monotone-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Monotone-devel] automate sync display branches?


From: Thomas Keller
Subject: Re: [Monotone-devel] automate sync display branches?
Date: Sun, 12 Sep 2010 22:48:43 +0200
User-agent: Mozilla/5.0 (Macintosh; U; Intel Mac OS X 10.5; de; rv:1.9.2.9) Gecko/20100825 Lightning/1.0b3pre Thunderbird/3.1.3

Am 12.09.10 21:17, schrieb Stephen Leake:
> Stephen Leake <address@hidden> writes:
> 
>> Thomas Keller <address@hidden> writes:
>>
>>> As I've written earlier on IRC it
>>> might be a good idea to team up with Tim (nvm.netsync) - see the
>>> discussion here:
>>>
>>> http://colabti.de/irclogger/irclogger_log/monotone?date=2010-09-11#l55
>>
>> That does look useful. Tim; should I work directly in nvm.netsync, or
>> branch from there (nvm.netsync.automate)?
> 
> I think it makes sense to not call the note_netsync_* Lua hooks from the
> automate version, since the intent is that the front-end will process
> the automate output, not the hooks. That way save some processing time,
> and we don't have to worry whether the hooks will screw up the automate
> output.

I'm not sure about that - out of the box these hooks do nothing bad
since they do not print out anything and if we follow your reasoning we
might start disabling other hooks which might interfer with the output
somehow. In the end its the user's problem to take care the hooks do not
interfer with automate's output and we already gave a set of tools into
the hands of the user / automate implementor

1) in guitone for example I start every mtn instance with
--no-builtin-rc (previously --norc) to avoid side effects

2) standard print()'s (and io:write()'s as well) are catched and
directed to the progress stream, which are not a defined part in the
automate interface anyways (otherwise we'd have to raise the automate
interface version number each time we change a string which might pop up
in one of the 'w', 'p' or 'e' streams)

And then I thought lately about very obscure use cases where I found it
useful that I could initiate a sync of a remote (server) database with
another database without having physical access to the machine itself
(read: no ssh, plain remote_stdio) and maybe people still like their
hooks to get fired then (for example to trigger a build process or some
notification). Maybe this is totally out of scope - but hey, unless
there isn't a very (very) good reason to restrict some functionality,
then I think we shouldn't do it.

Thomas.

-- 
GPG-Key 0x160D1092 | address@hidden | http://thomaskeller.biz
Please note that according to the EU law on data retention, information
on every electronic information exchange might be retained for a period
of six months or longer: http://www.vorratsdatenspeicherung.de/?lang=en

Attachment: signature.asc
Description: OpenPGP digital signature


reply via email to

[Prev in Thread] Current Thread [Next in Thread]