Re: Formatting Curmudgeons WAS: MMAP Buffers

On Tue, May 10, 2011 at 1:46 PM, Heikki Linnakangas <
heikki(dot)linnakangas(at)enterprisedb(dot)com> wrote:

> On 10.05.2011 04:43, Greg Smith wrote:
>
>> Josh Berkus wrote:
>>
>>> As I don't think we can change this, I think the best answer is to
>>> tell people
>>> "Don't submit a big patch to PostgreSQL until you've done a few small
>>> patches first. You'll regret it".
>>>
>>
>> When I last did a talk about getting started writing patches, I had a
>> few people ask me afterwards if I'd ever run into problems with having
>> patch submissions rejected. I said I hadn't. When asked what my secret
>> was, I told them my first serious submission modified exactly one line
>> of code[1]. And *that* I had to defend in regards to its performance
>> impact.[2]
>>
>> Anyway, I think the intro message should be "Don't submit a big patch to
>> PostgreSQL until you've done a small patch and some patch review"
>> instead though.
>>
>
> Well, my first patch was two-phase commit. And I had never even used
> PostgreSQL before I dived into the source tree and started to work on that.
> I did, however, lurk on the pgsql-hackers mailing list for a few months
> before posting, so I knew the social dynamics. I basically did exactly what
> Robert described elsewhere in this thread, and successfully avoided the
> culture shock.
>
>
Yeah, probably same for me, though I got a lot of support from existing
hackers during my first submission. But it was a tiring experience for sure.
I would submit a patch and then wait anxiously for any comments. I used to
get a lot of interesting and valuable comments, but would know that unless
one of the very few (Tom ?) members say something, good or bad, it won't go
anywhere and those comments did not come in the early days/months. I was an
unknown name and what I was trying to do was very invasive. So when I look
back now, I can understand the reluctance on other members to get excited
about the work. Most often they would see something in the design or the
patch which is completely stupid and they would loose all interest at the
very moment.

Since I had backing of EnterpriseDB and it was my paid job, it was much
easier to keep the enthusiasm, but I wouldn't be surprised if few others
would have turned their back to the project forever.

Fortunately, things have changed for better now. I think the entire commit
fest business is good. Also, we now have a lot more hackers with expertise
in different areas and with influential opinions. Its very likely that if
you submit an idea or a patch, you would get some
comment/suggestion/criticism very early.

Since HOT is mentioned often in these discussions, I thought I should share
my experience.

Thanks,
Pavan

--
Pavan Deolasee
EnterpriseDB http://www.enterprisedb.com