1 | initial version |

On 6.8.beta4, i have a different behaviour, with no assumption:

```
sage: f(x, y, z) = sqrt(x^2 + (y+1-z)^2)
sage: f.taylor(x, 0, 3)
(x, y, z) |--> -1/2*x^2/(y - z + 1) - y + z - 1
```

Which version are you using ?

2 | No.2 Revision |

On 6.8.beta4, i have a different ~~behaviour, ~~behaviour (which might be seen as a regression), with no assumption:

```
sage: f(x, y, z) = sqrt(x^2 + (y+1-z)^2)
sage: f.taylor(x, 0, 3)
(x, y, z) |--> -1/2*x^2/(y - z + 1) - y + z - 1
```

Which version are you using ?

3 | No.3 Revision |

On 6.8.beta4, i have ~~a different behaviour (which might be seen as a regression), ~~the following behaviour, with no assumption:

```
sage: f(x, y, z) = sqrt(x^2 + (y+1-z)^2)
sage: f.taylor(x, 0, 3)
(x, y, z) |--> -1/2*x^2/(y - z + 1) - y + z - 1
```

~~Which version are you using ?~~This can be considered as a bug of too much simplifying.

Copyright Sage, 2010. Some rights reserved under creative commons license. Content on this site is licensed under a Creative Commons Attribution Share Alike 3.0 license.