Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:120946
Feedback-ID: i8414410d:Fastmail
User-Agent: Cyrus-JMAP/3.9.0-alpha0-701-g9b2f44d3ee-fm-20230823.001-g9b2f44d3
Mime-Version: 1.0
Message-ID: <af2bd75b-ec38-44d6-afb4-31f019d9a280@app.fastmail.com>
In-Reply-To: 
 <CAAKU0ukfy5wDcDnkEgc2o3PhCSczdDxPWG+NcHAKiVHMSDVKZg@mail.gmail.com>
References: <ZOquxkEdpVa8GfcE@castor>
 <d2ca7e4a-99f9-4813-cbde-c4e92faa7b69@bastelstu.be>
 <CAAKU0ukfy5wDcDnkEgc2o3PhCSczdDxPWG+NcHAKiVHMSDVKZg@mail.gmail.com>
Date: Tue, 29 Aug 2023 15:18:52 +0000
To: "Levi Morrison" <levi.morrison@datadoghq.com>,
 =?UTF-8?Q?Tim_D=C3=BCsterhus?= <tim@bastelstu.be>
Cc: "Athos Ribeiro" <athoscribeiro@gmail.com>,
 "php internals" <internals@lists.php.net>
Content-Type: text/plain;charset=utf-8
Content-Transfer-Encoding: quoted-printable
Subject: Re: [PHP-DEV] [VOTE] Support optional suffix parameter in tempnam
From: larry@garfieldtech.com ("Larry Garfield")

On Tue, Aug 29, 2023, at 1:57 AM, Levi Morrison via internals wrote:
> On Sun, Aug 27, 2023 at 4:20=E2=80=AFAM Tim D=C3=BCsterhus <tim@bastel=
stu.be> wrote:
>>
>> Hi Athos
>>
>> On 8/27/23 04:02, Athos Ribeiro wrote:
>> > I am moving this RFC [1] to the voting phase.  Voting will be open =
for the
>> > next 2 weeks, until September 10th, as per https://wiki.php.net/rfc.
>> >
>> > [1] https://wiki.php.net/rfc/tempnam-suffix-v2
>> >
>>
>> I find this a useful feature in general, but I believe it not working=
 on
>> Windows completely nullifies the "could even provide more context for
>> software processing such files" argument in favor of this feature. It
>> will be unexpected for users if their code completely fails to work on
>> Windows, because the suffix is ignored.
>>
>> For that reason I voted "no".
>>
>> Best regards
>> Tim D=C3=BCsterhus
>>
>> --
>> PHP Internals - PHP Runtime Development Mailing List
>> To unsubscribe, visit: https://www.php.net/unsub.php
>
>
> I have voted no for a similar reason. It would be nice if there were
> os-specific packages in core that handled these kinds of things,
> because the functionality is definitely useful. But if you call an API
> like `FileSystem\Os\Unix\tempnam` then at least the platform specific
> behavior is obvious and understood.

I did as well.  I am in favor of the functionality, but having it silent=
ly not-work on Windows is not a good approach.  That may mean a larger r=
efactoring of tempnam is a good next step, to fix the issues noted in th=
e RFC and then support Windows consistently.  I would support that, and =
including a suffix either with that or as a follow-up.

(Side note: If messing about there, making prefix optional would also be=
 wise so that it can be skipped via named args.)

--Larry Garfield