Re: [Toybox] Did I mention why I didn't implement out of tree build?
> >> You may want to publish a DNS record > > It's dreamhost handling all that automatically. Most warnings are about your domain settings, manual configuration: https://help.dreamhost.com/hc/en-us/articles/215414867-How-do-I-add-custom-DNS-records- > If I had any control over it I'd > have https: on the lists site so I could perform moderation without sending > the > list password unencrypted through http. Https config is completely separate from domain settings. Dreamhost refuses to let users set it up unfortunately, but they do provide dns edit like above. > >> Listed in SORBS (last 28 days) http://www.sorbs.net/lookup.shtml > >> Yellow listed in Hostkarma http://ipadmin.junkemailfilter.com/remove.php To these, one writes personally mentioning your server IP and examples of your letters, they usually agree and mark you as okay. ___ Toybox mailing list Toybox@lists.landley.net http://lists.landley.net/listinfo.cgi/toybox-landley.net
Re: [Toybox] Did I mention why I didn't implement out of tree build?
On 11/14/19 6:58 PM, Denys Nykula wrote: >> The strange thing is that the binutils, gcc & musl lists don't seem to >> cause gmail to barf nearly as much as the toybox list does. Dreamhost _also_ sucks. It's the interaction of gmail and dreamhost that becomes truly pathological. > Mail-tester complains this way about this list server, rating its > likeliness of reaching destination as 32%: > >> Listed in SORBS (last 28 days) http://www.sorbs.net/lookup.shtml >> Yellow listed in Hostkarma http://ipadmin.junkemailfilter.com/remove.php > >> You may want to publish a DNS record (MX type) for the domain name >> lists.landley.net or use a different bounce email address. It's dreamhost handling all that automatically. If I had any control over it I'd have https: on the lists site so I could perform moderation without sending the list password unencrypted through http. (See also "every couple years the web archive stops working and they lose a month of messages resetting it", ala https://landley.net/dreamhost.txt and https://landley.net/dreamhost2.txt and...) Rob ___ Toybox mailing list Toybox@lists.landley.net http://lists.landley.net/listinfo.cgi/toybox-landley.net
Re: [Toybox] Did I mention why I didn't implement out of tree build?
> The strange thing is that the binutils, gcc & musl lists don't seem to > cause gmail to barf nearly as much as the toybox list does. Mail-tester complains this way about this list server, rating its likeliness of reaching destination as 32%: > Listed in SORBS (last 28 days) http://www.sorbs.net/lookup.shtml > Yellow listed in Hostkarma http://ipadmin.junkemailfilter.com/remove.php > You may want to publish a DNS record (MX type) for the domain name > lists.landley.net or use a different bounce email address. > You do not have a DMARC record, please add a TXT record to your domain > _dmarc.lists.landley.net with the following value: v=DMARC1; p=none > [SPF] lists.landley.net does not allow your server > to use toybox-boun...@lists.landley.net > You do not have a SPF record, please add the following one to your domain > lists.landley.net: v=spf1 a mx include:netblocks.dreamhost.com ~all > Your message is not signed with DKIM > Your message does not contain a List-Unsubscribe header Had to sort all these out for clients' mailing lists before their mail to major providers stopped breaking. Though it was more than a year ago, providers might have more rules now than mail-tester knows of. I also haven't seen any other mailing list sending its own additional copy when a recipient is included in CC. So providers might consider getting same letter twice weird, automatically blacklisting your server when it repeatedly does that. ___ Toybox mailing list Toybox@lists.landley.net http://lists.landley.net/listinfo.cgi/toybox-landley.net
Re: [Toybox] Did I mention why I didn't implement out of tree build?
On Fri, Nov 15, 2019 at 4:17 AM Rob Landley wrote: > > On 11/13/19 9:25 AM, Jarno Mäkipää wrote: > > It seems that gmail stopped receiving emails from list again. I got > > kicked out of list for excessive bounces yesterday... Somehow gmail > > worked for month or two after I complained about it last time. > > > > -Jarno > > Oh I still get unsubscribed by gmail weekly. Its spam filter remains > profoundly > broken. (And thanks for the reminder, I should trawl through my spam filter > again to see what it's false positived on before it's been a month and it > deletes it for me.) > > Just reply to the confirmation email and check the web archive to see what you > missed while unsubscribed. > > (Yeah, I should move to a real mail service, I know...) > If you find one that's as convenient as gmail can you let us know please! I also get unsubscribed from the toybox list at least once a month. The strange thing is that the binutils, gcc & musl lists don't seem to cause gmail to barf nearly as much as the toybox list does. Patrick ___ Toybox mailing list Toybox@lists.landley.net http://lists.landley.net/listinfo.cgi/toybox-landley.net
Re: [Toybox] Did I mention why I didn't implement out of tree build?
On 11/13/19 9:25 AM, Jarno Mäkipää wrote: > It seems that gmail stopped receiving emails from list again. I got > kicked out of list for excessive bounces yesterday... Somehow gmail > worked for month or two after I complained about it last time. > > -Jarno Oh I still get unsubscribed by gmail weekly. Its spam filter remains profoundly broken. (And thanks for the reminder, I should trawl through my spam filter again to see what it's false positived on before it's been a month and it deletes it for me.) Just reply to the confirmation email and check the web archive to see what you missed while unsubscribed. (Yeah, I should move to a real mail service, I know...) Rob ___ Toybox mailing list Toybox@lists.landley.net http://lists.landley.net/listinfo.cgi/toybox-landley.net
Re: [Toybox] Did I mention why I didn't implement out of tree build?
On 11/13/19 9:00 AM, Andrew Ilijic wrote: > Hi Rob, > > I saw this email and thought this patch might help make > `relative_path()` better. Ok, let's see... > https://www.mail-archive.com/toybox@lists.landley.net/msg06136.html > On 08 Nov 2019 18:27:44 -0800, Rob via Toybox wrote: >> In theory if you're using toybox cp you can use a relative path, which was >> part >> of why I did relative_path but darn it it's STILL broken: $ cd ~/toybox/toybox $ make cp $ mkdir sub $ cd sub $ ../cp -rs ../www/. . $ ls -l roadmap.html lrwxrwxrwx 1 landley landley 24 Nov 13 21:04 roadmap.html -> ../../www/./roadmap.html Nope, it's still a broken red link with an extra .. in front of it. Do you have a test case of what this fixes? (The current toybox realpath hasn't got --relative-to...) Rob ___ Toybox mailing list Toybox@lists.landley.net http://lists.landley.net/listinfo.cgi/toybox-landley.net
[Toybox] Did I mention why I didn't implement out of tree build?
It seems that gmail stopped receiving emails from list again. I got kicked out of list for excessive bounces yesterday... Somehow gmail worked for month or two after I complained about it last time. -Jarno > Hi Rob, > I saw this email and thought this patch might help make > `relative_path()` better. > > Commit Message: > While trying to add features to `readlink` and `realpath` I saw that > adding a "/" to the end of the `from` variable, made it possible for > me to get `realpath --relative-to=DIR` to work. > The change seems like a hack, but I saw Rob's email > https://www.mail-archive.com/toybox@lists.landley.net/msg06136.html > and though this might be able to help. This works for `realpath` > because `--relative-to` is only supposed to take a directory. > ~Andrew > PS: I hope this email looks correct. I made it by hand because of > Gmail deciding not to give me the original message. > https://www.mail-archive.com/toybox@lists.landley.net/msg06136.html > > On 08 Nov 2019 18:27:44 -0800, Rob via Toybox wrote: > > In theory if you're using toybox cp you can use a relative path, which was > > part > > of why I did relative_path but darn it it's STILL broken: ___ Toybox mailing list Toybox@lists.landley.net http://lists.landley.net/listinfo.cgi/toybox-landley.net
Re: [Toybox] Did I mention why I didn't implement out of tree build?
Hi Rob, I saw this email and thought this patch might help make `relative_path()` better. Commit Message: While trying to add features to `readlink` and `realpath` I saw that adding a "/" to the end of the `from` variable, made it possible for me to get `realpath --relative-to=DIR` to work. The change seems like a hack, but I saw Rob's email https://www.mail-archive.com/toybox@lists.landley.net/msg06136.html and though this might be able to help. This works for `realpath` because `--relative-to` is only supposed to take a directory. ~Andrew PS: I hope this email looks correct. I made it by hand because of Gmail deciding not to give me the original message. https://www.mail-archive.com/toybox@lists.landley.net/msg06136.html On 08 Nov 2019 18:27:44 -0800, Rob via Toybox wrote: > In theory if you're using toybox cp you can use a relative path, which was > part > of why I did relative_path but darn it it's STILL broken: From e8296ba33bf950894f7f289ec4106f842c222baf Mon Sep 17 00:00:00 2001 From: Andrew Ilijic Date: Wed, 13 Nov 2019 09:48:18 -0500 Subject: [PATCH] Change `relative_path()` so `from` ends with "/" While trying to add features to `readlink` and `realpath` I saw that adding a "/" to the end of the `from` variable, made it possible for me to get `realpath --relative-to=DIR` to work. The change seems like a hack but I saw Rob's email https://www.mail-archive.com/toybox@lists.landley.net/msg06136.html and though this might be able to help. This works for `realpath` because `--relative-to` is only supposed to take a directory. --- lib/lib.c | 4 1 file changed, 4 insertions(+) diff --git a/lib/lib.c b/lib/lib.c index 3bfee2ef..8460db35 100644 --- a/lib/lib.c +++ b/lib/lib.c @@ -1038,6 +1038,10 @@ char *relative_path(char *from, char *to) if (!(from = xabspath(from, -1))) return 0; if (!(to = xabspath(to, -1))) goto error; + size_t len = strlen(from); + from = xrealloc(from, len + 2); + from[len]='/'; from[len + 1]='\0'; + // skip common directories from root for (i = j = 0; from[i] && from[i] == to[i]; i++) if (to[i] == '/') j = i+1; -- 2.11.0 ___ Toybox mailing list Toybox@lists.landley.net http://lists.landley.net/listinfo.cgi/toybox-landley.net
[Toybox] Did I mention why I didn't implement out of tree build?
Because cp -s exists and works. git clone https://github.com/landley/toybox cp -sr $PWD/toybox walrus cd walrus make distclean defconfig toybox In theory if you're using toybox cp you can use a relative path, which was part of why I did relative_path but darn it it's STILL broken: cd ~/toybox mkdir bang cd bang cp -rs ../www/. . Sea of broken links with extra .. at the start. Grrr. Sigh, on the todo heap... (It works with absolute paths, which are what all my scripts still use because the host cp only supports absolute source paths with -s...) Rob ___ Toybox mailing list Toybox@lists.landley.net http://lists.landley.net/listinfo.cgi/toybox-landley.net
