While it sounds like a lot of work - if thats your itch I suppose its a good 
test of the work that Stephane mentions. By redoing it in a different 
environment it might iron out any accidental environment dependencies.

Presumably its just the VM you need to bootstrap - as I would view the image as 
being like a Word or PDF document - and would be surprised that they would 
expect you to bootstrap those too?

Good luck - and courage.

Tim

On Tue, 21 May 2024, at 9:47 AM, Nicolas Graves via Pharo-dev wrote:
> On 2024-05-21 09:34, stephane ducasse wrote:
>
>> Hi 
>>
>> Why can’t you use the latest Pharo vm and Pharo12?
>> We worked like mad over the years to get a much better system. 
>> Now why do you need to bootstrap Pharo because you can just use a 
>> bootstrapped image
>> and get pharo working and running?
>
> Guix is not using the FHS filesystem so it would be a mess to try and
> bootstrap from binaries. On top of that, they focus on
> reproducibility. The standard on this end is to try and not use any
> binary code if it's possible to build if from source, and yes, it
> requires working like mad over the years. The website is currently down,
> but there's an article about how they get all 30k packages from only a
> single tiny binary file.
>
> Some of the guys behind Guix are also involved in bootstrappable.org
>
> I understand and share that the current system is much better, it's just
> that to comply with their standard, that's what I have to do.
>
> What I understand from there it that it's feasible: 
> https://x.com/estebanlm/status/1042719107307712514
>
> I'll do the work if feasible, I just can't find the information where I
> should start from, hence my email.
>
> Best,
> Nicolas
>
>>
>> S
>>
>>> On 21 May 2024, at 09:31, Nicolas Graves via Pharo-dev 
>>> <pharo-dev@lists.pharo.org> wrote:
>>> 
>>> Hi Pharo community,
>>> 
>>> I'm trying to package pharo in the GNU Guix linux distribution. Since
>>> the distribution is focused on reproducibility and avoids using binary
>>> code when possible, for that I need to reproduce the bootstrapping steps
>>> that are embedded in Pharo. I understand from some developper online
>>> comment that this is possible.
>>> 
>>> I've started to work on this and it seems feasible, at least I get the
>>> dependency loop, what I've some difficulty is to get precisely when it's
>>> possible to get started.
>>> 
>>>      -------------------
>>>      |                 v
>>>  pharo VM            pharo
>>>      ∧                 |
>>>      -------------------
>>> 
>>> If I'm correct,
>>> - bootstrapping is very clear after version ~7.0.0, and with a proper
>>> build, just follow the distribution numbers in wget's with a cmake
>>> build-system.
>>> - for earlier versions, there's the repository
>>> https://github.com/pharo-project/PharoBootstrap
>>> that seems (git history) to be able to go back to Pharo 30.
>>> - I understand that at some point, Squeak and Pharo used to have the
>>> same VM. I have access to a squeak-vm 4.10.2.2614 in Guix, and I've seen
>>> this is in Pharo's ftp files, so I guess I can start there.
>>> 
>>> What I need:
>>> - Some indication about the first version of Pharo that I'm able to
>>> build purely from squeak-vm, and where to find its source code
>>> (git+version or commit if possible, I seem to only be able to find its
>>> binaries currently in phero's ftp files).
>>> - Any useful information for early bootstrap steps.
>>> 
>>> Some useful information on my side:
>>> - During the build process in Guix, we can't download data, so the
>>> entire bootstrap chain has to be explicited at once, hence the need to
>>> go back to the beginning in this case.
>>> 
>>> Thanks in advance!
>>> 
>>> -- 
>>> Best regards,
>>> Nicolas Graves
>>
>> Stéphane Ducasse
>> http://stephane.ducasse.free.fr
>> 06 30 93 66 73
>>
>> "If you knew today was your last day on earth, what would you do 
>> differently? ....ESPECIALLY if, by doing something different, today might 
>> not be your last day on earth.” Calvin & Hobbes
>>
>>
>>
>>
>>
>
> -- 
> Best regards,
> Nicolas Graves

Reply via email to