Even if it's not technically a regression against the specification as per
Benjamin's message above, breaking working code is still not very good.
Could you please report this on the OCaml bugtracker?

  http://caml.inria.fr/mantis/

On Sun, Sep 14, 2014 at 12:53 AM, Benjamin Greenman <blg59@cornell.edu>
wrote:

> Why do you match on the error string? String.sub is only documented to
> raise "Invalid_argument _".
>
> http://caml.inria.fr/pub/docs/manual-ocaml/libref/String.html
>
> On Sat, Sep 13, 2014 at 6:47 PM, Oliver Bandel <oliver@first.in-berlin.de>
> wrote:
>
>> Hello,
>>
>> is this a known bug?
>>
>> Needed to change code from
>>
>>    with Invalid_argument("Strings.sub")
>> to
>>    with Invalid_argument("Bytes.sub")
>>
>> The code now looks like this and then works again:
>>
>>
>>                   try
>>                     if parser_url_len > 0 && parser_url = String.sub url
>> 0 parser_url_len then parser_name else aux tl
>>                   with Invalid_argument("Bytes.sub") -> aux tl (* this
>> happens if url is shorter than parser_url *)
>>
>>
>> Ciao,
>>    Oliver
>>
>>
>> --
>> Caml-list mailing list.  Subscription management and archives:
>> https://sympa.inria.fr/sympa/arc/caml-list
>> Beginner's list: http://groups.yahoo.com/group/ocaml_beginners
>> Bug reports: http://caml.inria.fr/bin/caml-bugs
>>
>
>