Thread overview | ||||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
|
December 27, 2012 std.bitmanip improvements | ||||
---|---|---|---|---|
| ||||
Hi, I was looking into the BitArray part of std.bitmanip to implement various data compression routines like Huffmann, Lempel-Ziv etc.. and noticed that it could use some improvements, like more convenient constructors and functions (for ex. see issue 4717 ). How would I go about submitting those changes? Should I put all the changes in one big pull request or split it up according to the different enhancement requests on Bugzilla? Also is anyone else working on this module or is there something in other parts of the library that I should be aware of? Cheers, Marco |
December 27, 2012 Re: std.bitmanip improvements | ||||
---|---|---|---|---|
| ||||
Posted in reply to Marco Nembrini | Marco Nembrini:
> Also is anyone else working on this module or
Recently I remember some discussions in D.learn about someone working on that array of bits.
Bye,
bearophile
|
December 27, 2012 Re: std.bitmanip improvements | ||||
---|---|---|---|---|
| ||||
Posted in reply to Marco Nembrini | "Marco Nembrini" <marco.nembrini.co@gmail.com> wrote in message news:ezcalrajgcqlspefnlri@forum.dlang.org... > Should I put all the changes in one big pull request or split it up according to the different enhancement requests on Bugzilla? Split it up. |
December 27, 2012 Re: std.bitmanip improvements | ||||
---|---|---|---|---|
| ||||
Posted in reply to Marco Nembrini | 12/27/2012 1:38 PM, Marco Nembrini пишет: > Hi, > I was looking into the BitArray part of std.bitmanip to implement > various data compression routines like Huffmann, Lempel-Ziv etc.. and > noticed that it could use some improvements, like more convenient > constructors and functions (for ex. see issue 4717 ). How would I go > about submitting those changes? Should I put all the changes in one big > pull request or split it up according to the different enhancement > requests on Bugzilla? Judging by my experience small pulls that have no extra dependencies on other pulls are the fastest to get merged. The are also more throughly reviewed :) So I'd suggest do it step by step in a couple of pulls if possible. > Also is anyone else working on this module or is there something in > other parts of the library that I should be aware of? > IRC Era Scarecrow did an amazing job on it that covered a lot of enhancements including small string optimization, appending, slicing etc. I recall reviewing his code and it was a Phobos pull once but then it pulled out (for further enhancement? dunno). So definitely suggest you two to get in touch. > Cheers, > Marco -- Dmitry Olshansky |
December 27, 2012 Re: std.bitmanip improvements | ||||
---|---|---|---|---|
| ||||
Posted in reply to Marco Nembrini | On Thursday, December 27, 2012 10:38:35 Marco Nembrini wrote:
> Hi,
> I was looking into the BitArray part of std.bitmanip to implement
> various data compression routines like Huffmann, Lempel-Ziv etc..
> and noticed that it could use some improvements, like more
> convenient constructors and functions (for ex. see issue 4717 ).
> How would I go about submitting those changes? Should I put all
> the changes in one big pull request or split it up according to
> the different enhancement requests on Bugzilla?
> Also is anyone else working on this module or is there something
> in other parts of the library that I should be aware of?
I believe that there are a couple of threads from a few months ago in D.Learn talking about improvements to BitArray, and some folks having been working on that. I have no idea where that stands though.
In general, if you have incremental improvements for things, feel free to create pull requests for them on github. Unrelated requests should be separate. Please don't make huge pulls with lots of stuff in them. It makes reviewing them harder, and it's more problematic when some of your changes are acceptable and others aren't. If they're separate, they're much more easily managed.
However, if you're talking about seriously revamping the module as a whole rather than making a number of small improvements, then that may merit a full review in the newsgroup, which is a different beast altogether.
- Jonathan M Davis
|
December 27, 2012 Re: std.bitmanip improvements | ||||
---|---|---|---|---|
| ||||
Posted in reply to Jonathan M Davis | On Thursday, 27 December 2012 at 10:21:37 UTC, Jonathan M Davis wrote: > > I believe that there are a couple of threads from a few months ago in D.Learn > talking about improvements to BitArray, and some folks having been working on > that. I have no idea where that stands though. > Ok, I'll try to read up on that and get in touch with others involved. > In general, if you have incremental improvements for things, feel free to > create pull requests for them on github. Unrelated requests should be > separate. Please don't make huge pulls with lots of stuff in them. It makes > reviewing them harder, and it's more problematic when some of your changes are > acceptable and others aren't. If they're separate, they're much more easily > managed. > > However, if you're talking about seriously revamping the module as a whole > rather than making a number of small improvements, then that may merit a full > review in the newsgroup, which is a different beast altogether. > > - Jonathan M Davis I was thinking of mostly small convenience changes, I don't feel familiar enough with D yet to do a revamp of that, but maybe I can help get Era Scarecrow's version pulled. |
December 27, 2012 Re: std.bitmanip improvements | ||||
---|---|---|---|---|
| ||||
Posted in reply to Dmitry Olshansky | On Thursday, 27 December 2012 at 10:16:51 UTC, Dmitry Olshansky wrote:
>
> IRC Era Scarecrow did an amazing job on it that covered a lot of enhancements including small string optimization, appending, slicing etc.
>
> I recall reviewing his code and it was a Phobos pull once but then it pulled out (for further enhancement? dunno). So definitely suggest you two to get in touch.
>
Last time I was in touch with him, he was getting increasingly frustrated trying to learn to use github. I *think* (but don't take my word for it), that he may have abandoned the pull after having too much trouble trying to rebase it.
In any case, I'll try to contact him directly and make him aware of this thread.
|
December 28, 2012 Re: std.bitmanip improvements | ||||
---|---|---|---|---|
| ||||
Posted in reply to Dmitry Olshansky | On Thursday, 27 December 2012 at 10:16:51 UTC, Dmitry Olshansky wrote:
> IRC Era Scarecrow did an amazing job on it that covered a lot of enhancements including small string optimization, appending, slicing etc.
>
> I recall reviewing his code and it was a Phobos pull once but then it pulled out (for further enhancement? dunno). So definitely suggest you two to get in touch.
I'm glad for the praise... :)
Perhaps it seems silly, however I pulled it not because I needed to enhance it, but because it had trouble merging with other pulls. I looked at the error(s) and it was something like '13 white spaces mismatches'... How do you fix 13 white spaces? They're white space because you can't see them :P Had it been lines of code you can intelligently re-order or adjust them, but white spaces?? Sheesh...
Maybe being lazy but it was a headache trying to figure out the diff patch and how to fix it, so haven't tried too hard on it yet. If someone happens across that I'll give it another go and see if there's any other fixes/enhancements I can throw on it; Maybe get it in before the next release.
|
December 28, 2012 Re: std.bitmanip improvements | ||||
---|---|---|---|---|
| ||||
Posted in reply to Dmitry Olshansky | On Thursday, 27 December 2012 at 10:16:51 UTC, Dmitry Olshansky wrote: > > IRC Era Scarecrow did an amazing job on it that covered a lot of enhancements including small string optimization, appending, slicing etc. > > I recall reviewing his code and it was a Phobos pull once but then it pulled out (for further enhancement? dunno). So definitely suggest you two to get in touch. > I assume you are referring to the code in this thread: http://forum.dlang.org/thread/gtfxxmlqwgyadmpygftv@forum.dlang.org ? |
December 28, 2012 Re: std.bitmanip improvements | ||||
---|---|---|---|---|
| ||||
Posted in reply to Era Scarecrow | On 28.12.2012 11:48, Era Scarecrow wrote: > On Thursday, 27 December 2012 at 10:16:51 UTC, Dmitry Olshansky wrote: >> IRC Era Scarecrow did an amazing job on it that covered a lot of >> enhancements including small string optimization, appending, slicing etc. >> >> I recall reviewing his code and it was a Phobos pull once but then it >> pulled out (for further enhancement? dunno). So definitely suggest you >> two to get in touch. > > I'm glad for the praise... :) > > Perhaps it seems silly, however I pulled it not because I needed to > enhance it, but because it had trouble merging with other pulls. I > looked at the error(s) and it was something like '13 white spaces > mismatches'... How do you fix 13 white spaces? They're white space > because you can't see them :P Had it been lines of code you can > intelligently re-order or adjust them, but white spaces?? Sheesh... > > Maybe being lazy but it was a headache trying to figure out the diff > patch and how to fix it, so haven't tried too hard on it yet. If someone > happens across that I'll give it another go and see if there's any other > fixes/enhancements I can throw on it; Maybe get it in before the next > release. So the code was complete and working and the only problem was doing the pull on github? If so would you mind me trying to do that in your place? It seems worth spending some time on to get all those improvements in. Is the latest version this one: https://github.com/rtcvb32/phobos ? |
Copyright © 1999-2021 by the D Language Foundation