freenode/#lisp - IRC Chatlog
Search
15:39:46
dmiles[m]
my favorite way to search the clhs is the http://autocad.xarch.at/lisp/cormanlisp/all-lispchm.zip (windows chm files)
15:41:34
dmiles[m]
well what is good is as one is typing in the index box matching text comes up from everywhere
20:08:37
jmercouris
I asked yesterday, but I'll ask again, is someone willing to join a group to collectively review PRs for orphaned projects?
20:09:02
jmercouris
I would like to get as many individuals as possible in on this group to give it some inertia/carry-over into the future
20:14:53
warweasle_afk
Question about PIAIP: Can I create a lazy evaluating prolog style language using a topological sort instead of unification?
20:15:26
fourier
quicklisp-client has 11 open PRs, some of them without comments, with last update 10 months ago. Consider it as an abandoned project? :)
20:47:14
mfiano
sjl: Ah, I was just wondering what you use for vim CL development, since you are the only one I know that uses vim.
20:48:03
sjl
correct, you send text from a scratch buffer and review the results in a separate buffer
20:48:51
sjl
Vim has never really supported the "some parts of the buffer are modifiable but not others" model
20:50:12
mfiano
I used slimv for about 6 years before I moved to Emacs, but I am getting reacquainted with [n]vim recently.
20:51:15
mfiano
I haven't looked into the documentation for nvim's terminal yet, but at quick glance, it doesn't let me enter normal mode
20:53:36
mfiano
So what do you do to get a terminal running your implementation's REPL in the same context as the code you are sending to vlime's internal REPL?
20:54:18
sjl
so, the general idea with SLIME/SWANK is that you run a lisp process and fire up a swank server inside it, which listens for incoming connections
20:54:30
sjl
then in your editor, emacs/vim/whatever you connect to that server over tcp or whatever
20:54:50
sjl
in my case, I'm doing this, but I happen to be starting the Lisp process from within a Neovim split
20:55:35
mfiano
I see. I'll look into this. It's going to take me some getting used to away from Sly though (a more featureful SLIME).
20:56:27
sjl
but I can't get Vim out of my fingers at this point (evil mode was a pale imitation last time I tried it) so I make due with vlime
20:56:52
sjl
the real solution would be for me to take a year off and make my own terminal/ncurses editor in CL
20:57:01
mfiano
I am, but it's not practical on one of my devices (an old netbook which I don't even like running Xorg on due to the memory restrictions)
21:00:02
mfiano
I have a lot of experience developing CL in both editors, but I also prefer vim, mainly because it uses far less resources, and is very snappy on old hardware.
21:11:21
jfb4
sjl: so to deploy in production on a remote server, would you use the same mechanics of having swank running on remote?
21:20:12
jmercouris
phoe: I'm trying to organize a group of people on github to fork/share responsibility in accepting/reviewing PRs from orphaned repositories
21:24:00
jmercouris
jackdaniel: I assume you are correct about it's purpose, therefore this would seem to serve a different one
21:28:15
jmercouris
Xach: What are your criteria for forking/accepting new projects? What is your process?
21:28:48
Xach
jmercouris: if someone notices a project with no maintainer, or a maintainer who is actively giving up maintenance, we will add it to sharplispers on github.
21:29:02
Xach
There isn't much process. Sometimes it is done by email. Sometimes it is done through issues or irc.
21:30:06
Xach
jmercouris: "can i join sharplispers" and i respond "ok, what is your github user id" and then i do it.
21:30:54
jmercouris
Xach: can i join sharplispers? can you also make me an admin so that I may add new people?
21:31:27
Xach
In this case, I missed some context. Why do you want to join? Is there a specific project?
21:33:05
jmercouris
phoe: Yes, this is the discussion Xach and I right now, I am asking him to join the group with admin rights so I may add people
21:33:07
Xach
jmercouris: Is there a specific orphaned project you wish to maintain via sharplispers?
21:34:22
Xach
jmercouris: It would help. The goal of sharplispers is not to grow and expand sharplispers, but to keep useful things from falling entirely by the wayside into neglect.
21:35:22
jmercouris
Xach: In order to satisfy goal 1: Keep useful things from falling entirely by the wayside into neglect, it is necessary to 2. Have a large community of members that can sustain themselves over time and maintain a consistent knowledge pool
21:35:42
jmercouris
When you have such a small group of only 8 people, it only takes the loss of a few members for the percent churn to be so great that so much knowledge is immediately lost
21:35:47
galdor
from an outsider who has barely the time to hack some lisp for fun, I'm much more concerned about maintaining quality than having more developers
21:36:23
jmercouris
galdor: This isn't about expanding and developing new projects, one doesn't need to be an expert Lisper to review PR fixes and tests
21:36:51
jmercouris
Of course one must always be careful to avoid misstep, but it is certainly an easier task than programming novel systems
21:37:55
jmercouris
Xach: Sharplispers to me sounds more like a guild than anything else the way you are describing it
21:39:01
sjl
jfb4: if I wanted the thing to be debuggable in production, I could run a swank server in the production process and just make absolutely sure it's only listening on localhost, yes
21:40:22
rumbler31
sjl: I think by default its listening on localhost, you need to explicitly change a global to make swank listen on a specific interfaces
21:42:16
sjl
the only stuff I've "deployed" in CL so far has been some games running under a telnet server. for those I just disable the debugger entirely -- if they crash oh well
21:42:33
galdor
I now just fork what I need, and add everything as submodules in my "site-lisp" repository
21:43:00
Xach
jmercouris: Maybe a guild is a good description. The existing people are those who know and trust each other. People who have asked to join have had a good reputation and history.
21:43:22
galdor
in general you cannot really expect your PR to be merged as fast as you need it (i.e. right now), so having your fork and sending the PR separately seems to be a good idea
21:44:06
jmercouris
Xach: Then what you have, and what I would like to have are two different concepts, groups with a similar mission, but I would like to be significantly more inclusive
21:45:21
Xach
jmercouris: Would you still like to be added, and add cl-string-match and maintain it?
21:45:57
Xach
jmercouris: can you tell me more about the neglect of cl-string-match? Its author seems to be recently active on other projects.
21:46:37
jmercouris
Xach: Yeah, there's a several compiler warnings when loaded, and some incomplete sections of the API IIRC
21:47:38
sjl
jmercouris: https://bitbucket.org/vityok/cl-string-match/pull-requests/ doesn't have any open PRs I can see
21:48:24
jmercouris
and without an update in many many months, I am inclined to believe those things will not be completed
21:48:47
sjl
in the CL world I treat anything that's had commits in the past couple of years as "still active"
21:49:55
sjl
but yeah, if it's missing stuff you want, the first thing to do would be to PR the original repo
21:50:35
sjl
If it turns out that the author doesn't want to maintain it anymore, okay, fork it to an orphanage org (ideally with their blessing)
21:51:32
jmercouris
sjl: What do you think, do you prefer Xach's "guild" approach to a code orphanage, or do you think it should be a larger, growing organization that shares responsibility?
21:52:21
jmercouris
I guess one of the issues with github is there doesn't appear to be some sort of democratic pull request review (as far as I can tell) in which a group of people agreeing on consensus could merge changes, this makes it necessary to find "expert, trusted" reviewers
21:54:20
sjl
I generally prefer finding a particular person who's interested in a project and passing maintainership off to them
21:54:40
jmercouris
sjl: These "project" leaders could emerge in what is part of a project organizations
21:54:43
sjl
having a team of people means no one person owns anything, and there's always an air of "someone else will get around to reviewing it"
21:54:55
jmercouris
I think naturally individuals withn an organization will adopt sections of the "codebase" and take ownership
21:55:13
jmercouris
but in this case, the unit of granularity would be a "repository" within a group of repositories
21:56:21
jmercouris
sjl: After reviewing one or two PRs for a particular repository, I think a user will feel a sense of ownership towards that repository, while at the same time having a group of other individuals to support them
21:56:55
jmercouris
Plus if there is anything I know about Lisp users, I know they will not hesitate to tell you when they think you are wrong, so maybe the group review thing could work quite well anyway :P
22:02:10
jmercouris
I'm looking here, and there seems to be a few interesting features: https://help.github.com/articles/enabling-required-reviews-for-pull-requests/
22:04:19
jmercouris
More of the same topic: https://help.github.com/articles/defining-the-mergeability-of-pull-requests/
22:10:24
jmercouris
Using this API here: https://developer.github.com/v3/pulls/reviews/ I could program somethig that says something like "when there is 75% agreeance on merging a PR, and min 3 users, allow a merge" or something like that, thoughts?
22:26:30
Xach
jmercouris: To be frank, I think you are cuckoo and your ideas won't work out. But I also felt that way about drmeister when he started making a new CL, and that is proving me wrong for a long time now. So I guess I am more inclined to wait and see and cheer if things work out.
22:29:09
jmercouris
A lot of people already told me that nobody would use my browser, but I discovered somebody with a dot-file for my browser! So at there's at least one person using it :D
22:29:28
jmercouris
phoe: I haven't embarked on this journey quite yet, I am still thinking about it, but thank you nonetheless
22:29:40
phoe
Xach: a single person cannot feasibly maintain a central package repository for years with some degree of stability, you are completely bonkers
22:31:02
jmercouris
On one hand, I wish for a bigger Lisp community, on the other I like how it seems to be restricted mostly to experts, I don't wish for eternal september
22:31:35
jmercouris
However "small" the Lisp community is, I think there are enough members to maintain a core set of useful packages as part of an organization
22:35:38
fourier
I guess at the end of the day most used packages are maintained by the same group of lisp enthusiast who know each other for years (via internet of course)
22:37:56
fourier
also there were a couple of community-sharing-etc projects, last one I can remember was cl ultraspec
23:16:39
pjb
tom34: if you're already a programmer. If you're a total newbie, Common Lisp: A Gentle Introduction to Symbolic Computation http://www.cs.cmu.edu/~dst/LispBook/ http://www-cgi.cs.cmu.edu/afs/cs.cmu.edu/user/dst/www/LispBook/index.html
23:49:41
sjl
Is there any standard way to peek at a binary stream in CL? There's peek-char and unread-char for character streams, but I can't find anything for binary streams
0:41:49
sjl
I think I'll probably end up using flexi-streams. My next question is how to check whether a stream supports binary i/o.