract[i]=0;
-    // char* str==cptr_concat(strint, strfract);
-    // free(strint);
-    // return str;
-    return cptr_copy("<float>");
+#define _toString_float_impl(bufsize, maxPrecision) {\
+    char str[bufsize];\
+    if(precision>maxPrecision)\
+        throw("too big precision");\
+    if(precision==0)\
+        precision=toString_float_default_precision;\
+    int cn=sprintf(str, "%.*f", precision, n);\
+    if(withPostfix)\
+        str[cn++]= uppercase ? 'F' : 'f';\
+    str[cn]='\0';\
+    return cptr_copy(str);\
 }
 
+char* toString_float32(float32 n, uint8 precision, bool withPostfix, bool uppercase)
+    _toString_float_impl(48, toString_float32_max_precision)
+
+char* toString_float64(float64 n, uint8 precision, bool withPostfix, bool uppercase)
+    _toString_float_impl(512, toString_float64_max_precision)
+
 #define byte_to_bits(byte) {\
     str[cn++]='0' + (uint8)((byte>>7)&1); /* 8th bit */\
     str[cn++]='0' + (uint8)((byte>>6)&1); /* 7th bit */\
@@ -204,7 +201,7 @@ __toString_uint_def(64)
     switch(kp_fmt_dataFormat(f)){\
         case kp_f: ;\
             float##BITS n=*(float##BITS*)_n;\
-            return toString_float(n, kp_fmt_withPostfix(f), kp_fmt_isUpper(f));\
+            return toString_float64(n, toString_float_default_precision, kp_fmt_withPostfix(f), kp_fmt_isUpper(f));\
         case kp_b:\
             return toString_bin(_n, BITS/8, getEndian()==LittleEndian, kp_fmt_withPrefix(f));\
         case kp_h:\
diff --git a/src/base/type_system/base_toString.h b/src/base/type_system/base_toString.h
index 4f99771..3ea38ff 100644
--- a/src/base/type_system/base_toString.h
+++ b/src/base/type_system/base_toString.h
@@ -6,27 +6,37 @@ extern "C" {
 
 #include "../errors.h"
 
-// functions for base types
+// char and cstring
 // has different output for fmtChar and fmtString
 char* __toString_char(void* c, uint32 fmt);
+
+// bool
 char* __toString_bool(void* c, uint32 fmt);
 
+// signed int
 char* toString_int(int64 n);
 char* __toString_int8(void* n, uint32 fmt);
 char* __toString_int16(void* n, uint32 fmt);
 char* __toString_int32(void* n, uint32 fmt);
 char* __toString_int64(void* n, uint32 fmt);
 
+// unsigned int
 char* toString_uint(uint64 n, bool withPostfix, bool uppercase);
 char* __toString_uint8(void* n, uint32 fmt);
 char* __toString_uint16(void* n, uint32 fmt);
 char* __toString_uint32(void* n, uint32 fmt);
 char* __toString_uint64(void* n, uint32 fmt);
 
-char* toString_float(float64 n, bool withPostfix, bool uppercase); // very bad implimentation
+// float
+#define toString_float32_max_precision 6
+#define toString_float64_max_precision 15
+#define toString_float_default_precision 6
+char* toString_float32(float32 n, uint8 precision, bool withPostfix, bool uppercase); // uses sprintf
+char* toString_float64(float64 n, uint8 precision, bool withPostfix, bool uppercase); // uses sprintf
 char* __toString_float32(void* n, uint32 fmt);
 char* __toString_float64(void* n, uint32 fmt); 
 
+
 ///@param inverse set to true for little endian numbers (their bytes are in reverse order)
 char* toString_bin(void* bytes, uint32 size, bool inverse, bool withPrefix);
 ///@param inverse set to true for little endian numbers (their bytes are in reverse order)
diff --git a/src/kprint/kprintf.c b/src/kprint/kprintf.c
index cab22b5..baba98e 100644
--- a/src/kprint/kprintf.c
+++ b/src/kprint/kprintf.c
@@ -69,27 +69,13 @@ void kprintf(const char* format, ...){
                     break;
                 case 'f':
                     // float32 is promoted to float64 when passed through '...'
-                    argstr=toString_float(va_arg(vl, float64),0,0);
+                    argstr=toString_float64(va_arg(vl, float64), toString_float_default_precision,0,0);
                     break;
                case 'l':
                     l=true;
                     if((c=format[i++]))
                         goto format_escape_seq;
                     break;
-                    // switch (c) {
-                    //     case 'u':
-                    //         argstr=toString_uint(va_arg(vl, uint64),0,0);
-                    //         break;
-                    //     case 'i':
-                    //         argstr=toString_int(va_arg(vl, uint64));
-                    //         break;
-                    //     case 'f':
-                    //         argstr=toString_float(va_arg(vl, float64),0,0);
-                    //         break;
-                    //     default:
-                    //         throw(ERR_FORMAT);
-                    // }
-                    // break;
                 case 'p':
                 case 'x': ;
                     uint64 px=va_arg(vl, uint64);